**EnergyPlus Engineering Reference
**

The Reference to EnergyPlus Calculations

(incaseyouwantorneed to know)

Date: April 23, 2005

COPYRIGHT © 1996-2005 THE BOARD OF TRUSTEES OF THE UNIVERSITY OF ILLINOIS AND THE REGENTS OF THE UNIVERSITY OF CALIFORNIA THROUGH THE ERNEST ORLANDO LAWRENCE BERKELEY NATIONAL LABORATORY. ALL RIGHTS RESERVED. NO PART OF THIS MATERIAL MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS WITHOUT THE PRIOR WRITTEN PERMISSION OF THE UNIVERSITY OF ILLINOIS OR THE ERNEST ORLANDO LAWRENCE BERKELEY NATIONAL LABORATORY. ENERGYPLUS IS A TRADEMARK OF THE US DEPARTMENT OF ENERGY.

2

TABLE OF CONTENTS

Overview .......................................................................................................................1 Document Overview........................................................................................ 1 General Modeling Overview............................................................................ 1 Figure 1. EnergyPlus Program Schematic .......................................... 1 Simulation Manager ........................................................................................ 1 Table 1. Simulation Flags ..................................................................... 2 Integrated Solution Manager ........................................................................................3 Figure 2. Sequential Simulation Supply/Demand Relationship. .......... 3 Figure 3. Schematic of Simultaneous Solution Scheme....................... 4 Basis for the Zone and System Integration ..................................................... 4 Zone Volume Capacitance Multiplier ........................................................ 7 Summary of Predictor-Corrector Procedure.................................................... 7 System Control.......................................................................................... 7 Figure 4. Simplified single zone draw through system ........................ 9 Figure 5. Simplified Variable Volume System. ..................................... 9 Figure 6. Idealized variable volume system operation. ..................... 10 Moisture Predictor-Corrector......................................................................... 12 Moisture Prediction ................................................................................. 13 Moisture Correction................................................................................. 14 Zone Update Method .................................................................................... 15 Variable Time Step.................................................................................. 16 Simultaneous Solution of Plant/System Water Loop............................... 16 References.............................................................................................. 17 Surface Heat Balance Manager / Processes .............................................................18 Conduction Through The Walls..................................................................... 18 Conduction Transfer Function Module .................................................... 18

4/23/05 I

TABLE OF CONTENTS

Calculation of Conduction Transfer Functions ........................................ 19 Figure 7. Two Node State Space Example. ...................................... 21 EnergyPlus CTF Calculations ................................................................. 21 Figure 8. Multiple, staggered time history scheme ............................ 23 Figure 9. Sequential interpolation of new histories............................. 23 Figure 10. Master history with interpolation....................................... 24 References.............................................................................................. 24 Moisture Transfer Material Properties........................................................... 24 Typical Masonry Moisture Capacitance .................................................. 25 Figure 11. Masonary Moisture Capacitance Curves ......................... 25 Typical Wood Moisture Capacitance....................................................... 26 Figure 12. Wood Moisture Capacitance Curve.................................. 26 Linear Material Properties Example ........................................................ 26 Table 2. Max & Min Temperatures and Vapor Densities for Atlanta Summer Design Day Simulation ................................................................................... 26 Figure 13. Relative Humidity Curves at Wall Element Interfaces ...... 27 Figure 14. Moisture Capacitance Curve for Face Brick Layer............ 28 Figure 15. Moisture Capacitance for Fiber Glass Layer ..................... 28 Figure 16. Moisture Capacitance for Gypsum Drywall ....................... 29 References.............................................................................................. 29 Effective Moisture Penetration Depth (EMPD) MODEL ................................ 29 Overview ................................................................................................. 29 EMPD Model Description ........................................................................ 29 EMPD value determination ..................................................................... 31 Figure 17. Limit of effective penetration depth values for various vapor diffusivities at different ambient excitations............................................................... 32 EMPD Nomenclature .............................................................................. 32

4/23/05 II

TABLE OF CONTENTS

References.............................................................................................. 33 Outside Surface Heat Balance...................................................................... 33 Figure 18. Outside Heat Balance Control Volume Diagram ............... 33 External Shortwave Radiation................................................................. 34 External Longwave Radiation ................................................................. 34 Table 3. Nomenclature List of Variables............................................. 34 References.............................................................................................. 36 External Convection ................................................................................ 37 Table 4. Terrain-Dependent Coefficients (ASHRAE 2001). ............... 37 Table 5. Roughness Coefficients D, E, and F. .................................. 38 Table 6. Nomenclature List of Variables............................................. 38 Table 7. Surface Roughness Multipliers (Walton 1981). ................... 40 Table 9. Terrain Roughness Coefficients (Walton 1983)................... 42 Table 10. Nomenclature List of Variables.......................................... 42 Table 11. MoWiTT Coefficients (Yazdanian and Klems 1994).......... 43 Table 12. Nomenclature List of Variables.......................................... 43 Exterior Conduction................................................................................. 44 References.............................................................................................. 44 Inside Heat Balance ...................................................................................... 44 Figure 19. Inside Heat Balance Control Volume Diagram................. 45 Internal LW Radiation Exchange............................................................. 45 Internal SW Radiation ............................................................................. 47 Interior Conduction.................................................................................. 47 Interior Convection .................................................................................. 47 Figure 20. Ceiling Diffuser Correlation for Floors .............................. 49 Figure 21. Ceiling Diffuser Correlation for Ceilings ........................... 50

4/23/05 III

TABLE OF CONTENTS

Figure 22. Ceiling Diffuser Correlation for Walls ............................... 50 Transparent Insulation Material (TIM) ........................................................... 51 Introduction ............................................................................................. 51 Comparison of Opaque and Transparent Insulation ............................... 52 Figure 23. Energy flows of opaquely and transparently insulated walls (Wood and Jesch 1993). ....................................................................................... 52 Types of Transparent Insulation Materials .............................................. 52 Figure 24. Geometrical categories of classification for Transparent Insulation Material (Wood and Jesch 1993). .................................................................... 53 Basic Heat Balance -- TARP and FORTRAN Algorithm ............................... 53 Basic Heat Balance Cases...................................................................... 53 Figure 25. Thermal network for simple outside surface coefficient ... 55 Figure 26. Thermal network for detailed outside surface coefficient . 55 Figure 27. Thermal network for outside movable insulation .............. 55 Heat Balance Cases ............................................................................... 55 FORTRAN Algorithm Examples .............................................................. 57 Fortran Variable Descriptions.................................................................. 58 Table 13. Fortran Variables and Descriptions ................................... 58 TIM- Basic Mathematical Model .............................................................. 61 Figure 28. Cross Section of TIM and wall, showing energy flow ....... 61 Sample Test Run Cases: – Comparison................................................. 63 Table 14. TIM with Summer Conditions ............................................ 63 Table 15. TIM with Winter Conditions................................................. 64 References.............................................................................................. 64 Advanced Surface Concepts......................................................................................66 Exterior Naturally Vented Cavity .................................................................. 66 Baffle Heat Balance ................................................................................ 66

4/23/05 IV

TABLE OF CONTENTS

Figure 29. Baffle Surface Heat Balance ............................................ 66 Cavity Heat Balance................................................................................ 68 Figure 30. Plenum Heat Balance....................................................... 68 Underlying Heat Transfer Surface........................................................... 70 Solar and Shading Calculations .............................................................. 70 Local Wind Speed Calculations .............................................................. 70 Table 16. Terrain-Dependent Coefficients (ASHRAE 2001). ............. 70 Convection Coefficients .......................................................................... 71 Radiation Coefficients ............................................................................. 71 References.............................................................................................. 72 Sky and Solar/Shading Calculations ..........................................................................73 Sky Radiance Model ..................................................................................... 73 Figure 31. Schematic view of sky showing solar radiance distribution as a superposition of three components: dome with isotropic radiance, circumsolar brightening represented as a point source at the sun, and horizon brightening represented as a line source at the horizon. ...................................... 73 Table 17. Variables in Anisotropic Sky Model and Shadowing of Sky Diffuse Radiation ............................................................................................ 74 Sky Diffuse Solar Radiation on a Tilted Surface ..................................... 75 Table 18. Fij Factors as a Function of Sky Clearness Range. ........... 76 Shadowing of Sky Diffuse Solar Radiation.............................................. 76 Shadowing of Sky Long-Wave Radiation................................................ 77 Shading Module ............................................................................................ 77 Shading and Sunlit Area Calculations..................................................... 77 Figure 32. Overall Shadowing Scheme Depiction ............................. 78 Solar Position .......................................................................................... 78 Table 19. Relationship of Angles (degrees) to Time ......................... 79 Figure 33. Solar Position Illustration.................................................. 79

4/23/05 V

TABLE OF CONTENTS

Surface Geometry ................................................................................... 79 Figure 34. EnergyPlus Coordinate System ....................................... 80 Shadow Projection .................................................................................. 81 Figure 35. Basic shadowing concept structure.................................. 82 Figure 36. Illustration of Shadow Clipping ......................................... 83 Homogeneous Coordinates .................................................................... 83 Overlapping Shadows ............................................................................. 84 Figure 37. Point a – vertex of A enclosed by B ................................. 85 Figure 38. Surface A totally overlaps Surface B................................ 85 Figure 39. Figure formed from Intercept Overlaps between A and B 86 Figure 40. Complex overlapping condition ........................................ 86 Table 20. Surface / Area Characteristic / Convention ....................... 87 Figure 41. Multiple Shadow Overlaps................................................ 87 Solar Gains ............................................................................................. 87 Solar Distribution..................................................................................... 88 Figure 42. Illustration of Convex and Non-convex Zones.................. 89 Details of the Interior Solar Distribution Calculation ................................ 89 Figure 43. Vertical section through a two-zone building showing where transmitted beam solar falls. Some of the beam solar from exterior window EW is absorbed by the floor, D, interior wall, B, and interior window, IW. Some is transmitted by IW to the adjacent zone, Z2. Aoverlap is the irradiated area of a surface projected back onto the plane of EW. Beam reflected by D, B and IW contributes to the interior short-wave radiation flux in Z1. ............................................................................. 93 Ground Reflectances .............................................................................. 95 Ground Reflectances (Snow) .................................................................. 95 References.............................................................................................. 95 Solar Radiation Reflected from Exterior Surfaces .....................................................96 Figure 44. Examples of solar reflection from shadowing surfaces in the Surface:Shading series. ..................................................................... 97

4/23/05 VI

TABLE OF CONTENTS

Figure 45. Solar reflection from building surfaces onto other building surfaces. ............................................................................................................ 97 Figure 46. Shadowing by the building itself affects beam solar reflection from the ground. ............................................................................................... 98 Diffuse Reflection of Beam Solar and Sky Solar Radiation........................... 98 Receiving points...................................................................................... 98 Figure 47. Vertical rectangular exterior heat transfer surface showing location of receiving points for calculating incident solar radiation reflected from obstructions. ............................................................................................................ 99 Rays ........................................................................................................ 99 Figure 48. Two-dimensional schematic showing rays going outward from a point on a receiving surface. Rays 1-6 hit the ground, rays 7-11 hit an obstruction, and rays 1215 hit the sky. ................................................................................... 100 Sky Solar Radiation Diffusely Reflected from Obstructions .................. 100 Sky Solar Radiation Diffusely Reflected from the Ground..................... 101 Figure 49. Two-dimensional schematic showing rays going upward from a ground hit point.................................................................................................. 101 Beam Solar Radiation Diffusely Reflected from Obstructions ............... 102 Beam Solar Radiation Diffusely Reflected from the Ground ................. 102 Beam Solar Radiation Specularly Reflected from Obstructions ............ 103 Figure 50. Two-dimensional schematic showing specular reflection from an obstruction such as the glazed façade of a neighboring building. The receiving point receives specularly reflected beam solar radiation if (1) DB passes through specularly reflecting surface EF, (2) CD does not hit any obstructions (such as RS), and (3) AC does not hit any obstructions (such as PQ)...................................... 103 Daylighting and Window Calculations ......................................................................105 Daylighting Calculations.............................................................................. 105 Daylight Factor Calculation ......................................................................... 105 Table 21. Variables in Daylighting Calculations .............................. 105 Overview ............................................................................................... 109 Interior Illuminance Components .......................................................... 109 Daylight Factors .................................................................................... 109

4/23/05 VII

TABLE OF CONTENTS

Sky Luminance Distributions................................................................. 110 Figure 51. Angles appearing in the expression for the clear-sky luminance distribution. ....................................................................................... 112 Direct Normal Solar Illuminance............................................................ 112 Exterior Horizontal Illuminance ............................................................. 112 Direct Component of Interior Daylight Illuminance ................................ 113 Figure 52. Geometry for calculation of direct component of daylight illuminance at a reference point. Vectors Rref, W1, W2, W3 and Rwin are in the building coordinate system. ............................................................................................. 114 Internally-Reflected Component of Interior Daylight Illuminance .......... 115 Transmitted Flux from Sky and Ground ................................................ 116 Transmitted Flux from Direct Sun ......................................................... 117 Luminance of Shaded Window ............................................................. 118 Daylight Discomfort Glare ..................................................................... 118 Figure 53. Geometry for calculation of displacement ratios used in the glare formula. .......................................................................................................... 119 Table 22. Position factor for glare calculation.................................. 119 Time-Step Daylighting Calculation .............................................................. 120 Overview ............................................................................................... 120 Table 23. Variables in Time-Step Calculations................................ 120 Time-Step Sky Luminance .................................................................... 122 Interior Illuminance................................................................................ 123 Glare Index............................................................................................ 124 Lighting Control System Simulation ...................................................... 125 Figure 54. Control action for a continuous dimming system............ 126 Figure 55. Stepped lighting control with three steps........................ 127 References............................................................................................ 127 DElight Daylighting Calculations ................................................................. 128

4/23/05 VIII

TABLE OF CONTENTS

DElight Daylight Factor Calculation Differences from EnergyPlus Detailed Methods .............................................................................................................. 129 Figure 56. Bi-directional transmittance data. ................................... 129 DElight Time-Step Interior Daylighting Calculation Differences from EnergyPlus Detailed Methods ................................................................................................ 130 References............................................................................................ 130 Daylighting Devices..................................................................................... 130 Tubular Daylighting Devices ................................................................. 131 Figure 57. Tubular Daylighting Devices........................................... 131 Figure 58. Dome and diffuser surfaces. .......................................... 132 Figure 59. Discrete ray in a pipe...................................................... 132 Figure 60. Pipe transmittance comparison. ..................................... 133 Daylighting Shelves............................................................................... 138 Figure 61. Daylighting Shelf Diagram.............................................. 139 Figure 62. Window and outside shelf as adjacent perpendicular rectangles. Window Calculation Module........................................................................ 142 Optical Properties of Glazing ................................................................ 142 Table 24. Variables in Window Calculations ................................... 142 Glass Layer Properties.......................................................................... 144 Glass Optical Properties Conversion .................................................... 145 Glazing System Properties.................................................................... 145 Figure 63. Schematic of transmission, reflection and absorption of solar radiation within a multi-layer glazing system. .................................................. 146 Table 25: Solar spectral irradiance function. ................................... 147 Table 26: Photopic response function. ............................................ 148 Calculation of Angular Properties.......................................................... 148 Table 27: Polynomial coefficients used to determine angular properties of coated glass. ................................................................................................ 150

4/23/05 IX

141

................. 158 Figure 66...... 172 Edge-Of-Glass Effects ................................................ Slat cell showing arrangement of segments and location of source for calculation of diffuse-to-diffuse optical properties........................................ 150 Optical Properties of Window Shading Devices . 170 Figure 69......... 176 Table 32........ Fortran Variables used in Window/Frame and Divider calculations 176 4/23/05 X 169 175 .. ...................................................................................... ..... Side view of slats showing geometry for calculation of slat edge correction factor for incident direct radiation.......... 154 Figure 64....... reflection and absorption properties for sky and ground diffuse radiation................... . (b) side view of a cell showing case where some of the direct solar passes between adjacent slats without touching either of them...................... 151 Table 28..... ......... .................. 163 Table 29....................................... Comparison of blind optical properties calculated with the EnergyPlus and ISO 15099 methods........................... ........................ Glazing system with two glass layers showing variables used in heat balance equations... 164 Window Heat Balance Calculation ................................................. si. Fortran Variables used in Window Heat Balance Calculations The Glazing Heat Balance Equations ....... for the calculation of direct solar transmittance................. 172 Table 31........... Slat cell showing geometry for calculation of view factors between the segments of the cell.. 169 Table 30................ 155 Figure 65.................... 172 Figure 70: Different types of glass regions................................. Fortran Variables used in Edge of Glass calculations ............................................................................................. Slat Optical Properties...................................... 174 Apportioning of Absorbed Short-Wave Radiation in Shading Device Layers Window Frame and Divider Calculation . In this figure φs is the profile angle and φb is the slat angle............................................... 159 Figure 67.......................................................................................TABLE OF CONTENTS Calculation of Hemispherical Values.................. EnergyPlus values that differ by more than 12% from ISO 15099 values are shown in bold italics............................. 161 Figure 68........ Side view of horizontal slats in a vertical blind showing geometry for calculating blind transmission......... 170 Solving the Glazing Heat Balance Equations.......................... (a) Side view of a cell formed by adjacent slats showing how the cell is divided into segments.................

........................ 203 Figure 80.... 191 Heat Balance Equations for Shading Device and Adjacent Glass .................................... ........ 182 Figure 72................................. 196 Figure 78.................... 199 Airflow Windows.......... 198 Heat Balance Equations for Between-Glass Shading Device .......... Example of shadowing of reveal surfaces by other reveal surfaces................. Expression for area of shaded regions for different shadow patterns: (a) window without frame............................................... 183 Figure 73..... Al.... Not to scale.. .................... ....................................................................................... 204 Figure 81........ 193 Figure 76.TABLE OF CONTENTS Figure 71................................ . 191 Figure 75........ The opening areas Abot... 178 Beam Solar Reflection from Window Reveal Surfaces ......... 208 4/23/05 XI ....... 194 Figure 77.......... (b) window with frame .... (b) Vertical section through shading layer with openings at the top and bottom (side view)...................... Cross section through a window showing frame and divider (exaggerated horizontally)................................ Belgian Building Research Institute.. Glazing system with forced airflow between two glass layers showing variables used in the heat balance equations.................... Examples of openings for an interior shading layer covering glass of height H and width W................. (a) Horizontal section through shading layer with openings on the left and right sides (top view).................... Atop......... Glazing system with two glass layers and a between-glass shading device showing variables used in the heat balance equations........ Glazing system with two glass layers and an interior shading layer showing variables used in heat balance equations.. 1...............................” Version no........ 207 References............. Variation of gap air temperature with distance from the inlet for upward flow........... Vertical section (a) and perspective view (b) of glass layer and interior shading layer showing variables used in the gap airflow analysis............... Gap airflow configurations for airflow windows........................................................ 199 Figure 79.............................................. 192 Solving for Gap Airflow and Temperature . .... 187 Shading Device Thermal Model .............. From “Active facades............... Vertical section through a vertical window with outside and inside reveal showing calculation of the shadows cast by the top reveal onto the inside sill and by the frame onto the inside sill............... 185 Figure 74. Airflow window with between-glass shading device showing variables used in the heat balance equations....... June 2002............................. Ar and Ah are shown schematically.. ................................... 205 Figure 82....

..................................................................................................... Thermal vs....................................... 218 4/23/05 XII ..................................... 210 COMIS ..................................................................................... 215 Initialization Calculations.................. 210 Air Exchange.. 212 Simulation Method ........ 211 References................................................... 215 Central air system simulation ......................209 Convection from Surfaces....................................................................................... 218 Input data ....................................................................................................................................................................................................................................................................... 217 Zone Equipment Simulation ....... 213 Air Loop Simulation ................................................. 209 Infiltration.......... 214 Table 34............................. 209 Convection from Internal Sources................................................................................. Air Flow representations in Buildings.................................................................................................................. 214 Input data ....... 212 Iteration Scheme ................... 210 Table 33..................................... ManageAirLoop Code.......... 209 Infiltration/Ventilation............................... 212 Component Models ...........................................212 Air Loops .................... 213 Determination of Air Mass Flow Rates ............................................................................................... 211 Building System Simulation System Manager / Processes ........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ 213 Primary Air System Simulation....... 210 Calculation of Zone Air Temperature .. 209 Ventilation ......................................................................................... 212 Definition of Air Loop.................TABLE OF CONTENTS Air Heat Balance Manager / Processes .................................................................................................

............................................................................................................................................................ Branch Layout for Individual HVAC Sub-Loops ......... 222 Figure 84................... 233 Primary-Secondary Loop Systems................................................................... 237 Zone Design Loads and Air Flow Rates....... 223 Figure 85..... 228 Summary of Load Distribution Schemes ............... 231 Operation Schemes (Plant and Condenser) ..................... 238 Zone Design Data Arrays .............................................. 219 Simulation .................................................................................................................................................... 219 Direct Air ..................... 225 Figure 86......................... 221 Integration of System and Plant .......... ........ 232 Plant Operation Schemes ........ 230 Figure 87............................................................. 223 Plant Flow Resolver .................................. 221 Current Primary System Modeling Methodology....... 232 Condenser Operation Schemes .............................. 238 4/23/05 XIII . Example of a Heat Recovery Loop Simulation...................................................................................................... Example of a Primary-Secondary Nested Loop Simulation234 Heat Recovery Loop Systems.................................................... Plant/Condenser Supply Side Solution Scheme................TABLE OF CONTENTS Initialization Calculations...237 Sizing Manager ................................................................. ...... 234 Figure 88...................... Connections between the Main HVAC Simulation Loops and Sub-Loops............................................ Demand and Supply Side Loops..................... 230 Summary of Plant Loop Demand Calculation Schemes ..................................................................................................................................................... 235 Figure 89.............................................................. Load Distribution Scheme ... 238 Overview ......................................................................................... 235 Loop and Equipment Sizing............................ 220 Plant/Condenser Loops................................................................................................................................. 221 Figure 83.....................

............................................................................................................................................................. 263 4/23/05 XIV .................................................. 256 DX Coil Sizing ................................................................ 250 Component Sizing........ 250 Hot and Chilled Water Loop Sizing . 250 Introduction ................................................... 238 Zone Design Load Calculation .............................. 242 Table 36......... 260 Boiler Sizing ........................ 240 Updating and Adjusting the Zone Results.................................................................................................................... 260 Cooling Tower Sizing ................. 250 Introduction ..................................................................................................................................................................................... 250 Condenser Loop Sizing... 254 Sizing of Gas and Electric Heating Coils................................................................................................. Zone Sizing Data ................ 261 Fan Coil Unit Sizing................................................................. 250 Fan Sizing ............................................................................................. 262 Window Air Conditioner Sizing................................... 253 COIL:Water:SimpleHeating Sizing ....................... 241 System Design Loads and Air Flow Rates........................................................................................... 251 COIL:Water:DetailedFlatCooling Sizing ...................... 244 Plant Loop Sizing ....... 242 Overview ......................... System Sizing Data ................................................................................................................................................................................................. 251 COIL:Water:SimpleCooling Sizing .................................................................................................................................................. 259 Electric Chiller Sizing ...................................................................... 257 Pump Sizing .................... 242 System Design Flow Rate and Load Summation and Adjustment...................................................................... 242 System Design Data Arrays ....................................................................................TABLE OF CONTENTS Table 35........................

................................................. 268 UCSD Displacement Ventilation Room Air Model.................... Height versus temperature schematic for Mundt model...................................... Average velocity in the recirculation region.......................................................................... 271 References............. Summary of room air models available in EnergyPlus ...........................TABLE OF CONTENTS Single Duct Terminal Units.................................................................... Schematic View ........ Maximum flow rate in the recirculation region..................................................... 263 References............... 278 Figure 92............................................................................................................................. 283 Figure 95........... Top view of the flow structure in case R. Right: top view of one half of the velocity field (at height H/2). 277 Figure 91.................... ...........................................................................................................Room airflow dominated by jet ................ 292 4/23/05 XV ........................................... 265 Table 37..................................................... result of a CFD simulation using the geometry on the left (taking advantage of the system symmetry to simulate only one half of the room volume)..................... 285 Figure 97........... 280 Model Description ..... .................................... Differences between flat plate heat transfer and cross ventilation flow ..................................................... 268 Model Description ............ 265 Figure 90..... 287 Figure 98......................... Top view of the possible airflow patterns in cross-ventilation....... 276 UCSD Cross Ventilation Room Air Model ................................ 264 Alternative Modeling Processes ......................................... Average velocity in the main jet region........................................ 265 Mundt model ....................... 287 Figure 99..................................................................................... Results of the correlation for CV induced airflow velocity near the room surfaces........ 284 Figure 96...................................................................... 269 Implementation....279 Figure 93............................. Left: basic compartment geometry considered in the model..........265 Room Air Models................ 281 Figure 94........... 267 References:......................

.................................................. 295 Overview .................. 292 Implementation.... 293 References.......................................................................... Two Node State Space Example ................................. 304 Figure 102........ Resolution of Radiant System Response at Varying Time Steps 4/23/05 321 XVI .......................................................... 294 Air Distribution System (ADS) Model ................ Schematic ................ Nomenclature for Cross-Ventilation RoomAir Model........................................................................... 319 Figure 109.... 313 Low Temperature Radiant System Controls ... 300 References.......................................................................................a combination of jet and recirculation flow .............. 317 Figure 106............................................. Variable Temperature Low Temperature Radiant System Component Details..................... 295 Model Description .......................... 318 Figure 108....................Room airflow -........................ 309 Extension of Time Series Solutions to Include Heat Sources and Obtain Internal Temperatures............................................. 302 Radiant System Models........ Single Layered Building Element .......... 304 Figure 103.................... 292 Nomenclature for Cross-Ventilation RoomAir Model ......................................................... 309 Figure 104.................................................... Multilayered Building Element ..............................303 Low Temperature Radiant System Model................. 295 Integration of the ADS Model ... 303 Figure 101.............................................................................................................................................................. Two Layer Example for Deriving the Laplace Transform Extension to Include Sources and Sinks .................................................................. Variable Temperature Low Temperature Radiant System Controls ............................................... 293 Table 38................................................. Variable Flow Low Temperature Radiant System Controls318 Figure 107................................ Two Node State Space Example with a Heat Source ..... 310 Figure 105..............................................................TABLE OF CONTENTS Figure 100.... 303 One Dimensional Heat Transfer Through Multilayered Slabs ..........................................................................................................................

................................................... 331 Table 39...................................... Summary of output variables of the ECHILLER SPARK model Figure 113............................ 329 Simulation Models – Encyclopedic Reference........ Summary of parameter variables in the ECHILLER SPARK model 343 Table 43........................................ 336 Chillers: Electric EIR Chiller .... 331 Boilers: SPARK Simple Boiler............................. 333 Chillers: Absorption Chiller.................... 321 High Temperature Radiant Heater Model ........................................... 327 Figure 110....... 333 Chillers: Constant COP Chiller..........................................331 Boilers ...................................... Schematic of the ECHILLER model with feedback loop..............TABLE OF CONTENTS Heat Exchanger Formulation for Hydronic Systems ................................ Summary of input variables in the ECHILLER SPARK model344 Table 44................... Summary of output variables of the SBOILER SPARK model Chillers .................................................................................................................................................. Volumetric Flow for PISSIM1 model..................... 337 Overview ..................................... 337 Model Description .................................... 333 Chillers: Combustion Turbine Chiller.. Diagram of Chiller:Electric with Heat Recovery.................... 345 4/23/05 XVII 331 332 344 ..................................... 335 Chillers: Electric Chiller with Heat Recovery Options (Plant) .....342 Table 42......... 336 Figure 111.................................................................................... 331 Boilers: Simple Boiler ................................................................... Input Description for High Temperature Radiant Heaters328 References............................................ 342 Figure 112...................................................... 337 Chillers: SPARK Electric Chiller ..................................................................................... Summary of parameter variables in the SBOILER SPARK model Table 40.......................................... Summary of input variables in the SBOILER SPARK model332 Table 41.............................

.............................. 383 Coil Model -..................................... Field Data Showing the Net Impact of Part-Load Operation on Sensible Heat Ratio........................... 367 Coil Model -........... 356 Table 48..................... 346 Figure 115... Operating Conditions (From Nodes -...... Evaporator and Condenser Temperature Profiles............ Surface Area Fraction Wet Vs Air mass Flow Rate...... 356 Table 47...........................................................not user inputs)..........................................Electric Heating (HVAC) ......................................................Detailed Cooling Coil (HVAC) ............................................................................DX Heating Coil Model (HVAC) ....................................................... Design Inputs (User Inputs).. 357 Figure 118.............. 349 Figure 116............................. 347 Table 45......... Coil Geometry and flow variables for coils .. 366 Figure 120........DX Cooling Coil Model (HVAC) .. 349 Table 46. 378 Coil Model -...................... 350 Coil Model – Water Cooling Coil (HVAC) .Desuperheater Heating Coil ........... Concepts of Moisture Buildup and Evaporation ............. 357 Table 49..................TABLE OF CONTENTS Figure 114.......................................... 377 Figure 123.. 368 Figure 122................. ......................................... 381 Coil Model – Gas Heating (HVAC) ... 348 Coils ........ 377 Figure 124.... Air Outlet Temperature Vs Air Mass Flow Rate ................. Simplified Schematic of Cooling/Dehumidifying Coil....................... 349 Coil Model -................................... Sensible Load variations Vs Air mass Flow Rate............. Refrigerant Mass Flow ............ 355 Figure 117...... 391 4/23/05 XVIII ................................ 382 Coil Model -......................... 366 Figure 119...................... List of break variables in ECHILLER model....... Simplified Schematic of Cooling/Dehumidifying Coil..................... Total and Sensible Load variations Vs Air mass Flow Rate367 Figure 121..... UA Descriptions of Model ................. 367 References..... Transient Sensible and Latent Capacity of a Cooling Coil Over an Operating Cycle.....................

.......................................................... Cooling Tower Schematic ................... Desuperheater heating coil used as a reheat coil with DXSystem:AirLoop ..........................................Stand Alone (HVAC) ..Heat Exchanger Assisted Cooling Coils (HVAC) .................................................................... 411 Energy Recovery Ventilator..... 407 Figure 131................................................................................................................................................... Heating ...................... Demand Control Ventilation -........... 397 Coil Model – Water to Air HeatPump.... Schematic of a heat exchanger assisted cooling coil................................................ 408 References.... 400 Figure 129........ 392 Coil Model -............... 401 Model Description ............... Cooling ..............................Air Flow Paths........ 398 Controller:Outside Air........ 397 Controller:Simple............................... 406 Demand Controlled Ventilation .......................................................................... Schematic of the Energy Recovery Ventilator:Stand Alone compound object ..............................TABLE OF CONTENTS Figure 125........................... Two Set Point managers used in Controller:Simple................. 401 Overview ....................................... 411 Electric Load Center Generators.............. 400 Cooling Towers (Plant)...................... 396 Figure 126.. Simple Controller used with Central Chilled Water Coil 397 Figure 128........................................................... 411 4/23/05 XIX ............................................. 397 Controllers (HVAC) ......................................................................................................................................................................................................... 404 Calculating the Actual Exiting Water Temperature and Fan Power ......... 401 Figure 130.................................................................................... 400 Controller:Stand Alone ERV..................................... 396 Coil Model – Water to Air HeatPump..... 410 Electric Load Center Distribution Manager ............... 405 References................... 397 Figure 127...... 404 Method for Calculating Steady-State Exiting Water Temperature...................................................................................................

...................................................................................................................... 416 Figure 135...... Two Stage Evaporative Cooler......... Evaporative Cooler -....... Secondary Air Process – Indirect Wet Coil Evap Cooler421 Two Stage: Indirect Staged with a Direct ............................................................... 423 Figure 141.............................................................. Secondary Air Process -................. 420 Figure 138........ 423 Model ............................. 418 Figure 136....................................................................................................................................................TABLE OF CONTENTS Evaporative Coolers.................................. 422 Figure 140........ Psychrometric Chart -............Indirect Dry Coil Evap Cooler419 EvapCooler:InDirect:WetCoil ........ 434 UnitarySystem : HeatPump : AirToAir (HVAC)......................... 416 Figure 134..........................Constant Enthalpy ........................................................... 423 Fans ........................... Direct Stage Evaporative Cooler .. 419 Figure 137.................................................................................... 420 Figure 139...... Schematic for Blow Thru Furnace with High Humidity Control UnitarySystem : BlowThru : HeatOnly or HeatCool (HVAC) ................ 425 Furnaces...................... 426 Furnace : BlowThru : HeatOnly or HeatCool (HVAC) ................. 428 Figure 144..................................................... 423 Overview ....................... 434 4/23/05 XX 433 .... 415 Figure 133......... Schematic of EnergyPlus DX System in Air Loop for a Blow-Thru Application ................................................................ Heat Pumps and Unitary Systems ... 428 Figure 143........................... 415 EvapCooler:Direct:CelDekPad.............................................Indirect Wet Coil ...................... 424 References.............. 425 DXSystem:AirLoop.............................. Graph of Saturation Efficiency......................... Evaporative Cooler -.... 425 Figure 142..... Thermodynamic process for Supply Air .........................Indirect Dry Coil............................................ Schematic of the EnergyPlus Blowthru Furnace ....... 418 EvapCooler:InDirect:CelDekPad...............

Air to Air Heat Exchanger with Supply Air Temperature Control References.......................................... Ground ................................. 462 References............................................................. 465 Model ............................................ 442 Heat Exchanger(s): Air............... Schematic of a Blowthru Air-to-Air Heat Pump .................................................... Pond model component configuration (Chiasson 1999)457 Surface Ground Heat Exchanger ................... Water. 462 Model ................................... 442 Heat Exchanger:Air to Air:Generic ......... Information Flowchart for model implementation (Jin 2002) 440 Figure 148........................................................................................................................................................ 455 Figure 151............................... 443 Table 50................................. Operating Conditions for Defining Heat Exchanger Performance Figure 150......................................................................................................... 437 Water to Water Heat Pumps .......................................................................... Schematic of the Generic Air-to-Air Heat Exchanger .................... 465 Constant Volume Single Duct:4 Pipe ............................................................................................................... 454 Pond Ground Heat Exchanger ...... 460 Humidifiers (HVAC)........................ Flow diagram of EnergyPlus Water to Water HeatPump implementation .......................................................................................... 439 Parameter estimation procedure:......... 434 Unitary System: HeatPump: Water To Air (HVAC) ................................... 442 Figure 149................. 461 Electric Steam Humidifier........................................................................ 465 4/23/05 XXI 443 454 .............................. 461 Overview ........................................................................................ Heat transfer mechanisms in a Pond (Chiasson 1999) 455 Figure 152.................................................................... 438 Figure 146........... 465 Induction Units ........TABLE OF CONTENTS Figure 145....... Schematic indicator diagram for a reciprocating Compressor(Jin 2002) .............................................................. 439 Figure 147.......................................

.................................................. Control of Outdoor Temperature Controlled Baseboard Heat Distribution of Radiant Gains .................. 471 Photovoltaic Arrays ..... 471 Table 51..................................... 483 Pump Heat Addition to the Loop ................................................................................................ 472 Table 52............................TABLE OF CONTENTS References............................................... 484 4/23/05 XXII 470 ..................... 470 References....... 467 Sources and Types of Gains .................................................................. 471 Simple Model ............................. 483 Purchased Chilled Water ........................................ 467 Heat Gain from Lights ...................................................................... Nomenclature for Sandia PV model .............. 473 Figure 155................................................................................. 484 Set Point Managers (HVAC) ... 467 Internal Gains (Heat Balance)........................................................................................................................................................................................................................................................................................................................ 481 Variable Speed Pump ................................................... 474 Sandia Photovoltaic Performance Model .................................. Sensible Heat Gain from People Correlation ................................................................................... 477 Table 53. 469 Heat Gain from Baseboard Heat ................................................................. Equivalent circuit in the four parameter model ........................ 484 Purchased Hot Water.... 478 Pumps ............................................................................. 482 Constant Speed Pump ............... 469 Figure 154.............................................. 472 Equivalent One-Diode Model ..................................................... Nomenclature for Simple Photovoltaic model...... 467 Figure 153...................................................................................................................................................................................................... 467 Heat Gain from People... 484 Overview ......................... General Nomenclature for the PV model...........

.................................................................................................................................. 486 Model Description ...................... 505 Table 55... ................................................. 487 Solar Collectors................................ Collector Heat Balance..................................... Seven point Thermal Sensation Scale ....................................................................................... 504 Thermal Comfort (Heat Balance) . Plenum Air Heat Balance .................. 502 References....................................... 484 Single Zone Reheat ..................................................................... 495 Figure 157.......... 491 Unglazed Transpired Solar Collectors .. 487 Outside Air Pretreat.............................. 507 4/23/05 XXIII ................ 486 Mixed Air ....................................................................................................................................................... 487 Coldest .......................... 486 Single Zone Max Hum... 484 Outside Air ........................................................... Nine point Thermal Sensation Scale ..................................................................................TABLE OF CONTENTS Scheduled ............................... 485 Single Zone Heating.......................................... 506 Table 56................................... General Nomenclature list for Thermal Comfort Models ............................................. 507 Table 57.............................................................. 491 Figure 156................................................................................................................................................................................................................ 487 Flat-Plate Solar Collectors ........................ 485 Single Zone Min Hum................................................................................. 505 Mathematical Models for Predicting Thermal Comfort ........................................................... Terrain-Dependent Coefficients (ASHRAE 2001)........... 485 Single Zone Cooling..................................................................................... 505 Background on Thermal Comfort Models....... 499 Table 54............................................................... 488 References.......................................................................................... 487 Warmest............

.............................. 508 Pierce Two-Node Model...... 525 Figure 159................... 517 MRT Calculation............................... Building with Trombe Wall... 526 Figure 160................................. 530 Controls and Outside Air ........................................................................................................ 507 Table 58............................................................ 517 Table 60........................................ 525 Passive Trombe Wall ................................................................ 530 Unit Ventilator (HVAC) .......... 522 References........................................................................................................................................................................ 524 Figure 158............... 522 Table 61....... Passive Trombe Wall Summer.......... 525 Results ................ 511 Table 59.......................................................................................................................................... 528 Results ........................................... Passive Trombe Wall Winter .............................. Nomenclature list for KSU Two-Node model.............................. 529 Controls.......... 527 Active Trombe Wall ....................... Nomenclature list for Pierce Two-Node model ..................... Nomenclature list for Fanger model ........................................................................................................................................................................................................................................ 527 Input File .............................................. Active Trombe Wall Winter.. 525 Input File ....................................TABLE OF CONTENTS Fanger comfort Model ........................................................ 511 KSU Two-Node Model ........................................... 527 References............................................................................................................................................. Nomenclature and variable list for MRT calculation ..................................................................................................................................... 530 VAV Terminal Units: Variable Speed Fan ....... 532 4/23/05 XXIV ................... 529 Unit Heater (HVAC)............................................... 523 Trombe Walls ............ 528 Figure 161...........................................................................

TABLE OF CONTENTS

Overview ............................................................................................... 532 Model .................................................................................................... 532 References............................................................................................ 534 Water Heaters ............................................................................................. 534 Mixed Water Heater .............................................................................. 534 Simple Water Heater............................................................................. 538 Figure 162. Hot Water Heater in Plant Loop Context ....................... 539 Water Mains Temperatures......................................................................... 540 References............................................................................................ 540 Zone Controls (HVAC) ................................................................................ 541 Thermostatic Zone Control.................................................................... 541 Zone Thermostats ................................................................................. 541 Humidistat ............................................................................................. 543 Economics Calculations ...........................................................................................545 Component Costs ....................................................................................... 545 Line Item Costs ..................................................................................... 545 Special Modules/Reporting.......................................................................................551 Report Environmental Impact Factors......................................................... 551 Types of Pollutants................................................................................ 551 Figure 163. Example Annual Atmospheric Pollutants and Water Consumption 552 Figure 164. Example Annual Total Carbon Equivalent for Major Greenhouse Gases .......................................................................................................... 552 Carbon Equivalent................................................................................. 552 Table 62. Carbon Equivalents (IPCC 2001) ..................................... 553 Fossil Fuel Emissions Factors .............................................................. 553 Table 63. Emission Factors for Natural Gas.................................... 553 Table 64. Emission Factors for Residual Fuel Oil (No. 4 and No. 6 Fuel Oil)

4/23/05

554

XXV

TABLE OF CONTENTS

Table 65. Emission Factors for Distillates (No. 1 and No. 2 Fuel Oil)554 Table 66. Emission Factors for Residential Oil Furnace ................. 555 Table 67. Emission Factors for LPG (butane and propane) ............ 555 Table 68. Emission Factors for Gasoline and Diesel....................... 556 Table 69. Emission Factors for Coal ............................................... 556 Off-Site Electricity Generation Emissions ............................................. 557 Table 70. United States National Average Emission Factors for Electricity Generation .......................................................................................................... 557 Table 71. U. S. State Average Greenhouse Gas Emission Factors for Electricity Generation, in g/MJ .......................................................................... 558 Table 72. U. S. State Average Criteria Pollutant Emission Factors for Electricity Generation, in g/MJ .......................................................................... 559 Table 73. United States National Average Water Consumption Factorsa 560

Table 74. U.S. State Average Water Consumption Factors for Electricity Generationa .......................................................................................................... 561 Other Energy-Related Pollutants and Sources of Other Information .... 562 References............................................................................................ 562 Theoretical Models ...................................................................................................564 Indirect Evaporative Cooler Model -- RDDSpecial ...................................... 564 Figure 165. RDDSpecial Evap Cooler ............................................. 564 Water Consumption .............................................................................. 566 Figure 166. RDDSpecial Evap Cooler, with relief valve .................. 566

4/23/05

XXVI

OVERVIEW

DOCUMENT OVERVIEW

Overview

Document Overview

This document is organized to give you the best possible look into the EnergyPlus calculations. First, the concepts of modeling in EnergyPlus are presented. These include descriptions of the heat balance process, air loop/plant loop processes as well as other important processes for the building simulation. Discussions during the modeling process may reference specific “object names” as found in the Input/Output Reference document. The remainder of the document focuses on individual models.

**General Modeling Overview
**

The EnergyPlus program is a collection of many program modules that work together to calculate the energy required for heating and cooling a building using a variety of systems and energy sources. It does this by simulating the building and associated energy systems when they are exposed to different environmental and operating conditions. The core of the simulation is a model of the building that is based on fundamental heat balance principles. Since it is relatively meaningless to state: “based on fundamental heat balance principles”, the model will be described in greater detail in later sections of this document in concert with the FORTRAN code which is used to describe the model. It turns out that the model itself is relatively simple compared with the data organization and control that is needed to simulate the great many combinations of system types, primary energy plant arrangements, schedules, and environments. The next section shows this overall organization in schematic form. Later sections will expand on the details within the blocks of the schematic.

Figure 1. EnergyPlus Program Schematic

Simulation Manager

The simulation manager of EnergyPlus is contained in a single module. The main subroutine is shown below. Flow within the entire program is managed using a series of flags. These paired flags, in order (from the highest to the lowest) are:

4/23/05 1

OVERVIEW

SIMULATION MANAGER

Table 1. Simulation Flags Begin SimulationFlag End SimulationFlag BeginEnvironmentFlag EndEnvironmentFlag(one to many days) BeginDayFlag EndDayFlag BeginHourFlag EndHourFlag BeginTimeStepFlag EndTimeStepFlag There is also a WarmupFlag to signal that the program is in warmup state. The operation of these flags can be seen in the following subroutine. The advantage of using the flag system is that any subroutine throughout the code can determine the exact state of the simulation by checking the status of the flags.

SUBROUTINE ManageSimulation ! Main driver routine for this module BeginSimFlag = .TRUE. EndSimFlag = .FALSE. CALL OpenOutputFiles CALL GetProjectData CALL GetEnvironmentInfo ! Get the number and type of Environments DO Envrn = 1, NumOfEnvrn ! Begin environment loop ... BeginEnvrnFlag = .TRUE. EndEnvrnFlag = .FALSE. WarmupFlag = .TRUE. DayOfSim = 0 DO WHILE ((DayOfSim.LT.NumOfDayInEnvrn).OR.(WarmupFlag)) ! Begin day loop ... DayOfSim = DayOfSim + 1 BeginDayFlag = .TRUE. EndDayFlag = .FALSE. DO HourOfDay = 1, 24 ! Begin hour loop ... BeginHourFlag = .TRUE. EndHourFlag = .FALSE. DO TimeStep = 1, NumOfTimeStepInHour ! Begin time step (TINC) loop ... BeginTimeStepFlag = .TRUE. EndTimeStepFlag = .FALSE. ! Set the End__Flag variables to true if necessary. Note that each flag builds on ! the previous level. EndDayFlag cannot be .true. unless EndHourFlag is also .true., etc. ! Note that the EndEnvrnFlag and the EndSimFlag cannot be set during warmup. ! Note also that BeginTimeStepFlag, EndTimeStepFlag, and the ! SubTimeStepFlags can/will be set/reset in the HVAC Manager. IF ((TimeStep.EQ.NumOfTimeStepInHour)) THEN EndHourFlag = .TRUE. IF (HourOfDay.EQ.24) THEN EndDayFlag = .TRUE. IF ((.NOT.WarmupFlag).AND.(DayOfSim.EQ.NumOfDayInEnvrn)) THEN EndEnvrnFlag = .TRUE. IF (Envrn.EQ.NumOfEnvrn) THEN EndSimFlag = .TRUE. END IF END IF END IF END IF CALL ManageWeather CALL ManageHeatBalance BeginHourFlag = .FALSE. BeginDayFlag = .FALSE. BeginEnvrnFlag = .FALSE. BeginSimFlag = .FALSE. END DO ! ... End time step (TINC) loop. END DO ! ... End hour loop. END DO ! ... End day loop. END DO ! ... End environment loop. CALL CloseOutputFiles RETURN END SUBROUTINE ManageSimulation

4/23/05

2

INTEGRATED SOLUTION MANAGER

SIMULATION MANAGER

**Integrated Solution Manager
**

EnergyPlus is an integrated simulation. This means that all three of the major parts, building, system, and plant, must be solved simultaneously. In programs with sequential simulation, such as BLAST or DOE-2, the building zones, air handling systems, and central plant equipment are simulated sequentially with no feedback from one to the other. The sequential solution begins with a zone heat balance that updates the zone conditions and determines the heating/cooling loads at all time steps. This information is fed to the air handling simulation to determine the system response; but that response does not affect zone conditions. Similarly, the system information is passed to the plant simulation without feedback. This simulation technique works well when the system response is a well-defined function of the air temperature of the conditioned space. For a cooling situation, a typical supply and demand situation is shown schematically in the Figure 2. Here, the operating point is at the intersection of the supply and demand curves.

Figure 2. Sequential Simulation Supply/Demand Relationship. However, in most situations the system capacity is dependent on outside conditions and/or other parameters of the conditioned space. The simple supply and demand situation above becomes a more complex relationship and the system curve is not fixed. The solution should move up and down the demand curve. This doesn’t happen in sequential simulation methods and the lack of feedback from the system to the building can lead to nonphysical results. For example, if the system provides too much cooling to a conditioned space the excess is reported by the program as "overcooling". Other categories of unmatched loads exist and are similarly reported by the program. While this kind of reporting enables the affected system or plant components to be properly sized, the system designer would, in most cases, prefer to see the actual change in zone temperature. The same mismatches can occur between the system and plant simulations when they are simulated sequentially. To obtain a simulation that is physically realistic, the elements have to be linked in a simultaneous solution scheme. The entire integrated program can be represented as a series of functional elements connected by fluid loops as shown in Figure “Schematic of Simultaneous Solution Scheme”. In EnergyPlus all the elements are integrated and controlled by the Integrated Solution Manager. The loops are divided into supply and demand sides, and the solution scheme generally relies on successive substitution iteration to reconcile supply and demand using the Gauss-Seidell philosophy of continuous updating.

4/23/05

3

INTEGRATED SOLUTION MANAGER

BASIS FOR THE ZONE AND SYSTEM INTEGRATION

Figure 3. Schematic of Simultaneous Solution Scheme In the sections which follow, the various individual functions of the integrated solution will be described.

**Basis for the Zone and System Integration
**

The basis for the zone and system integration, Taylor.(1990, 1991), incorporates a shortened simulation time step, typically to between 0.1 and 0.25 hours, and uses a time-marching method having the zone conditions lagged by one time step. The error associated with this approach depends significantly on the time step. The smaller the step size the smaller the error, but the longer the computation time. To permit increasing the time step as much as possible while retaining stability, zone air capacity was also introduced into the heat balance. The resulting method is called “lagging with zone capacitance”. Although requiring substantially more time to execute than sequential simulation methods, the improved realism of the simultaneous solution of loads, systems and plants simulation is desirable. This method was fully implemented in the program IBLAST (Integrated Building Loads Analysis and System Thermodynamics) (Taylor 1996) that was used as a basis for EnergyPlus. The method of lagging with zone capacitance uses information from previous time steps to predict system response and update the zone temperature at the current time. In older sequential programs, one hour is used frequently as a time step because it is convenient for record keeping purposes and it keeps computation time reasonable. But dynamic processes in the zone air can occur on a much shorter time scale than one hour. The time constant, τ , for a zone is on the order of:

τ≈

ρVc p

Qload + Qsys

(1)

where the numerator is the zone air heat capacitance and the denominator is the net rate of heat energy input. Clearly, the value of τ can vary because the zone load and system output change throughout the simulation. Therefore, a variable adaptive time step shorter than one hour is used for updating the system conditions. For stability reasons it was necessary to derive an equation for the zone temperature that included the unsteady zone capacitance term and to identify methods for determining the zone conditions and system response at successive time steps. The formulation of the solution scheme starts with a heat balance on the zone.

4/23/05

4

**INTEGRATED SOLUTION MANAGER
**

N sl dT Cz z = ∑ Qi + dt i =1 N surfaces N zones i =1

BASIS FOR THE ZONE AND SYSTEM INTEGRATION

∑

i =1

i

hi Ai (Tsi − Tz ) +

∑ m C (T

i p

zi

− Tz ) + minf C p (T∞ − Tz ) + Qsys

(2)

where:

∑Q

i =1 N surfaces

N sl

= sum of the convective internal loads

∑

i =1

hi Ai (Tsi − Tz ) = convective heat transfer from the zone surfaces

**minf C p (T∞ − Tz ) = heat transfer due to infiltration of outside air
**

N zones i =1

∑ m C (T

i p

zi

− Tz ) = heat transfer due to interzone air mixing

Qsys = system output.

Cz

dTz = energy stored in zone air dt

If the air capacitance is neglected, the steady state system output must be: - Qsys =

∑ Qi +

i =1

N sl

N surfaces

∑

i =1

hi Ai (Tsi − Tz ) +

N zones i =1

∑ m C (T

i p

zi

− Tz ) + minf C p (T∞ − Tz )

(3)

Air systems provide hot or cold air to the zones to meet heating or cooling loads. The system energy provided to the zone, Qsys , can thus be formulated from the difference between the supply air enthalpy and the enthalpy of the air leaving the zone as in Equation (4):

Qsys = msys C p (Tsup − Tz )

(4)

This equation assumes that the zone supply air mass flow rate is exactly equal to the sum of the air flow rates leaving the zone through the system return air plenum and being exhausted directly from the zone. Both air streams exit the zone at the zone mean air temperature. The result of substituting Equation (4) for Qsys in the heat balance Equation (2) is shown in Equation (5):

Cz

dTz Nsl = ∑ Qi + dt i =1

N surfaces

∑

i =1

hi Ai (Tsi − Tz ) +

N zones i =1

∑ m C (T

i p

zi

− Tz )

+ minf C p (T∞ − Tz ) + msys C p (Tsup − Tz )

(5)

The sum of zone loads and system output now equals the change in energy stored in the zone. Typically, the capacitance Cz would be that of the zone air only. However, thermal masses assumed to be in equilibrium with the zone air could be included in this term. In order to calculate the derivative term a finite difference approximation may be used, such as:

dT −1 = (δ t ) (Tz t − Tz t −δ t ) + Ο (δ t ) dt

4/23/05

(6)

5

INTEGRATED SOLUTION MANAGER

BASIS FOR THE ZONE AND SYSTEM INTEGRATION

The use of numerical integration in a long time simulation is a cause for some concern due to the potential build-up of truncation error over many time steps. In this case, the finite difference approximation is of low order that further aggravates the problem. However, the cyclic nature of building energy simulations should cause truncation errors to cancel over each daily cycle so that no net accumulation of error occurs, even over many days of simulation (Walton, 1990). The Euler formula, Equation (6), was employed in Equation (5) to replace the derivative term. All the terms containing the zone mean air temperature were then grouped on the left hand side of the equation. Since the remaining terms are not known at the current time, they were lagged by one time step and collected on the right hand side. This manipulation resulted in Equation (7), the formula for updating the zone mean air temperature:

N zones ⎛ Nsurfaces ⎞ Tzt − Tzt −δ t t Cz + Tz ⎜ ∑ hi Ai + ∑ mi C p + minf C p + msys C p ⎟ = ⎜ i =1 ⎟ dt i =1 ⎝ ⎠ N zones ⎛ Nsurfaces ⎞ t Qit + msys C pTsupply + ⎜ ∑ hi ATsi + ∑ mi C pTzi + minf C pT∞ ⎟ ∑ i ⎜ i =1 ⎟ i =1 i =1 ⎝ ⎠ N sl t −δ t

(7)

One final rearrangement was to move the lagged temperature in the derivative approximation to the right side of the equation. The explicit appearance of the zone air temperature was thus eliminated from one side of the equation. An energy balance equation that includes the effects of zone capacitance was then obtained by dividing both sides by the coefficient of Tz:

N zones ⎛ Tz N surfaces ⎞ + ⎜ Cz + ∑ hi ATsi + ∑ mi C pTzi + minf C pT∞ ⎟ Q + msys C T ∑ i ⎜ δt ⎟ i =1 i =1 i =1 ⎝ ⎠ t Tz = N surfaces N zones ⎞ Cz ⎛ + ⎜ ∑ hi Ai + ∑ mi C p + minf C p + msys C p ⎟ ⎟ δ t ⎜ i =1 i =1 ⎝ ⎠ N sl t i t p supply t −δ t

(8)

Equation (8) could be used to estimate zone temperatures, however it was found to severely limit the time step size under some conditions. To correct this, higher order expressions for the first derivative, with corresponding higher order truncation errors, were developed. The goal of this approach was to allow for the use of larger time steps in the simulation than would be possible using the first order Euler form, without experiencing instabilities. Approximations from second through fifth order were tried as reported by Taylor, et al (1990) with the conclusion that the third order finite difference approximation, shown below, gave the best results:

**dTz 3 1 −1 ⎛ 11 ⎞ ≈ (δ t ) ⎜ Tzt − 3Tzt −δ t + Tzt − 2δ t − Tzt −3δ t ⎟ + O (δ t 3 ) dt t 2 3 ⎝6 ⎠
**

When this form for the derivative is used, equation (7) changes to:

(9)

3 1 −1 ⎛ 11 ⎞ Cz (δ t ) ⎜ Tzt − 3Tzt −δ t + Tzt − 2δ t − Tzt −3δ t ⎟ = 2 3 ⎝6 ⎠

∑ Qi +

i =1

N sl

N surfaces

∑

i =1

hi Ai (Tsi − Tz ) +

N zones i =1

∑ m C (T

i p

zi

− Tz )

(10)

**+ minf C p (T∞ − Tz ) + msys C p (Tsup − Tz )
**

6

4/23/05

INTEGRATED SOLUTION MANAGER

SUMMARY OF PREDICTOR-CORRECTOR PROCEDURE

and the zone temperature update equation becomes:

Tzt =

∑Q + ∑

i i =1 i =1

N sl

N surfaces

hi ATsi + i

N zones

∑ mC T

i i =1

p zi

**+ minf C pT∞ + msys C pTsupply − ⎜
**

N surfaces

⎛ 11 ⎞ C z + ⎜ ⎟ ⎝ 6 ⎠ δt

∑

i =1

hi A +

N zones

3 t − 2 δ t 1 t − 3δ t ⎞ ⎛ Cz ⎞⎛ t −δ t ⎟ ⎜ −3Tz + Tz − Tz ⎟ 2 3 ⎠ (11) ⎝ δt ⎠⎝

∑ mC

i i =1

p

+ minf C p + msys C

This is the form currently used in EnergyPlus. Since the load on the zone drives the entire process, that load is used as a starting point to give a demand to the system. Then a simulation of the system provides the actual supply capability and the zone temperature is adjusted if necessary. This process in EnergyPlus is referred to as a Predictor/Corrector process. It is summarized below.

Code Reference: the ZoneTempPredictorCorrector module performs the calculations.

Zone Volume Capacitance Multiplier If the Zone Volume Capacitance Multiplier = 1.0, this represents just the capacitance of the air volume in the specified zone. If the value is not defined it is set to 1.0. This multiplier can be greater than 1.0 if the zone air capacitance needs to be increased for stability of the simulation. This multiplier increases the capacitance of the air volume by increasing the zone volume that is used in the zone predictor-corrector algoriothm in the simulation. This can be done for numerical reasons, such as to increase the stability that will decrease the air temperature deviations at the time step level. Or it can be increased to try and account for the additional capacitance in the air loop not specified in the zone, i.e. Dampers, diffusers, duct work, etc., to see the effect on the dynamics of the simulation. In the source code below we see how the ZoneVolCapMultp increases the zone volume used for the air ratio at the time step in the system. This is constant throughout the simulation.

AIRRAT(ZoneNum) = Zone(ZoneNum)%Volume*ZoneVolCapMultp* & PsyRhoAirFnPbTdbW(OutBaroPress,REAL(MAT(ZoneNum)),ZoneAirHumRat(ZoneNum))* & PsyCpAirFnWTdb(ZoneAirHumRat(ZoneNum),REAL(MAT(ZoneNum)))/(TimeStepZone*SecInHour)

**Summary of Predictor-Corrector Procedure
**

The predictor-corrector scheme can be summarized as follows: • • Using equation (3), an estimate is made of the system energy required to balance the equation with the zone temperature equal to the setpoint temperature. With that quantity as a demand, the system is simulated to determine its actual supply capability at the time of the simulation. This will include a plant simulation if necessary. The actual system capability is used in equation (11) to calculate the resulting zone temperature.

•

System Control Previously, the formulation of a new heat balance equation with an unsteady zone capacitance term was discussed Equation (4). In this equation the updated zone temperature was calculated by removing its explicit dependence from the right hand side and lagging, by one time step, the unknown terms on that side. However, the right hand side still contains implicit dependencies on the zone temperature through the system control logic; the need for heating or cooling in the zones, is based on zone temperature. In real buildings the control system consists of one or more sensing units in the zone, such as a wall thermostat that samples the air temperature and sends signals to a control unit. The controller looks at the difference between the actual zone temperature and the desired temperature to ascertain

4/23/05 7

INTEGRATED SOLUTION MANAGER

SUMMARY OF PREDICTOR-CORRECTOR PROCEDURE

if heating or cooling is required and then sends appropriate signals to the system components to drive the zone temperature closer to the desired value. Although many control systems use only the zone temperature to control the system, most modern energy management systems consider many other variables, such as outside environment conditions. Simulating such controllers would seem to be relatively straightforward in a simulation especially since some of the more complex control problems, such as managing duct pressures and flow rates, are not modeled. However, real controllers have an advantage because they can sample zone conditions, and thus update system response, on a time scale much shorter than any characteristic time of the system or zone. Thus the feedback between zone and system usually results in steady or, at worst, slowly oscillating zone conditions and system operation unless the system is grossly oversized. On the other hand, the numerical model is only able to sample zone conditions at discrete time intervals. In the interest of minimizing computation time, these intervals need to be as long as possible. Frequently, they are of the order of, or longer than, the characteristic times of the system and zones, except in the case of small system capacity in relation to zone capacitance. This situation has the potential for unstable feedback between zone and system, resulting in an oscillatory or diverging solution. Prior to implementing the new heat balance method in IBLAST, several system control strategies were considered. The primary objective was selection of a control method that would: be numerically stable over a reasonable range of conditions, realistic from the standpoint of looking and operating like an actual system controller, and flexible enough to be applied to all current and projected systems. The method actually implemented in IBLAST, and later EnergyPlus, took advantage of the computational model's "knowledge" of how much energy enters or leaves the zone as a function of zone temperature i.e., the zone load. The real controller, on the other hand, does not have this information. The net zone load is given by Equation (12):

Qload = ∑ Qi +

i =1

N sl

N surfaces

∑

i =1

hi Ai (Tsi − Tz ) +

N zones i =1

∑ m C (T

i p

zi

− Tz ) + minf C p (T∞ − Tz )

(12)

This is Equation (4) without the term due to the system. In addition, Tz is now the desired zone temperature as defined by the control system setpoints that must be specified for each zone. An assumption was made that if the system has sufficient capacity (based on the desired zone temperature) to meet the zone conditioning requirements (i.e. Qsys = Qload ) at the desired zone temperature then those requirements will be met. On the other hand, if the system can not provide enough conditioning to the zone to maintain the desired temperature, then the system provides its maximum output to the zone and the zone temperature is allowed to "float." Equation (12) was used to calculate the system output required to maintain the desired zone temperature; the actual zone temperature update was accomplished using Equation (8). This method was called predictive system energy balance. It has many characteristics of a predictor-corrector method since the system response is first approximated based on a predicted zone temperature and then the actual change in zone temperature is determined from that system response. The predictive system energy balance method required that the system controls on air mass flow rate, supply air temperature, etc., be formulated as a function of the zone temperature. However, this was not a serious drawback. The first example considered was a single zone draw through system. Typically, such systems have a cooling coil and heating coil in series, and constant air volume flow rate. Single zone draw through systems run at maximum capacity when turned on so the only way to regulate net system output and keep the zone temperature within the desired range is to turn the system on and off. A simplified schematic of this system type is shown in Figure “Simplified single zone draw through system”.

4/23/05

8

INTEGRATED SOLUTION MANAGER

SUMMARY OF PREDICTOR-CORRECTOR PROCEDURE

OUTSIDE AIR MIXING BOX RELIEF AIR C/C H/C CONSTANT VOLUME FAN RETURN AIR ZONE

Figure 4. Simplified single zone draw through system The amount of heating or cooling provided by the system in relation to the desired zone temperature is given by:

Qsys = msys C pη (Tsup − Tz , desired )

(13)

where η is the fraction of the time step that the system is turned on and varies between 0 and 1. The supply air temperature is also implicitly limited by the effectiveness of the coils and the operating parameters of the central plant components. These interactions are discussed later. A far more complex, though again simplified, system is the variable air volume (VAV) system, shown in Figure “Simplified Variable Volume System. In VAV systems, the supply air temperature as well as the supply air volume are continuous functions of zone temperature. As shown in Figure “Idealized variable volume system operation., when the zone temperature is between Tcl and Tcu, cooling is required and the system varies the supply air flow rate while maintaining a constant supply air temperature. When the zone temperature is between Thl and Thu, heating is required and air is supplied at a constant minimum flow rate while the supply air temperature is varied.

OUTSIDE AIR MIXING BOX RELIEF AIR DAMPER RETURN AIR ZONE H/C C/C VARIABLE VOLUME FAN

Figure 5. Simplified Variable Volume System. The next figure (Idealized variable volume system operation) shows idealized behavior of a VAV system; in practice, the air flow rate and temperature are not exact linear functions of zone temperature.

4/23/05

9

the VAV system becomes a constant volume flow rate system with a variable supply air temperature. When a zone requires cooling.0 COLD DECK TEMPERATURE MIN. from the numerator and denominator of Equation (11) but with the system related terms omitted. As long as a VAV system has sufficient capacity. (14) Qslope = N surfaces ∑ i =1 hi Ai + ∑ mC i + minf C p (15) Equations (14) and (15) are derived. Also excluded from these expressions are the effects of zone capacitance. when heating is required in a zone. This means that the desired zone temperature. Again.ηc . The amount of cooling is matched to the load by dampers in the supply air duct that vary the air volume flow rate of being supplied to the zone.min + (1 − ηc . assuming the heat energy output varies linearly with zone temperature and normalizing with respect to the maximum coil output gives the following result: 4/23/05 10 . upper − Tc . For the purposes of this calculation. or the throttling range of the reheat coil. The dampers are set to provide air to the zone at the minimum supply air fraction. when the system is heating. the volume flow rate of supply air normalized to the maximum flow rate. Conversely.0 ⎟ ⎠ (16) Normally.INTEGRATED SOLUTION MANAGER SUMMARY OF PREDICTOR-CORRECTOR PROCEDURE REHEAT THROTTLING RANGE HOT DECK ∆Th ∆Tc FLOW RATE 1. which effectively alters the coil’s heating capacity.min must be greater than zero to ensure a supply of fresh air sufficient to eliminate contaminants from the zone. the zone temperatures can be expected to vary within the limits defining the range of operation of the air damper. T hl T hu T cl T cu ZONE TEMPERATURE Figure 6. or supply air fraction.min ) ⎜ ⎛ Tz − Tc . Idealized variable volume system operation. when cooling. Throttling the hot water supply to the reheat coil. the minimum supply air fraction ηc. is variable and must be calculated in order to determine the system output. lower ⎜ Tc . lower ⎝ ⎞ ⎟ . is given by: ηc = ηc . used to predict the system response. Assuming that the volume flow rate varies linearly with zone temperature. the VAV system is designed to provide air to that zone at a constant supply air temperature. the following definitions were found useful: Q0 = ∑ Qi + i =1 N sl N surfaces ∑ i =1 hi ATsi + i N zones i =1 N zones i =1 ∑ mC T i p p zi + minf C pT∞ VOLUME FLOW RATE FRACTION VAV DAMPER RANGE SUPPLY AIR TEMP. min ≤ ηc ≤ 1. modulates the supply air temperature. respectively.

lower − ηc .min − Tc . pred . heat = Qh / c .upper − Th .INTEGRATED SOLUTION MANAGER SUMMARY OF PREDICTOR-CORRECTOR PROCEDURE ηh = ⎜ ⎜ ⎛ Th .upper − Tc . upper − Th .max + C p ρVmin + Qslope Th .cool ) (18) (19) Equation (18) is valid for zone temperatures below Th.min Tc . while Equation (19) is valid for all temperatures above this value. upper − Tz ⎝ Th . as given by Equation (12).lower Qslope C p ρVmax (24) C2 = (25) 4/23/05 11 . max + C p ρVmin (Tc / c − Tz .upper Th . pred . lower ⎞ ⎟ . (23) C1 = 1 − ηc .0 ≤ η h ≤ 1.lower (20) Tz .cool = where. B1 + B12 + B2 2 (21) B1 = Tc / c + Tc . These expressions are as shown in Equations (18) and (19): Qsys .min − C2 C1 (22) ⎛C ⎛η ⎞⎞ B2 = 4 ⎜ 3 + Tc / c ⎜ c . the zone is supplied with air at the cooling coil outlet temperature at the minimum air fraction.maxTh .lower + Q0 + C p ρVminTc / c Qh / c .heat ) Qsys . the definitions of ηc and ηh were then used to develop expressions for the predicted zone temperature in the cases of heating and cooling: Tz .0 ⎟ ⎠ (17) Observe that when ηh is equal to zero.h = ηh Qh / c .upper − Th .c = C p ρ (ηcVmax ) (Tc / c − Tz . pred . pred .upper. two equations are necessary to describe the system output in terms of ηh and ηc.lower ⎟ ⎟ ⎜C ⎟ ⎝ C1 ⎠⎠ ⎝ 1 and. Because the control strategies of the VAV system are different whether the system is heating or cooling. Equating the system output to the zone load.

i. the system provides air at the minimum volume flow rate and at a temperature given by: Tsup ply = Tc / c + η hQh / c . When heating is required by the zone.e. Therefore. either in practice or in simulations. an iteration scheme was developed that solved these equations simultaneously with the coil performance models. Writing the transient mass balance equation with the change in the zone humidity ratio = sum of internal scheduled latent loads + infiltration + system in & out + convection to the zone surfaces in the equation below. Moisture Predictor-Corrector To preserve the stability of the calculation of the zone humidity ratio.max C p ρVmin (28) The reheat coil capacity fraction ηh is determined by using Equation (17). ρ air Vz dWz = ∑ kg massSched Loads + minf (W∞ − Wz ) + msysWsysin − msysout Wz dt + ∑ Ai hmi ρ airz Wsurfsi − Wz i =1 surfs ( ) Then the third order derivative derived from a Taylor Series expansion is defined as: ( 11 Wzt − 3Wzt −δ t + 32 Wzt −2δ t − 13 Wzt −3δ t ) + O δ t 3 . When a zone requires cooling the system supply air temperature is constant at the cooling coil outlet temperature and the volume flow rate is given by: Vsup ply = ηcVmax (27) where the supply air fraction ηc is computed from Equation (16). has been used. These values are then used in Equation (8) to calculate the updated zone temperature. or update.INTEGRATED SOLUTION MANAGER MOISTURE PREDICTOR-CORRECTOR C3 = Q0 C p ρVmax (26) Once the predicted zone temperature has been calculated from Equations (20) and (21). Once Equation (27) or (28). the zone humidity ratio. the supply air flow rate and temperature are known. the system response may be determined. if the coil has infinite capacity. dWz 6 ≈ ( ) δt dt t 4/23/05 12 . The third order differential approximation derived by a Taylor Series was used in the calculation of the next time steps zone air temperature This idea was applied to predict and correct. a similar methodology was used for the mass balance as was used by temperature in the heat balance as explained above in the Summary of Predictor-Corrector Procedure. The equations describing VAV system operation may be solved without iteration if the cooling coil outlet temperature is constant. and if the reheat coil capacity varies linearly with zone temperature. when realistic coil models are used. This is not the case.

PredictedSystemLoad = msysout Wzt − msysWsysin Massflow * HumRat = kg air/sec * kgWater/kg Air = kgWater/sec Then solving the mass balance for the predicted system load or response is: surfs ⎡ρ V ⎤ PredictedSystemLoad [kgWater / sec] = ⎢ air z ( 11 ) + minf + ∑ Ai hmi ρ airz ⎥ *Wzt − 6 i =1 ⎣ δt ⎦ surfs ⎡ ⎤ ρ air Vz t −δ t t − 2δ t 3 + 1 Wzt −3δ t ) ⎥ 3 ⎢ ∑ kg massSched Loads + minf W∞ + ∑ Ai hmi ρ airz Wsurfsi + δ t ( 3Wz − 2 Wz i =1 ⎣ ⎦ Then using the following substitutions the mass balance equation becomes: A = minf + ∑ Ai hmi ρ airz i =1 surfs B = ∑ kg massSched Loads + minf W∞ + ∑ Ai hmi ρ airz Wsurfsi i =1 surfs C= ρ air Vz δt 13 4/23/05 .INTEGRATED SOLUTION MANAGER MOISTURE PREDICTOR-CORRECTOR The coefficients of the approximated derivative are very close to the coefficients of the analogous Adams-Bashforth algorithm. Moisture Prediction For the moisture prediction case the equation is solved for the anticipated system response as shown below. This third order derivative zone humidity ratio update increases the number of previous time steps that are used in calculating the new zone humidity ratio. and decreases the dependence on the most recent. one way for the predict step and one way for the correct step. surfs ρ air Vz 11 t t t ( 6 )Wz + minf Wz + msys Wz + ∑ Ai hmi ρ air Wzt = δt i =1 out z ∑ kgmassSched Loads + minfW∞ + msysWsysin + ∑ Ai hmi ρairzWsurfsi i =1 surfs − ρ air Vz ( −3Wzt −δ t + 32 Wzt −2δ t − 13 Wzt −3δ t ) δt This gives us the basic mass balance equation that will be solved two different ways. Then the approximated derivative is substituted into the mass balance and the terms with the humidity ratio at past time steps are all put on the right hand side of the equation. The higher order derivative approximations have the potential to allow the use of larger time steps by smoothing transitions through sudden changes in zone operating conditions.

After the system simulation is completed the actual response from the system is used in the moisture correction of step. When the system is operating the mass flow for the system out includes the infiltration mass flow rate. surfs ⎡ ⎢ ∑ kg massSched Loads + minf W∞ + ∑ Ai hmi ρ airz Wsurfsi i =1 ⎢ ⎢ ρ air Vz t −δ t t − 2δ t 3 + 1 Wzt −3δ t ) 3 ⎢ + δ t ( 3Wz − 2 Wz t Wz = ⎢ surfs ρ air Vz 11 ⎢ ( ) + minf + ∑ Ai hmi ρ airz ⎢ δt 6 i =1 ⎢ ⎢ ⎢ ⎣ ⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ 4/23/05 14 . therefore the infiltration mass flow rate is not included as a separate term in the mass balance equation. For example. dehumidification processes will try to remove moisture at the specified negative predicted moisture load to meet the relative humidity set point. then the zone should stabilize at the outside air humidity ratio and upon system startup the added mass and heat in the zone air is handled by the system. Therefore if the system is shut down for a longer time. This ensures that the mass and energy brought into the zone due to infiltration is conditioned by the coils and is part of their load.INTEGRATED SOLUTION MANAGER MOISTURE PREDICTOR-CORRECTOR PredictedSystemLoad [kgWater / sec] = ⎡( 11 ) ∗C + A⎤ ∗ WSetPo int − ⎣ 6 ⎦ ⎡ B + C ∗ ( 3Wzt −δ t − 3 Wzt − 2δ t + 1 Wzt −3δ t ) ⎤ 2 3 ⎣ ⎦ At the prediction point in the simulation the system mass flows are not known. For system operating msysout = minf + msysin . therefore the system response is approximated. But when the system is off. the infiltration mass flow in is then exhausted out of the zone directly. Likewise. which is shown next. The system simulation components that have moisture control will try to meet this predicted moisture load. The predicted system moisture load is then he used in the system simulation to achieve the best results possible. Moisture Correction For the correct step you solve the expanded mass balance equation for the final zone humidity ratio at the current time step. humidifiers will look for positive moisture loads and add moisture at the specified rate to achieve the relative humidity set point. ⎡ ∑ kg massSched Loads + minf W∞ + msysWsysin ⎤ ⎢ surfs ⎥ ρ air Vz ⎢ t −δ t t − 2δ t t − 3δ t ⎥ 3 + 1 Wz ) ⎥ 3 ⎢ + ∑ Ai hmi ρ airz Wsurfsi + δ t ( 3Wz − 2 Wz t i =1 Wz = ⎢ ⎥ surfs ρ air Vz 11 ⎢ ⎥ ( ) + msysout + ∑ Ai hmi ρ airz ⎢ ⎥ δt 6 i =1 ⎢ ⎥ ⎢ ⎥ ⎣ ⎦ For system off. the mass flow in due to infiltration is exhausted from the zone.

Zones can interact with each other thermally through adjacent surfaces and by intermixing of zone air. In EnergyPlus. and C parameters from the prediction step modified with actual zone mass flows with the system ON and OFF: If (ZoneMassFlowRate > 0. This moisture update equation is used for the Conduction Transfer Function (CTF) case in EnergyPlus. We have previously documented the method used to update the zone 4/23/05 15 . and the only difference will be the additional moisture capacitance of the zone surfaces for the Moisture Transfer Function (MTF) case. However.0) Then B = ∑ kg massSched Loads + minf W∞ + msysin Wsys + ∑ Ai hmi ρ airz Wsurfsi i =1 surfs A = minf + msysout + ∑ Ai hmi ρ airz i =1 surfs C= ρ air Vz δt Else If (ZoneMassFlowRate <= 0. because heat transfer through each zone's surfaces and interzone mixing of air still occur. in addition to the moisture cases. can be grouped together.0) Then B = ∑ kg massSched Loads + ( minf + mExhaust ) W∞ + ∑ Ai hmi ρ airz Wsurfsi i =1 surfs A = minf + mExhaust + ∑ Ai hmi ρ airz i =1 surfs End If Then inserting them in the mass balance equation. even though conditions in one zone may be changing much more rapidly than conditions in the other zones. subsequently. the conditions in each zone are updated by Equation (8). the new space temperatures must be computed at the same simulation time and on the same time step in all zones. B. This means that EnergyPlus does not have to iterate to find a self consistent solution of the updated zone conditions. Zone Update Method A zone is not necessarily a single room but is usually defined as a region of the building or a collection of rooms subject to the same type of thermal control and having similar internal load profiles that. The equations are identical except that the convection to the zone surfaces is non-zero for the moisture cases.INTEGRATED SOLUTION MANAGER ZONE UPDATE METHOD and then using the same A. This moisture update allows both methods to be updated in the same way. which uses previously calculated values of the zone conditions. it simplifies to: ⎡ B + C ∗ ( 3Wzt −δ t − 3 Wzt − 2δ t + 1 Wzt −3δ t ) ⎤ 2 3 ⎥ W =⎢ 11 ( 6 )∗C + A ⎢ ⎥ ⎣ ⎦ t z This is implemented in the Correct Zone Air Humidity Ratio step in EnergyPlus.

This approach can be justified because the internal loads. This method. the adaptive time step could not be applied easily to the surface heat transfer calculations. the temperature of the return water from the coils must be equal to the chiller or boiler entering water temperature. ∆t is the user specified time step and δt is the adaptive time step that is always less than or equal to ∆t. The coil capacity was controlled by adjusting the water flow rate. Simultaneous Solution of Plant/System Water Loop Simultaneous solution of the system and plant operating parameters required that the temperature of the water entering the coils must be the same as the temperature leaving the chillers or boilers. In addition. This meant that the user had to specify the coil water inlet temperature and the maximum coil inlet water flow rate to run the simulation. infiltration. In this two time level scheme. capacity. In practice so long as the plant is not out of capacity the leaving water temperature from chillers and boilers is constant and equal 4/23/05 16 . The first step in integrating the plants was to implement the detailed coil models and coil control strategies without actually adding the plant models themselves. an adaptive time step was required.25 hours (15 minutes) was found to give stable results without a large increase in computation time. the contributions to the zone loads from the surfaces. The zone temperature update was made using Equation (29) for each adaptive time step. If the maximum zone temperature change is more than a preset maximum of 1°C the system and zone updates are performed using a new time step ∆t. Variable Time Step Prior to the integration of the central plant simulation in IBLAST. as the time step got shorter the time required to execute the program got prohibitively high. a time step ∆t for the zone temperature update of 0. This would shorten the time step to maintain stability of the zone air energy balance calculation when zone conditions were changing rapidly and expand it to speed computation when zone conditions were relatively unchanging. # zones ⎛ ⎞ ⎛ Cz ⎞ + ⎜ Tz + msys C pTsup ply ⎟ ⎜ ∑ Qc + ∑ hi ATsi + ∑ m j C pTzj + minf C pT∞ ⎟ i ⎝ δt ⎠ i =1 j =1 ⎝ ⎠ t (29) Tz = ( t −∆t ) ( t −δ t ) # zones ⎛ # surf . ⎞ ⎛C ⎞ + ⎜ z + msys C p ⎟ ⎜ ∑ hi Ai + ∑ m j C p + minf C p ⎟ ⎝ δt ⎠ j =1 ⎝ i =1 ⎠ ( t −∆t ) In Equation (29). A second variable time step is used to update the system response and the zone mean air temperature. but the effect of the plant on the chilled water temperature was eliminated. The problem of updating the zone temperature was resolved by using a two time step approach in which the zone air temperature is updated using an adaptive time step that ensures stability. even using interpolation methods to determine new temperature and flux histories.INTEGRATED SOLUTION MANAGER ZONE UPDATE METHOD temperature at each time step. has proved to be satisfactory over a wide range of conditions. called the predictor corrector method. This adaptive time step is initially set to ∆t/2 and is successively halved until the maximum zone temperature change is less than the allowable maximum change. The maximum temperature change experienced by a zone is then evaluated on a system by system basis. The real life analogy would be a chiller of very large. This time step is selected by first calculating the system response and updating the zone temperature using the user specified time step ∆t. After implementation of this step. Clearly. and user specified internal loads are updated at the default or user specified time step that is constant. But. experience with the program showed that updating the zone temperatures on a fixed time step frequently resulted in instabilities unless a very short time step was used. which is just a different form of Equation (8): ( t −δ t ) # surf . surface temperatures. infiltration and mixing vary on a different and longer time scale than the system response and the zone air temperature. mixing. However. though not infinite.

INTEGRATED SOLUTION MANAGER ZONE UPDATE METHOD to the design value. Ph. R.D. 2. University of Illinois. An Improved Root-Finding Procedure for Use in Calculating Transient Heat Flow Through Multilayered Slabs. Liege.D. No. Calculating Building Heating and Cooling Loads Using the Frequency Response of Multilayered Slabs. D. University of Illinois. 11. Hittle. 26. R.E. C.. Ph. Vol. L. Pedersen. 1. December 3-5. Thesis. 26. D.K. H. International Journal of Heat and Mass Transfer. R. J. Vol. Ouyang. Thesis. Strand. Simultaneous Simulation of Buildings and Mechanical Systems in Heat Balance Based Energy Analysis Programs. and R. if either the chiller or boiler plant was overloaded then the temperature of the water leaving the plant was not equal to the design value and the maximum output of the plant could change because of the off-design conditions. Heat Source Transfer Functions and Their Application to Low Temperature Radiant Heating Systems. T. 1987. D. Liesen. ASME Journal of Heat Transfer. August 20-22. University of Wisconsin.O. Long-time Solutions to Heat Conduction Transients with Time-Dependent Inputs. IL. D. R. An iterative scheme using the secant method to predict successive updates to the plant leaving water conditions was therefore employed to solve for the water loop conditions with the plant operating at its maximum capacity. Building and Environment. No. Nice. 1983. Taylor. Pedersen. Proceedings of the 3rd International Conference on System Simulation in Buildings. Liesen. 1991. 1991. The simulation of the system and plant loops is described in greater detail in the later sections. Fisher. Seem. WI. pp.O. Lawrie. C. 173-177. Volume 102. E. A Procedure for Calculating Thermal Response Factors of Multi-layered Walls--State Space mMethod. Ph. Belgium. Hittle. However. No iteration was required to match system and plant boundary conditions. pp. Lawrie. pp.. and G. Thesis..D.D. Taylor. No. References Ceylan..E. IL. 1990. Conference Proceedings IBPSA Building Simulation '91. K. C. France. Madison. and F. Modeling of Heat Transfer in Buildings. 115-120. 1995.K. J. R. L. 4/23/05 17 . Fisher. 1685-1693. J. Haghighat.. 1979. E. D. Impact of Simultaneous Simulation of Buildings and Mechanical Systems in Heat Balance Based Energy Analysis Programs on System Response and Control. 1980. Urbana. Bishop. Urbana. C. K. Myers.

t − ∑ Y jTi .t − jδ j =1 j =1 j =1 nz nz nq (32) for the outside heat flux (q″=q/A) where: Xj = Outside CTF coefficient.t + ∑ X jTo ....t − jδ + YoTo . j= 0. Note that the first terms in the series (those with subscript 0) have been separated from the rest in order to facilitate solving for the current temperature in the solution scheme. These equations state that the heat flux at either face of the surface of any generic building element is linearly related to the current and some of the previous temperatures at both the interior and exterior surface as well as some of the previous flux values at the interior surface.t − ∑ Z jTi .. 4/23/05 18 .t − jδ + ∑ Φ j qki .t − jδ j =0 j =0 ∞ ∞ (30) where q” is heat flux.1.. The basic form of a conduction transfer function solution is shown by the following equation: ′′ ′′ qki (t ) = − Z oTi .t − jδ + ∑ Φ j qko .t − jδ j =1 j =1 j =1 nz nz nq (31) for the inside heat flux.nq.nz.t − jδ + X oTo .t − jδ − ∑ Y jTi .. the similarity of higher order terms can be used to replace them with flux history terms. While in most cases the terms in the series decay fairly rapidly...1. Zj = Inside CTF coefficient.. o signifies the outside of the building element. T is temperature.. the infinite number of terms needed for an exact response factor solution makes it less than desirable. j= 0. Φϕ = Flux CTF coefficient. Yj = Cross CTF coefficient.. Fortunately.t + ∑ Y jTo .2.nz.SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS Surface Heat Balance Manager / Processes Conduction Through The Walls Conduction Transfer Function Module The most basic time series solution is the response factor equation which relates the flux at one surface of an element to an infinite series of temperature histories at both sides as shown by Equation (30): ′′ qko (t ) = ∑ X jTo . j = 1..1. t represents the current time step. j= 0. The new solution contains elements that are called conduction transfer functions (CTFs). and ′′ ′′ qko (t ) = −YoTi .nz. Ti = Inside face temperature To = Outside face temperature ′′ qko = Conduction heat flux on outside face q′′ = Conduction heat flux on inside face The subscript following the comma indicates the time period for the quantity in terms of the time step δ. i signifies the inside of the building element. and X and Y are the response factors.

older method used Laplace transformations to reach the solution. and A. Another common. Ouyang and Haghighat 1991).. With a single. The formulation is valid for any surface type and does not require the calculation or storage of element interior temperatures. the environmental temperatures (interior and exterior) are the inputs. and D are coefficient matrices. linear equation with constant coefficients. The only storage of data required are the CTFs themselves and a limited number of temperature and flux terms. Hittle & Bishop. the Laplace method was used in BLAST (Hittle. This formulation can be used to solve the transient heat conduction equation by enforcing a finite difference grid over the various layers in the building element being analyzed. the state variables are the nodal temperatures. the state space representation with finite difference variables would take the following form: ⎡ T1 ⎤ d⎢ ⎥ ⎢ ⎥ ⎡ T1 ⎤ ⎢ ⎥ ⎣Tn ⎦ = A ⎢ ⎥ + B ⎡ Ti ⎤ [ ] ⎢ ⎥ [ ] ⎢T ⎥ dt ⎣ o⎦ ⎢Tn ⎥ ⎣ ⎦ ⎡ T1 ⎤ ⎡ q "i ⎤ ⎡ Ti ⎤ ⎢ ⎥ ⎢ q " ⎥ = [C] ⎢ ⎥ + [ D] ⎢T ⎥ ⎣ o⎦ ⎣ o⎦ ⎢Tn ⎥ ⎣ ⎦ where T1. Seem 1987. the vector of state variables (x) can be eliminated from the system of equations. The coefficients (CTFs) in the equation are constants that only need to be determined once for each construction type. Thus. In this case.. B. Seem (1987) shows that for a simple one layer slab with two interior nodes as in Figure 7 and convection at both sides the resulting finite difference equations are given by: 4/23/05 19 . and the resulting heat fluxes at both surfaces are the outputs. t is time. Ti and To are the interior and exterior environmental temperatures. u is a vector of inputs. C. y is the output vector. relatively simple. Tn-1. the conduction heat transfer through an element can be calculated. and the output vector (y) can be related directly to the input vector (u) and time histories of the input and output vectors. Through the use of matrix algebra. and qi′′ and qo are the heat fluxes (desired output). Tn are the finite difference nodal temperatures. . n is the number of ′′ nodes. 1979.SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS The final CTF solution form reveals why it is so elegant and powerful.. 1983). The basic state space system is defined by the following linear matrix equations: d [ x] = [ A ][ x ] + [ B ][u ] dt [ y ] = [C][ x] + [ D][u ] where x is a vector of state variables. T2. Calculation of Conduction Transfer Functions The basic method used in EnergyPlus for CTF calculations is known as the state space method (Ceylan and Myers 1980.

In matrix format: 1 ⎡ dT1 ⎤ ⎡ 1 hA ⎤ ⎡ hA ⎢ dt ⎥ ⎢ − RC − C ⎥ ⎡ T1 ⎤ ⎢ C RC +⎢ ⎢ ⎥=⎢ ⎥ 1 1 hA ⎥ ⎢T2 ⎥ ⎢ ⎣ ⎦ ⎢ dT2 ⎥ ⎢ − − 0 ⎢ dt ⎥ ⎢ ⎢ RC RC C ⎥ ⎣ ⎦ ⎣ ⎦ ⎣ ⎡ q "o ⎤ ⎡ 0 −h ⎤ ⎡ T1 ⎤ ⎡ 0 h ⎤ ⎡To ⎤ ⎢ q " ⎥ = ⎢ h 0 ⎥ ⎢T ⎥ + ⎢ −h 0 ⎥ ⎢ T ⎥ ⎦⎣ 2⎦ ⎣ ⎦⎣ i ⎦ ⎣ i⎦ ⎣ ⎤ 0 ⎥ ⎡To ⎤ ⎥⎢ ⎥ hA ⎥ ⎣ Ti ⎦ C⎥ ⎦ 4/23/05 20 .SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS C dT1 T −T = hA (To − T1 ) + 2 1 dt R dT2 T −T = hA (Ti − T2 ) + 1 2 dt R C q "i = h (Ti − T2 ) q "o = h (T1 − To ) where: R= C= kA ρcp A 2 . and A is the area of the surface exposed to the environmental temperatures. .

While more time consuming than calculating CTFs using the Laplace Transform method. Another difference between the Laplace and State Space methods is the number of coefficients required for a solution. Note that as with the Laplace method that the actual number of terms will vary from construction to construction. the State Space method requires more coefficients.SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS To 1 hA T1 R C C T2 1 hA Ti Figure 7. B. conduction transfer function series become progressively more unstable as the 4/23/05 21 . The accuracy of the state space method of calculating CTFs has been addressed in the literature. Two distinct advantages of the State Space method over the Laplace method that are of interest when applying a CTF solution for conduction through a building element are the ability to obtain CTFs for much shorter time steps and the ability to obtain 2. and D matrices. In addition.and 3-D conduction transfer functions. However. The important aspect of the state space technique is that through the use of matrix algebra the state space variables (nodal temperatures) can be eliminated to arrive at a matrix equation that gives the outputs (heat fluxes) as a function of the inputs (environmental temperatures) only. Seem (1987) summarizes the steps required to obtain the CTF coefficients from the A. Their results showed that for an adequate number of nodes the state space method computed a heat flux at the surface of a simple one layer slab within 1% of the analytical solution. In addition. they found almost no difference in the response factors calculated by each method. the resulting matrix form has more physical meaning than complex functions required by the Laplace transform method. both Seem (1987) and Strand (1995) have demonstrated the effectiveness of the State Space method in handling these situations that can have important applications in buildings. EnergyPlus CTF Calculations Conduction transfer functions are an efficient method to compute surface heat fluxes because they eliminate the need to know temperatures and fluxes within the surface. For a wall composed of insulation between two layers of concrete. C. the number of temperature and flux history terms is identical (nz=nq). Ceylan and Myers (1980) compared the response predicted by the state space method to various other solution techniques including an analytical solution. the matrix algebra (including the calculation of an inverse and exponential matrix for A) is easier to follow than root find algorithms. This eliminates the need to solve for roots in the Laplace domain. Ouyang and Haghighat (1991) made a direct comparison between the Laplace and state space methods. In general. Two Node State Space Example. While not implemented in the Toolkit.

Several ways of doing this are described below. t+∆t/3. and convective heat transfer from the zone surfaces. BLAST uses a conduction transfer function CTF method attributed to Hittle (1980) to solve the transient conduction problem for each surface. requiring a large number of terms in the CTF series. determining the series of CTF coefficients is a one time calculation. making the method much faster than finite difference calculations. and roof. etc. However. the transient heat conduction in the material between the surfaces must be solved. air exchange between zones. actually. This phenomenon was most apparent for thermally massive constructions with long characteristic times and. A problem with CTF methods is that the series time step is fixed. t-2 hours. with enough terms. The method shown in the Figure 8 maintains two sets of histories out of phase with each other. In addition. 1987) were considered to address this problem. the number of terms in the series grows. In addition. Seem's method did give better accuracy and stability at short time steps than the current BLAST technique but. that is. This became a problem as investigations into short time step computational methods for the zone/system interactions progressed because. of the minimum time step allowed for the CTF calculations. continuous functions between those discrete points. The method results in a time series of weighting factors that. 1980. Even though CTF methods require that values of the surface temperatures and fluxes be stored for only a few specific times before the current time. Methods that develop CTF series from finite difference approximations to the heat conduction equation (Meyers. More sets of temperature and flux histories could be used.1 hour time step without restricting the types of surfaces that could be used. Hittle's method was replaced by Seem's method (1987) in IBLAST. as the time step used to calculate the CTF series gets shorter. this instability caused the entire simulation to diverge. correspondingly. This resulted in some improvement in stability at shorter time steps. As time advances the oldest term in the input series is dropped and the data moved back one time step to allow the newest value to be added to the series.SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS time step decreases. In an attempt to overcome this problem. the series becomes unstable due to truncation and round-off error. eventually. t+∆t/4. the temperature and flux histories are. 1/4. air exchange with the outside environment. The zone heat gains consist of specified internal heat gains. there is no way to calculate information at these intermediate times once a series has been initialized. This procedure would allow a CTF series computed for a time step ∆t. allowing the simulation time step to take on values: 1/3. The terms in the temperature and flux histories are out of phase with these points. and computes conditions at the current time t. gives the current inside and outside heat flux. so long as the surface temperatures and flux values were still ∆t apart. a CTF series computed for a one hour time step takes information at t-1 hours. However. floor. Eventually. to be used to compute information at times t+∆t/2. or any other arbitrary fraction of the time step. Of these. when multiplied by previous values of the surface temperatures and fluxes and the current inside and outside surface temperatures. The method is easily applied to multilayered constructions for which analytical solutions are unavailable. the surface convection load requires the most complicated calculations because a detailed energy balance is required at the inside and outside surface of each wall.. the method still had difficulty computing stable CTF series for time steps of less than 1/4 hour for the heaviest constructions in the BLAST library. such as slab-on-grade floors (12" heavyweight concrete over 18" dirt). Seem. The figure shows how this would work for two sets of histories out of phase by one half of a time step. This solution gives the inside and outside temperatures and heat fluxes that must be known in order to calculate the convection component to the zone load for each zone surface. For convenience. the time step used to determine the CTF series should be the same as the time step used to update the zone mean air temperature in the zone energy balance. 1/5. have accuracy and stability problems at time steps as large as 0. This indicates that the problem is related to roundoff and truncation error and is in no way an indictment of the CTF method itself.5 hours when modeled by Hittle's CTF method. but not enough to allow IBLAST to run at a 0. they can be calculated by shifting the phase of the temperature and flux histories by only a fraction of a time step. But. etc. The time step between inputs to the CTF series would be the 4/23/05 22 . Heavy constructions.

Sequential interpolation of new histories A final method. temporary history was interpolated from the master values. at concurrent simulation times current values of history terms were different form previous "in phase" history terms. This method proved to be the best of the three options described because it eliminated propagation of information backwards in time and only required concurrent storage of two sets of temperature and flux histories. was something of a hybrid of the previous two methods. This was unacceptable from. staggered time history scheme Another method is shown in Figure “Sequential interpolation of new histories” that uses successive interpolations to determine the next set of temperature and flux histories. Cycling through each history. This method was subsequently incorporated into the IBLAST program in conjunction with Seem's procedure for calculating the coefficients of the CTF series. However. This scenario is illustrated in this figure for two separate sets of temperature and flux histories. significant memory was required to store all the sets of histories. This method required permanent storage for only one set of temperature and flux histories at a time. One "master" history set was maintained and updated for all time. Multiple. allowed calculations of the zone energy balance to be performed with updated surface information at a shorter time step than one CTF history series would otherwise allow. As a result.SURFACE HEAT BALANCE MANAGER / PROCESSES CONDUCTION THROUGH THE WALLS smallest convenient interval at which the CTF series is stable. 4/23/05 23 . a physical point of view. shown in Figure “Master history with interpolation”. if the smallest time step for a stable CTF series was large compared to the zone temperature update time step. this solved the problem of current events propagating information backwards in time. but smoothed out temperature and flux data as more interpolations were performed. x o x o x o dt x o x o x x o x o x x o x o x x x x o o x o history 1 history 2 x time Figure 8. x o x o x' o' dt x o x' o' x" x o x' o' x" x o x' o' x" x' o' x" x" time x o x' o' o" x o x' o' o" history 1 history 2 history 3 history 4 history 5 Figure 9. When surface fluxes needed to be calculated at times out of phase with this master history a new. because it allowed current information to change data from a previous time. in order. The current history is interpolated directly from the previous history set using the required time phase shift between the two. This method required no interpolation between the series once each set of histories was initialized.

Belgium. C. and specific heat.SURFACE HEAT BALANCE MANAGER / PROCESSES MOISTURE TRANSFER MATERIAL PROPERTIES x o x o x o dt x o x o x x o x o x x o x o x x time x o x o x x o history 1 history 2 Figure 10. C. Myers. Ouyang. Building and Environment. Calculating Building Heating and Cooling Loads Using the Frequency Response of Multilayered Slabs. IL.D. K. 26. we frequently accept this simplification. even thermal conductivity. ASME Journal of Heat Transfer. Volume 102.K. 2. R. Bishop. J. and they are all functions of moisture content. Seem. University of Wisconsin. Modeling of Heat Transfer in Buildings. No. Madison.E. R. L. Thesis. Vol. Long-time Solutions to Heat Conduction Transients with Time-Dependent Inputs. and F. 1983. University of Illinois. 4/23/05 24 . Impact of Simultaneous Simulation of Buildings and Mechanical Systems in Heat Balance Based Energy Analysis Programs on System Response and Control. Liesen. Hittle. Hittle. Urbana. December 3-5. Fisher. No. Thesis. E. 1987. 1991. Heat Source Transfer Functions and Their Application to Low Temperature Radiant Heating Systems.O. and G. 1990. 173-177. J. Pedersen. The additional material properties needed are the porosity. University of Illinois. WI. D. 11.D.. pp. Master history with interpolation References Ceylan. D. Almost all material properties change over a large enough range of physical conditions.. Ph. The moisture capacitance coefficients are the most non-linear.D. Haghighat. pp. E. T.. Fisher. L. Lawrie. An Improved Root-Finding Procedure for Use in Calculating Transient Heat Flow Through Multilayered Slabs. 1979. Liesen. and R.D. Nice. IL.K. even though the thermal properties change as a function of moisture content. International Journal of Heat and Mass Transfer. Proceedings of the 3rd International Conference on System Simulation in Buildings. Simultaneous Simulation of Buildings and Mechanical Systems in Heat Balance Based Energy Analysis Programs. 1685-1693. D. Vol.O. R. 1991. Urbana. 1980. Strand. Pedersen. R. A Procedure for Calculating Thermal Response Factors of Multi-layered Walls--State Space mMethod. H. C. The problem is that the additional material properties needed for the mass transfer equations are even stronger functions of temperature and moisture content. J. the amount of moisture in the material for that amount of material [kg moisture/kg dry solid] and is dimensionless. Ph. D. With heat transfer only solutions. R.E.. Liege. pp. Lawrie. Conference Proceedings IBPSA Building Simulation '91. Ph. Taylor. Thesis. 26. density. 115-120. and are mainly functions of vapor density and temperature. C. France. and coefficients to represent the moisture capacitance. August 20-22. water vapor diffusivity. No. D. 1. K. Moisture Transfer Material Properties Moisture transfer properties are non-linear over the entire range from dry to saturated. 1995. Taylor..

Figure 11. Investigating these typical curves show that there are linear planes in these curves that can be used. high moisture content with continuous threads of moisture in the pores. Masonary Moisture Capacitance Curves The figure above shows us that there are some linear planes in the moisture capacitance curves. There are normally 3 distinct regions: funicular state. masonry and wood. The moisture capacitance coefficients that are used in the MTF formulation are a numerical fit of the linear plane in the 3-D graph for that material. Most buildings are in the pendular state. as appropriate. and the dry state. where the layer of moisture in the pores is essentially gone and all the moisture is in vapor. The graph above shows there is a large linear plane that is in the pendular state. The bottom part of the figure above also shows a 2-D moisture isotherm from the 3-D graph. for a building simulation. the micropore surface is covered with thin molecular layers of moisture and intraporous capillary liquid bodies have reduced in significance. 4/23/05 25 . pendular state. Building elements with a typical diurnal cycle for the outside boundary conditions and typical thermostat settings for the inside boundary conditions can stay within these linear planes over a period of time. In the 2-D graph it is clear that there is a large portion in the pendular state where a transfer function analysis would be valid.SURFACE HEAT BALANCE MANAGER / PROCESSES MOISTURE TRANSFER MATERIAL PROPERTIES Typical Masonry Moisture Capacitance The next 2 figures are curves for two groups of major construction materials used in buildings.

902025 26. Again.SURFACE HEAT BALANCE MANAGER / PROCESSES MOISTURE TRANSFER MATERIAL PROPERTIES Typical Wood Moisture Capacitance Figure 12.755466 Min Temperature 0.315024 26. the outside temperature ranges from 21.018213 Max Vapor Density 0. In this case.316999 23. The example in the next section illustrates this assumption. as determined from the simulation.01685 0.7°C to 24.25°C.034845 23.417565 24.012331 From the Table above we can roughly determine the endpoints of the path that the moisture material properties traveled on the moisture capacitance curve. Wood Moisture Capacitance Curve The same linear plane can be seen for wood in the pendular state in the figure above. 4/23/05 26 .245834 47.0162 to 0. A linear fit of the plane in the pendular state is determined and the coefficients are used in a simulation. The vapor densities range from 0.01685 kg moisture/m3 of air on the outside and 0. Linear Material Properties Example The example is for an exterior wall building element with an exterior bricklayer. Max & Min Temperatures and Vapor Densities for Atlanta Summer Design Day Simulation Outside Out-Brick Brick-Insul Insul-Gyp 33.016816 0.012284 Min Vapor Density 0. and finally a gypsum interior layer.01051 0. while the inside ranges from 22.695833 23. Table 2.633353 22.016458 0. The complete table for all the layer interfaces is shown in the table below. First the environment weather extremes must be determined. The weather environment used for this example was an Atlanta Summer Design Day.016768 0.3°C.012621 0.63116 40. then fiberglass insulation.020962 0.689005 0. here it is very clear that there is a linear portion in the pendular state where a transfer function analysis would be valid.016215 0.016281 Gyp-Inside Inside 25.0123 to 0.014503 Max Temperature 21.0167 kg moisture/m3 of air on the inside.69°C to 33.

0 (100%). In = Interface facing the zone interior Figure 13. and Gypsum Drywall in the figures shown below. which means no condensation took place in this example. A rough straight-line path can be determined on the moisture capacitance curves. but the actual path should be in a band around the linear path. This linear path is obviously not the actual path. as shown in the figure above. 4/23/05 27 . the actual interface relative humidity is never greater than 0.SURFACE HEAT BALANCE MANAGER / PROCESSES MOISTURE TRANSFER MATERIAL PROPERTIES Out = Interface facing the outdoors. As determined from the simulation data. Relative Humidity Curves at Wall Element Interfaces The figure above shows that none of the interfaces were above RH = 1.8 (80%).0 (100%). the actual path did not enter the invalid region of RH > 1. The dark dashed line shows a linear path between the 2 endpoints for each layer for the wall of Brick. Fiberglass Insulation.

0 (100%). Figure 15. and is an invalid region.SURFACE HEAT BALANCE MANAGER / PROCESSES MOISTURE TRANSFER MATERIAL PROPERTIES Figure 14. Moisture Capacitance Curve for Face Brick Layer The approximate paths for all the wall materials are in or at the upper edges of the large linear plane that encompasses the pendular state. or the wet and dry extremes. and further justifies the appropriateness of the linear assumptions for building analysis. It should be noticed that the pendular state contains the majority of the moisture capacitance curves except as conditions approach RH = 1 (100%) and RH = 0 (0%). Moisture Capacitance for Fiber Glass Layer 4/23/05 28 . Also. Engineering judgment will have to be used when the temperature and vapor density conditions approach the upper and lower edges of the pendular state. the upper right hand corner of each moisture capacitance curve corresponds to RH > 1.

R. R.D.e. the linear sections of the moisture capacitance curve can be used in building simulations and a simulation using these moisture extensions can be a useful tool to examine many moisture problems. and Pedersen. For short periods where the cyclic integral of the total moisture adsorption and desorption is near zero (i.SURFACE HEAT BALANCE MANAGER / PROCESSES EFFECTIVE MOISTURE PENETRATION DEPTH (EMPD) MODEL Figure 16. With the current implementation. the EMPD concept has been shown to be a reasonable approximation of reality (Kerestecioglu et al. to assume that the only moisture that effects cooling system performance is contained in the room air is a false. it is difficult to use the MTF simulation approach for annual simulations without careful consideration of the moisture capacitance coefficients due to the moisture non-linearities. If one assumes that all building moisture is contained in the room air. it is very important to know the moisture conditions of the building. lumped approach to simulate surface moisture adsorption and desorption. 1989). “Development of a Response Factor Approach For Modeling The Energy Effects of Combined Heat And Mass Transfer With Vapor Adsorption In Building Elements.J. References Liesen. University of Illinois at UrbanaChampaign. In other words. with proper selection. Thus. ASHRAE Transactions. C. but a profound effect on the performance of air conditioning systems. there is no net moisture storage). EMPD Model Description The EMPD concept assumes that a thin layer (δM) close to the wall surface behaves dynamically and exchanges moisture with the air domain when exposed to cyclic air moisture pulses.J. The EMPD (Effective Moisture Penetration Depth) model is a simplified. Moisture Capacitance for Gypsum Drywall This example shows that. linens.) store and release moisture. furnishings. Liesen. Part 1.. etc. and it can lead to significant error in the prediction of room moisture conditions and cooling system loads.O. Thesis. In order to accurately describe building performance during periods when cooling is needed.g.” Ph. then one ignores the fact that the materials that bound the room (e. 1994. Theory". Department of Mechanical and Industrial Engineering.. 1999 Effective Moisture Penetration Depth (EMPD) MODEL Overview Moisture has little effect on heating system performance. "Modeling the Energy Effects of Combined Heat and Mass Transfer in Building Elements. wall surfaces. the following constraint must be met: 4/23/05 29 .

rather.45) ⎦ ⎣ The lumped mass transfer equation for the i-th solid domain may be written as (40) 4/23/05 30 . (36). the moisture content may be differentiated with respect to time in the following manner: du ∂U dW * ∂U dT * dW * dT * = + * = AT − Bρ dτ ∂W * dτ ∂T dτ dτ dτ (38) where AT and Bρ are the isothermal moisture capacity and thermo-gradient coefficient. From Eqs.45 ⎠ ⎝ 1 * (37) Given that U=U(W*.T*). The EMPD model assumes no spatial distribution of moisture content across the thickness (L) of the solid. This may be mathematically stated as: ∫ L 0 U ( x)dx = U δ M (34) For most building materials. respectively. a thin layer (δM) of uniform moisture content (U) is assumed to represent the total moisture content of the solid. (36) and (37). they can be expressed as: AT = and abϕ b + cdϕ d W* (39) ⎡1 ⎤ 4111 Bρ = − ⎢ * − * *(abϕ b + cdϕ d ) 2⎥ T (T − 35. τ2-τ1 denotes the finite time interval over which the equation holds.7093 − * ⎟ Rv ρ aT T − 35. the equilibrium moisture sorption isotherm can be defined by the following general equation (Kerestecioglu et al.SURFACE HEAT BALANCE MANAGER / PROCESSES τ2 1 EFFECTIVE MOISTURE PENETRATION DEPTH (EMPD) MODEL ∫τ dU dτ = 0 dτ (33) where. 1988): U = aϕ b + cϕ d where (35) W* ϕ≈ Wsat * and (36) Wsat * = 4111 ⎞ ⎛ exp ⎜ 23.

7 ) where (45) ξ= ∆ϕ ∆τ (46) Figure 17 gives the EMPD values to be used for various vapor diffusivities evaluated at different ambient excitations. EMPD value determination An effective moisture penetration depth may be determined from either experimental or detailed simulation data by using actual surface areas and moisture vapor diffusivity. (40) and (41).0211* Dv 0.i Ai (Wr − Wi* ) dτ (41) Using Eqs.i Ai (Wr − Wi* ) + ( Aρbδ M Bρ )i i dτ dτ (42) The energy equation for the envelope contains the surface temperature and is given by the conduction equation ρC p dT = ∇ ⋅ ( k ∇T ) dτ (43) with the boundary conditions at interior surface − k ∇T = − qT "+ hT (T * − Tr ) + λ hM (W * − Wr ) (44) A more detailed account of the numerical solution procedure can be found in Kerestecioglu et al. ( Ai ρbδ M AT )i dWi* dT * = hM . 4/23/05 31 . (1988). (38).567024 − 12. An empirical function derived from the detailed simulation may be used to determine the EMPD value (Kerestecioglu et al. one obtains the final equation needed for closure moisture transfer at internal surface. (39).SURFACE HEAT BALANCE MANAGER / PROCESSES EFFECTIVE MOISTURE PENETRATION DEPTH (EMPD) MODEL ( Aρbδ M )i dU i = hM .21373*exp ( −267. 1989): δ M = 12.7 * ξ −0.

[kg/m2-s] = Convective heat transfer coeff. Dv (m /s x 10 ) 2 6 ξ=0.SURFACE HEAT BALANCE MANAGER / PROCESSES EFFECTIVE MOISTURE PENETRATION DEPTH (EMPD) MODEL 12 Effective penetration depth (mm) 10 8 ξ=0. Limit of effective penetration depth values for various vapor diffusivities at different ambient excitations.16 15 20 Figure 17.52 J/kg-K] = Temperature [K] = Moisture content [kg/kg] = Humidity ratio [kg/kg] Greek letters = Effective penetration depth for moisture equation [m] = Heat of vaporization [J/kg] = Density [kg/m3] = Time [s] = Relative humidity [0 to 1] = Ambient moisture excitation rate [1/h] 4/23/05 32 .K] = Convective mass transfer coeff.02 6 4 2 0 0 5 10 Vapor diffusivity.08 ξ=0.04 ξ=0. EMPD Nomenclature A AT Bρ Cp hM hT k L q"T Rv T U W δM λ ρ τ φ ξ = Area [m2] = Isothermal moisture capacity [m3/kg] = Thermo-gradient coefficient [kg/kg-K] = Specific heat [J/kg. [W/m2-K] = Thermal conductivity [W/m-K] = Length [m] = Imposed heat flux [W/m2] = Ideal gas constant [461.

which is traditionally taken to be positive in the direction from outside to inside of the wall.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE Subscripts and superscripts a b * i = Air = Bulk = Surface = i-th surface References Kerestecioglu. V. ′′ qconv = Convective flux exchange with outside air. reflected. including direct.. FL Outside Surface Heat Balance Shortwave radiation. Swami.. M. L. ′′ qko = Conduction heat flux (q/A) into the wall. 1988. Brahma. 1989. “FSEC 1. P. Cape Canaveral. A. Cape Canaveral. Dabir. A. and diffuse sunlight Longwave radiation from the environment Convective exchange with outside air Outside Face Conduction into wall. Gu. N. P. Each of these heat balance components is introduced briefly below. ′′ qLWR = Net long wavelength (thermal) radiation flux exchange with the air and surroundings.” Florida Solar Energy Center. Chandra. Florida Solar Energy Center. R. "Theoretical and Computational Investigation of Simultaneous Heat and Moisture Transfer in Buildings: Effective Penetration Depth Theory. M. Swami. 1989.1 User’s Manual.. and S. A. FL. Kamel. Kerestecioglu." ASHRAE Winter Meeting. 4/23/05 33 . Simplified procedures generally combine the first three terms by using the concept of a sol-air temperature. P.. Swami and A. Fairey.. Razzaq. "Theoretical and Computational Investigation of Algorithms for Simultaneous Heat and Moisture Transport in Buildings." FSEC-CR-191-88. Kerestecioglu. and Fairey. Atlanta. qko Wall Figure 18. GA.. All terms are positive for net flux to the face except the conduction term.. M. Outside Heat Balance Control Volume Diagram The heat balance on the outside face is: ′′ ′′ ′′ ′′ qα sol + qLWR + qconv − qko = 0 where: (47) ′′ qα sol = Absorbed direct and diffuse solar (short wavelength) radiation heat flux.

These assumptions are frequently used in all but the most critical engineering applications. The radiation heat flux is calculated from the surface absorptivity. Nomenclature List of Variables. and wavelength for each participating surface. surface facing angle and tilt. External Longwave Radiation ′′ qLWR is a standard radiation exchange formulation between the surface. However. and sky and ground view factors. τ = 0. surface temperature. The relevant material properties of the surface. • the medium within the enclosure is non-participating. angle. are complex functions of temperature. emissivity ε and absorptivity α. it is generally agreed that reasonable assumptions for building loads calculations are (Chapman 1984. etc. sky and ground temperatures. Mathematical variable q"LWR Description Exterior surface long wave radiation flux Linearized radiative heat transfer coefficient to air temperature Surface Outside face temperatures Outside air temperature Environmental ground surface temperature Sky Effective temperature view factor of wall surface to ground surface View factor of wall surface to sky Units W/m2 Range - hr W/(m2 K) - Tsurf Tair Tgnd K K K - Tsky Fgnd K - 0~1 Fsky - 0~1 4/23/05 34 . the sky. This is influenced by location. spatial relationships between surfaces and surroundings. and the ground. ρ = 1. Table 3. weather conditions. surface face material properties.ε) each surface is at a uniform temperature energy flux leaving a surface is evenly distributed across the surface. Lienhard 1981): • • • each surface emits or reflects diffusely and is gray and opaque (α = ε. and material properties of the surfaces. The longwave radiation heat exchange between surfaces is dependent on surface temperatures.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE External Shortwave Radiation ′′ qα sol is calculated using procedures presented later in this manual and includes both direct and diffuse incident solar radiation absorbed by the surface face.

we can determine the long wave radiative heat flux at the building exterior surface (Walton 1983. Using the assumptions above.sky (Tsurf − Tsky ) + hr .0000000 567 Consider an enclosure consisting of building exterior surface. gnd = 4 4 εσ Fgnd (Tsurf − Tgnd ) Tsurf − Tgnd (51) 4/23/05 35 .SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE Fair ε σ View factor of wall surface to air Surface long-wave emissivity Stefan-Boltzmann constant W/m2K4 0~1 0~1 0. and air. surrounding ground surface. The total longwave radiative heat flux is the sum of components due to radiation exchange with the ground. and sky. " ′′ ′′ qLWR = q′′ + qsky + qair gnd (48) Applying the Stefan-Boltzmann Law to each component yields: " 4 4 4 4 4 4 qLWR = εσ Fgnd Tsurf − Tgnd + εσ Fsky Tsurf − Tsky + εσ Fair Tsurf − Tair ( ) ( ) ( ) (49) where ε =long-wave emittance of the surface σ =Stefan-Boltzmann constant Fgnd = view factor of wall surface to ground surface temperature Fsky = view factor of wall surface to sky temperature Fair =view factor of wall surface to air temperature Tsurf = outside surface temperature Tgnd = ground surface temperature Tsky = sky temperature Tair = air temperature Linearized radiative heat transfer coefficients are introduced to render the above equation more compatible with the heat balance formulation. gnd (Tsurf − Tgnd ) + hr . ′′ qLWR = hr .air (Tsurf − Tair ) where (50) hr . sky. McClellan and Pedersen 1997).

McClellan. A Heat Transfer Textbook. 103. 1993. Investigation of Outside Heat Balance Models for Use in a Heat Balance Cooling Load Calculation. N. 1997. Thermal Analysis Research Program Reference Manual. 1983. and C. NBSSIR 832655. Inc. ASHRAE Transactions. Lienhard.SURFACE HEAT BALANCE MANAGER / PROCESSES 4 4 εσ Fsky (Tsurf − Tsky ) OUTSIDE SURFACE HEAT BALANCE hr . National Bureau of Standards. New York: Macmillan Publishing Company. A. J.5 (1 + cos φ ) Fsky = 0. 1993 ASHRAE Handbook – Fundamentals. Vol. sky = Tsurf − Tsky 4 4 εσ Fair (Tsurf − Tair ) (52) hr . Refrigerating.. pp. J.5 (1 + cos φ ) (56) The ground surface temperature is assumed to be the same as the air temperature. Inc. N. 4th Edition. gnd = 4 4 εσ Fgnd (Tsurf − Tair ) Tsurf − Tair 4 4 εσ Fsky β (Tsurf − Tsky ) (57) hr . G. The final forms of the radiative heat transfer coefficients are shown here. H.5 (1 − cos φ ) (54) (55) where φ is the tilt angle of the surface. Heat Transfer. O.J. and Air-Conditioning Engineers. T. hr . Walton.air = Tsurf − Tair (59) References ASHRAE. sky = Tsurf − Tsky 4 4 εσ Fsky (1 − β ) (Tsurf − Tair ) (58) hr . M. 1981. 1984.: Prentice-Hall. Chapman. 469-484. Atlanta: American Society of Heating. 4/23/05 36 . Englewood Cliffs.air = Tsurf − Tair (53) The longwave view factors to ground and sky are calculated with the following expressions (Walton 1983): Fgnd = 0. The view factor to the sky is further split between sky and air radiation by: β = 0. Pedersen. Part 2.

Use the “Other Side Coefficients” object in the input file to set heat transfer coefficients and temperatures on surfaces. Terrain-Dependent Coefficients (ASHRAE 2001). is the convection coefficient. the exterior surfaces (exposed to wind) are assumed to be wet. you will see 1000 as its value. and a and δ are terrain-dependent coefficients. EnergyPlus offers a choice of six algorithms: Simple. Specific details are given in the Input/Output reference document.To) where hco. the local wind speed must be individually calculated for each surface. open country Rough. TARP.14 270 0. Both of the options can be applied using general schedules. Terrain 1 2 3 4 5 4/23/05 Description Flat.) In addition to the correlation choices described below. δ is the boundary layer thickness for the given terrain type. forest Exponent. and DOE-2. with much disparity between them (Cole and Sturrock 1977.22 370 0. a δ (m) 0. The convection coefficient is set to a very high number (1000) and the outside temperature used for the surface will be the wet bulb temperature. it is also possible to override the convection coefficients on the outside of any surface by two means: • • Use the “Convection Coefficients” object in the input file to set the convection coefficient value on either side of any surface. Local Wind Speed Calculation All of the outside convection algorithms (except for TARP) use the same calculation to determine the local wind speed at the heat transfer surface. See the “Outside Convection Algorithm” object in the Input Output Reference document. The wind speed in the weather file is assumed to be measured at a meteorological station located in an open field at a height of 10 m. Note that when the outside environment indicates that it is raining. BLAST.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE External Convection ′′ Convection is modeled using the classical formulation: qconv = hco(Tair . Detailed.22 370 37 . The wind speed is modified from the measured meteorological wind speed by the equation (ASHRAE 2001): ⎛δ ⎞ Vz = Vmet ⎜ met ⎟ ⎝ zmet ⎠ amet ⎛z⎞ ⎜ ⎟ ⎝δ ⎠ a (60) where z is the height of the centroid of the surface. zmet is the height of the standard metereological wind speed measurement. Since the 1930's there have been many different methods published for calculating this coefficient. The values of a and δ are shown in the following tables: Table 4. industrial. Layer Thickness. To adjust for different terrain at the building site and differences in the height of building surfaces. (If you choose to report this variable.10 210 0. MoWiTT.33 460 0. wooded country Towns and cities Ocean Urban. Yazdanian and Klems 1994). Substantial research has gone into the formulation of models for estimating the exterior surface convective heat transfer coefficient.

E.33 F 0.49 10.036 Example Material Stucco Brick Concrete Clear pine Smooth Plaster Glass Note that the simple correlation yields a combined convection and radiation heat transfer coefficient.23 10. and F are shown in the following table: Table 5. the roughness coefficients D.065 4. and F. open country (terrain class 1) at a height of 10 m. Radiation to sky.028 0. BLAST. These are described in detail below.057 0. Page 22. MoWiTT. To accomodate one additional terrain in EnergyPlus. and air is calculated automatically by the program.23 E 5. Detailed Algorithm Table 6. Since the meteorological measurement site is assumed to be flat.0 0.58 12.79 8. Simple Algorithm The simple algorithm uses surface roughness and local surface windspeed to calculate the exterior heat transfer coefficient. Roughness Index 1 (Very Rough) 2 (Rough) 3 (Medium Rough) 4 (Medium Smooth) 5 (Smooth) 6 (Very Smooth) D 11. E. Roughness Coefficients D. and air is included in the exterior convection coefficient for this alogrithm. Variable A 4/23/05 Description Surface area of the surface Units m 2 Range /= 0 38 . Specifically. and DOE-2 algorithms. The TARP algorithm uses a different set of coefficients that are unique for all five terrain classes. ground.22 8.5863.0 -0.0 -0. the wind speed calculation becomes: ⎛z⎞ Vz = Vmet β ⎜ ⎟ ⎝δ ⎠ a (61) where β is a constant value of 1. ASHRAE Handbook of Fundamentals (ASHRAE 1989). All other algorithms yield a convection only heat transfer coefficient. Nomenclature List of Variables.1 3. ground. The basic equation used is: h = D + EVz + FVz 2 (62) The roughness correlation is taken from Figure 1.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE The terrain-dependent coefficients are only defined for terrain classes 1-4 in the Detailed. Radiation to sky.894 4. terrain 5 is mapped to the similar terrain 2.192 4.4.0 3.

3=medium rough.5863 Roughness Index - 1~6 4/23/05 39 .SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE hc Surface exterior convective heat transfer coefficient Forced convective heat transfer coefficient Natural convective heat transfer coefficient Perimeter of surface Surface roughness multiplier Environmental air temperature Outside surface temperature Temperature difference between the surface and air. 1=very rough) W/(m2K) - hf hn P Rf Tair Tso ∆T W/(m2K) W/(m2K) m °C °C °C - Vmet m/s - Vz Wf z m/s m /= 0 zmet m - φ degree - a β - = 1. 4=medium smooth. 2=rough. Wind speed measurement at standard meteorological conditions Windspeed modified for height z above ground Wind direction modifier Height of the centroid of the surface above ground Height at which standard metereological wind speed measurements are taken Angle between the ground outward normal and the surface outward normal Terrain-dependent coefficients Terrain-dependent coefficients Surface roughness index (6=very smooth. 5=smooth.

Ramsey.382 + cos φ (for downward heat flow) (67) Note that Equations (66) and (67) are equivalent when the wall is vertical. Surface Roughness Multipliers (Walton 1981). hc = h f + hn (63) The forced convection component is based on a correlation by Sparrow. In all three models.5 for leeward surfaces (65) Leeward is defined as greater than 100 degrees from normal incidence (Walton 1981). BLAST.537 ⋅ W f R f where PVz A (64) Wf = 1. The surface roughness multiplier Rf is based on the ASHRAE graph of surface conductance (ASHRAE 1981) and may be obtained from the following table: Table 7. convection is split into forced and natural components (Walton 1981).238 − cos φ (for upward heat flow) (66) hn = 1.17 1.0 for windward surfaces or Wf = 0.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE The Detailed. The total convection coefficient is the sum of these components. 4/23/05 40 . and TARP convection models are very similar.11 1. Roughness Index 1 (Very Rough) 2 (Rough) 3 (Medium Rough) 4 (Medium Smooth) 5 (Smooth) 6 (Very Smooth) Rf 2. the natural convection component hn is: hn = 9.67 1.482 ⋅ or 3 ∆T 7.52 1.810 ⋅ 3 Ts − To 1.13 1. and Mass (1979): h f = 2.00 Example Material Stucco Brick Concrete Clear pine Smooth Plaster Glass Based on the ASHRAE Handbook of Fundamentals (ASHRAE 1993).

1. Previously the only difference was that the assumed height of the meteorological wind speed measurement was 9. However.14 m instead of 10 m. TARP uses a 4/23/05 41 . Variable hc Description Surface exterior convective heat transfer coefficient Forced convective heat transfer coefficient Natural convective heat transfer coefficient Outside surface temperature Temperature difference between the surface and air.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE BLAST Algorithm As of version 1. and natural convection coefficients given by Equation (66) or (67).2. the BLAST algorithm is identical to the ASHRAE Detailed algorithm.0 α - - β - - φ degree - The TARP detailed convection model is very similar to the BLAST detailed model with forced convection coefficients given by Equation (64). which are then summed using Equation (63). Windspeed modified for height z above ground Height at which standard metereological wind speed measurements are taken Terrain-dependent coefficients Terrain-dependent coefficients Angle between the ground outward normal and the surface outward normal Units W/(m2K) Range - hf W/(m2K) - hn W/(m2K) - Tso ∆T °C/K °C/K - Vz m/s - zmet m = 10. TARP Algorithm Table 8. Nomenclature List of Variables.

47 1. Terrain Roughness Coefficients (Walton 1983). 4/23/05 42 . open country Rough.67 Range - Tso ∆T °C/K °C/K - The MoWiTT model is based on measurements taken at the Mobile Window Thermal Test (MoWiTT) facility (Yazdanian and Klems 1994).20 0. using Equation (68) to replace Equation (60) (Walton 1983): ⎛ z ⎞ Vz = Vmet ⋅ β ⋅ ⎜ ⎟ ⎝ zmet ⎠ α (68) α and β are the terrain-dependent coefficients shown in the following table: Table 9. Units W/(m2K(m/s)b W/(m K ) W/(m2K) 2 4/3 Description Flat. b and turbulent natural convection constant Ct are given in Table 11. The correlation applies to very smooth. Variable A b Ct hc Description Constant Constant Turbulent natural convection constant Surface exterior convective heat transfer coefficient Outside surface temperature Temperature difference between the surface and air.g.35 0.00 0. window glass) in low-rise buildings and has the form: 1 2 ⎡ ⎤ hc = ⎢Ct ( ∆T ) 3 ⎥ + ⎡ aVzb ⎤ ⎣ ⎦ ⎣ ⎦ 2 (69) Constants a. industrial.25 β 1. Terrain 1 2 3 4 5 MoWiTT Algorithm Table 10.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE slightly different equation for calculating the modified wind speed.10 0. wooded country Towns and cities Ocean Urban.85 0.30 0. Nomenclature List of Variables. vertical surfaces (e.15 0. forest α 0.

Table 11.84 a W/m2K(m/s)b 2. glass) is calculated as: 4/23/05 43 .38 2. Nomenclature List of Variables. MoWiTT Coefficients (Yazdanian and Klems 1994).89 0. high-rise surfaces.617 Range - hc.SURFACE HEAT BALANCE MANAGER / PROCESSES OUTSIDE SURFACE HEAT BALANCE NOTE: The MoWiTT algorithm may not be appropriate for rough surfaces.glass W/(m2K) - hn W/(m2K) - Rf Tso ∆T °C/K °C/K - φ radian - Description of the Model and Algorithm The DOE-2 convection model is a combination of the MoWiTT and BLAST Detailed convection models (LBL 1994).84 0. or surfaces that employ movable insulation. Variable a b hc Description Constant Constant Surface exterior convective heat transfer coefficient Convective heat transfer coefficient for very smooth surfaces (glass) Natural convective heat transfer coefficient Surface roughness multiplier Outside surface temperature Temperature difference between the surface and air.86 b 0.g. The convection coefficient for very smooth surfaces (e. Wind Direction (Units) Windward Leeward DOE-2 Algorithm Table 12. Angle between the ground outward normal and the surface outward normal Units W/(m2K(m/s)b W/(m K) 2 Ct W/m2K4/3 0.

Atlanta: American Society of Heating. References ASHRAE. S. R. Inside Heat Balance The heart of the heat balance method is the internal heat balance involving the inside faces of the zone surfaces. Cole. Inc. and Air-Conditioning Engineers. 1989 ASHRAE Handbook – Fundamentals. 1981 ASHRAE Handbook – Fundamentals. and Air-Conditioning Engineers.1E-053 source code. the convection coefficient is modified according to the equation hc = hn + R f (hc . ASHRAE Transactions. Refrigerating. ASHRAE. The incident short wave radiation is from the solar radiation entering the zone through windows and emittance from internal sources such as lights. 1993 ASHRAE Handbook – Fundamentals. Atlanta: American Society of Heating. Exterior Conduction (71) ′′ The conduction term. Vol. National Bureau of Standards. 1981. Inc. Thermal Analysis Research Program Reference Manual. M. p. E. 1983. 2) convection to the air. 2001 ASHRAE Handbook – Fundamentals. 3) short wave radiation absorption and reflectance and 4) long wave radiant interchange. The Convective Heat Exchange at the External Surface of Buildings. and Air-Conditioning Engineers. Sturrock. Technical Report. 1989. 1994. 1994. Atlanta: American Society of Heating. 12. IL. This heat balance is generally modeled with four coupled heat transfer components: 1) conduction through the building element. 1087. NBSSIR 832655. ASHRAE. 204. H.. G. United States Army Construction Engineering Research Laboratory. Air Flow Around Buildings. 207. J. ASHRAE. and Air-Conditioning Engineers. Yazdanian. Klems. glass − hn ) where Rf is the roughness multiplier given by Table 7. M. DOE2. Mass. 1977. Vol. Measurement of the Exterior Convective Film Coefficient for Windows in Low-Rise Buildings. and N. Inc. glass = hn2 + ⎡ aVzb ⎤ ⎣ ⎦ 2 (70) hn is calculated using Equation (66) or Equation (67). The long wave radiation interchange 4/23/05 44 . Sparrow. Refrigerating. Ramsey. For less smooth surfaces. 1979. 1993. J. Part 1. Chapter 3. Refrigerating. G. Heat Transfer. Atlanta: American Society of Heating. I-P & S-I Editions. Walton. p. A. Building and Environment. Champaign.SURFACE HEAT BALANCE MANAGER / PROCESSES INSIDE HEAT BALANCE hc . qko . Journal of Heat Transfer. 2001. Refrigerating. 1981. N. W. and E. 100. Chapter 16.. Lawrence Berkeley Laboratory (LBL). Constants a and b are given in Table 11. Walton. 101. p. Typically in EnergyPlus. the Conduction Transfer Function (CTF) method is used. Inc. and J. N. Vol. Effect of Finite Width on Heat Transfer and Fluid Flow about an Inclined Rectangular Plate. can in theory be calculated using a wide variety of heat conduction formulations. Passive Solar Extension of the Building Loads Analysis and System Thermodynamics (BLAST) Program. The available models are described in this section: Conduction Through The Walls.

It also permits separating the radiant and convective parts of the heat transfer at the surface. equipment. The zone air completely absorbs LW radiation from the surfaces within the zone. Wall Room Shortwave radiation from solar and internal sources Longwave radiation exchange with other surfaces in zone Conduction from outside. it oversimplifies the zone surface exchange problem. The heat balance on the inside face can be written as follows: ′′ ′′ ′′ ′′ q′′ + qSW + q′′ + qki + qsol + qconv = 0 LWX LWS where: (72) q′′ = Net long wave radiant exchange flux between zone surfaces. 4/23/05 45 . The model. LWX ′′ = Net short wave radiation flux to surface from lights. Inside Heat Balance Control Volume Diagram Internal LW Radiation Exchange LW Radiation Exchange Among Zone Surfaces There are two limiting cases for internal LW radiation exchange that are easily modeled: The zone air is completely transparent to LW radiation. qki Longwave radiation from internal sources Convective heat exchange with zone air Figure 19. qSW q′′ = Long wave radiation flux from equipment in zone. LWS ′′ = Conduction flux through the wall. This means that it does not participate in the LW radiation exchange among the surfaces in the zone. Each of these heat balance components is introduced briefly below. qki ′′ qsol = Transmitted solar radiation flux absorbed at surface. which considers room air to be completely transparent. ′′ qconv = Convective heat flux to zone air. and people. The limiting case of completely absorbing air has been used for load calculations and also in some energy analysis calculations. is reasonable physically because of the low water vapor concentrations and the short mean path lengths. the heat balance formulation in EnergyPlus treats air as completely transparent. This model is attractive because it can be formulated simply using a combined radiation and convection heat transfer coefficient from each surface to the zone air. such as all other zone surfaces. and as a result. However. which is an important attribute of the heat balance method.SURFACE HEAT BALANCE MANAGER / PROCESSES INSIDE HEAT BALANCE includes the absorption and emittance of low temperature radiation sources.

Because the approximate view factors may not satisfy the basic requirements of reciprocity (two surfaces should exchange equal amounts of heat in each direction). This procedure relies on a matrix of exchange coefficients between pairs of surfaces that include all exchange paths between the surfaces. No surface facing within 10 degrees of another surface is seen by the other surface. This model is based on the “ScriptF” concept developed by Hottel (Hottel and Sarofim.0). the long wave radiant exchange is calculated for each surface using: qi . only reciprocity is enforced. If the area of one surface in a zone is greater than the sum of the areas of all other surfaces. Once the ScriptF coefficients are determined. The ScriptF coefficients are developed by starting with the traditional direct radiation view factors. absorptions and re-emissions from other surfaces in the enclosure are included in the exchange coefficient. for very large surfaces. then the enclosure will be treated as normal. but in some special situations they are not. floors and ceilings (subject to the preceding facing direction constraint). 4/23/05 46 . The other limitation is that the exact calculation of direct view factors is computationally very intensive even if the positions of all surfaces are known. Thermal Mass and Furniture Furniture in a zone has the effect of increasing the amount of surface area that can participate in the radiation and convection heat exchanges.j is the ScriptF between surfaces i and j. EnergyPlus uses a procedure to approximate the direct view factors. The suggested action for the second case (the extra large surface) is to divide the large surface into several smaller surfaces. McGraw Hill. that enforcement becomes impossible. These two changes both affect the response to temperature changes in the zone and also affect the heat extraction characteristics. and the results should be examined very carefully to ascertain that they are reasonable. and special rules are applied. Warning messages are produced for both of these cases. All surfaces see thermal mass surfaces. Accordingly. reciprocity only is enforced. In other words all reflections. there are several complicating factors in finding the direct view factors-–the main one being that the location of surfaces such as thermal mass representing furniture and partitions are not known. Chapter 3. and completeness (every surface should have a direct view factor sum of 1. and the view factors are modified so that only the large surface is seen by very small surfaces. Approximate the direct view factor from surface 1 to surface 2 as the ratio of the area of surface 2 to the total area “seen” by surface 1. In the case of building rooms and zones. 1967). Normally both of the requirements are satisfied. Radiative Transfer. which is called ScriptF. but sometimes. It also adds participating thermal mass to the zone. The procedure has two steps: • • Determine the total area of other surfaces “seen” by a surface. EnergyPlus does a view factor fix operation before they are used in the ScriptF determination. Both assumptions are reasonable for building zone interchange. j = Ai Fi . The major assumptions are that all surface radiation properties are grey and all radiation is diffuse.SURFACE HEAT BALANCE MANAGER / PROCESSES INSIDE HEAT BALANCE EnergyPlus uses a grey interchange model for the long wave radiation among zone surfaces. The determination of the “seen” area has several constraints: • • • • No surface sees itself. j (Ti 4 − T j4 ) where Fi. All surfaces see roofs. • • If a user includes less than four surfaces in a zone.

4/23/05 47 . Specific details are given in the Input/Output reference document. but the heat balance formulation allows the effect to be modeled in a realistic manner by including the furniture surface area and thermal mass in the heat exchange process.SURFACE HEAT BALANCE MANAGER / PROCESSES INSIDE HEAT BALANCE The proper modeling of furniture is an area that needs further research. The values can be specified directly or with schedules. of course. a mixed / forced convection model. and a trombe wall convection model. but that would involve partial surface irradiation. is not a completely realistic model. This represents the heat transfer to the inside face of the building element. Reference “Inside Convection Algorithm” object in the Input Output Reference document and the inside convection field for each zone. a CTF formulation is used to determine this heat flux. Convection to Zone Air The convection flux is calculated using the heat transfer coefficients as follows: ′′ qconv = hc (Ta − Ts ) (73) The inside convection coefficients (hc) can be calculated using one of several models. These models are explained in the following sections. In addition to the correlation choices described below. Internal SW Radiation SW Radiation from Lights The short wavelength radiation from lights is distributed over the surfaces in the zone in some prescribed manner. qki shown in Equation (31). Since the heat balance approach can deal with any distribution function. it is also possible to override the convection coefficients on the inside of any surface by using the “Convection Coefficients” object in the input file to set the convection coefficient value on the inside of any surface. The radiative part is then distributed over the surfaces within the zone in some prescribed manner. This. which is inconsistent with the rest of the zone model that assumes uniform conditions over an entire surface. it is virtually impossible to treat this source in any more detail since the alternative would require knowledge of the placement and surface temperatures of all equipment. It would be possible to calculate the actual position of beam solar radiation. An overall default for the simulation is selected in the “Inside Convection Algorithm” object and can be overridden by selecting a different option in a zone description. However. LW Radiation From Internal Sources The traditional model for this source is to define a radiative/convective split for the heat introduced into a zone from equipment. it is possible to change the distribution function if it seems appropriate. Transmitted Solar Transmitted solar radiation is also distributed over the surfaces in the zone in a prescribed manner. and it departs from the heat balance principles. Interior Convection Four inside convection models are included in EnergyPlus: two natural convection models. The current procedures incorporate a set of prescribed distributions. Again. Currently the implementation uses coefficients based on natural convection correlations (3 options) and mixed and forced convection (1 option). Interior Conduction ′′ This contribution to the inside surface heat balance is the wall conduction term.

correlates the convective heat transfer coefficient to the surface orientation and the difference between the surface and zone air temperatures (where ∆T = Surface Temp.482 ∆T 1 3 7.0 AND an downward facing surface) a reduced convection correlation is used. then assigns heat transfer coefficients as follows: For a horizontal surface with reduced convection h = 0.040 For a vertical surface: h = 3.0 AND an downward facing surface) an enhanced convection correlation is used.810 ∆T 1 3 1.31 ∆T 1 3 (74) For (∆T < 0. which is based on flat plate experiments.948 For a horizontal surface with enhanced convection h = 4.281 For a Tilted surface with Enhanced Convection 4/23/05 48 . For (∆T > 0. Simple Natural Convection Algorithm The simple convection model uses constant coefficients for each of three heat transfer configurations as follows. h = 1. – Air Temp.283 − cos Σ (75) where Σ is the surface tilt angle.382 + cos Σ (76) where Σ is the surface tilt angle. h= 9.0 AND an upward facing surface) OR (∆T > 0.076 For a Tilted surface with Reduced Convection h = 2. The model uses a criteria similar to the detailed model criteria to calculate reduced and enhanced convection scenarios.). h= 1. The following algorithm is used: For no temperature difference OR a vertical surface the following correlation is used.SURFACE HEAT BALANCE MANAGER / PROCESSES INSIDE HEAT BALANCE Detailed Natural Convection Algorithm The detailed natural convection model.0 AND an upward facing surface) OR (∆T < 0.

SURFACE HEAT BALANCE MANAGER / PROCESSES

INSIDE HEAT BALANCE

h = 3.870

Ceiling Diffuser Algorithm The ceiling algorithm is based on a room outlet temperature reference. The correlations shown in the figures below. For Floors:

**h = 3.873 + 0.082 ∗ ACH 0.98
**

The correlation for floors is illustrated in the following figure:

Floor Correlation 14 12 h (W/m**2-K) 10 8 6 4 2 0 0 50 ACH 100 150

(77)

exp. data correlation

Figure 20. Ceiling Diffuser Correlation for Floors For ceilings:

**h = 2.234 + 4.099 ∗ ACH 0.503
**

The correlation for ceilings is illustrated in the following figure:

(78)

4/23/05

49

SURFACE HEAT BALANCE MANAGER / PROCESSES

INSIDE HEAT BALANCE

Ceiling Correlation 50 40 h (W/m**2-K) 30 20 10 0 0 50 ACH 100 150

exp. data correlation

Figure 21. Ceiling Diffuser Correlation for Ceilings For Walls:

**h = 1.208 + 1.012 ∗ ACH 0.604
**

The correlation for walls is illustrated in the following figure:

Wall Correlation 25 h (W/m**2-K) 20 15 10 5 0 0 50 ACH 100 150

(79)

exp. data correlation

Figure 22. Ceiling Diffuser Correlation for Walls Trombe Wall Algorithm The Trombe wall algorithm is used to model convection in a "Trombe wall zone", i.e. the air space between the storage wall surface and the exterior glazing. (See the later sections on Passive and Active Trombe Walls below for more information about Trombe walls.) The algorithm is identical to the convection model (based on ISO 15099) used in Window5 for convection between glazing layers in multi-pane window systems. Validation of this model for use with a Trombe wall has not yet been completed. This algorithm gives the convection coefficients for air in a narrow vertical cavity that is sealed and not ventilated. This applies both to the air gap in between panes of a window or to the air gap between the Trombe wall glazing and the inner surface (often a selective surface). These convection coefficients are really the only difference between a normal zone and a Trombe zone. The rest of the zone heat balance is the same, e.g., transmitted solar, long-wave radiation between surfaces, etc.

4/23/05 50

SURFACE HEAT BALANCE MANAGER / PROCESSES

TRANSPARENT INSULATION MATERIAL (TIM)

For a vertical cavity, the correlation from ISO 15099 is:

**NU1=0.0673838i Ra 3 NU1=0.028154Ra 0.4134 NU1=1+1.7596678E-10iRa 2.2984755
**

NU2 = 0.242i Ra

1

for 5E4 < Ra < 1E6 for 1E4 < Ra < 5E4 for Ra <= 1E4

(

A)

0.272

**NU=MAX(NU1,NU2)
**

where Nu = Nusselt number Ra = Rayleigh number A = aspect ratio of cavity This is then used in EnergyPlus as follows: Net convection coefficient from glazing to wall is:

hnet = k i NU

(

L

)

where k = conductivity of air L = air gap thickness Convection coefficient applied to each wall separately and actually used in the zone heat balance is:

hc = 2ihnet

References ISO15099. Thermal performance of windows, doors and shading devices: detailed calculations. Draft, July 18, 2001.

**Transparent Insulation Material (TIM)
**

Introduction Transparent Insulation Materials (TIM) were originally designed for use in solar collector systems, where there was a need to increase the insulation in the solar collector without dramatically reducing solar energy transmittance. Transparent Insulation provides both these properties, insulation from heat loss and transmittance of solar energy. The combination of these properties is achieved, because Transparent Insulation is a transmitter of short wave radiation but a barrier to long wave radiation. Therefore short wave solar radiation passes through the Transparent Insulation and long wave heat radiation is insulated by the transparent insulation. Incident solar energy falling on the transparent insulation is reflected and re-reflected within the material and eventually falls on the absorber. In addition,

4/23/05

51

SURFACE HEAT BALANCE MANAGER / PROCESSES

TRANSPARENT INSULATION MATERIAL (TIM)

transparent insulation materials also have increase thermal resistance due to conduction in comparison to standard glass. Transparent Insulation is now used in the housing industry as a passive solar feature. It is attached to the walls of houses for insulation and solar energy gains are transmitted to the house during the right ambient conditions. The walls of the house act as a thermal mass, absorbing the sunlight at the surface and converting it into heat which is slowly transmitted to the inside of the house. Comparison of Opaque and Transparent Insulation A qualitative comparison between the performance of Transparent Insulation and opaque insulation is shown diagrammatically in the figure below. The upper half of the figure represents approximate heat transfer through the wall cross-section for both transparent and opaque insulation cases. The lower half of this figure shows representative temperature variations through the wall cross-sections for different solar conditions.

Figure 23. Energy flows of opaquely and transparently insulated walls (Wood and Jesch 1993). While both types of insulation reduce energy losses from the building via conduction through the building surfaces, transparent insulation allows solar radiation to penetrate deeper into the surface construction. This increases the construction internal temperature and can result in heat being conducted into the building under the proper weather conditions. This can be seen in the lower half of the above figure during a sunny day. The temperature plot shows a maximum between the transparent insulation and the rest of the surface construction. As a result, the temperature gradient results in heat transfer from this point into the interior space, causing a heating effect on the zone. Thus, the advantage of transparent insulation is that, like opaque insulation, it reduces winter heat transfer losses during low or no solar conditions and has the possibility of providing heating during sunny winter days. It should be noted that this same effect in summer could be detrimental to the cooling loads of a building since the introduction of solar radiation closer to the space will increase the solar heating within the zone. Most systems counteract this with a shading device or with sophisticated transparent insulation systems. Types of Transparent Insulation Materials Transparent insulation can be classified into four general categories: Absorber Parallel Covers Cavity Structures

4/23/05 52

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

Absorber Vertical Covers Quasi-Homogeneous Structures Cross-sections of each of these types is shown in the figure below. The arrows in these diagrams indicate solar rays and the path these rays trace as the are transmitted through the transparent insulation layer. The most advantageous set-up (see absorber-parallel below) would send most of the rays downward towards the interior of the building while minimizing the rays that are reflected back to the exterior environment.

Figure 24. Geometrical categories of classification for Transparent Insulation Material (Wood and Jesch 1993).

**Basic Heat Balance -- TARP and FORTRAN Algorithm
**

Basic Heat Balance Cases A heat balance must exist at the outside surface-air interface. The incoming conductive, convective, and radiative fluxes must sum up to zero:

Conductive + Convective + Radiative = 0

(80)

In contrast to the internal surface heat balance that treats all surfaces simultaneously, the external thermal balance for each surface is performed independent of all other surfaces. This implies that there is no direct interaction between the individual surfaces. TARP includes four possible representations for the basic outside surface heat balance. The first two depend on which of the optimal surface conductance algorithms the user selects. The simple outside surface conductance that includes both the convective and thermal interchange between the surface and the environment in a single coefficient, is represented by the thermal network in Figure 25. Equation (80) can also be expressed as:

**[ KOPt + Y0 iTIt − X 0 iTO t ] + ⎡ HOi( Ta − TOt ) ⎤ ⎣ ⎦
**

4/23/05

+QSO = 0

(81)

53

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

This can be solved for the outside surface temperature.

⎡ KOPt + QSO + Y0 iTI t +HOiTa ⎤ TO t = ⎢ ⎥ X 0 +HO ⎣ ⎦

(82)

The detailed outside surface conductance model considers convection and radiant interchange with the sky and with the ground as separate factors. Its use in the outside thermal balance is shown in Figure 26. In this case, equation (80) can be expanded to give

**[ KOPt +Y0 iTI t -X 0 iTO t ] + ⎡ HAi( Ta -TO t ) +HSi( Ts -TO t ) +HGi( Tg -TO t )⎤ ⎣ ⎦
**

This can be solved for the outside surface temperature:

+QSO = 0

(83)

⎡ KOPt + QSO + Y0 iTI t +HAiTa +HSiTs +HG iTg ⎤ TO t = ⎢ ⎥ X 0 +HA+HS+HG ⎣ ⎦

(84)

The third and fourth representations occur when the outside surface has been covered with movable insulation. The insulation has a conductance of UM. The thermal network in Figure 27 represents this case. The insulation must be mass-less because it is not generally possible to perform a correct thermal balance at the juncture of two surfaces each modeled by CTF. The equation for the thermal balance between the surface and the insulation is

**⎡ KOPt + Y0 iTIt − X 0 iTO t + UMi( TM-TO t ) ⎤ +QSO = 0 ⎣ ⎦
**

Which can be rewritten to solve for TO :

(85)

⎡ KOPt + QSO + Y0 iTI t +UM iTM ⎤ TO t = ⎢ ⎥ X 0 +UM ⎣ ⎦

(86)

Depending on whether or not the detailed or simple algorithm for surface conductance is being used, there are two expressions for TM, the outside temperature of the insulation. For the simple conductance:

**⎡ QSM+UM iTO t +HOiTa ⎤ TM = ⎢ ⎥ UM+HO ⎣ ⎦
**

For the detailed conductance:

(87)

⎡ QSM + UM iTO t +HAiTa +HSiTs +HG iTg ⎤ TO t = ⎢ ⎥ UM+HA+HS+HG ⎣ ⎦

(88)

In this case the values of HA, HS and HG must be found by using an estimated value of TM in place of TO.

4/23/05

54

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

QSO

KOP

TA 1/Ho

TO 1/Yo

TI

Figure 25. Thermal network for simple outside surface coefficient

TS

QSO

1/HS KOP

TA 1/HA

TO 1/Yo

TI

TG

1/HG

Figure 26. Thermal network for detailed outside surface coefficient

QSO

KOP

TM 1/UM

TO 1/Yo

TI

Figure 27. Thermal network for outside movable insulation Heat Balance Cases TOt and TIt are related through the Y0 CTF. However TIt is also unknown. While it is possible to combine the outside and the inside surface heat balances to compute TOt and TIt simultaneously, TARP uses a simpler procedure where TOt is based on a previous value of TI. When Y0 is small, as occurs in well insulated or very massive surfaces, TIt can be replaced by TIt -1 (which is known for the previous hour’s heat balance) without significantly effecting the value of TOt . When Y0 is large, TO and TI can so strongly be coupled that separate outside and inside heat balances do not work because the environment and zone temperatures have negligible influence on the heat balances. The TARP uses the inside surface heat balance to couple TOt with TZ and TR. These two temperatures are less strongly influenced by TO and allow a reasonable heat balance. On the first heat balance iteration, TZ and TR are the values at time t-1. The user may optionally require that TOt be recomputed with every iteration of TIt . In this case TZ and TR have values from the previous iteration and a true simultaneous solution is achieved. In most conventional constructions,

4/23/05 55

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

recomputing TOt does not significantly change the computed zone loads and temperatures. The inside surface heat balance is given by

⎡ KIPt + QSI + HC iTZ + HRiTR + Y0 iTO ⎤ TI t = ⎢ ⎥ Z 0 + HC + HR ⎣ ⎦

(89)

The surface heat balances can be combined in eight ways according to conditions for calculations of the outside surface temperature

⎡ ⎤ Y0 F1 = ⎢ ⎥ ⎣ Z 0 + HI + HR ⎦

(90)

⎡ UM ⎤ F2 = ⎢ ⎣UM + HO ⎥ ⎦ UM ⎡ ⎤ F3 = ⎢ ⎣UM + HA + HS + HG ⎥ ⎦

Case1: Y0 small, simple conductance, no movable insulation: From Equation (82)

(91)

(92)

**⎡ KOPt + QSO + Y0 iTI t-1 +HOiTa ⎤ TO t = ⎢ ⎥ X 0 +HO ⎣ ⎦
**

Case2: Y0 not small, simple conductance, no movable insulation: From Equations (82) and (89)

(93)

**⎡ KOPt + QSO + HOiTa + F1 i( KIPt +QSI+HIiTZ+HR iTR ) ⎤ TO t = ⎢ ⎥ X 0 +HO-F1 iY0 ⎣ ⎦
**

Case3: Y0 small, detailed conductance, no movable insulation: From Equation (83)

(94)

**⎡ KOPt + QSO + Y0 iTI t-1 +HA iTa + HSiTs + HG iTg ⎤ TO t = ⎢ ⎥ X 0 +HA+HS+HG ⎣ ⎦
**

Case4: Y0 not small, detailed conductance, no movable insulation: From Equations (83) and (89)

(95)

⎡ KOPt + QSO + HAiTa + HSiTs + HG iTg + F1 i( KIPt +QSI+HIiTZ+HR iTR ) ⎤ TO t = ⎢ ⎥ X 0 +HA+HS+HG-F1 iY0 ⎣ ⎦

(96)

4/23/05

56

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

Case5: Y0 small, simple conductance, with movable insulation: From Equations (86) and (87)

⎡ KOPt + QSO + Y0 iTI t-1 +F2 i( QSM+HOiTM ) ⎤ TO t = ⎢ ⎥ X 0 +UM-F2 i UM ⎣ ⎦

(97)

Case6: Y0 not small, simple conductance, with movable insulation: From Equations (86), (87) and (89)

**⎡ KOPt + QSO + F2 i( QSM+HOiTa ) + F1 i( KIPt +QSI+HIiTZ+HR iTR ) ⎤ TO t = ⎢ ⎥ X 0 + UM-F2 i UM-F1 iY0 ⎣ ⎦
**

Case7: Y0 small, detailed conductance, with movable insulation: From Equations (86) and (88)

(98)

**⎡ KOPt + QSO + Y0 iTI t-1 +F3 ( QSM+HAiTa +HSiTs +HG iTg ) ⎤ TO t = ⎢ ⎥ X 0 +UM-F3 i UM ⎢ ⎥ ⎣ ⎦
**

Case8: Y0 not small, detailed conductance, with movable insulation: From Equations (86), (88) and (89)

(99)

**⎡ KOPt + QSO + F1 i( KIPt +QSI+HIiTZ+HR iTR ) +F3 ( QSM+HA iTa +HSiTs +HG iTg ) ⎤ TO t = ⎢ ⎥ X 0 +UM-F3 i UM-F1 iY0 ⎢ ⎥ ⎣ ⎦
**

FORTRAN Algorithm Examples Case5: Y0 small, simple conductance, with movable insulation: From Equation (97)

! Outside heat balance case: Movable insulation, slow conduction, simple convection F2 = DBLE(HmovInsul) / ( DBLE(HmovInsul) + DBLE(HExtSurf(SurfNum)) ) TH(SurfNum,1,1) = (-CTFConstOutPart(SurfNum) & +DBLE(QRadSWOutAbs(SurfNum) ) & +Construct(ConstrNum)%CTFCross(0)*TempSurfIn(SurfNum) & +F2* ( DBLE(QRadSWOutMvIns(SurfNum)) & + DBLE(HExtSurf(SurfNum))* DBLE(TempExt) ) ) & /( Construct(ConstrNum)%CTFOutside(0) + DBLE(HmovInsul) & - F2* DBLE(HMovInsul))

(100)

**Case6: Y0 not small, simple conductance, with movable insulation: From Equation (98)
**

! Outside heat balance case: Movable insulation, quick conduction, simple convection F2 = DBLE(HmovInsul) / ( DBLE(HmovInsul) + DBLE(HExtSurf(SurfNum)) ) TH(SurfNum,1,1) = (-CTFConstOutPart(SurfNum) & + DBLE(QRadSWOutAbs(SurfNum)) & +F2*( DBLE(QRadSWOutMvIns(SurfNum)) & +DBLE(HExtSurf(SurfNum))* DBLE(TempExt) ) &

4/23/05 57

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

+F1*( CTFConstInPart(SurfNum) & + DBLE(QRadSWInAbs(SurfNum)) & + DBLE(QRadThermInAbs(SurfNum)) & + DBLE(HConvIn(SurfNum))*MAT(ZoneNum) & + DBLE(NetLWRadToSurf(SurfNum)) ) ) & /( Construct(ConstrNum)%CTFOutside(0) + DBLE(HmovInsul) & -F2* DBLE(HMovInsul )- F1*Construct(ConstrNum)%CTFCross(0) )

**Case7: Y0 small, detailed conductance, with movable insulation: From Equation (99)
**

! Outside heat balance case: Movable insulation, slow conduction, convection F2 = DBLE(HMovInsul)/ ( DBLE(HMovInsul) + DBLE(HExtSurf(SurfNum)) +DBLE(HSky) + DBLE(HGround) ) TH(SurfNum,1,1) = (-CTFConstOutPart(SurfNum) +DBLE(QRadSWOutAbs(SurfNum)) +Construct(ConstrNum)%CTFCross(0)*TempSurfIn(SurfNum) +F2*( DBLE(QRadSWOutMvIns(SurfNum)) +DBLE(HExtSurf(SurfNum))*DBLE(TempExt) +DBLE(HSky)*DBLE(SkyTemp) +DBLE(HGround)*DBLE(OutDryBulbTemp) ) ) /( Construct(ConstrNum)%CTFOutside(0) +DBLE(HMovInsul) - F2*DBLE(HMovInsul) ) detailed & & & & & & & & &

**Case8: Y0 not small, detailed conductance, with movable insulation: From Equation (100)
**

! Outside heat balance case: Movable insulation, quick conduction, detailed convection F2 = DBLE(HMovInsul)/ ( DBLE(HMovInsul) + DBLE(HExtSurf(SurfNum)) & +DBLE(HSky) + DBLE(HGround) ) TH(SurfNum,1,1) = (-CTFConstOutPart(SurfNum) & +DBLE(QRadSWOutAbs(SurfNum)) & +F1*( CTFConstInPart(SurfNum) & +DBLE(QRadSWInAbs(SurfNum)) & +DBLE(QRadThermInAbs(SurfNum)) & +DBLE(HConvIn(SurfNum))*MAT(ZoneNum) & +DBLE(NetLWRadToSurf(SurfNum)) ) & +F2*( DBLE(QRadSWOutMvIns(SurfNum)) & +DBLE(HExtSurf(SurfNum))*DBLE(TempExt) & +DBLE(HSky)*DBLE(SkyTemp) & +DBLE(HGround)*DBLE(OutDryBulbTemp) ) & /( Construct(ConstrNum)%CTFOutside(0) & +DBLE(HMovInsul) - F2*DBLE(HMovInsul) & -F1*Construct(ConstrNum)%CTFCross(0) )

Fortran Variable Descriptions Table 13. Fortran Variables and Descriptions FORTRAN Variable TH(SurfNum,1,1) Description Temperature History(SurfNum,Hist Term,In/Out), where: Hist Term (1 = Current Time, 2MaxCTFTerms = previous times), Tarp Variable TOt Units C Description Temperature of outside of surface I at time t

4/23/05

58

SURFACE HEAT BALANCE MANAGER / PROCESSES

BASIC HEAT BALANCE -- TARP AND FORTRAN ALGORITHM

In/Out (1 = Outside, 2 = Inside) Construct(ConstrNum)%CTFCross(0) Construct(ConstrNum)%CTFInside(0) Construct(ConstrNum)%CTFOutside(0) CTFConstInPart(SurfNum) Cross or Y term of the CTF equation Inside or Z terms of the CTF equation Outside or X terms of the CTF equation Constant inside portion of the CTF calculation Y0 Z0 X0 KIPt W/m2K W/m2K W/m2K W/m2 Cross CTF term Inside CTF term Outside CTF term Portion of inward conductive flux based on previous temperature and flux history terms Portion of outward conductive flux based on previous temperature and flux history terms Radiation interchange factor between surfaces C Temperature of ground at the surface exposed to the outside environment Inside convection coefficient Overall outside surface conductance Radiative conductance (outside surface to ground temperature Conductance fM bl

59

CTFConstOutPart(SurfNum)

Constant Outside portion of the CTF calculation

KOPt

W/m2

F1, F2, F3

Intermediate calculation variables

F1, F2, F3

GroundTemp

Ground surface temperature

Tg

HConvIn(SurfNum)

Inside convection coefficient Outside Convection Coefficient

HI

W/m2K

HExtSurf(SurfNum)

HO, HA

W/m2K

HGround

Radiant exchange (linearized) coefficient

HG

W/m2K

HmovInsul

4/23/05

Conductance or "h" l f bl

UM

W/m2K

TARP AND FORTRAN ALGORITHM value of movable insulation HSky Radiant exchange (linearized) coefficient HS W/m2K of Movable insulation Radiative conductance (outside surface to sky radiant temperature Temperature of zone air Net surface to surface radiant exchange Short wave radiant flux absorbed at inside of surface Short wave radiant flux absorbed at outside of surface Short wave radiant flux absorbed at surface of movable insulation Long wave radiant flux from internal gains Sky temp Temperature of external surface of movable insulation or outside ambient air temperature Temperature of inside of surface I at time t-1 MAT(ZoneNum) NetLWRadToSurf(SurfNum) Zone temperature Net interior long wave radiation to a surface from other surfaces Short-wave radiation absorbed on inside of opaque surface TZ HR*TR C W/m2 QRadSWInAbs(SurfNum) QSI W/m2 QRadSWOutAbs(SurfNum) Short wave radiation absorbed on outside opaque surface QSO W/m2 QRadSWOutMvIns(SurfNum) Short wave radiation absorbed on outside of movable insulation QSM W/m2 QRadThermInAbs(SurfNum) Thermal Radiation absorbed on inside surfaces Sky temperature Exterior surface temperature or exterior air temperature Ts TM. Ta W/m2 SkyTemp TempExt C C TempSurfIn(SurfNum) Temperature of inside surface for each heat transfer surface TIt-1 C 4/23/05 60 .SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -.

TARP AND FORTRAN ALGORITHM TIM. not throughout the material. α Ss + I s i Fss + I g i Fsg ) S (102) = solar absorptance of the surface • = angle of incidence of the sun’s rays S = area of the surface Ss = sunlit area of the surface Ib = intensity of the beam (direct) radiation Is = intensity of the sky diffuse radiation Ig = intensity of the beam (direct) radiation Fss = angle factor between the surface and the sky Fsg = angle factor between the surface and the ground Now. α wall + ρ wall = 1 (103) The model for TIM is simplified in that it assumes that absorption of solar radiation takes place at the inside and outside of the TIM only. the model assumes that the solar radiation absorbed during the first pass through the TIM affects the outside surface of the TIM while the solar radiation reflected at the outer wall surface that 4/23/05 61 . showing energy flow Mathematical model to calculate amount of energy absorbed at the surface of Movable Insulation (TIM) and at the Outside surface of the Wall.Basic Mathematical Model TIM TIM Incident Solar Gain TIM QSM Wall WALL WALL OUTSIDE QSO INSIDE Figure 28. In addition.SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -. Cross Section of TIM and wall. QSM = αTIM i Incident Solar (101) The total solar gain on any exterior surface is a combination of the absorption of direct and diffuse solar radiation given by Incident Solar = ( I b icos θ i Where.

Thus. the heat absorbed at the outside of the TIM is as shown in Equation (101). QSM QSO αTIM τTIM αWALL ⎫ ⎞ ⎪ ⎟ + (1 − α wall ) ⎬ ⎪ ⎠ ⎭ (106) = Short wave radiant flux absorbed at surface of Movable Insulation = Short wave radiant flux absorbed at surface of Wall. = Absorptance of TIM = Transmittance of TIM. = Absorptance of Wall.TARP AND FORTRAN ALGORITHM gets absorbed during the back reflection will affect the inside TIM surface (which is also the outside surface of the wall). ρ WALL = Reflectance of Wall surface Following is the FORTRAN Code used in the HeatBalanceSurfaceManager module. to determine the short wave radiation absorbed on outside of movable insulation and the short wave radiation absorbed on outside of opaque surface of the wall.SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -. The heat absorbed at the inside of the TIM/outside of the wall includes two components. Thus. This is characterized by the following equation: First pass solar absorbed by wall = (τ TIM i Incident Solar )i(α wall ) The amount of solar absorbed by the TIM and aggregated at the inside surface of the TIM (outside wall surface) is: Amount of back reflection absorbed by TIM = (τ TIM i Incident Solar )i(1 − α wall )iαTIM (104) The heat absorbed at the interface between the wall and the TIM includes both of these components. QSO is equal to: QSO = (τ TIM i Incident Solar )i{α wall + (1 − α wall )iα TIM } Substituting the definition for QSM into this equation and rearranging results in: (105) ⎧τ iQSM ⎫ QSO = ⎨ TIM ⎬i{α wall + (1 − α wall )iα TIM } ⎩ αTIM ⎭ ⎧⎛ α ⎪ QSO = (τ TIM iQSM )i⎨⎜ wall ⎪⎝ αTIM ⎩ Where. The first component is the amount of solar that is transmitted through the TIM and absorbed at the inside of the wall. 4/23/05 62 .

The winter design day was also modified in some runs to have a clearness of 1.90. ! Visible reflectance at normal incidence: back side 0. The Transparent Insulation material is conceived by applying a Movable Insulation on the exterior. The second “winter table” shows winter conditions with clearness=0 (the typical default for a winter design day) and clearness=1 (to illustrate solar radiation with other winter conditions).84. south. TIM with Summer Conditions Conductivity EXTWALL80 Construction 4/23/05 Thickness. The building was a very simple box with walls facing north. The main purpose of these runs was to verify that the transparent insulation model was predicting results that were reasonable using a simple test case. ! Thickness {m} 0.05.84. Test cases included no movable insulation. The program was run for this fictional 1 zone building located in Chanute AFB IL. the TIM had the following thermal properties: 0.0. ! Visible reflectance at normal incidence: front side 0. and west. Savings reported are heating and cooling loads for the design days only. ! IR transmittance at normal incidence 0. In the test cases.0 so that the effect that solar radiation during winter-time conditions could be studied.04. east. ! Visible transmittance at normal incidence 0.05.90.HMovInsul. The first “summer table” illustrates the execution of a summer design day. ! Solar reflectance at normal incidence: front side 0. Transparent Insulation Material has been applied to the south wall (except as noted in the table below).031.031.SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -. all of which are exterior walls. (21st June and 21st January).TARP AND FORTRAN ALGORITHM IF (Surface(SurfNum)%MaterialMovInsulExt. movable opaque insulation. and TIM on the exterior (south wall unless otherwise noted). 4” Common Brick and ¾” Plaster or Gypboard.AbsExt) IF (HMovInsul > 0) THEN ! Movable outside insulation in place QRadSWOutMvIns(SurfNum) = QRadSWOutAbs(SurfNum)*AbsExt & /Material(Construct(ConstrNum)%LayerPoint(1))%AbsorpSolar ! For Transparent Insulation QRadSWOutAbs(SurfNum) = Material(Surface(SurfNum)%MaterialMovInsulExt)%Trans & *QRadSWOutMvIns(SurfNum)* & ( (Material(Construct(ConstrNum)%LayerPoint(1))%AbsorpSolar/AbsExt) +(1-Material(Construct(ConstrNum)%LayerPoint(1))%AbsorpSolar) & ) Sample Test Run Cases: – Comparison A series of test cases were run in EnergyPlus to test the TIM model. The following two tables shows data for two series of runs.05. ! Solar reflectance at normal incidence: back side 0.GT.0) & CALL EvalOutsideMovableInsulation(SurfNum.RoughIndexMovInsul. ! Solar transmittance at normal incidence 0. for two design days. Sensible Cooling Energy Saved 63 . ! Conductivity {W/m-K} The Wall Construction is defined as an EXTWALL80 composed of 1” Stucco. The results showed that the TIM model was performing reasonably well and was producing results that were within expectations. ! IR emissivity: front side 0. Table 14. ! IR emissivity: back side 0.

57E+07 -2.040 0.31E+08 2.0. National Bureau of Standards (now National Institute of Standards and Technology). 4/23/05 64 .01E+07 -1.93E+08 6. TIM with Winter Conditions Conductivity EXTWALL80 Construction [W/m-K] 0.040 0.05E+07 2.025 0.040 [m] 0.040 0.00E+00 Sensible Heating Energy [J] 1.63E+08 5.70E+08 9.TARP AND FORTRAN ALGORITHM Normal case Without any Insulation With Dense Insulation Present With Dense Insulation Present With Dense Insulation Present With TIM Present With TIM Present With TIM Present With TIM Present -R value = ( 0.49E+08 3.10E+08 2.050 0.36E+08 2.040 0.040 0.22E+09 1.100 0.52E+08 -1.050 0.10E+08 5.050 0.025 0.43E+08 Walton.02E+08 4.SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -.05m.000 0.26E+09 1.05E+09 Energy Saved Winter Clearness=1 [J] 0.47E+09 1.000 0.Sensible ness.84E+08 6.12E+08 -2.30E+09 1.09E+08 3.10E+08 7.26E+08 -2.040 0.41E+08 3.00E+00 2.24E+09 1.26E+09 1.04W/m-K) With TIM Present (EAST WALL) With TIM Present (NORTH WALL) With TIM Present (WEST WALL) [W/m-K] 0.0.17E+08 3.100 0.050 1.040 0.035 Thick.040 0.044 0.30E+08 8.50E+08 9.14E+08 2.040 0.050 Energy [J] 3.50E+08 2.84E+07 3.035 0.60E+07 4.74E+08 1.025 0.00E+00 7..09E+08 6.22E+09 1.70E+07 7. G.19E+08 3.27E+08 Table 15. February 1983.000 0.66E+08 4.040 0.040 0.05m.10E+08 9.044 [J] 1.10E+08 Normal case Without any Insulation With Dense Insulation Present With Dense Insulation Present With Dense Insulation Present With TIM Present With TIM Present With TIM Present With TIM Present -R value = ( 0.050 0.37E+08 3.40E+07 7.26E+09 Energy Saved Winter Clearness=0 [J] 0.40E+08 0.050 0.53E+07 -9.100 0.70E+08 2.025 0.040 0.04W/m-K) With TIM Present (EAST WALL) With TIM Present (NORTH WALL) With TIM Present (WEST WALL) References 1.050 0.27E+08 4.100 0.050 0.89E+08 4.26E+08 -1.040 0.050 0.63E+08 4.040 0.N.30E+09 1.76E+08 2.07E+08 3.040 0.000 0.30E+08 7.24E+09 2.57E+08 4.040 0.26E+09 1. Heating Energy [m] 0. “The Thermal Analysis Research Program Reference Manual Program (TARP)”.64E+08 [J] 0.040 0.63E+08 5.73E+08 2.

http://www.15 sto.htm#bf1. Oltmannsstr. http://www.O. Brauchwasservorerwarmmung mit transparent gedammten Bauteilen ( Hybridsystem ). Jesch.aee.ise. Goetzberger. Arbeitsgemeinschaft Erneuerbare energie. and W.Lindauer. Braun.Platzer.bwk. E. Wood and L.fhg.de/Projects/Solbuild/materials.html Werner J.de/Projects/development99/art4.J.Stahl.html Two 0-Energy Houses.F. Thermotropic materials and Systems for Overheating Protection. Hens.or. Transparent Insulation of Building Facades.TARP AND FORTRAN ALGORITHM P. Volker Wittwer.Institut fur Bauphysik. 4/23/05 65 .Steps from Research to Commercial applications. D-79100 Freiburg. Oltmannsstrasse 22. D-79100 Freiburg.htm G.be/bwk/sr99/bwf. Transparent Insulation: an alternative solution for summer discomfort. Fraunhofer Institute for Solar Energy Systems. Oltmannsstr.htm Advanced Building Technologies – Transparent Insulation Materials ( TIM ). 5.ise. Transparent Insulation materials: a review..html Transparent Insulation. http://www. The use of Transparent Insulation Materials and Optical Switching Layers in Window Systems. H. A. Fraunhofer Institute for Solar Energy Systems. 5. Fraunhofer Institute for Solar Energy Systems.at/service/info-material/diverse/twd/index_e.SURFACE HEAT BALANCE MANAGER / PROCESSES BASIC HEAT BALANCE -.smartarch.Areas of Application.kuleuven.at/verz/english/tin. http://www. Transparent insulation technology: a technical note. http://www. Germany. http://www. Germany. Schmid. Ambient Press Limited 1993.nl/smartgrid/items/oo5_chur.com/advancedtech/transp. Façade Modules with back-ventilated Transparent Insulation. Germany.system StoTherm Solar. Verbeeck.enermodal.fhg. M. H. Society for Renewable Energy.Research and Development toward Series Application. Die neue Transparenz: Warmedamm-Verbund.esv. D-7800 Freiburg. http://www.Leonhardt. Transparent Insulation.ac.. Fraunhofer.html Robert Hausner.

Env + qconv . The modeling here assumes that the heat capacity in the outer baffle can be neglected since it is much lower than the underlying mass surface. This object is used with the Surface:HeatTransfer object where the Heat Transfer surfaces are referred to as the underlying surfaces. Baffle Heat Balance The baffle is assumed to be sufficiently thin and high-conductivity so that it can be modeled using a single temperature (for both sides and along its area).baff is determined by formulating a heat balance on a control volume that just encapsulates the collector surface. and the characteristics of the cavity and openings for natural ventilation.cav + qconv . From the thermal envelope’s point of view. This temperature Ts . The actual outer surface is referred to as the “baffle”. The heat balances are diagrammed in the following figure. This model uses the OtherSideConditionsModel to pass boundary conditions to the heat transfer modeling for the underlying surfaces.ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY Advanced Surface Concepts Exterior Naturally Vented Cavity The Exterior Naturally Vented Cavity is provided as a special case for the outside boundary conditions of heat transfer surfaces with a multi-skin exterior. the presence of a vented cavity on the outside of the surface modifies the conditions experienced by the underlying heat transfer surfaces.cav + qsource = 0 LWR where: (107) 4/23/05 66 . The constructions and materials for the Surface:HeatTransfer object should reflect the construction of just the underlying surface. Figure 29. Env + q′′ . The Surface:HeatTransfer:ExteriorNaturalVentedCavity object is used to describe the detached layer. or baffle. Baffle Surface Heat Balance The heat balance on the baffle surface’s control volume is: ′′ ′′ ′′ ′′ ′′ qα sol + qLWR . This exterior cavity acts as a radiation and convection baffle situated between the exterior environment and the outside face of the underlying heat transfer surface.

All terms are positive for net flux to the baffle. the sky. gnd Tamb + hr . The hco is treated in the same way as an outside face with ExteriorEnvironment conditions. External SW Radiation ′′ qα sol is calculated using procedures presented elsewhere in this manual and includes both direct and diffuse incident solar radiation absorbed by the surface face. In addition. is the convection coefficient. gnd + hr . when it is raining outside.cav + hc . the LWR ground. linearized coefficients.cav + qsource ) (h co + hr . The radiation heat flux is calculated from the surface absorptivity.cav ) (108) where. External Convection ′′ ′′ qconv . we assume the baffle gets wet and model the enhanced surface heat transfer using a large value for hco .cav is modeled using the classical formulation: qconv = hcp(Tair . Env is modeled using the classical formulation: qconv = hco(Tair . Each of these heat balance components is introduced briefly below. Env is a standard radiation exchange formulation between the surface.baff yields the following equation: Ts . surface temperature.cav = surface convection flux exchange with cavity air.cavTso + hc . air + hr . shading surfaces. surface facing angle and tilt. sky + hr . is the convection coefficient.ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY ′′ qα sol is absorbed direct and diffuse solar (short wavelength) radiation heat flux. and sky and ground view factors. Substituting models into (107) and solving for Ts .To) where hco. and ground temperatures. ′′ qconv . etc. ′′ qsource is a source/sink term that accounts for energy exported out of the control volume when the baffle is a hybrid device such as a photovoltaic panel.To) where hcp. Env = surface convection flux exchange with outside air. This is influenced by location.atmTamb + hr . External LW Radiation q′′ . surface face material properties. Cavity Convection Radiation is modeled using ′′ ′′ qconv . sky. 4/23/05 67 . q′′ .baff = (I α + h s co amb T ′′ + hr .cavTa . Cavity LW Radation q′′ . Env is net long wavelength (thermal) radiation flux exchange with the air and LWR surroundings. weather conditions. air. and the atmosphere. Radiation is modeled using linearized coefficients.cav is net long wavelength (thermal) radiation flux exchange with the outside face of the LWR underlying surface(s). q′′ .skyTsky + hr . ′′ qconv .cav is a standard radiation exchange formulation between the baffle surface and the LWR underlying heat transfer surface located across the cavity. The value for hcp is obtained from correlations used for window gaps from ISO (2003) standard 15099.

hco is the convection coefficient for the outdoor environment [W/m2·K]. (109) 4/23/05 68 . The uniform temperature of the cavity air. gnd is the linearized radiation coefficient for the ground [W/m2·K].cav is the convection coefficient for the surfaces facing the plenum [W/m2·K]. the heat balance on the plenum air control volume is: Qvent + Qco + Qc . hr .cav is the drybulb temperature for air in the cavity [ºC]. Figure 30.atm is the linearized radiation coefficient for the surrounding atmosphere [W/m2·K]. Plenum Heat Balance When the UTSC is passive. and Ta . The cavity air is modeled as well-mixed.cav .baff = 0 where. Ta . α is the solar absorptivity of the collector [dimensionless]. hr .ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY I s is the incident solar radiation of all types [W/m2].cav is the linearized radiation coefficient for the underlying surface [W/m2·K]. Tsky is the effective sky temperature [ºC]. Tso is the temperature of the outside face of the underlying heat transfer surface [ºC]. hr . also assumed for ground surface [ºC]. Cavity Heat Balance The cavity is the volume of air located between the baffle and the underlying heat transfer surface. hr . Tamb is the outdoor drybulb from the weather data. is determined by formulating a heat balance on a control volume of air as diagrammed below. sky is the linearized radiation coefficient for the sky [W/m2·K]. hc .

mvent = ρ Vtot where.baff (h c . This value is available for user input. we elect to use correlations for natural ventilation presented as equations (29) and (30) in Chapter 26. of ASHRAE HOF (2001).ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY Qvent is the net rate of energy added from natural ventilation – where outdoor ambient air exchanges with the cavity air. so we have: 4/23/05 69 . This value is available for user input.cav ) / Tamb (if Tamb > Ta .cav > Tamb ) Vthermal = CD Ain 2 g ∆H NPL (Tamb − Ta .baff is the net rate of energy added by surface convection heat transfer with the collector. Reasoning that the configuration is similar to single-side natural ventilation.25 to 0. Qc . and Vtot = Vwind + Vthermal is the total volumetric flow rate of air ventilating in and out of the cavity. Simplistic engineering models are used to model mvent resulting from natural buoyancy and wind forces. cav A + mvent c p + hc . Mass continuity arguments lead to modeling the area of the openings as one half of the total area of the openings. Vwind = Cv AinU ∞ Vthermal = CD Ain 2 g ∆H NPL (Ta . Qco is the net rate of energy added by surface convection heat transfer with the underlying surface. (h c .cav − Tamb ) / Ta .cav A ) ) mvent is the air mass flow from natural forces [kg/s] Modeling natural ventilation air exchanges in a general way is challenging.cav and baffle is vertical) Cv is the effectiveness of the openings that depends on opening geometry and the orientation with respect to the wind. ρ is the density of air [kg/m3]. CD is the discharge coefficient for the opening and depends on opening geometry. cav ATso + mvent c pTamb + hc .cav ATs . And substituting into (109) yields the following equation: Ta .cav = where.6. ASHRAE HoF (2001) indicates values ranging from 0.cav (if Ta .

14 270 0. zmet is the height of the standard metereological wind speed measurement.22 370 0.cav . Local Wind Speed Calculations The outdoor wind speed affects terms used in modeling. If the cavity is horizontal and Tamb > Ta .81 [m/s2]. and Ta . Terrain 1 2 3 4/23/05 Description Flat.cav . The exterior baffle and cavity system is coupled to the underlying surface using the OtherSideConditionsModel mechanism. a δ (m) 0. These native EnergyPlus Heat Balance routines are used to calculate Tso . The wind speed is modified from the measured meteorological wind speed by the equation (ASHRAE 2001): ⎛δ ⎞ U ∞ = Vmet ⎜ met ⎟ ⎝ zmet ⎠ amet ⎛z⎞ ⎜ ⎟ ⎝δ ⎠ a (110) where z is the height of the centroid of the system. Layer Thickness. wooded country Towns and cities Exponent. and a and δ are terrain-dependent coefficients. as well as radiation reflected from the ground and adjacent surfaces. Terrain-Dependent Coefficients (ASHRAE 2001). Underlying Heat Transfer Surface The exterior baffle and cavity are applied to the outside of a heat transfer surface.33 460 70 . hc . To adjust for different terrain at the building site and differences in the height of building surfaces.baff . This surface is modeled using the usual EnergyPlus methods for handling heat capacity and transients – typically the CTF method. Solar radiation incident on the surface includes beam and diffuse radiation. The UTSC model provides values for hr . This is value is available for user input. δ is the boundary layer thickness for the given terrain type. The values of a and δ are shown in the following tables: Table 16. Solar and Shading Calculations The exterior vented cavity model uses standard EnergyPlus surfaces in order to take advantage of the detailed solar and shading calculations.cav then Vthermal = 0 because this is a stable situation.cav for use with the Heat Balance Model calculations for the outside face of the underlying surface (described elsewhere in this manual). is also taken into account. the local wind speed is calculated for each surface. open country Rough. ∆H NPL is the height from midpoint of lower opening to the Neutral Pressure Level.ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY Ain = A 2 g is the gravitational constant taken as 9. Shading of the collector by other surfaces. such as nearby buildings or trees. The wind speed in the weather file is assumed to be measured at a meteorological station located in an open field at a height of 10 m. Ts .

. These coefficients are defined in the classic way by: hc = Tair − Tsurf ′′ qconv First. Whereas radiation calculations usually use temperature raised to the fourth power.22 370 The exterior vented cavity can be defined such that it has multiple underlying heat transfer surfaces. this greatly complicates solving heat balance equations for a single temperature. is used to model thermal radiation between the collector surface and the outside face of the underlying heat transfer surface. Convection Coefficients Exterior cavity modeling requires calculating up to three different coefficients for surface convection heat transfer. Radiation Coefficients Exterior vented cavity modeling requires calculating up to four different linearized coefficients for radiation heat transfer. 4/23/05 71 . Elsherbiny et. We assume a view factor of unity. It is calculated using: hr . The convection coefficient is modeld in the same way used in EnergyPlus to model air gaps in windows. Winkelmann).cav is the convection coefficient for baffle surfaces facing the cavity. Linearized radiation coefficients have the same units and are used in the same manner as surface convection coefficients and introduce very little error for the temperature levels involved.baff s . Wright. The formulations are documented in ISO (2003) standard 15099.f90 (by F. ebaff is the longwave thermal emmittance of the baffle. This coefficient is applied to both the baffle and the underlying surface.cav . Second.cav = σ SB ebaff eso (T (T 4 s .10 210 5 Urban. It is modeled in exactly the same way as elsewhere in EnergyPlus and will depend on the user setting for Outside Convection Algorithm – Outside Surface Heat Balance entry elsewhere in this document. These correlations vary by Rayleigh number and surface tilt and are based on the work of various research including Hollands et. industrial. hco is the convection coefficient for the baffle surface facing the outdoors. hc . The radiation coefficient. all temperatures are converted to Kelvin.ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY 4 Ocean 0. σ SB is the Stefan-Boltzmann constant. forest 0. al. which itself was derived from Window5 subroutine “nusselt”.baff 4 − Tso − Tso ) ) (111) where. The routines were adapted from Subroutine NusseltNumber in WindowManager. hr .. and Arnold. The centroid heights for each surface are area-weighted to determine the average height for use in the local wind calculation. al. and eso is the longwave thermal emmittance of the underlying heat transfer surface.

2001 ASHRAE Fundamentals Handbook. Altanta GA. gnd are used elsewhere in EnergyPlus for the outside face surface heat balance and are calculated in the same manner as equation (111). [This is accomplished by calling subroutine InitExteriorConvectionCoeffs in the file HeatBalanceConvectionCoeffs. American Society of Heating Refrigeration and Air-Conditioning Engineers. 2003. and shading devices – Detailed calculations. hr . Thermal performance of windows. ISO.f90. sky . hr . doors.atm . 4/23/05 72 . International Organization for Standardization. and hr . ISO 15099:2003.ADVANCED SURFACE CONCEPTS EXTERIOR NATURALLY VENTED CAVITY The three other coefficients. ] References ASHRAE HOF 2001.

AnisoSkyMult is determined by surface orientation and sky radiance distribution. clearness factor and brightness factor. which are determined from sun position and solar quantities from the weather file. Schematic view of sky showing solar radiance distribution as a superposition of three components: dome with isotropic radiance. In actuality. 4/23/05 73 . the horizon brightening is highest at the horizon and decreases in intensity away from the horizon. and accounts for the effects of shading of sky diffuse radiation by shadowing surfaces such as overhangs. (2) A circumsolar brightening centered at the position of the sun. circumsolar brightening represented as a point source at the sun. the diffuse sky irradiance on a surface is given by AnisoSkyMult(SurfNum) * DifSolarRad where DifSolarRad is the diffuse solar irradiance from the sky on the ground and SurfNum is the number of the surface. for clear skies. 1990. Isotropic dome Circumsolar brightening (concentrated at center of sun) Horizon brightening (concentrated at horizon) Figure 31. as described in Perez et al. defined below. The sky radiance distribution is based on an empirical model based on radiance measurements of real skies.SKY AND SOLAR/SHADING CALCULATIONS SKY RADIANCE MODEL Sky and Solar/Shading Calculations Sky Radiance Model In EnergyPlus the calculation of diffuse solar radiation from the sky incident on an exterior surface takes into account the anisotropic radiance distribution of the sky. The horizon brightening is assumed to be a linear source at the horizon and to be independent of azimuth. It does not account for reflection of sky diffuse radiation from shadowing surfaces. which is characterized by two quantities. For overcast skies the horizon brightening has a negative value since for such skies the sky radiance increases rather than decreases away from the horizon. The circumsolar brightening is assumed to be concentrated at a point source at the center of the sun although this region actually begins at the periphery of the solar disk and falls off in intensity with increasing angular distance from the periphery.. and horizon brightening represented as a line source at the horizon. For this distribution. (3) A horizon brightening. In this model the radiance of the sky is determined by three distributions that are superimposed (see Figure 31) (1) An isotropic distribution that covers the entire sky dome. The proportions of these distributions depend on the sky condition.

b F1.SKY AND SOLAR/SHADING CALCULATIONS SKY RADIANCE MODEL Table 17.041 for Z in radians Brightening coefficient factors Shadowing factor for circumsolar radiation Shadowing factor for sky dome radiation Shadowing factor for horizon radiation Sky radiance Azimuth angle of point in sky Altitude angle of point in sky Irradiance on surface from a horizon element Irradiance on surface from a sky dome element Sunlit fraction Units W/m2 W/m2 W/m2 W/m2 W/m2 radians radians radians W/m 2 FORTRAN variable Surface(SurfNum)%Tilt*DegTo Radians F1. Variables in Anisotropic Sky Model and Shadowing of Sky Diffuse Radiation Mathematical variable Isky Ihorizon Idome Icircumsolar Ih S a. F12R. etc. F2 IncAng ZenithAng Delta Epsilon AirMass Material%Thickness F11R. F2 α Z ∆ ε m IO I κ Fij Rcircumsolar Rdome Rhorizon E θ φ Ii Iij SF 4/23/05 Description Solar irradiance on surface from sky Solar irradiance on surface from sky horizon Solar irradiance on surface from sky dome Solar irradiance on surface from circumsolar region Horizontal solar irradiance Surface tilt intermediate variables Circumsolar and horizon brightening coefficients Incidence angle of sun on surface Solar zenith angle Sky brightness factor Sky clearness factor relative optical air mass Extraterrestrial solar irradiance Direct normal solar irradiance constant = 1. SunLitFrac DifShdgRatioIsoSky DifShdgRatioHoriz Theta Phi FracIlluminated 74 W/m2 radians-3 W/m2 radians radians W/m 2 W/m2 - .

and the sky clearness factor is ε= where 4/23/05 (Ih + I ) / Ih + κ Z 3 1+ κ Z 3 75 . In the above equations: Ih = horizontal solar irradiance (W/m2) S = surface tilt (radians) a = max(0. cosZ) F1 = circumsolar brightening coefficient F2 = horizon brightening coefficient where α = incidence angle of sun on the surface (radians) Z = solar zenith angle (radians).SKY AND SOLAR/SHADING CALCULATIONS SKY RADIANCE MODEL I’ Sky solar irradiance on surface with shadowing W/m2 - Sky Diffuse Solar Radiation on a Tilted Surface The following calculations are done in subroutine AnisoSkyViewFactors in the SolarShading module. In the absence of shadowing. Isky.087. The brightening coefficients are a function of sky conditions. they are given by F1 = F11 (ε ) + F12 (ε )∆ + F13 (ε ) Z F2 = F21 (ε ) + F22 (ε )∆ + F23 (ε ) Z Here the sky brightness factor is ∆ = Ihm / Io where m = relative optical air mass Io = extraterrestrial irradiance (taken to have an average annual value of 1353 W/m2). the sky formulation described above gives the following expression for sky diffuse irradiance. on a tilted surface: I sky = I horizon + I dome + I circumsolar where I horizon I dome = irradiance on surface from sky horizon = irradiance on surface from sky dome = I h F2 sin S = I h (1 − F1 )(1 + cos S ) / 2 I circumsolar = irradiance on surface from circumsolar region = I h F1a / b AnisoSkyMult is then Isky /DifSolarRad.cosα) b = max(0.

SFi is calculated using the beam solar shadowing method as though the sun were located at the ith horizon point.1326077 -1.1299457 0.1874525 -0.0596012 0.2877813 -0. These values have higher precision than those listed in Table 6 of Perez et al.0012448 2. Fij Factors as a Function of Sky Clearness Range.0620636 -0.065-1.065 -0.2367284 -0.8230357 0.5682053 0.3765031 0.1519229 -0.950-2.950 0. The sky is assumed to be a superposition of the three Perez sky components described above.0659650 -0.1272340 0.500-6.0601591 -1.800-4.500-1.3616149 0.230 0.800 0.1513752 -0. ε Range F11 F12 F13 F21 F22 F23 1.1561313 -1. … .4868735 -0.6777470 -0.4620442 0.0083117 0.3296958 0. Perez.8730280 -0.2506212 Shadowing of Sky Diffuse Solar Radiation Sky diffuse solar shadowing on an exterior surface is calculated as follows in subroutine SkyDifSolarShading in the SolarShading module.230-1.200 1. 345O αi = incidence angle on surface of radiation from θi The corresponding ratio for the isotropic sky dome is given by Rdome Irradiance from dome with obstructions = = Irradiance from dome without obstructions ∑∑ I i =1 j =1 24 6 24 6 ij SFij ij ∑∑ I i =1 j =1 The Fij values in this table were provided by R.500 0. For the horizon source the following ratio is calculated by dividing the horizon line into 24 intervals of equal length: Rhoriz Irradiance from horizon with obstructions = = Irradiance from horizon without obstructions ∑ I SF i =1 24 i 24 i ∑I i =1 i where Ii is the unobstructed irradiance on the surface from the ith interval.3589221 0. and the sums are over intervals whose center lies in front of the surface.2255647 -0.4118494 0.5999137 -0.000-1.0639588 -0.5877285 -0. 5/21/99.3272588 -0. 1990.6825954 -0.0260542 1.1 Table 18.0554140 -0.0558651 4.0189325 0.2642124 -1.1310694 > 6.0721249 -0.2951290 0. Here I i = E (θi )dθ cos α i where E (θi) = radiance of horizon band (independent of θ) dθ = 2π/24 = azimuthal extent of horizon interval (radians) θi = 0O.0139754 1.0220216 1. private communication.2210958 0. 15O.041 for Z in radians The factors Fij are shown in the following table.3920403 -0.500 1.2504286 0. SFi is the sunlit fraction from radiation coming from the ith interval. 4/23/05 76 1 .SKY AND SOLAR/SHADING CALCULATIONS SKY RADIANCE MODEL I = direct normal solar irradiance κ = 1.0288748 1.1088631 -0.200 0..

If obstructions such as overhangs are present the sky long-wave incident on a surface is multiplied by the isotropic shading factor.φj) = sky radiance (independent of θ and φ for isotropic dome) dθ = 2π/24 = azimuthal extent of sky element (radians) dφ = (π/2)/6 = altitude extent of sky element (radians) θi = 0O. The total sky diffuse irradiance on the surface with shadowing is then ′ I sky = Rhorizon I horizon + Rdome I dome + Rcircumsolar I circumsolar Rhorizon and Rdome are calculated once for each surface since they are independent of sun position. described above. The sun is to the right so that walls 1 and 3 and windows a and c are completely shaded.j) is a grid of 144 points (6 in altitude by 24 in azimuth) covering the sky dome. the figure below shows a flat roofed.9. Rdome. 22. the circumsolar ratio is Rcircumsolar = Irradiance from circumsolar region with obstructions = SFsun Irradiance from circumsolar without obstructions where SFsun is the beam sunlit fraction. With shadowing we then have: AnisoSkyMult = I’sky /DifSolarRad. Wall 2 and window b are partially shaded. … . Iij is th the unobstructed irradiance on the surface from the sky element at the ij point. 15O. it is necessary to know how much of each part of the building is shaded and how much is in direct sunlight. … . The sunlit area of each surface changes as the position of the sun changes during the day. Shadowing of Sky Long-Wave Radiation EnergyPlus calculates the sky long-wave radiation incident on exterior surfaces assuming that the sky long-wave radiance distribution is isotropic. The long-wave radiation from these obstructions is added to the long-wave radiation from the ground. The purpose of the EnergyPlus shadow algorithm is to compute such sunlit 4/23/05 77 . L-shaped structure with a window in each of the visible sides.φj) Because the circumsolar region is assumed to be concentrated at the solar disk. and the sum is over points lying in front of the surface.5O. Here I ij = E (θi .5O. and wall 4 and window d are completely sunlit. SFij is the th sunlit fraction for radiation coming from the ij element. 345O φj = 7.5O αij = incidence angle on surface of radiation from (θi. Shading Module Shading and Sunlit Area Calculations When assessing heat gains in buildings due to solar radiation. in this calculation both obstructions and ground are assumed to be at the outside air temperature and to have an emissivity of 0.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE where (i. As an example. φ j ) cos φ j dθ dφ cos α ij where E (θi. 82.

The following procedure is used to determine the direction cosines. (Note that 1 4/23/05 78 . the hour angle is calculated from: HourAngle = 15i(12 − (TimeValue + EquationOfTime ) ) + (TimeZoneMeridian − Longitude ) ( ) TimeZoneMeridian is the standard meridian for the location’s time zone {GMT +/-}.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE areas. are based on Astronomical Algorithms. The shadow algorithm is based on coordinate transformation methods similar to Groth and Lokmanhekim and the shadow overlap method of Walton. Figure 32. Solar declination is a function of local/site latitude. It is common astronomical practice to express the hour angle in hours. ε. and the equation of time. Meeus. You can convert the hour angle displayed from EnergyPlus to time by dividing by 15. The values of the solar declination angle. minutes and seconds of time rather than in degrees. The fractional year is calculated. in radians: γ= 2π (day _ of _ year ) 366 From this fractional year. For each time step (time value = fractional hour). negative after noon. Solar HourAngle (H) gives the apparent solar time for the current time period (degrees). HourAngle is positive before noon. Overall Shadowing Scheme Depiction Solar Position Current solar position is described in terms of three direction cosines that are convenient for determining the angle of incidence of the sun’s rays on a building surface. the equation of time and solar declination angle are calculated. δ. Predecessors to the EnergyPlus shadowing concepts include the BLAST and TARP shadowing algorithms.

the Yaxis pointing north. This Right-hand coordinate system has the X-axis pointing east. The azimuth angle (ψ) of a surface is the angle from the north axis to the projection onto the X-Y plane of a normal to the surface (clockwise positive).819719 hours 4 minutes 4 seconds 0. Surfaces are described by the coordinates of their vertices in a three dimensional Cartesian coordinate system.066667 seconds The Solar Altitude Angle (β) is the angle of the sun above the horizontal (degrees). Relationship of Angles (degrees) to Time Unit of Angle 1 radian 1 degree 1 arcmin 1 arcsec Equivalent time 3. This is shown more clearly in the following figure. Solar Position Illustration Surface Geometry Shadow calculations first require that the building surfaces be described geometrically.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE hour is equivalent to 15 degrees. 360° of the Earth’s rotation takes place every 24 hours. 4/23/05 79 . The vertices are recorded in counter-clockwise sequence (as the surface is viewed from outside its zone).) The relationship of angles in degrees to time is shown in the following table: Table 19. and the Z-axis pointing up (see figure below). The surface tilt angle (φ) is the angle between the Z-axis and the normal to the surface. The Solar Azimuth Angle (φ) is measured from the North (clockwise) and is expressed in degrees. Figure 33.

Of pertinent interest here is that the user may specify the vertices in either “relative” or “world” coordinates. when vertices are reported. The global coordinates of the surface origins are calculated by: X so = X zo + X zr icosψ z − Yzr isinψ z Yso = Yzo + X zr isinψ z − Yzr icosψ z X so = X zo + X zr icosψ z − Yzr isinψ z 4/23/05 (115) (116) (117) 80 . they are reported in world coordinates. starting at the upper-left-corner (4-sided surface) and are listed counter-clockwise. Relative Coordinate Transformation When vertices are specified in “relative” coordinates. there can be a “building” north axis as well as a “zone” north axis. The global coordinates of zone origins are related to the building relative coordinates by: X zo = X br icosψ b − Ybr isinψ b Yzo = Ybr isinψ b − Ybr icosψ b Z zo = Z br Where zo – represents Zone Origin br – represents the Zone Origin as input (relative to building origin) (112) (113) (114) The zone may also be rotated ψz degrees relative to the building coordinates. Origins of zone surfaces are then given relative to the zone coordinate system. The building north axis/coordinate system is a rotation of ψb degrees from the global/world coordinate system. EnergyPlus Coordinate System The SurfaceGeometry object specifies to EnergyPlus how the surface vertices will be presented in the input file. Regardless of input specifications.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE Figure 34.

2 and 3 are in shade. 3. and 5 are in sunlight. casts a shadow. which is composed of a set of enclosing plane polygons. 4 and 5. The sun is positioned behind and to the right of the structure and a shadow is cast onto the horizontal surface (the ground). It is possible to think of the structure's shadow as the combination of shadows cast by surfaces 1. The global coordinates of the surface vertices are given by: X = X so + X sr icosψ − Ysr isinψ icos φ Y = Yso + X sr isinψ − Ysr icosψ icos φ Z = Z so + Ysr isin φ World Coordinates Relative Coordinates (119) (120) (121) Vertices in the global coordinate system can be transformed to the coordinate system relative to a given surface by X ' = X − X so Y ' = Y − Yso Z ' = Z − Z so X sr = − X ' icosψ + Y ' isinψ Ysr = − X ' isinψ icos φ + Y ' icosψ icos φ + Z ' isin φ Z sr = − X ' isinψ isin φ + Y ' icosψ isin φ + Z ' icos φ Shadow Projection (122) (123) (124) (125) (126) (127) All architectural forms are represented by plane polygons. 2. This last combination of shadow casting surfaces is the simplest.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE A surface azimuth angle relative to the zone coordinate system (ψs) is converted to a global azimuth by: ψ = ψ s +ψ z +ψ b (118) The surface tilt angle (φ) is not changed by these rotations about the Z-axis. In the EnergyPlus shadow algorithm every surface is considered to be one of the 4/23/05 81 . The X-axis of the surface coordinate system is a horizontal line through the second vertex. The coordinates of the surface vertices are given in a coordinate system in the plane of the surface relative to the second vertex as shown for surfaces in Figure 34. or by surfaces 2 and 3. The structure consists of a top (surface 1) and four vertical surfaces (2 and 3 visible to the observer and 4 and 5 not visible). Surfaces 1. This can give good accuracy even for curved surfaces: a sphere can be approximated by the 20 nodes of an icosohedron with only 3 percent error in the shadow area cast by the sphere. Consider how a solid object. 4. 4 and 5 or by 1. Figure 35 shows a box shaped structure on a horizontal surface.

cos θ = CS1 iCW1 + CS2 iCW2 + CS3 iCW3 If cos θ is less than zero. and only those surfaces that are not sunlit at a given hour are considered shadowing surfaces. 4/23/05 82 . the sun is behind the surface. (131) A shadow is projected from the vertices of the shadowing polygon (SP) along the direction of the sun's rays to the plane of the shadow receiving polygon (RP). Figure 35. The "submerged" portion of the SP must be clipped off before projection. Basic shadowing concept structure The expressions in equation (127) are the direction cosines of the surface: CW1 = sinψ icos φ CW2 = cosψ isin φ CW3 = cos φ (128) (129) (130) The cosine of the angle of incidence of the sun's rays on the surface are given by the dot product of surface and sun direction cosines.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE surfaces that enclose a solid. a false shadow is cast as in Figure 36. If any vertices of the SP are below the plane of the RP (z < 0).

which intersect the plane of the RP. y1 . w1) to point (x2. These points become new vertices of the SP. Y) is represented by a three element vector (x. A line is also represented by a three element vector (a. c) = ( x1 . through linear interpolation. and w is any real number except zero. the points on the perimeter of the SP. y1. y. y. c). c) from point (x1. b. w2) is given by: (a. y2 . which together with the other positive vertices define a clipped SP that casts only a real shadow.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE Figure 36. z) relative to the RP coordinate system casts a shadow to a point in the plane of the RP given by x' = x − y' = y − where z ia cos θ z ib cos θ (132) (133) a = sinψ iCS1 − cosψ iCS2 and b = − cosψ icos φ iCS1 − sinψ icos φ iCS2 + sin φ iCS3 Homogeneous Coordinates Two-dimensional homogeneous coordinate techniques are used to determine the vertices of shadow overlaps. z2 ) 4/23/05 (134) 83 . Illustration of Shadow Clipping This is done by finding. w) where x = w*X. In homogeneous coordinates. y = w*Y. A point (X. b. The directed line (a. points and lines are represented by a single form that allows simple vector operations between those forms [Newman-Sproul]. y2. z1 ) ⊗ ( x2 . b. A vertex located at (x.

y. or on a line. b2 . 1). The vertices that define the overlap between two convex polygons. Then if (135) (a. Any non-convex polygon can be constructed as a sum of convex ones. for determining intercepts. b. c2 ) (137) Note that the use of homogeneous coordinates as outlined above provides a consistent method and notation for defining points and lines.1) > 0 (136) the point is to the left of the line. and for determining whether a point lies to the left. The overlap of A and B is the polygon a-b-c-d. w) can be transformed to the Cartesian point with coordinates (x/w. b1. Thus. w) lie on a line (a. if (X. w) of line (a1. its homogeneous coordinates are (X. y. the point is to the right of the line. point c is the result of rule 2. the basic job of the shadow algorithm is to determine the area of the overlap between the polygons representing the shadows and the polygon representing the receiving surface. y / w. Y. consist of: • • • the vertices of A enclosed by B the vertices of B enclosed by A and the intercepts of the sides of A with the sides of B In Figure 37. Overlapping Shadows After transforming the shadows onto the plane of the receiving surface. c) • ( x / w. Coordinate and projection transformations of a convex polygon produce another convex polygon. Similarly. w) = (a1 . y. c1) and line (a2. The intercept (x. Figure 40 shows a more complex overlap.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE The sequence in the cross product is a convention to determine sign. The condition that a point (x. Figure 39 shows an overlap defined only by the intercepts of A and B. Y) is a Cartesian coordinate pair. The overlap between two convex polygons is another convex polygon. If it is less than zero. y. the homogeneous coordinates (x. Normalization provides the means for transforming to and from homogeneous notation and Cartesian coordinates. A and B. w) = 0 The point is normalized by dividing by w. c1 ) ⊗ (a2 . point a is the result of rule 1. b. y/w). c2) is given by: ( x. c) • ( x. b. y. to the right. There is considerable simplification if only convex (no interior angle > 180 ) polygons are considered. b2. c) is that (a. b1 . and points b and d result from rule 3. Figure 38 shows an overlap where all of the vertices of B are enclosed by A. 4/23/05 84 .

The intercept of two sides may not lie beyond the ends of either side. These are "line segments" rather than "lines". Figure 38. Surface A totally overlaps Surface B. A point is enclosed by a clockwise. The sequence of vertices of the receiving polygons should be reversed so it and all shadow polygons will have the same sequence. It is possible to tell if line segments A and B intercept within their end points by noting that the ends of A must lie on both sides of B. 4/23/05 85 . and the ends of B must lie on both sides of A.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE Figure 37. while a projection reverses the order. Point a – vertex of A enclosed by B Coordinate transformation retains the order of the vertices of a polygon. This should be done before the intercept is calculated. convex polygon if the point lies to the right of all sides (or does not lie to the left of any side) of the polygon.

Given a closed. they may also overlap each other as in Figure 41. Y2) …. y1) The area is positive if the vertices are counter-clockwise and negative if they are clockwise. its area is given: Area = 1 2 ∑ (x y i =1 i n i +1 − xi +1 yi ) (138) where (xn+1. y1). they must be sorted into clockwiseorder for the area to be computed. 4/23/05 86 . Figure formed from Intercept Overlaps between A and B Once the vertices are determined. The total sunlit area can be expressed as the sum of all polygon areas given a proper sign on each of the areas. Complex overlapping condition If two shadows overlap the receiving surface. Figure 40. Yn).SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE Figure 39. (xn. The areas of all overlaps can be computed. (x2. The vertices of this overlap can be computed. planar polygon of n sequential vertices (x1.yn+1)= (x1.

The shadow overlapping is performed and the sunlit portion of the gross wall area is computed.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE The following convention was adopted: Table 20.τi) for all overlap polygons. the wall is considered a simple rectangle and the window on it is ignored. Solar Gains The total solar gain on any exterior surface is a combination of the absorption of direct and diffuse solar radiation given by 4/23/05 87 . Surface / Area Characteristic / Convention Surface Characteristic receiving surface overlap between shadow and receiving overlap between two shadows and so on through multiple overlaps where signs of the overlapping areas. Then the τ of every overlap of polygons i and j is the product of τi and τj The shaded area is then computed by summing Ai•(1 . The sunlit area of the window is subtracted from the gross wall sunlit area to determine the net wall sunlit area. It is easy to determine the sunlit area of a window once all the shadow and overlap vertices on the wall have been computed. Then the window rectangle is overlapped with the shadow to determine its sunlit area. Let τ of the receiving polygon be one. First. Area Convention positive (A) negative (B & C) positive (D) the sign of the overlap area is the product of the Figure 41. because they were precisely determined on the wall. During this calculation it is not necessary to recompute the shadows. Consider wall 2 of Figure 32. Multiple Shadow Overlaps Partially transparent shadowing surfaces can also be modeled by giving a transparency (τ) to every shadowing polygon.

shadowing by window and door reveals is also calculated. If no floor is present in the zone. the incident beam solar radiation is absorbed on all interior surfaces according to their absorptances. The zone heat balance is then applied at each surface and on the zone's air with the absorbed radiation being treated as a flux on the surface. which is assumed to be uniformly distributed on all interior surfaces. Beam solar radiation entering the zone is treated as for MinimalShadowing. wings. the solar distribution field value determines how EnergyPlus will treat beam solar radiation entering a zone through exterior windows. 4/23/05 88 . FullExterior In this case. where it is absorbed according to the floor's solar absorptance. there is no exterior shadowing except from window and door reveals. shadow patterns on exterior surfaces caused by detached shading. As for MinimalShadowing. There are three choices: MinimalShadowing. All beam solar radiation entering the zone is assumed to fall on the floor. Solar Distribution As discussed in the Input Output Reference (Object:Building).SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE S ⎛ ⎞ Qso = α i⎜ I b icos θ i s + I s i Fss + I g i Fsg ⎟ S ⎝ ⎠ where a =solar absorptance of the surface A =angle of incidence of the sun's rays S =area of the surface Ss = sunlit area Ib =intensity of beam (direct) radiation Is =intensity of sky diffuse radiation Ig =intensity of ground reflected diffuse radiation Fss = angle factor between the surface and the sky Fsg = angle factor between the surface and the ground For the surface of a building located on a featureless plain (139) Fss = and 1 + cos φ 2 (140) Fsg = 1 − cos φ 2 (141) If the surface is shaded the program modifies Fss by a correction factor that takes into account the radiance distribution of the sky (see “Shadowing of Sky Diffuse Solar Radiation”). It is up to the user to estimate the effect of this shading and modify the input value of Fsg accordingly. MinimalShadowing In this case. Any reflected by the floor is added to the transmitted diffuse radiation. overhangs. and exterior surfaces of all zones are computed. Shading of ground diffuse solar radiation is not calculated by the program. FullExterior and FullInteriorAndExterior.

This can be determined by viewing the eplusout. that daylighting through interior windows is not calculated. The most common non-convex zone is an L-shaped zone. diffuse solar radiation. Figure 42. walls and windows. In this calculation the effect of an interior or exterior shading device. is accounted for. If this option is used. you should be sure that the surfaces of the zone totally enclose a space. taking into account the effect of exterior shadowing surfaces and window shading devices. The program determines the amount of this radiation that is (1) 4/23/05 89 . and how much is transmitted to the outside. Examples of convex and non-convex zones are shown in Figure 42. Note. If you use FullInteriorAndExterior the program will calculate how much beam radiation falling on an interior window is absorbed by the window. by projecting the sun's rays through the exterior windows.dxf file with a program like AutoDesk’s Volo View Express. You should also be sure that the zone is convex. however. and how much is transmitted into the adjacent zone. Illustration of Convex and Non-convex Zones Details of the Interior Solar Distribution Calculation EnergyPlus calculates the distribution of short-wave radiation in the interior of each thermal zone. This radiation consists of beam solar radiation. If you use FullInteriorAndExterior the program will also calculate how much beam radiation falling on the inside of an exterior window (from other windows in the zone) is absorbed by the window. how much is reflected back into the zone. how much is reflected back into the zone. and short-wave radiation from electric lights. FullInteriorAndExterior This is the same as FullExterior except that instead of assuming all transmitted beam solar falls on the floor the program calculates the amount of beam radiation falling on each surface in the zone. (A formal definition of convex is that any straight line passing through the zone intercepts at most two surfaces. including floor. FulInteriorAndlExteriorWith Reflections This case is the same interior distribution as the preceding option but uses exterior reflections as well (see Solar Radiation Reflected from Exterior Surfaces for further explanation). if present.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE FullExteriorWith Reflections This case is the same interior distribution as the preceding option but uses exterior reflections as well (see Solar Radiation Reflected from Exterior Surfaces for further explanation).) If the zone’s surfaces do not enclose a space or if the zone is not convex you should use Solar Distribution = FullExterior instead of FullInteriorAndExterior. (Interior windows are assumed to have no shading device).

Interior Solar Radiation Absorbed by Opaque Surfaces The short-wave radiation absorbed on the inside face of an opaque surface (floor. Since Qsw is ultimately absorbed or transmitted by zone heat transfer surfaces. (2) absorbed in the glass and shading device layers of the zone’s exterior and interior windows. (3) transmitted through the zone’s interior windows to adjacent zones. wall or ceiling) is given by QRadSWInAbs( SurfNum) = QS ( ZoneNum) * AbsIntSurf ( SurfNum) + AISurf ( SurfNum) * BeamSolarRad [W/m 2 ] where (142) SurfNum = surface number ZoneNum = number of zone that surface belongs to QS(ZoneNum) = short-wave diffuse irradiance in the zone [W/m2] AbsIntSurf(SurfNum) = inside solar absorptance of the surface AISurf(SurfNum) = inside beam solar irradiance factor for the surface [-] BeamSolarRad = outside beam normal solar irradiance [W/m2] Interior Diffuse Radiation QS is assumed to be uniformly distributed throughout the zone. for a window. and (4) transmitted back out of the exterior windows. 90 4/23/05 . = back diffuse transmittance plus back diffuse system absorptance of glass layers and shading device layers (if present) Solving this equation for QS gives: QS ( ZoneNum) = SW N surf i =1 Q ( ZoneNum) i i ∑ AbsInsSurf A = QSW ( ZoneNum) *VMULT ( ZoneNum) (143) where 2 The program does not allow shading devices on interior windows.2 Most of this calculation is done in subroutine CalcInteriorSolarDistribution in the SolarShading module.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE absorbed on the inside face of opaque surfaces. It is calculated as follows. Let Qsw be the total diffuse short-wave radiation entering the zone or originating in the zone. or. summing over these surfaces gives the following energy balance equation: N surf i =1 ∑ QS (ZoneNum) *α A = Q i i SW ( ZoneNum) where i = zone surface number counter Nsurf = number of heat transfer surfaces in zone Ai = surface area [m2] αi = inside solar absorptance for an opaque surface. The effects of movable shading devices on the exterior windows are taken into account.

GndSolarRad = ground-reflected solar radiation = GndReflectance*(DifSolarRad + BeamSolarRad*SOLCOS(3)). The incident diffuse solar in this case consists of sky diffuse and. The incident diffuse solar in this case consists of ground diffuse solar radiation. which in turn is produced by sky diffuse and beam solar reflecting from the ground. DBZone(ZoneNum) is the diffuse solar radiation originating from beam solar that passes through the exterior windows in the zone and reflects diffusely from inside zone surfaces plus beam solar entering the zone as diffuse radiation from windows with shading devices or diffusing glass (all divided by BeamSolarRad) [m2] DSZone(ZoneNum) is the sky-related diffuse solar incident on the zone’s exterior windows and transmitted into the zone (divided by DifSolarRad). sky diffuse reflected from exterior obstructions and the ground.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE VMULT ( ZoneNum) = 1 N surf i =1 [m-2 ] i i ∑ AbsIntSurf * A Qsw is given by QS SW = QD( ZoneNum) + ZoneIntGain( ZoneNum)%QLTSW + ZoneIntGain( ZoneNum)%T _ QLTSW where [W] ZoneIntGain(ZoneNum)%QLTSW = short-wave radiation into zone from general (overhead) electric lighting [W] ZoneIntGain(ZoneNum)%T_QLTSW = short-wave radiation into zone from task electric lighting [W] QD(ZoneNum) = diffuse solar radiation entering or originating in zone [W] QD(ZoneNum) is given by: QD( ZoneNum) = DBZone( ZoneNum) * BeamSolarRad + DSZone( ZoneNum) * DifSolarRad + DGZone( ZoneNum) * GndSolarRad [W ] where DifSolarRad = unobstructed horizontal diffuse solar irradiance from the sky [W/m2]. if Calculate Solar Reflection From Exterior Surfaces = Yes in the Building object. DGZone(ZoneNum) is ground-related diffuse solar radiation incident on the zone’s exterior windows and transmitted into the zone (divided by GndSolarRad). DBZone(ZoneNum) is calculated as: DBZone( ZoneNum) = BTOTZone − BABSZone where [m 2 ] 4/23/05 91 . where GndReflectance is the ground solar reflectance and SOLCOS(3) is cosine of the beam solar incidence angle on the ground [W/m2].

4 See “Beam Solar Reflection from Window Reveal Surfaces. depending on slat angle.3 BABSZone = total beam solar absorbed inside the zone. For diffusing glass.” 5 If Solar Distribution = FullInteriorAndExterior in the Building object. there can also be beam-to-beam transmission.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE BTOTZone = total beam solar incident on the zone’s exterior windows that is transmitted as beam or diffuse. all beam solar is assumed to fall on the floor. Otherwise. and. BABSZone is given by the following sum (see Figure 43): BABSZone = Beam absorbed by opaque inside surfaces5 + Beam transmitted through the zone’s interior windows + + Beam transmitted back out of the zone’s exterior windows + + Beam absorbed by the zone’s exterior and interior windows + + Beam absorbed by inside daylighting shelves For beam incident on an exterior window we have the following: For transparent glass with no shade or blind there is only beam-to-beam transmission. or if a window shade is in place. etc. the program calculates where beam solar from exterior windows falls inside the zone. BTOTZone is given by: BTOTZone = N extwin i =1 ∑ TBmAll * SunlitFract * CosInc * Area * InOutprojSLFracMult i i i i i + Diffuse entering zone from beam reflected by window inside reveal surfaces4 + Diffuse transmitted by windows from beam reflected by outside reveal surfaces – Beam absorbed by window inside reveal surfaces Here.. there is only beam-to-diffuse transmission. TBmAll = beam-to-beam plus beam-to-diffuse transmittance of window SunlitFract = fraction of window irradiated by sun CosInc = cosine of solar incidence angle on window Area = glazed area of window [m2] InOutProjSLFracMult = shadowing factor due to inside and outside projections of window frame and/or divider (= 1. solar profile angle. If a window blind is in place there is beam-to-diffuse transmission.0 if there is no frame or divider). 4/23/05 92 3 .

ZoneNum) * [QD(OtherZoneNum) + ZoneIntGain(OtherZoneNum)%QLTSW + ZoneIntGain(OtherZoneNum)%T _ QLTSW ] where FractDifShortZtoZ(OtherZoneNum. This factor is calculated in subroutine ComputeDifSolExcZonesWIZWindows taking into account multiple reflection between zones. Beam reflected by D. 4/23/05 93 6 . and interior window. etc. interior wall. For example. QS(ZoneNum) is modified to take into account short-wave radiation received from the other zones through these windows: QS ( ZoneNum) → QS ( ZoneNum) + other zones ∑ FractDifShortZtoZ (OtherZoneNum. B. Aoverlap is the irradiated area of a surface projected back onto the plane of EW. Some of the beam solar from exterior window EW is absorbed by the floor. Some is transmitted by IW to the adjacent zone. Vertical section through a two-zone building showing where transmitted beam solar falls. any beam solar radiation absorbed by a surface is assumed to be uniformly distributed over the surface even though in reality it is likely to be concentrated in one or more discrete patches on the surface. and some of this is in turn reflected back to Zone2. D. Interior Beam Radiation The inside beam solar irradiance factor6 in (142) is given by: AbsIntSurf ( SurfNum) Nextwin AISurf ( SurfNum) = ∑ TBmi * Aoverlapi (SurfNum) * CosInci A( SurfNum) i =1 where i = exterior window number Nextwin = number of exterior windows in zone CosInci = cosine of angle of incidence of beam on exterior window i For the purposes of the surface heat balance calculation.ZoneNum) = “diffuse solar exchange factor” = fraction of short-wave radiation in OtherZoneNum that is transmitted to ZoneNum. If zone ZoneNum shares interior windows with other zones. IW. for two zones means that some of the radiation transmitted from Zone1 to Zone2 is reflected back to Zone1. B and IW contributes to the interior short-wave radiation flux in Z1. Z2.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE Aoverlap(IW) Aoverlap(B) Aoverlap(D) EW Z1 B IW Z2 D Figure 43.

the geometry of the window. the geometry of the interior surfaces. etc. solar profile angle.SKY AND SOLAR/SHADING CALCULATIONS SHADING MODULE TBmi = beam-to-beam transmittance of exterior window i at CosInci 7 Aoverlapi(SurfNum) = beam solar irradiated area of surface SurfNum projected back onto the plane of exterior window i (the Aoverlap’s for an exterior window sum up to the glazed area of the window).. AbsIntSurf(SurfNum) = inside face solar absorptance of surface SurfNum A(SurfNum) = area of surface SurfNum [m2] Interior Solar Radiation Absorbed by Windows The interior short-wave radiation absorbed by layer l (glass. TBmi can be > 0 if a blind is present and the slat angle.back A(SurfNum) = glazing area [m2] Interior Solar Radiation Transmitted by Interior Windows Interior Diffuse Radiation Transmitted by Interior Windows The interior diffuse short-wave radiation transmitted by an interior window to the adjacent zone is given by QS ( ZoneNum) *τ dif ( SurfNum) * A( SurfNum) where [W] τ dif ( SurfNum) = diffuse transmittance of the interior window Interior Beam Radiation Transmitted by Interior Windows The interior beam solar radiation transmitted by an interior window to the adjacent zone is 7 TBmi is zero if the window has diffusing glass or a shade. shade or blind) of a window is equal to: QS ( ZoneNum) *α ldif ( SurfNum) + .back α lbeam = the system beam solar absorptance of layer l for irradiance from the back side . These overlap areas (Figure 43) are determined with the EnergyPlus shadowing routines by considering a zone’s exterior window as a “sending” surface and the inside faces of the zone’s other surfaces as “receiving” surfaces (see “Shading Module”). are such that some beam passes between the slats. and the location of the window with respect to the interior surfaces.back A( SurfNum) ∑ TBm * Aoverlap ( SurfNum) * CosInc extwin i =1 i i i [W/m 2 ] α ldif = the system diffuse solar absorptance of layer l for irradiance from the back side . The overlap areas for a particular exterior window depend on the sun position. 94 4/23/05 .back BeamSolarRad * where α lbeam ( SurfNum) N .

2000). ASHRAE Transactions. (Ref Object: Snow Ground Reflectance Modifiers). Ground Reflected Solar Modifier and Daylighting Reflected Solar Modifier. Polygon area derived from Green’s Theorem. This fractional amount (entered monthly) is used in the following equation: GroundReflectedSolar=(BeamSolar • COS(SunZenithAngle)+DiffuseSolar) • GroundReflectance Of course. Ground Reflectances (Snow) When snow is on the ground. (McGraw-Hill. National Bureau of Standards (now National Institute of Standards and Technology). ground reflectances may change. "The Application of Homogeneous Coordinates to Shadowing Calculations".. DaylightingGroundReflectance used = GroundReflectance • DaylightingModifierSnow References Groth. R.SKY AND SOLAR/SHADING CALCULATIONS N extwin i =1 SHADING MODULE BeamSolarRad *τ beam ( SurfNum) ∑ TBmi * Aoverlapi ( SurfNum) * CosInci [W] where τ ( SurfNum) is the beam-to-beam transmittance of the interior window at the angle of incidence of beam solar from the exterior window on the interior window. January 22-24. Values can range from 0.. The program does not track where this radiation falls in the adjacent zone: it is counted as diffuse radiation in that zone. "Shadow .N. (Honolulu. M. M..0 to 1. 1973). the Ground Reflected Solar is never allowed to be negative.0. F.0 to 1. Newman. Vol 84 (ASHRAE.. Meeus.0. G. and Sproul. C. Graphic Gems repository." Second Hawaii International Conference on System Sciences.0 for each month. Walton. beam QS ( ZoneNum) → QS ( ZoneNum) + [beam solar from adjacent zones]*VMULT ( ZoneNum) Ground Reflectances Ground reflectance values (Ref Object: GroundReflectances) are used to calculate the ground reflected solar amount. This object allows the user to specify two values. 4/23/05 95 . Astronomical Algorithms. 1978). Principles of Interactive Computer Graphics. and Lokmanhekim. GroundReflectance used = GroundReflectance • SolarModifierSnow Daylighting Reflected Solar Modifier is used to modified the basic monthly ground reflectance when snow is on the ground (from design day input or weather data values). C. (Willmann-Bell. Walton. The Snow Ground Reflectance Modifier can further modify the ground reflectance when snow is on the ground. Values can range from 0. N. Jean. February 1983.. no ground reflected solar is used.. Part I. HI. “The Thermal Analysis Research Program Reference Manual Program (TARP)”. G. If the user enters 0.A New Technique for the Calculation of Shadow Shapes and Areas by Digital Computer. W. Ground Reflected Solar Modifier is used to modified the basic monthly ground reflectance when snow is on the ground (from design day input or weather data values). 1969). Therefore.

or Surface:Shading:Attached. 4/23/05 96 8 . In this case one section of the building reflects solar radiation onto another section (and vice-versa). for example) are assumed to be diffusely reflecting. These are surfaces like overhangs or neighboring buildings entered with Surface:Shading:Detached:Fixed. But in this case the ground plane is considered unobstructed.e. Examples are shown in Figure 44. The ground surface. Beam solar and sky solar reflection from the ground is calculated even if “withReflections” is not used (the default). The reflectance values for windows and glass doors are calculated by the program from the reflectance properties of the individual glass layers that make up surface’s construction assuming no shading device is present and taking into account interreflections among the layers.8 The reflecting surfaces fall into three categories: Shadowing surfaces. See Figure 45.SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACES SHADING MODULE Solar Radiation Reflected from Exterior Surfaces EnergyPlus has an option to calculate beam and sky solar radiation that is reflected from exterior surfaces and then strikes the building. Exterior building surfaces. This calculation occurs if “withReflections” is used on the SolarDistribution option in the Building object. These surfaces can have diffuse and/or specular (beam-to-beam) reflectance values that are specified with the Shading Surface Reflectance object. A different method from that described here is used for calculating reflections from daylighting shelves (see “Daylighting Shelves”). The reflectance values for opaque surfaces are calculated by the program from the Absorptance:Solar and Absorptance:Visible values of the outer material layer of the surface’s construction. Surface:Shading:Detached:Building. the shadowing of the ground by the building itself or by obstructions such as neighboring buildings is ignored. In this case the user-input value of ground view factor is not used. This shadowing is taken into account only if “WithReflections” is used in the SolarDistribution field (Building Object) (Figure 46). i. Windows and glass doors are assumed to be specularly reflecting.. For zones with detailed daylighting. Opaque building surfaces (walls. these reflections are also considered in the daylight illuminance calculations.

(c) Beam-to-beam (specular) reflection from the façade of an adjacent highly-glazed building represented by a vertical shadowing surface.SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACES SHADING MODULE Solar Reflection from Obstructions Sky diffuse Beam Beam Beam (a) (b) (c) Figure 44. Solar reflection from building surfaces onto other building surfaces. Wall Window Figure 45. (b) Diffuse reflection of sky solar radiation from the top of an overhang. dashed arrows are diffuse solar radiation. The reflection from the wall is diffuse. (a) Diffuse reflection of beam solar radiation from the top of an overhang. In this example beam solar reflects from a vertical section of the building onto a roof section. Examples of solar reflection from shadowing surfaces in the Surface:Shading series. The reflection from the window is specular. Solid arrows are beam solar radiation. 4/23/05 97 .

0.0) and (5. Shadowing by the building also affects sky solar reflection from ground (not shown).75. (1.25.0.75).25).0. Beam-to-diffuse reflection from the ground onto the building occurs only for sunlit areas.0. added to the contribution from other hit points.3). 4/23/05 98 .0. Separate calculations are done for beam-to-diffuse and sky solar reflection from all obstructions and beam-to-diffuse and sky solar reflection from the ground.) Receiving points An N-sided surface is assigned N receiving points with the following coordinates. called here “receiving surfaces.0. (3.75.3).75) and (3. The radiance at the hit point from reflection of incident beam or sky solar is determined and the contribution of this radiance to the receiving surface is calculated. not for shaded area. = otherwise 2N 2N For example. A and C.25).2. B. windows and doors). expressed in terms of the surface vertex coordinates: Pij = ∑ aki vkj .0).SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION Beam Beam Ground A B C Ground Figure 46. and averaged over the receiving points. roofs. ground or an obstruction. this expression gives receiving points at (1.3. Then it determinines whether each ray hits the sky. Shadowing by the building itself affects beam solar reflection from the ground.0. i = 1. Diffuse Reflection of Beam Solar and Sky Solar Radiation A ray-tracing method is used to calculate beam solar and sky solar radiation that is diffusely reflected onto each of a building’s exterior surfaces (walls. (0.0. for a vertical 3m by 5m rectangle with vertices (0. = 1/5 otherwise N +1 1 If N > 3: akj = if k = i.3 k =1 N where Pij = jth coordinate of the ith receiving point vkj = jth coordinate of the kth surface vertex If N = 3: akj = 3/5 if k = i.” below.0.0.25. (5.2. j = 1. as shown in Figure 47. (For beam-to-beam reflection see “Beam Solar Radiation Specularly Reflected from Obstructions.” The calculation begins by generating a set of rays proceeding into the outward hemisphere at each receiving point on a receiving surface.

Element of solid angle associated with ray ( dΩi ) 4. and the surface unit vector at the hit point pointing into the hemisphere containing the receiving point 6. : 1. the coordinates of the ground hit point 7. the following is determined for each ray. Cosine of angle between ray and plane of receiving surface ( cos α i ) 3. In subroutine InitSolReflRecSurf. An upgoing ray may hit an obstruction or the sky. If the ray hits one or more obstructions. Unit vector in direction of ray 2.SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION 5m 3m Receiving points for reflected solar radiation Exterior heat transfer surface Figure 47. Distance from receiving point to hit point 4/23/05 99 . If the ray is downgoing and hits the ground. the solar reflectance ( ρ obs . Rays A total of 90 rays are sent out into the exterior hemisphere surrounding each receiving point. the coordinates of the hit point on the obstruction nearest the receiving point 5.i if an obstruction). i. A downgoing ray may hit an obstruction or the ground. For the surface containing the hit point: the surface number. Vertical rectangular exterior heat transfer surface showing location of receiving points for calculating incident solar radiation reflected from obstructions. See Figure 48.

and assuming that the radiance of the sky is uniform. Hitobs. Rays 1-6 hit the ground.i = 1 if ray i hits an obstruction. rays 7-11 hit an obstruction. In the present case this means that the irradiance at the hit point due to reflection of sky radiation from the ground or from other obstructions is not considered. It is given by: ReflFacSkySolObs(RecSurfNum) = 1 N rec where ∑∑ Hit 1 i =1 N rec N ray obs . DifShdgRatioIsoSkyobs. N rec is the number of receiving points. and rays 12-15 hit the sky.i DifShdgRatioIsoSkyobs . Sky Solar Radiation Diffusely Reflected from Obstructions The factor for reflection of sky radiation from obstructions onto a receiving surface is calculated in subroutine CalcSkySolDiffuseReflFactors.i = (obstructed sky irradiance on obstruction)/(unobstructed sky irradiance on obstruction) In this equation the product ViewFacSky*DifShdgRatioIsoSky is the sky irradiance at the hit point divided by the horizontal sky irradiance taking into account shadowing of sky diffuse radiation on the obstruction by other obstructions. The above reflection factor is used in the timestep calculation to find the irradiance on a receiving surface due to sky radiation reflected from obstructions: QRadSWOutIncSkyDiffReflObs(RecSurfNum) = DifSolarRad * ReflFacSkySolObs(RecSurfNum) (W/m2) 4/23/05 100 .i ρobs .i ViewFacSkyobs . = 0 otherwise. “obs. Two-dimensional schematic showing rays going outward from a point on a receiving surface.i = unobstructed sky view factor of the obstruction = (1 + cos tiltobs ) / 2 .i” denotes the obstruction hit by ray i. Note that we ignore secondary reflections here and in the following sections.i cos α i / π RecSurfNum is the receiving surface number. N ray is the number of rays.SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION 13 12 11 10 9 8 7 Ground plane 6 14 15 Rays from the receiving point Obstruction Receiving surface (window) One of the window’s receiving points 5 4 3 21 Figure 48. ViewFacSkyobs.

It is given by: ReflFacSkySolGnd(RecSurfNum) = 1 N rec N ray ⎡ ⎤ ∑∑ ⎢( Hit gnd . α j (i ) is the angle of incidence of ray j (i ) with respect to the ground plane. j (i ) cos α j (i ) d Ω j (i ) / π ⎥ 1 i =1 ⎣ j (i ⎦ N rec N ray where j (i ) denotes an upgoing ray from the ground point hit by ray i from the receiving point. = 0 otherwise. Sky Solar Radiation Diffusely Reflected from the Ground If a downgoing ray from a receiving point hits the ground (for example. To do this. the program calculates the radiance at the ground hit point due to sky diffuse solar reaching that point. For this calculation it is assumed that the radiance of the sky is uniform. dΩ j (i ) is the solid angle element associated with ray j (i ) . Rays from the ground hit point 10 7 6 Obstruction 4 3 2 1 Ground plane 5 8 9 11 12 13 14 15 Receiving surface (window) One of the window’s receiving points One of the ground hit points Figure 49. rays 6-10 in Figure 49).SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION where DifSolarRad is the horizontal sky irradiance on an unobstructed horizontal plane (W/m2). Hitsky. The factor for reflection of sky radiation from the ground onto a receiving surface is calculated in subroutine CalcSkySolDiffuseReflFactors. This factor is used in the timestep calculation to find the irradiance on a receiving surface due to sky radiation reflected from the ground: 4/23/05 101 . Two-dimensional schematic showing rays going upward from a ground hit point. rays are sent upward from the ground hit point and it is determined which of these rays are unobstructed and so go to the sky (for example.j(i) = 1 if ray j(i) hits the sky.i d Ωi cos αi / π )∑) Hitsky . rays 1-6 in Figure 48).

Beam Solar Radiation Diffusely Reflected from the Ground This calculation is the same as that for beam solar diffusely reflected from obstructions except that only rays from a receiving point that hit the ground are considered. The calculation is done for the hourly sun positions on each of the design days. It is also done for hourly sun positions on selected days during the weather file run period (the same days for which the shadowing calculations are done). to find the radiance at a hit point on an obstruction a line is drawn from the hit point to center of the sun. in which case it is assumed that no beam solar reaches the hit point. sun = 1 if the line from ray i’s hit point to the sun is unobstructed. The factor for diffuse reflection of beam solar from the ground onto a receiving surface is calculated in subroutine CalcBeamSolDiffuseReflFactors. = 0 otherwise.i .HourOfDay) + WeightPreviousHour * ReflFacBmToDiffSolObs(RecSurfNum.i Hitobs . which is assumed to be uniform over the ground plane but may vary monthly (because of snow cover. sun d Ωi cos α gnd . From this line it is determined (1) if there is an obstruction between the hit point and the sun.i cos α sun . for example). The factor for diffuse reflection of beam solar radiation from obstructions onto a receiving surface is calculated in subroutine CalcBeamSolDiffuseReflFactors.i .i cos α sun . sun d Ωi cos α i ρobs .PreviousHour)) where BeamSolarRad is the timestep value of beam normal solar intensity (W/m2). and (2) if beam solar does reach the hit point.i where IHr = hour number Hitobs . what the the solar angle of incidence at that point is.i = 1 if ray i from the receiving point hits an obstruction. = 0 otherwise. This factor is used in the timestep calculation to find the diffuse irradiance on a receiving surface due to beam solar diffusely reflected from obstructions: QRadSWOutIncBmToDiffReflObs(RecSurfNum) = BeamSolarRad * (WeightNow * ReflFacBmToDiffSolObs(RecSurfNum.obs .IHr) = 1 N rec ∑ ∑ Hit 1 i =1 N rec N ray obs .IHr) = N rec where 4/23/05 ∑∑ Hit 1 i =1 N rec N ray gnd . and WeightNow and WeightPreviousHour are time-averaging factors. Hitobs .obs .SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION QRadSWOutIncSkyDiffReflGnd(RecSurfNum) = DifSolarRad * ρ gnd * ReflFacSkySolGnd(RecSurfNum) (W/m2) where ρ gnd is the solar reflectance of the ground.i Hit gnd . Beam Solar Radiation Diffusely Reflected from Obstructions This calculation is similar to that for sky solar reflected from obstructions. It is given by: ReflFacBmToDiffSolObs(RecSurfNum. α sun . It is given by: 1 ReflFacBmToDiffSolGnd(RecSurfNum.i is the angle of incidence of the sun on the obstruction.i . However. gnd 102 .

Two-dimensional schematic showing specular reflection from an obstruction such as the glazed façade of a neighboring building. If no.HourOfDay) + WeightPreviousHour * ReflFacBmToDiffSolGnd(RecSurfNum.SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION Hit gnd . Hit gnd . find intersection point C and construct ray CD. If no. sun = 1 if the line from ray i’s hit point ot the sun is unobstructed. Select specularly reflecting surface EF. 6. 4. 7. (2) CD does not hit any obstructions (such as RS). 3.i = 1 if ray i hits the ground. B. of the sun with respect to the plane of EF and construct ray DB. Find the mirror image.PreviousHour)) Beam Solar Radiation Specularly Reflected from Obstructions Figure 50 shows schematically how specular (beam-to-beam) reflection from an obstruction is calculated. and (3) AC does not hit any obstructions (such as PQ). which could be important for snow-cover conditions. This factor is used in the timestep calculation to find the diffuse irradiance on a receiving surface due to beam solar diffusely reflected from the ground: QRadSWOutIncBmToDiffReflGnd(RecSurfNum) = BeamSolarRad * ρ gnd * (WeightNow * ReflFacBmToDiffSolGnd(RecSurfNum. α sun . Check if DB passes through EF. Select receiving point D on receiving surface JK.i . 4/23/05 103 9 . 2. Check if CD is obstructed. 9 Mirror image of sun about EF B P E Sun A Q C R J D S K Receiving point Receiving surface Specularly reflecting obstruction F Figure 50. if yes. gnd = angle of incidence of sun on ground (= solar zenith angle). 5. construct ray AC and check if it is obstructed. = 0 otherwise. = 0 otherwise. The program could be improved by adding a ground surface specular component. The calculation procedure is as follows: 1. The receiving point receives specularly reflected beam solar radiation if (1) DB passes through specularly reflecting surface EF. find reflected beam irradiance (W/m2) at D: The ground surface is assumed to be diffusely reflecting so there is no specular reflection from the ground.

Fraction of shading surface that is glazed 4. then f C .PreviousHour)) 4/23/05 104 .SOLAR RADIATION REFLECTED FROM EXTERIOR SURFACESDIFFUSE REFLECTION OF BEAM SOLAR AND SKY SOLAR RADIATION bm I D .IHr) = specularly reflecting surfaces ∑ ⎡ 1 ⎢ ⎣ N rec ∑f 1 N rec C . Diffuse solar reflectance of the unglazed part of the shading surface 2.0 unless the window has dividers). α D = angle of incidence of ray CD on JK. This object contains values for: 1. glazed ρ spec (α C ) cos α D ⎥ ⎦ ⎤ The program assumes that specular reflection from a surface is due to glazing. The factor for specular reflection of beam solar from obstruction onto a receiving surface is calculated in subroutine CalcBeamSolSpecularReflFactors. It is given by: ReflFacBmToBmSolObs(RecSurfNum. glazed is “Fraction of shading surface that is glazed” and ρ spec (α C ) is the front reflectance of the indicated glazing construction as a function of beam solar incidence angle.refl = BeamSolarRad * ρ spec (α C ) cos α D where α C = angle of incidence of beam solar at point C of the obstruction. If the reflecting surface is a window belonging to the building itself (as in Figure 45). the glazed façade of a neigboring building) the surface reflection information is entered with the Shading Surface Reflectance object. Diffuse visible reflectance of the unglazed part of the shading surface 3. The above specular reflection factor is used in the timestep calculation to find the beam irradiance on a receiving surface due to beam-beam reflection from obstructions: QRadSWOutIncBmToBmReflObsRecSurfNum) = BeamSolarRad * (WeightNow * ReflFacBmToBmSolObs(RecSurfNum. If the surface is a shading surface (that represents. for example. glazed is the fraction of the window that is glazed (which is 1. Name of glazing construction In this case f C . ρ spec (α C ) = reflectance of obstruction as a function of the angle of incidence.HourOfDay) + WeightPreviousHour * ReflFacBmToBmSolObs(RecSurfNum.

which are used to determine glare.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING CALCULATIONS Daylighting and Window Calculations Daylighting Calculations The EnergyPlus daylighting model. In this calculation the illuminance at the reference points in each zone is found by interpolating the stored daylight factors using the current time step’s sun position and sky condition. The daylighting calculation has three main steps: 1. (2) In EnergyPlus the clear-sky daylight factors are calculated for hourly sun-path sun positions several times a year whereas in DOE-2 these daylight factors are calculated for a set of 20 sun positions that span the annual range of sun positions for a given geographical location. intermediate. These factors are calculated for the hourly sun positions on sun-paths for representative days of the run period. Dividing daylight illuminance or luminance by exterior illuminance yields daylight factors. window management in response to solar gain and glare. the zone lighting electric reduction factor is passed to the thermal calculation. A similar option uses window shades to automatically control solar gain. window size and orientation. if available. site conditions. in DOE-2 only two sky types are used—clear and overcast. and various lighting control strategies. 1985]. and the contribution of light that reflects from the walls. 2. Daylight factors. The user specifies the coordinates of one or two reference points in each daylit zone. Taken into account are such factors as sky luminance distribution. and external obstructions. which uses this factor to reduce the heat gain from lights. are also calculated. and overcast. glazing transmittance. The electric lighting control system is simulated to determine the lighting energy needed to make up the difference between the daylighting illuminance level and the design illuminance. EnergyPlus then integrates over the area of each exterior window in the zone to obtain the contribution of direct light from the window to the illuminance at the reference points. in conjunction with the thermal analysis. inside surface reflectances. then multiplying by the exterior horizontal illuminance. determines the energy impact of daylighting strategies based on analysis of daylight availability. 3. which is described in [Winkelmann.sky Description Exterior horizontal illuminance due to light from the sky Units lux FORTRAN variable GILSK 4/23/05 105 . to decrease glare below a specified comfort level. If glare control has been specified. 1983] and [Winkelmann and Selkowitz. The EnergyPlus daylighting calculation is derived from the daylighting calculation in DOE2. which are ratios of interior illuminance or luminance to exterior horizontal illuminance. Finally. floor and ceiling before reaching the reference points. clear turbid. are calculated and stored. Daylight Factor Calculation Table 21. sun control devices such as movable window shades. the program will automatically deploy window shading. A daylighting calculation is performed each heat-balance time step when the sun is up. There are two major differences between the two implementations: (1) In EnergyPlus daylight factors are calculated for four different sky types—clear.1E. Window luminance and window background luminance. Variables in Daylighting Calculations Mathematical variable Eh.

sun related light Window background luminance factor due to sky. WLUMSU Lw. sun related light Window luminance factor due to sky.sun Exterior horizontal illuminance due to light from the sun Interior illuminance factor due to sky. φsky ψcs ψts ψis ψos φsun γ radians cd/m2 cd/m2 cd/m2 cd/m2 radians or degrees radians NWD THSKY. dsun wsky. sun related light Number of exterior windows in a zone Azimuth and altitude angles of a point in the sky Clear sky luminance distribution Clear turbid sky luminance distribution Intermediate sky luminance distribution Overcast sky luminance distribution Altitude angle of the sun Angle between point in the sky and the sun.shade cd/m2 WLUMSK. PHSKY PHSUN G Lz m h Eh. wsun bsky. SFACSU BFACSK. WLUMSU 4/23/05 106 . or angle between vertical and ray from reference point to window element Sky zenith luminance Optical air mass of the atmosphere Building altitude Exterior horizontal illuminance for sky type k Number of azimuth. bsun cd/lm cd/lm DFACSK.k Nθ. NPH Rref Rwin dΩ Lw m m steradians cd/m2 RREF RWIN DOMEGA WLUMSK. Nφ cd/m2 m m lux ZENL AM Elevation NTH. DFACSU SFACSK. BFACSU N θsky.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION Eh. altitude steps for sky integration Vector from zone origin to reference point Vector from zone origin to window element Solid angle subtended by window element Luminance of a window element as seen from reference point Luminance of window element with shade in place lux GILSU dsky.

DWY DIS - ˆ Rray ˆ Wn ˆ W21 ˆ W23 τvis - RAY WNORM cd/m lm lm lm lux m2 2 W21 W23 TVISB ELUM. EINTSU ATOT ZoneDaylight%AveVisDiffR eflect TH. PH HitPointLum-Surface%Area - L ΦFW ΦCW F1 ρFW ρCW Er A ρ θ.φ) Aw β radians cd/m2 m2 radians T(β) TVISBR 4/23/05 107 .φ) Area of glazed part of window Angle of incidence. φ L(θ. of light from a sky or ground element Glazing visible transmittance at incidence angle β lux - dx. pointing away from zone Unit vector along window yaxis Unit vector along window xaxis Glass visible transmittance Luminance of sky or obstruction Downgoing luminous flux from a window Upgoing luminous flux from a window First-reflected flux Area-weighted reflectance of floor and upper part of walls Area-weighted reflectance of ceiling and upper part of walls Average internally-reflected illuminance Total inside surface area of a zone Area-weighted average reflectance of zone interior surfaces Azimuth and altitude angle of a sky or ground element Luminance of sky or ground element at (θ. dy D B m m radians DWX.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION dEh Horizontal illuminance at reference point from window element Size of window element Distance from reference point to window element Angle between window element’s outward normal and ray from reference point to window element Unit vector from reference point to window element Unit vector normal to window element. at center of window. FLFW-FLCW-SurfaceWindow%RhoFloor Wall SurfaceWindow%RhoCeilin gWall EINTSK.

unsh. θmax φw Φsh.sh lm - ΦCW. dΦCW lm - θmin.sh. modified to take direction of occupant view into account lm - dΦ lm - dΦFW.unsh lm - f - SurfaceWindow%FractionU pgoing Φinc fsunlit Φ Lsh Lb G Gi ω lm cd/m2 cd/m2 steradians SunLitFrac BLUM GTOT SolidAngAtRefPt Ω steradians SolidAngAtRefPtWtd 4/23/05 108 . radians radians lm THMIN. Φunsh. ΦFW. ΦFW.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION dΦinc Luminous flux incident on window from sky or ground element Luminous flux from sky or ground element transmitted through window Luminous flux from sky or ground element transmitted through window and goind downwar. without shade Upgoing and downgoing portions of transmitted flux through window with shade Upgoing and downgoing portions of transmitted flux through window without shade Fraction of hemisphere seen by the inside of window that lies above the window midplane Flux incident on glazing from direct sun Fraction of glazing that is sunlit Transmitted flux from direct sun Luminance of window with shade Window background luminance Discomfort glare constant Discomfort glare constant from window i Solid angle subtended by window with respect to reference point Solid angle subtended by window with respect to reference point. THMAX - ΦCW. upward Azimuth angle integration limits Window normal altitude angle Transmitted flux through window with shade.

the sun-related interior daylight will be proportional to the exterior horizontal solar illuminance. and (2) light that originates from the sun and reaches the window directly or by reflection from exterior surfaces. Light from the window reaches the workplane directly or via reflection from the interior surfaces of the room. Calculate exterior horizontal daylight illuminance from sky and sun for standard (CIE) clear and overcast skies. Interior Illuminance Components To calculate daylight factors. the sky-related interior daylight will be proportional to the exterior horizontal illuminance. Daylight Factors The following daylight factors are calculated: d sky = Illuminance at reference point due to sky-related light Eh . To calculate these factors the following steps are carried out for each hourly sun position on the sun paths for the design days and for representative days10 during the simulation run period: 1. due to light from the sky. sky 10 The sun positions for which the daylight factors are calculated are the same as those for which the solar shadowing calculations are done. for overcast sky and for standard clear sky. window luminance factors. Similarly. window luminance and window background luminance for each window/reference-point combination. for bare and for shaded window conditions (if a shading device has been specified). sky condition (clear or overcast) and room geometry. Calculate interior illuminance. 2. and window background luminance factors. 109 4/23/05 . Eh.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION Nx.sky. Divide by exterior horizontal illuminance to obtain daylight factors. For fixed sun position. Ny Number of elements in x and y direction that window is divided into for glare calculation Position factor for horizontal and vertical displacement ratios xR and yR Hopkinson position factor Window background luminance Illuminance on window background Total internally-reflected component of daylight illuminance Illuminance setpoint Glare index - NWX. NWY p(xR. 3. Eh. daylight incident on a window is separated into two components: (1) light that originates from the sky and reaches the window directly or by reflection from exterior surfaces.sun. yR) - DayltgGlarePositionFactor pH Lb Eb Er cd/m lm lm 2 DayltgGlarePositionFactor BLUM - Es GI lm - IllumSetPoint GLINDX Overview There are three types of daylight factors: interior illuminance factors.

91 + 10e −3( −φsun ) 2 −0. In the calculation of daylight factors. last sun-up hour ⎥ ⎢ Intermediate sky. sky-condition/sun-position and shading device: ⎡Clear sky.. sun wsky = wsun = bsky = bsun = For a daylit zone with N windows these six daylight factors are calculated for each of the following combinations of reference point. sun Window background luminance due to sky-related light Eh ... #2 ⎥ ⎢ ⎥ ⎥ ⎢ Clear sky. sky Average window luminance due to sun-related light Eh . last sun-up hour .DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION d sun = Illuminance at reference point due to sun-related light Eh . first sun-up hour ⎥ ⎢ ⎥ ⎢ Intermediate sky. sun Average window luminance due to sky-related light Eh . CIE. #1⎤ ⎢ Window #2 ⎥ ⎢ . Clear Sky The clear sky luminance distribution has the form [Kittler. first sun-up hour ⎥ ⎢ ⎥ ⎡ Window #1 ⎤ ⎢Overcast sky. ⎣ ⎦ ⎥⎢ (if shade assigned) ⎥ ⎢ ⎥⎢ ⎦ ⎥⎣ ⎣ Window #N ⎦ ⎢ Clear/turbid sky. 1965. The standard skies are as follows. 1973] ψ cs (θ sky . 1990]. φsky ) = Lz (0.. first sun-up hour ⎥ ⎡ Unshaded window ⎤ ⎥⎢ Ref. first sun-up hour ⎤ ⎢Clear/turbid sky.e.. last sun-up hour ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ ⎦ Sky Luminance Distributions The luminance distribution of the sky is represented as a superposition of four standard CIE skies using the approach described in [Perez et al. ⎡ ⎢ ⎥⎢ ⎥ ⎢ Shaded window ⎥ ⎢ Ref. last sun-up hour ⎥ ⎢ Overcast sky. the luminance of the sky at a point directly overhead).91 + 10e −3γ + 0.32cosecφsky ) π + 0.45sin 2 φsun ) Here. window. Lz is the zenith luminance (i.27385(0.. pt. pt. sky Window background luminance due to sun-related light Eh .45cos 2 γ )(1 − e 0. which are ratios of interior and exterior illumination 4/23/05 110 .

008) + 0.856 + 16e −3γ + 0. are shown in Figure 51. The angle.59φsky − 0.316) + φsky + 4.6298(φsun − 0.326 ⎣ ⎦ Z 2 = exp ⎡ −0. φsky ) = Lz (0.316) + 2.99224sin(2.3cos 2 γ )(1 − e 0. The various angles. between sun and sky element is given by γ = cos −1 ⎡sin φsky sin φsun + cos φsky cos φsun cos(θ sky − θ sun ) ⎤ ⎣ ⎦ The general characteristics of the clear-sky luminance distribution are a large peak near the sun.059) + 0. which are defined in the building coordinate system.812}⎤ ⎣ ⎦ Z 3 = 0. 1987] ψ ts (θ sky .3sin 2 φsun ) Intermediate Sky The intermediate sky luminance distribution has the form [Matsuura.73852 Z 4 = exp ⎡ −0. φsky ) = Lz Z1Z 2 /( Z 3 Z 4 ) where Z1 = ⎡1.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION quantities that are both proportional to Lz. in the vertical plane containing the sun.27385(0.563γ {(φsun − 0.35(sin(3.812}⎤ 2 ⎣ ⎦ 4/23/05 111 .563( π − φsun ){2.009) + 2. γ.0 for all sky luminance distributions. 1987] ψ is (θ sky .32cosecφsky ) π + 0.799 ⎤ / 2. and an increase in luminance as the horizon is approached. For this reason we will use Lz = 1. the zenith luminance cancels out.31) sin(2. a minimum at a point on the other side of the zenith from the sun. Clear Turbid Sky The clear turbid sky luminance distribution has the form [Matsuura.008)(φsky + 1.856 + 10e −3( −φsun ) 2 −0.6φsun + 0.6φsun + 0.

Direct Normal Solar Illuminance For purposes of calculating daylight factors associated with beam solar illuminance. Overcast Sky The overcast sky luminance distribution has the form [Moon & Spencer. Angles appearing in the expression for the clear-sky luminance distribution. the direct normal solar illuminance is taken to be 1. Note that at fixed solar altitude the zenith ( φsky = π / 2 ) is three times brighter than the horizon ( φsky = 0 ). is used in the time-step calculation. 1942] ψ os (φsky ) = Lz 1 + 2sin φsky 3 Unlike the clear sky case. The actual direct normal solar illuminance.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION Figure 51.0 W/m2. determined from direct normal solar irradiance from the weather file and empiricallydetermined luminious efficacy. the overcast sky distribution does not depend on the solar azimuth or the sky azimuth. Exterior Horizontal Illuminance The illuminance on an unobstructed horizontal plane due to diffuse radiation from the sky is calculated for each of the four sky types by integrating over the appropriate sky luminance distribution: 4/23/05 112 .

The integral is evaluated as a double summation: Eh . due to a window element is dEh = Lw d Ω cos γ where Lw is the luminance of the window element as seen from the reference point.DAYLIGHTING AND WINDOW CALCULATIONS 2π π / 2 DAYLIGHT FACTOR CALCULATION Eh . Direct Component of Interior Daylight Illuminance The direct daylight illuminance at a reference point from a particular window is determined by dividing the window into an x-y grid and finding the flux reaching the reference point from each grid element. The subtended solid angle is approximated by dΩ = where dxdy cos B D2 (144) D = Rwin − Rref CosB is found from ˆ ˆ cos B = Rray ⋅ Wn where Rray = ( Rwin − Rref ) / D 4/23/05 113 .k = ∑∑ψ k (θ sky (i ). Rref .k = ∫ ∫ψ 0 0 k (θ sky . is or os. φsky ) sin φsky cos φsky dθ sky dφsky where k = cs. ts.k . The horizontal illuminance at the reference point. The geometry involved is shown in Figure 52. φsky ( j ) ) sin φsky ( j ) cos φsky ( j )∆θ sky ∆φsky i =1 j =1 Nθ Nφ where θ sky (i ) = (i − 1/ 2)∆θ sky φsky ( j ) = ( j − 1/ 2)∆φsky ∆θ sky = 2π / Nθ ∆φsky = π / 2 Nφ Nθ = 18 and Nφ = 8 were found to give a ±1% accuracy in the calculation of Eh .

window elements that lie below the workplane ( cos γ < 0 ) are omitted since light from these elements cannot reach the workplane directly. Figure 52. the window luminance is Lw = Lτ vis (cos B) 4/23/05 114 . Vectors Rref. the luminance of the window element is found by projecting the ray from reference point to window element and determining whether it intersects the sky or an exterior obstruction such as an overhang.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION W − W2 W3 − W2 ˆ ˆ ˆ Wn = window outward normal = W21 × W23 = 1 × W1 − W2 W3 − W2 Equation (144) becomes exact as dx / D and dy / D → 0 and is accurate to better than about 1% for dx ≤ D / 4 and dy ≤ D / 4 . W1. W3 and Rwin are in the building coordinate system. W2. Unshaded Window For the unshaded window case. Geometry for calculation of direct component of daylight illuminance at a reference point. The net illuminance from the window is obtained by summing the contributions from all the window elements: Eh = window elements ∑ Lw d Ω cos γ (145) In performing the summation. If L is the corresponding luminance of the sky or obstruction.

It therefore works best for rooms 4/23/05 115 . Φ FW (lumens). and the building’s own wall and roof surfaces) but can be transmitting (like a tree or translucent awning). ceiling and windows in the room. The total reflected flux absorbed by the room surfaces (or lost through the windows) is AEr(1-ρ). and ρ is the area-weighted average reflectance of the room surfaces. 1954]. position on shade. F1. sh . where A is the total inside surface area of the floor. and an upward-going flux. is described in [Winkelmann. which falls on the floor and portions of the walls below the imaginary horizontal plane passing through the center of the window (window midplane). The illuminance contribution at the reference point from a shaded window element is then given by Eq. walls. The calculation of the window luminance with the shade in place. To find the final average internally-reflected illuminance.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION where τ vis is the visible transmittance of the glass for incidence angle B. that strikes the ceiling and portions of the walls above the window midplane. neighboring buildings.. Interior obstructions are assumed to be opaque (τobs = 0). In this method the daylight transmitted by the window is split into two parts—a downward-going flux. (145) with Lw = Lw. on the room surfaces (which in this method is uniform throughout the room) a flux balance is used. which usually have non-uniform luminance characteristics.e. is approximated by F1 = Φ FW ρ FW + Φ CW ρCW where ρFW is the area-weighted average reflectance of the floor and those parts of the walls below the window midplane. overhangs. Closely-woven drapery fabric and translucent roller shades are closer to being perfect diffusers than Venetian blinds or other slatted devices. A fraction of these fluxes is absorbed by the room surfaces. including windows. Exterior obstructions are assumed to be non-reflecting. Φ CW . Er. Shaded Window For the window-plus-shade case the shade is assumed to be a perfect diffuser. From conservation of energy AEr (1 − ρ ) = F1 or Er = Φ FW ρ FW + Φ CW ρCW A(1 − ρ ) This procedure assumes that the room behaves like an integrating sphere with perfectly diffusing interior surfaces and with no internal obstructions.sh. 1983]. then L = Lskyτobs. and angle of incidence of solar radiation falling on the shade. [Lynes. The remainder.. Exterior obstructions are generally opaque (like fins. i. If Lsky is the sky luminance and τobs is the transmittance of the obstruction (assumed independent of incidence angle). Lw. Internally-Reflected Component of Interior Daylight Illuminance Daylight reaching a reference point after reflection from interior surfaces is calculated using the split-flux method [Hopkinson et al. and ρCW is the area-weighted average reflectance of the ceiling and those parts of the walls above the window midplane. the luminance of the shade is independent of angle of emission of light. 1968]. the first-reflected flux.

φ )T ( β ) cos β cos φ dθ dφ (148) The downgoing and upgoing portions of this flux are Φ FW . sh = Φ (1 − f ) Φ CW . θ max π /2 For a window with a diffusing shade the total transmitted flux is Φ sh = Aw θ min π / 2 −φw ∫ ∫ L(θ . For an unshaded window the total downgoing transmitted flux is obtained by integrating over the part of the exterior hemisphere seen by the window that lies above the window midplane. and subtending a solid angle cos φ dθ dφ is d Φ inc = Aw L(θ .unshaded = Aw θ max π / 2 ∫ ∫ θ min L(θ . φ )T ( β ) cos β cos φ dθ dφ (147) φw is the angle the window outward normal makes with the horizontal plane. have matte surfaces (which is usually the case). is given by f = 0. can lead to substantial inaccuracies in the split-flux calculation. φ ) . φ ) . φ ) cos β cos φ dθ dφ The transmitted flux is d Φ = d Φ incT ( β ) where T(β) is the window transmittance for light at incidence angle β. Transmitted Flux from Sky and Ground The luminous flux incident on the center of the window from a luminous element of sky or ground at angular position (θ . This gives Φ FW .unshaded = Aw where θ max θ min π / 2 −φw ∫ ∫ 0 L(θ . such as would be the case for rooms whose depth measured from the window-wall is more than three times greater than ceiling height. sh = Φf where f. the fraction of the hemisphere seen by the inside of the window that lies above the window midplane. φ )T ( β ) cos β cos φ dθ dφ (146) 0 The upgoing flux is obtained similarly by integrating over the part of the exterior hemisphere that lies below the window midplane: Φ CW .DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION that are close to cubical in shape. of luminance L(θ . and have no internal partitions.5 − φw / π 4/23/05 116 . Deviations from these conditions. This transmittance depends on whether or not the window has a shade.

and down-going transmitted fluxes are equal: Φ FW . (147) and (148) depend on 12 . sh = Φ. Φ CW . sh = Φ / 2 . For a horizontal skylight ( φw = π / 2 ): Φ FW . 1983] we have φ . sh = 0 The limits of integration of θ in Equations (146). if present). sh = Φf 4/23/05 117 . sh = Φ CW . From [Figure sin α = sin( A − π / 2) = which gives sin φ tan φw cos φ A = cos −1 (tan φ tan φw ) Thus θ min = − cos −1 (− tan φ tan φw ) θ max = cos −1 (− tan φ tan φw ) Transmitted Flux from Direct Sun The flux incident on the window from direct sun is Φ inc = Aw EDN cos β f sunlit The transmitted flux is Φ = T ( β )Φ inc where T is the net transmittance of the window glazing (plus shade.unsh = 0 For a window with a diffusing shade Φ FW . sh = Φ (1 − f ) Φ CW . For an unshaded window all of the transmitted flux is downward since the sun always lies above the window midplane.unsh = Φ Φ CW .Winkelmann.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION For a vertical window ( φw = 0 ) the up. Therefore Φ FW .

DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHT FACTOR CALCULATION Luminance of Shaded Window The luminance of a shaded window is determined at the same time that the transmitted flux is calculated.8 w Lb + 0. 1954] that accounts for the decrease in visual excitation as the luminous element moves away from the line of sight. as is done for calculating the direct component of interior illuminance. yR ) j =1 i =1 N y Nx where p is a “position factor” [Petherbridge & Longmore.5 Lw G = discomfort glare constant Lw = average luminance of the window as seen from the reference point Ω = solid angle subtended by window. The modified solid angle is Ω = ∑∑ dω (i.j) with respect to the reference point.j) is the solid angle subtended by element (i. given by 4/23/05 118 . 1972]: G= where L1. j ) Nx N y where Lw(i.j) as seen from the reference point.6 Ω0. 1970] and [Hopkinson. we have Lw = ∑∑ L j =1 i =1 N y Nx w (i. j ) p ( xR . xR and yR (Figure 53). modified to take direction of occupant view into account Lb = luminance of the background area surrounding the window By dividing the window into Nx by Ny rectangular elements. ω = ∑∑ dω (i. This factor depends on the horizontal and vertical displacement ratios.07ω 0.j) is the luminance of element (i. Similarly. φ )T ( β ) cos β cos φ dθ dφ Daylight Discomfort Glare The discomfort glare at a reference point due to luminance contrast between a window and the interior surfaces surrounding the window is given by [Hopkinson. j ) j =1 i =1 N y Nx where dω(i. It is given by Lsh = 1 θ max π θ min π / 2 −φw ∫ ∫ π /2 L(θ .

057 0. j ) = YD / RR where ˆ ˆ RR = D( Rray ⋅ vview ) A2 = D 2 − ( RR) 2 YD = Rwin (3) − Rref (3) Figure 53.492 0.00 0. 1966].5 Vertical Displacement 1.081 0. the Hopkinson position factor [Hopkinson.0 0.014 0.DAYLIGHTING AND WINDOW CALCULATIONS 1/ 2 DAYLIGHT FACTOR CALCULATION xR (i.0 Factor 1.25 Interpolation of this table is used in EnergyPlus to evaluate p at intermediate values of xR and yR.008 1.123 0.008 2.5 0.226 0.119 0.061 0.019 0.019 0.016 0.5 0.0 0 0 0 0 119 . The factor p can be obtained from graphs given in [Petherbridge & Longmore.026 0.043 0.008 2.023 0. Position factor for glare calculation xR: Horizontal Displacement Factor 0 yR : 0. 1. Geometry for calculation of displacement ratios used in the glare formula.011 0.026 0.5 0.029 0.128 0.008 0.006 >3. since p = pH . 1954] or it can be calculated from tabulated values of pH.008 1.0 0.5 4/23/05 0 1.0 0. j ) = A2 − (YD) 2 RR yR (i.065 0.011 0. Table 22.006 3. The values resulting from the latter approach are given in Table 22.

dir Sh.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION 0 0 2. Variables in Time-Step Calculations Mathematical variable Snorm.0 The background luminance is 0.003 0 0.0 0 0 >2. If glare control has been specified window shading (by movable shading devices or switchable glazing) is deployed to reduce glare. A factor of two variation in Eb generally produces a change of only 0. Es ) where Er is the total internally-reflected component of daylight illuminance produced by all the windows in the room and Es is the illuminance set point at the reference point at which glare is being calculated. Finally the illuminance at each reference point for the final window and shade configuration is used by the lighting control system simulation to determine the electric lighting power required to meet the illuminance set point at each reference point. DifSolarRad 4/23/05 120 . and.003 0 0. A precise calculation of Eb is not required since the glare index (see next section) is logarithmic. for sun-related interior illuminance. for sky-related interior illuminance. then. the net illuminance and glare due to all of the windows in a zone are found. Table 23. The interior illuminance at each reference point is found for each window by interpolating the daylight illuminance factors for the current sun position.5 to 1. Glare Index The net daylight glare at a reference point due to all of the windows in a room is expressed in terms of a glare index given by number of windows GI = 10 log10 ∑ i =1 Gi th where Gi is the glare constant at the reference point due to the i window Time-Step Daylighting Calculation Overview A daylighting calculation is performed each time step that the sun is up for each zone that has one or two daylighting reference points specified. multiplying by the exterior horizontal solar illuminance that time step. multiplying by the exterior horizontal illuminance from the appropriate sky types that time step. The exterior horizontal illuminance from the sun and sky is determined from solar irradiance data from the weather file. By summation.003 0 0.0 in the glare index.dif Description Direct normal solar irradiance Exterior diffuse horizontal solar irradiance Units W/m2 W/m 2 FORTRAN variable BeamSolarRad SDIFH.003 0 0 0 Lb = Eb ρb where ρb is approximated by the average interior surface reflectance of the entire room and Eb = max( Er .003 0 0.

k Eh. for sky of type k Weighting factor for time step interpolation - wj - iL iS Itot Btot. BFSKHR WeightNow bsun. ηdir Iwin Swin Bwin dsun. k’ sk. for sky of type k Window luminance factor for sun. DirLumEff DaylIllum SourceLumFromWinAtRefP t BACLUM DaylIllFacSun. SFSUHR. bsky. dsky.k Window background luminance factor for sun. DaylSourceFacSky. SFSKHR DaylBackFacSun. B Reference point index Window shade index Total daylight illuminance at reference point Total window background luminance lux cd/m2 IL IS DaylIllum BLUM 4/23/05 121 .dir Z m ∆ ε k. BFSUHR.k’ ψk.sky Eh.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION Sh. DaylIllFacSky. for sky of type k W/m2 radians cd/m2 SDIRH Zeta AirMass SkyBrightness SkyClearness ISky SkyWeight - fk Eh. DaylBackFacSky. DFSUHR DaylSourceFacSun.k’ Exterior direct horizontal solar irradiance Solar zenith angle Relative optical air mass Sky brightness Sky clearness Sky type index Interpolation factor for skies k and k’ Sky luminance distribution formed from linear interpolation of skies k and k’ Fraction of sky that is type k Horizontal illuminance from sky type k Exterior horizontal illuminance from sky Exterior horizontal illuminance from sun Luminous efficacy of diffuse and direct solar radiation Interior illuminance from a window Window luminance Window background luminance Interior illuminance factor for sun.k cd/m lux lux lm/W lux cd/m2 cd/m2 2 HorIllSky HISKF HISUNF DiffLumEff.k wsun. DFSUHR.sun ηdif. wsky.

dif 1+ κ Z 3 +κZ3 where Sh.dif m / S norm . max[0.is )ψ is where ψts is the clear turbid sky luminance distribution and 4/23/05 122 .DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION Iset fL fP NL MP Illuminance setpoint Fractional electric lighting output Fractional electric lighting input power Number of steps in a stepped control system Lighting power multiplier lux - ZoneDaylight%IllumSetPoin t FL FP LightControlSteps ZonePowerReductionFactor Time-Step Sky Luminance The sky luminance distribution..dir ext where m is the relative optical air mass and S norm . ψts .osψ is + (1 − sis . and sis .05) / 4]} If 1. The interpolation factors are a function of sky clearness and sky brightness [Perez et al. Snorm.isψ ts + (1 − sts .dif is the diffuse horizontal solar irradiance.os = min {1. (ε − 1) / 0.dif + S norm . Z is the solar zenith angle and κ is a constant equal to 1. Sky clearness is given by Sh .dir ε= Sh . 1990]. Sky brightness is given by ext ∆ = S h . If ε ≤ 1.os )ψ os where ψis is the intermediate sky luminance distribution. (∆ − 0.2<ε ≤ 3 ψ ts .041 for Z in radians.” The two sky types that are interpolated depend on the value of the sky clearness.2. ψos is the overcast sky luminance distribution. ψis and ψos — described above under “Sky Luminance Distributions. for a particular time step is expressed as a linear interpolation of two of the four standard skies — ψcs.is = sts .dir is the direct normal solar irradiance.2 ψ is . ψ.os = sis . dir is the extraterrestrial direct normal solar irradiance.

iS ) f k ′ ]Eh . j / N t ] where Nt is the number of time steps per hour. iS .dir where Z is the solar zenith angle. iS .k (iL . the sun. First. sun = ηdir Snorm .2) /1.k ′ (iL . sky = ηdif S h .ts = min[1. ih ) + (1 − w j )d sun (iL . and k is the sky type index. from a window is calculated as follows by multiplying daylight factors and exterior illuminance. (ε − 3) / 3] Interior Illuminance For each time step the interior illuminance. respectively. iS ) = d sun Eh . ih is the hour number. The efficacies are calculated from direct and global solar irradiance using a method described in [Perez et al. The interior illuminance from a window is calculated as I win (iL . ηdif is the luminous efficacy (in lumens/Watt) of diffuse solar radiation from the sky and ηdir is the luminous efficacy of direct radiation from the sun. sun + [d sky . iS ) = w j d sun (iL .ts = scs . 2 for shaded window). iS ) f k + d sky .) The fractions fk and fk’ are given by 4/23/05 123 . sky where Eh.k (iL .dir cos Z Eh .sun and Eh. ih ) + (1 − w j ) d sky .k (iL .is = (ε − 1. iS .tsψ cs + (1 − scs . respectively.and sky-related daylight illuminance factors for the time step are determined by interpolation of the hourly factors: d sun (iL .k (iL . and fk and fk’ are the fraction of the exterior horizontal illuminance from the sky that is due to sky type k and k’.8 If ε > 3 ψ cs . 1990]. ih + 1) where iL is the reference point index (1 or 2). iS ) = w j d sky . Iwin.sky are the exterior horizontal illuminance from the sun and sky. The horizontal illuminance from sun and sky are given by Eh . ih + 1) d sky . iS is the window shade index (1 for unshaded window.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION sts . iS . the time-step interpolation weight is given by w j = 1 − min[1.ts )ψ ts where ψcs is the clear sky luminance distribution and scs . For the j th time step in an hour.

respectively (see “Exterior Horizontal Luminance.k + (1 − sk . For example.k ′ = scs .k’ are the horizontal illuminances from skies k and k’. k = cs (clear sky). sky The total illuminance at a reference point from all of the exterior windows in a zone is I tot (iL ) = windows in zone ∑ I win (is .k (iL . and sk. iL ) where iS = 1 if the window is unshaded and iS = 2 if the window is shaded that time step. k’ = ts (clear turbid sky) and sk .) Similarly.8 ∑ 0.k and Eh.k ′ Eh . iS ) = wsun Eh .k’ is the interpolation factor for skies k and k’ (see “Time-Step Sky Luminance. (ε − 3) / 3] Similarly.k ′ ) Eh . the total background luminance is calculated: Btot (iL ) = windows in zone ∑ Bwin (is . sun + [ wsky . sun + [bsky . if ε > 3. iS ) f k + bsky .” above).k ′ ) Eh .6 Ω(iL )0. iS )1. ρb I set (iL )) In the last relationship. and window background luminance. the window source luminance.ts = min[1.k ′ where Eh. iS ) f k ′ ]Eh .k ′ (iL .k sk .k ′ ) Eh .07ω (iL ) where B (iL ) = max( Bwin (iL ).k ′ (iL . iS ) f k ′ ]Eh . for a window are calculated from S win (iL . sky Bwin (iL . Swin.k ′ Eh .k ′ (1 − sk . iS ) f k + wsky .k ′ sk . iL ) Glare Index The net glare index at each reference point is calculated as GI (iL ) = 10 log10 in zone S win (iL .k + (1 − sk .k (iL . the background luminance is approximated as the larger of the background luminance from daylight and the average background luminance that would be 4/23/05 124 .5 S win (iL . (Before the illuminance calculation is done the window shading control will have been simulated to determine whether or not the window is shaded. Bwin.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION fk = fk′ = sk .k ′ Eh . iS ) windows B (iL ) + 0.” above). iS ) = bsun Eh .

and the glare at the first point stays below GI. fP. Glare Control Logic If glare control has been specified and the glare index at either reference point exceeds a user-specified maximum value.0 (Figure 54). shade the window if it decreases glare at both points. then: If glare is too high at both points. If glare is too high only at the second point.max.min ≤ f L ≤ 1 125 .min for fL<fL.max. shade the window if the glare at the second point decreases. shade a window if it is unshaded and shading it decreases the glare. it will be left in the shaded state. B(iL) would be better approximated as an areaweighted luminance of the surfaces surrounding a window. say to control solar gain. The relationship between fP and fL depends on the lighting control type. even if it does not decrease the glare below GI.0 as fL increases from fL. shade the window if the glare at the first point decreases.min ⎪ 1 − f L .min and that fP increases linearly from fP. 2) If there are two reference points. corresponding to fL is then calculated. fL. then the windows in the zone are shaded one by one in attempt to bring the glare at both points below GI.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION produced by the electric lighting at full power if the illuminance on the room surfaces were equal to the setpoint illuminance. 3) Shades are closed in the order of window input until glare at both points is below GI. set L ⎥ I set (iL ) ⎣ ⎦ Here.min for f L < f L . the electric lighting control is simulated. Note that if a window has already been shaded. In a more detailed calculation. required to meet the setpoint at reference point iL is given by ⎡ I (i ) − I tot (iL ) ⎤ f L (iL ) = max ⎢ 0. This gives ⎧ f P . Continuous Dimming Control For a continuously-dimmable control system. taking into account the luminance contribution from the electric lights.max.max. Lighting Control System Simulation Once the final daylight illuminance value at each reference point has been determined.) The following logic is used: 1) If there is only one reference point.min to 1.max.min ⎩ 4/23/05 for f L . The fractional electric lighting output. where the luminance of each room surface is separately determined. (Each time a window is shaded the glare and illuminance at each reference point is recalculated. Iset is the illuminance setpoint and Itot is the daylight illuminance at the reference point. GI. it is assumed that fP is constant and equal to fP. or until there are no more windows left to shade. The fractional electric lighting input power. If glare is too high only at the first point. and the glare at the second point stays below GI.min ⎪ f P = ⎨ f L + (1 − f L ) f P .max.min − f L . This relationship assumes that the electric lights at full power produce an illuminance equal to Iset at the reference point.min to 1.

Continuous/Off Dimming Control A “continuous/off” dimming system has the same behavior as a continuous dimming system except that the lights switch off for fL < fL. This gives ⎧0. if fP < 1. NL (Figure 55). is specified.0 Increasing daylight illuminance Fractional light output Zero daylight illuminance Minimum light output fraction 0 0 1.min. 4/23/05 126 . fP exceeds pL. This can be used to simulate the uncertainty associated with manual switching of lights.min rather than staying at fP. Control action for a continuous dimming system.0 Fractional input power Minimum input power fraction Figure 54. if f L = 1 ⎩ If a lighting control probability.DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION 1. if f L = 0 ⎪ ⎪ int( N L f L ) + 1 fP = ⎨ . fP is set one level higher a fraction of the time fP + 1/NL if a random number between 0 and 1 equal to 1-pL. fP takes on discrete values depending on the range of fL and the number of steps. for 0 < f L < 1 NL ⎪ ⎪1. Stepped Control For a stepped control system. pL. Specifically.

DAYLIGHTING AND WINDOW CALCULATIONS TIME-STEP DAYLIGHTING CALCULATION Step 1 1. Glare from Daylighting in Buildings. 1968. P. Heinnemann. and D. Applied Science Publishers. Principles of Natural Lighting. 98 (1970). Hopkinson. J. Daylighting. Stepped lighting control with three steps. the net lighting power multiplier. J. R. An Empirical Formula for the Computation of the Indirect Component of Daylight Factors. Hopkinson. R. Standardization of the Luminance Distribution on Clear Skies. Illuminating Engineering 37. Proc. Longmore. p. No. 1965. For this fraction the electric lighting is unaffected and the power multiplier is 1.G. R. 707-726 (1942). Illumination from a Nonuniform Sky. 22. Spencer. R. Applied Ergonomics 3. p. MP. Lynes. 322.2. Lighting Power Reduction Using the value of fP at each reference point and the fraction fZ of the zone controlled by the reference point. 1973... Kittler. Hopkinson. CIE Pub. London. K. Eng.G.0 Step 2 Fractional input power Step 3 0 0 Daylight illuminance Illuminance set point Figure 55. 129. Petherbridge. Standardization of Outdoor Conditions for the Calculation of the Daylight Factor with Clear Skies. Ltd. Soc.A. Longmore and P. CIE Technical Report of TC 3-09 (1987). 206 (1972). 201 (1954). CIE Inter-Session Meeting on Sunlight.G. References CIE Technical Committee 4. for the entire zone is calculated. R. Glare from Windows.. Construction Research and Development Journal 2.0. Newcastle-Upon-Tyne. this value multiplies the lighting power output without daylighting. 4/23/05 127 . Paris. (London) 19. Luminance Distributions of Various Reference Skies. Hopkinson. Trans. Petherbridge and J. 2 2 ⎛ ⎞ M P = ∑ f P (iL ) f Z (iL ) + ⎜ 1 − ∑ f Z (iL ) ⎟ iL =1 ⎝ iL =1 ⎠ In this expression. Commission Internationale d’Eclairage. Illum. Matsuura.G. Moon. P. London. the term to the right in the parentheses corresponds to the fraction of the zone not controlled by either reference point. 1966.

C. and [Modest. and J. resulting in a final total illuminance at each reference point. Jan. J. Petherbridge. Daylight factors are calculated for each reference point. and Superlite. For each point in time.173 (1954) Winkelmann. Daylighting Simulation in the DOE-2 Building Energy Analysis Program. refer to the section in this documentation entitled “Daylighting Calculations” and to [Winkelmann. 5. The DElight daylighting calculation methods are derived from the daylighting calculations in DOE-2. and to the luminance at subdivided nodal patches of interior surfaces. and exterior reflecting surfaces). DElight first calculates the contribution of light transmitted through all simple and complex fenestration systems in the zone to the illuminance at each reference point. Solar Energy 44. Daylighting Calculation in DOE-2. F. Seals. Modeling Daylight Availability and Irradiance Components from Direct and Global Irradiance. prismatic or holographic glass).e. are pre-calculated and stored for later use. 1983. P.. including exterior light sources. for a set of sun positions and sky conditions that are representative of the building location. [Winkelmann and Selkowitz.. 271-286 (1985) DElight Daylighting Calculations EnergyPlus includes the DElight method of analyzing daylighting from both simple apertures (i. for a given exterior luminous environment (including sky. DElight calculates the interior daylighting illuminance at user specified reference points and then determines how much the electric lighting can be reduced while still achieving a combined daylighting and electric lighting illuminance target. 1982].C. 1983]. and visible light transmittance of simple and complex fenestration systems. sun. roof monitors) and/or optically complicated glazings (e. The effect of inter-reflection of this initial light between interior reflecting surfaces is then calculated. In this calculation the illuminance at the reference points in each zone is found by interpolating the stored daylight factors using the current time step sun position and sky condition. Daylight Factor Calculation: Daylight factors. Electric Lighting Control Calculation: The electric lighting control system is simulated to determine the proportion of lighting energy needed to make up the difference between the daylighting illuminance level at the given time step. windows and skylights) and complex fenestration systems that include geometrically complicated shading (e.DAYLIGHTING AND WINDOW CALCULATIONS DELIGHT DAYLIGHTING CALCULATIONS Perez. Longmore. The subsequent reduction of electric lighting depends on daylight illuminance level. Michalsky and R.1E (as are the Detailed daylighting calculations in EnergyPlus).g. Lawrence Berkeley Laboratory report no.g. R. 271-289 (1990). P. design illuminance set point. F. Solid Angles Applied to Visual Comfort Problems. Winkelmann.. The engineering documentation included here focuses on the details of these differences from methods documented elsewhere. 1985]. fraction of zone controlled by reference point. and S. which uses this factor to reduce the heat gain from lights. The daylight illuminance level in a zone depends on many factors. reflectance of interior surfaces. Finally. R. 4/23/05 128 . which are ratios of interior illuminance to exterior horizontal illuminance. and type of lighting control. the zone lighting electric reduction factor is passed to the thermal calculation. Ineichen. LBL-11353. with several key modifications. and location of reference points. This total illuminace is then divided by the exterior horizontal illuminance for the given exterior environment to give a daylight factor. Selkowitz. then multiplying by the exterior horizontal illuminance. and the design illuminance level. Light and Lighting 47. Time-Step Interior Daylighting Calculation: A daylighting calculation is performed for each heat-balance time step when the sun is up. 6. location. size. The DElight daylighting calculation has three main steps: 4. Stewart.. The user spcifies the coordinates of one or more reference points in each daylit zone. Energy and Buildings 8. For the details of the heritage calculations.

and to the luminance at each gridded nodal patch of interior surfaces. non-horizontal). using ray-tracing techniques) or pre-measured. Bi-directional transmittance data. This calculation uses the same formula as EnergyPlus Detailed modified for arbitrarily oriented surfaces (i. As illustrated in Figure 56.e. windows and skylights) in the zone to the illuminance at each reference point. of the light transmitted by complex fenestration systems (CFS). A BTDF is a set of data for a given CFS. The analysis of a CFS within DElight is based on the characterization of the system using bi-directional transmittance distribution functions (BTDF). The twodimensional CFS “surface” and directional hemispheres are “abstract” in that they may not literally correspond to actual CFS component geometric details. At this time all reference points are assumed to be oriented on a horizontal virtual surface “facing” toward the zenith and “seeing” the hemisphere above the horizontal plane... whereas DElight calculates the total contribution from all apertures to each reference point. which gives the ratios of incident to transmitted light for a range of incoming and outgoing directions. prior to analysis by DElight. Once a specific instance of a CFS aperture has been positioned within a building. Complex Fenestration System Calculation: DElight calculates the contribution to the initial interior illuminance at each reference point.g. which must be either pre-calculated (e.DAYLIGHTING AND WINDOW CALCULATIONS DELIGHT DAYLIGHTING CALCULATIONS DElight Daylight Factor Calculation Differences from EnergyPlus Detailed Methods Initial Interior Illuminance/Luminance Calculation: DElight calculates the total initial contribution of light transmitted through all simple fenestration systems (i. varying portions of that light are transmitted at multiple outgoing directions across the interior hemisphere of the CFS. The pre-calculated or pre-measured BTDF for a CFS is independent of its final position and orientation within a building. and to the luminance at each gridded nodal patch of interior surfaces. a BTDF can be thought of as collapsing a CFS to a “black box” that is represented geometrically as a flat two-dimensional light-transmitting surface that will be treated as an aperture surface in the daylit zone description. This differs from the EnergyPlus Detailed daylighting calculations (henceforth referred to as “EnergyPlus Detailed” in two ways. Figure 56. The first is that EnergyPlus Detailed calculates initial illuminace values at reference points for each pair of reference point and aperture (window/skylight) in the zone. The second difference from EnergyPlus Detailed is that the initial luminance of interior surface nodal patches is calculated to support the inter-reflection calculation described below. Reference Points: DElight allows up to 100 reference points to be arbitrarily positioned with a daylighting zone.e. For each incoming direction across the exterior hemisphere of the CFS. the incident light from all exterior sources across the CFS 4/23/05 129 . and to calculate luminance rather than illuminance..

as does EnergyPlus Detailed. 1983. F. LBL-12599A. Jan. Daylighting Calculation in DOE-2. the remaining fraction is assumed to have no lighting control. and S. Any non-zero fraction is thus the equivalent of a relative weighting given to that reference point’s influence on the overall electric lighting control. 271-286 (1985) Daylighting Devices Daylighting devices are used to improve daylighting in a zone. DElight only calculates daylight factors for the total contribution from all windows/skylights and CFS to each defined reference point. but which do not control the electric lighting.0 for this relative weighting to make physical sense. The sum of all fractions for defined reference points must less than or equal to 1.C. Daylighting Simulation in the DOE-2 Building Energy Analysis Program. daylighting devices also have a thermal impact on the zone heat balance. reflecting surfaces. References Modest. As a result the simulation of daylighting devices is tightly integrated into both the daylighting model and the zone heat balance. The ability to input up to 100 reference points supports a more complete assessment of this distribution. Winkelmann. (1982). 4/23/05 130 . LBL-11353. Winkelmann. Visual Quality: DElight does not currently calculate a measure of visual quality such as glare due to daylighting. Selkowitz. Besides their contribution to illuminance. The algorithms for this BTDF treatment of CFS in DElight are still under development.C. Values of 0. and Lawrence Berkeley Laboratory report no.0. This method subdivides each reflecting surface in the zone into nodal patches and uses view factors between all nodal patch pairs in an iterative calculation of the total contribution of reflected light within the zone.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES exterior hemisphere can be integrated over all incident directions for each relevant transmitted direction to determine the light transmitted by the CFS surface in that direction. One or more of these reference points must be included in the control of the electric lighting system in the zone. which allows the definition of reference points for which interior illuminance levels are calculated. The light transmitted by the CFS aperture is then distributed to surfaces in the zone according to its non-uniform directionality. 66-79. M. Energy and Buildings 8. resulting in a more accurate calculation of the varied distribution of inter-reflected light throughout the zone. There are two types of daylighting device in EnergyPlus: tubular daylighting devices and daylighting shelves. A General Model for the Calculation of Daylighting in Interior Spaces.0 are valid. 1982]. DElight Time-Step Interior Daylighting Calculation Differences from EnergyPlus Detailed Methods Interior Illuminance Calculation: As discussed above. If the sum is less than 1. DElight does calculate luminance on nodal patches of all interior. This method replaces the split-flux method used in EnergyPlus Detailed. Inter-reflected Interior Illuminance/Luminance Calculation: The effect of inter-reflection of the initial interior illuminance/luminance between interior reflecting surfaces is calculated using a radiosity method derived from Superlite [Modest. Electric Lighting Control Calculation: Up to 100 reference points can be defined within a DElight daylighting zone. Energy and Buildings 5. Thus DElight does not support dynamic control of fenestration shading during the subsequent time-step calculations. A variety of visual quality metrics could be calculated from these data in future implementations. and are subject to change in the future. Lawrence Berkeley Laboratory report no. F. Each reference point input includes the fraction of the zone controlled by that point.

The pipe channels the daylight from the dome to the diffuser via multiple internal reflections. Tubular Daylighting Devices The dome is typically a hemisphere made of clear plastic. 4/23/05 131 . phenomena: • • Daylighting Solar gains • Conductive/convective gains Solar gains and conductive/convective gains are simulated by the zone heat balance. are constructed of three components: a dome. The diffuser is typically a flat frosted plastic cover.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES Tubular Daylighting Devices Tubular daylighting devices (TDDs). Together the dome and diffuser become "receiver" and "transmitter". and a diffuser. but related. Daylighting is simulated independently. a pipe. also known as tubular skylights or light pipes. radiation entering the dome ends up exiting the diffuser. In EnergyPlus the TDD model includes three different. The pipe is assumed to be a smooth cylinder with a highly reflective inside surface. It allows daylight into the pipe while keeping exterior weather out. For both daylighting and heat balance simulations. The surface is usually either bare polished metal or a special reflective sheet adhered to the inside. Figure 57.e. The diffuser evenly distributes the daylight to the zone. i. the dome and diffuser are treated as special window surfaces to take advantage of many of the standard daylighting and heat transfer routines.

e. Dome and diffuser surfaces. Visible (i. Measured reflectivities for commercial TDDs range from 0. respectively. Although the actual surface reflectivity is slightly dependent on the incident angle. Harrison 1998). daylighting) and solar absorptivities give visible and solar reflectivities. For an opaque surface. the total transmittance of the TDD is most compatible with the EnergyPlus daylighting and heat balance code. the model assumes a constant reflectivity for all angles.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES Figure 58.99. The transmittance of a discrete ray through a pipe is dependent on the reflectivity of the inside pipe surface. Calculation of the transmittance of the pipe and the TDD for different types of radiation is fundamental to all phenomena covered by the model. Figure 59. The full analytical expression for the transmittance of a beam of light in a highly reflective pipe has been developed by Swift and Smith and verified by experiment (1994). The pipe is simulated by a separate code module. the reflectivity is: ρ = 1−α where α = surface absorptivity. and the point of entry into the pipe. the incident angle of the ray.90 to 0. the aspect ratio of the pipe. Pipe Beam Transmittance The transmittance of beam radiation is derived from the integration of the transmittance of many discrete rays. Discrete ray in a pipe. By integrating over all rays incident on the pipe entrance. they find the transmittance of a beam of collimated radiation to be: 4/23/05 132 . While several different measures for characterizing TDD performance are in use (Zhang 2002.

6 R=0. A polynomial fit was also considered but it was found that interpolation gave superior results. L/D Figure 60. 0. values are calculated over a range of incident angles and stored in a table.98 FSEC R=0.0 4.98 Transmittance.5 0.0 0.0 9. 4/23/05 133 .0 2.0 Aspect Ratio.7 R=0.94 R = 0.92 EnergyPlus R=0. interpolated values from EnergyPlus are compared to the results of ray tracing simulations performed at the Florida Solar Energy Center for an incident angle of 30 degrees (McCluney 2003).0 8.0 7.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES τ= π∫ 4 1 s2 1− s 2 s =0 ρ INT [a tanθ / s] (1 − (1 − ρ ) ( a tan θ / s − INT [ a tan θ / s ]) ) ds where a = L/D. the aspect ratio of the TDD ρ = surface reflectivity θ = incident angle s = entry point This integral does not have an analytical solution and must be calculated numerically. It was found that a large number of points (100.8 R=0. In the graph below. The results are subsequently used in the daylighting simulation and heat balance simulation respectively. the program integrates and tabulates values for the visible and solar transmittance of the pipe.92 FSEC 0.9 R = 0.0 3. 1.98 EnergyPlus R = 0.0 1.94 FSEC R=0. The tabulated values are interpolated to rapidly give the transmittance at any incident angle. During initialization of each unique TDD. Pipe transmittance comparison. Since the integration is time consuming and the transmittance of the pipe must be utilized many times at every time step.92 0.0 5.94 EnergyPlus 0.0 6.000) were necessary to achieve an acceptable accuracy.

The use of a frosted diffuser on the TDD. Diffuse transmittance is always assumed for the diffuser because multiple internal reflections in the pipe scatter the beam with a diffusing effect. Solar Gains Solar radiation incident on a window is calculated separately as sun. For beam radiation the TDD beam transmittance τTDD(θ) for the solar spectrum is used as described above. the TDD model replaces the window glazing transmittance with the appropriate TDD transmittance and converts all transmitted light to diffuse. sky. it can be approximated as diffuse. however. TDD Beam Transmittance The beam transmittance of the TDD takes into account the dome and diffuser transmittances in addition to the pipe transmittance. However. 4/23/05 134 . A different transmittance must be applied for each type of radiation. Although the light exiting the pipe is not isotropic. sky. it can be approximated by the user by appropriately decreasing the transmittance of the diffuser material. and ground. The model does not take into account refraction due to the curvature of the dome surface. The illuminance due to sky radiation and ground reflected radiation is calculated with the normal daylighting model integration over the sky and ground within the viewable hemisphere of the dome. The details are discussed in the following sections. The transmittance of each sky or ground element is also found using the TDD beam transmittance at the incident angle of the sky or ground element relative to the dome. Daylighting The daylighting simulation of the TDD treats the diffuser surface as a regular window illuminated from the outside by sun. as described above τdiffuser = diffuse transmittance of the diffuser glazing The dome transmittance is calculated for a flat window. Light from the diffuser is converted to diffuse inside the zone in the same way as an interior shade. Recent research (Zhang 2002) has suggested that a 30 degree bend has a 20% loss in transmitted light. ensures that the light delivered to the zone is very close to isotropic diffuse. For sky and ground radiation a diffuse transmittance for the TDD must be developed. The illuminance due to the direct beam of the sun is found using the TDD beam transmittance τTDD(θ) as described above. τ TDD (θ ) = τ dome (θ )τ pipe (θ )τ diffuser where τdome(θ) = beam transmittance of the dome glazing at the incident angle τpipe(θ) = beam transmittance of the pipe at the incident angle. The incident angle θ is relative to the dome surface. If the effect of bends must be simulated. The transmittance of diffuse radiation can be defined as the total transmitted flux divided by the total incident flux. The calculation of TDD diffuse transmittance is considerably more complex and is handled differently in the daylighting simulation and the heat balance simulation. and ground radiation.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES The effect of bends in the pipe on beam transmittance is not included in this model.

iso ∫ = θ π /2 =0 τ (θ ) cos θ sin θ dθ =0 ∫θ π /2 cos θ sin θ dθ For a given pipe or TDD. derived below It is important to note that transmittances above are for the total TDD.aniso = ∑ ∑I I trans .aniso = where τ diff .iso once during initialization using a numerical integration.iso I inc .circumsolar + I trans .circumsolar + I inc . While the daylighting model is capable of calculating the luminance of any position in the sky. for determining the transmittance of sky radiation. The diffuse isotropic transmittance is useful. the solar code only calculates the ultimate irradiance on a surface. horiz I inc . τdiff. sky radiation has an anisotropic distribution modeled as the superposition of three simple distributions: a diffuse isotropic background.circumsolar + I inc . The program calculates τdiff.iso + I inc . As described in the Sky Radiance Model section. horiz Substituting in the appropriate transmittances: τ diff .aniso inc . The transmittance of the dome and diffuser must be included to account for their angular dependences as well. aniso = I trans . horiz I inc . but not sufficient.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES τ diff = ∑ ∑I I trans inc Swift and Smith (1994) suggest a weighted integral of the beam transmittance over the hemisphere for an arbitrary angular distribution: τ diff where ∫ = θ π /2 =0 τ (θ ) P (θ ) sin θ dθ =0 ∫θ π /2 P(θ ) sin θ dθ P(θ) = angular distribution function For isotropic diffuse radiation P(θ) is the cosine of the incident angle θ. τ diff .iso is a constant.iso + I trans . horiz τdiff.horiz = diffuse transmittance of the horizon.iso + τ (θ ) I inc . τ diff . For this reason it is not possible to integrate over an angular distribution function for sky radiance.circumsolar + τ diff .iso = diffuse isotropic transmittance τ(θ) = beam transmittance at incident angle θ of sun τdiff. a circumsolar brightening near the sun.iso + I inc . Instead the three sky distributions must be handled piecewise. 4/23/05 135 .horiz I inc . and a horizon brightening.

diff. a visible diffuse transmittance is not required for the TDD daylighting simulation. The result is a constant that is calculated once during initialization. and ground radiation is calculated as normal for a window but uses the respective transmittances for the TDD. τ diff . For reporting purposes only. the actual value of the radiance cancels out. Solar The beam transmittance of a TDD is calculated in the same way for both daylighting and solar gains. Ground radiation is assumed to be isotropic diffuse. horiz inc . sky f skymultτ TDD .DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES The beam transmittance is used as an approximation for all circumsolar radiation. τ diff . sky. If the visible and solar properties (i. The transmittance of ground radiation is the diffuse isotropic transmittance. sky ) Fsgτ TDD . diff . the reported TDD beam transmittances are equal. Since the daylighting model calculates the diffuse illuminance using a sky and ground integration of many discrete elements.diff. because the daylighting and solar models treat diffuse radiation differently.diff .sky = total horizontal diffuse solar radiation due to the sky θ = incident angle of the beam on the dome fsunlit = sunlit beam fraction of the dome area fskymult = anisotropic sky view multiplier (see AnisoSkyMult) Fsg = view from ground to dome τTDD(θ) = TDD beam transmittance τTDD. However.iso The solar flux transmitted by a TDD due to beam. sky = I h . the TDD diffuse transmittances reported for visible and solar radiation will not necessarily be equal. gnd = ( I sun cos θ + I h .e. ′′ qTDD −trans . and therefore constant across the entire horizon.iso = TDD isotropic diffuse transmittance Daylighting vs.aniso = TDD anisotropic sky transmittance τTDD.horiz ∑I = ∑I trans .beam = ( I sun cos θ ) f sunlitτ TDD (θ ) ′′ qTDD −trans . the visible diffuse transmittance is estimated concurrent with the sky and ground integration using: 4/23/05 136 . even though the properties may be the same. absorptances in the input file) are the same.iso where Isun = solar beam intensity of the sun Ih.aniso ′′ qTDD −trans . horiz ∫ = θ π /2 =−π / 2 π /2 τ (θ ) cos θ dθ cos θ dθ ∫θ =−π / 2 Since the radiance of the horizon is isotropic. gnd = τ diff . The diffuse horizon transmittance is found by integrating the beam transmittance over the arc of the horizon.

Every reflection in the pipe leaves behind some solar radiation according to the surface absorptance. The solar incident on the diffuser q"inc. reflects a measurement from outside air temperature to inside air temperature.in + Qabs .83 m in length. glazing The inward bound solar absorbed by the pipe is the difference between the solar transmitted by the dome and the solar incident on the diffuser.diffuser Adiffuser The solar transmitted by the dome q"trans. Normal exterior and interior convection and IR radiation exchange occurs for both surfaces.e. however. 5) Inward flowing absorbed solar in dome and diffuser glazing Qabs . ′′ ′′ qinc . aniso + q′′ TDD . Solar radiation is inevitably absorbed by the TDD before it reaches the zone.33 m in diameter by 1. Rays incident at a greater angle make more reflections and leave behind more absorbed solar in the pipe wall.out = ⎜ qrefl . ′′ ′′ Qabs . TDDs are treated as one entity with an effective thermal resistance (i.279 m2 K/W for a commercial TDD measuring 0.dome Adome − qinc . The temperatures are then copied to the inside face of the dome and the outside face of the diffuser. Although little research has been done on the thermal characteristics of TDDs.diffuser = qbeam τ TDD . This value. The model assumes an effective thermal resistance from outside surface temperature to inside surface temperature.diffuser (1 − τ TDD ) ⎟ Adiffuser zoneSW ⎜ ⎟ τ diffuser ⎝ ⎠ 4/23/05 137 .dome is calculated as usual for a window.iso gnd τ diffuser τ diffuser τ diffuser The outward bound solar absorbed by the pipe is given by: ⎛ ⎞ (1 − τ TDD ) + q′′ ′′ Qabs .out + Qabs .diffuser is more complicated because each component must be treated separately.in = qtrans . one experiment (Harrison 1998) reports an average effective thermal resistance of 0.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES τ diff = ∫τ TDD (θ )d Φ inc inc ∫ dΦ Conductive/Convective Gains For conductive and convective heat gain. The outside face temperature of the dome and the inside face temperature of the diffuser are calculated as usual by the outside and inside heat balances respectively. pipe = Qabs . etc. R-value) between the outside and inside surfaces.beam (θ ) τ ( Hour ) τ ′′ + qsky TDD . The total absorbed solar radiation in the TDD is the sum of the following gains: 1) Inward bound solar absorbed by multiple pipe reflections 2) Outward bound solar absorbed by multiple pipe reflections due to: 3) Reflection off of diffuser surface (inside of TDD) 4) Zone diffuse interior shortwave incident on the diffuser from lights.

Ottawa. 159-168. Proceedings of International Daylight Conference. P.. "Rating of Tubular Daylighting Devices for Visible Transmittance and Solar Heat Gain -. are constructed of up to three components: a window.dome Adome 2 + ′′ qabs . R. "Cylindrical Mirror Light Pipes". Solar Energy Materials and Solar Cells 36 (1995). January 15. pp.diffuser Adiffuser 2 All absorbed solar radiation in the TDD is distributed among the transition zones that the pipe passes through between dome and diffuser. and an outside shelf. J. diffuser − qtrans .. and ′′ ′′ ′′ ′′ qrefl . 2 (2002). but this is lost to the outside. Any exterior length of pipe also receives a proportional amount of heat. pp. X. McCurdy. FL 32922. Lighting Research & Technology 34. The inside shelf and outside shelf are both optional. Muneer. Canada. "Preliminary Evaluation of the Daylighting and Thermal Performance of Cylindrical Skylights". 1679 Clearlake Rd. D.Final Report". and Kubie. 149-169. lights. G. Cooke. an inside shelf. Swift. "A Design Guide For Performance Assessment of Solar Light-Pipes". Zhang. Cocoa. diffuser − qabs . 4/23/05 138 .. R. McCluney. G. and Smith. 205-212. The upper window interacts with the daylighting shelf but the lower window does not. All light reflected from the outside shelf also goes onto the zone ceiling.diffuser The inward flowing portion of solar absorbed in the dome and diffuser glazing is: Qabs . Florida Solar Energy Center. However. The window is divided into two window surfaces: an upper window and a lower window... 2003. References Harrison. T. S. and ambient surface reflections. or simply light shelves. G.. B. pp. if neither shelf is specifed. The inside shelf acts to reflect all transmitted light from the upper window onto the ceiling of the zone as diffuse light. Daylighting Shelves Daylighting shelves.diffuser = qinc . FSEC-CR-1385-03. The transition zone heat gain is proportional to the length of the zone. The outside shelf changes the amount of light incident on the window. 1998. except to receive shading from the outside shelf. glazing = ′′ qabs .DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES where q"zoneSW is the zone interior diffuse shortwave flux from window. the daylighting shelf object has no effect on the simulation. J.

the shelf surface is allowed to interact convectively and radiatively with the zone air and other zone surfaces. The remaining light is entirely interreflected sky-related and interreflected sun-related upgoing flux. It is assumed that no beam or downgoing flux can pass the end of the shelf regardless of the shelf's position or orientation. Regardless of the solar distribution selected in the input file. The 4/23/05 139 . The general model is similar in both cases. but the details vary. The beam radiation not absorbed is reflected throughout the zone as diffuse shortwave radiation. e. the surface area is doubled by the program during initialization. The zone interior solar distribution is modified by the inside shelf. all beam solar radiation transmitted by the upper window is incident on one side (half the doubled surface area) of the shelf surface. the upper window cannot be seen from anywhere in the zone).g.e. partition. the shelf is essentially equivalent to internal mass. Inside Shelf Heat Balance In the heat balance simulation the inside shelf is defined as an interzone heat transfer surface.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES Outside Shelf Inside Shelf Daylit Zone Window Figure 61. Daylighting Shelf Diagram Daylighting shelves are simulated separately for daylighting and the zone heat balance. Because the shelf surface has two sides that participate in the zone heat balance. Inside Shelf Daylighting The inside shelf is modeled in the daylighting simulation by converting all light transmitted by the upper window into diffuse upgoing flux. Since the inside shelf external boundary condition is required to refer to itself. it is implied that the upper window cannot contribute any direct illuminance (i. In the daylighting simulation this is accomplished by forcing all the transmitted flux to be upgoing: Φ CW = Φ Φ FW = 0 where ΦCW = upgoing flux ΦFW = downgoing flux Φ = total flux Since it is assumed that all light falls on the inside shelf. Like internal mass.

both are added directly to the zone diffuse shortwave. sky = exterior horizontal illuminance due to light from the sky θ = incident angle of the beam on the shelf fsunlit = sunlit beam fraction of the shelf surface area fskymult = anisotropic sky view multiplier (see AnisoSkyMult) ρvis = shelf surface reflectivity in the visible spectrum Fws = view factor from window to shelf τdiff.vis Φ shelf . Fws. The sunlit beam fraction fsunlit takes into account the effect of shading due to other surfaces. the shelf view factor to the sky. The luminance of any intervening exterior or interior surfaces is assumed to be zero. Outside Shelf Daylighting In the daylighting model the luminous flux transmitted by the upper window is determined by integrating over the sky and ground and summing the luminance contribution of each sky or ground element. sun + Φ shelf .0 is assumed until a better model is developed. However. The total beam.vis and Esun = exterior illuminance due to light from the sun Eh. The additional shelf upgoing flux is the sum of sun-related and sky-related flux: Φ shelf . vis = diffuse window transmittance in the visible spectrum The sunlit beam fraction fsunlit and the anisotropic sky view multiplier fskymult are borrowed from the heat balance solar calculations. sky. sky = Eh . After the integration is complete. thereby reducing the window transmitted flux due to diffuse ground luminance. The view factor to the outside shelf. For this reason the sky-related flux may be over-predicted for some building and shelf geometries. sky where Φ shelf . the flux is simply proportional to the sunlit fraction. This value is utilized in the heat balance simulation for solar calculations but is not currently available in the daylighting simulation. The anisotropic sky view multiplier fskymult takes into account the anisotropic distribution of sky radiation. for clear sky conditions the sun-related flux is dominant and the resulting error is small.CW = Φ shelf . A value of 1. and shading. Therefore. sky f skymult ρ vis Fwsτ diff . and ground radiation transmitted by the upper window does not change. Although shadows on the shelf surface change the luminance distribution of the shelf.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES treatment of sky and ground radiation is unchanged. if not specified by the user in the input object. 4/23/05 140 . there is no angular dependence because diffuse properties are assumed. the program calculates the amount of diffuse light that is reflected through the window from the outside shelf and adds it as a lump sum to the upgoing flux transmitted by the window. As a shading surface. sun = ( Esun cos θ ) f sunlit ρvis Fwsτ diff . is an exact view factor calculated for adjacent perpendicular rectangles. the effect of the outside shelf during the integration is to block part of the view of the ground.

the heat balance calculation for the outside shelf is very similar to the daylighting calculation. View factors to sky and ground are used instead. The solar flux incident on the upper window due to the shelf is given by: ′′ ′′ ′′ qshelf −inc = qshelf −inc . Window and outside shelf as adjacent perpendicular rectangles.DAYLIGHTING AND WINDOW CALCULATIONS DAYLIGHTING DEVICES Window H Shelf W L Figure 62. sky = total horizontal diffuse solar radiation due to the sky θ = incident angle of the beam on the shelf fsunlit = sunlit beam fraction of the surface area fskymult = anisotropic sky view multiplier (see AnisoSkyMult) 4/23/05 141 . The main difference is that the incident flux on the upper window is calculated first and reported. sun + qshelf −inc . The transmitted and absorbed fractions are subsequently determined. sky = I h . For this geometry the view factor is given by (Mills 1995): 1/ 2 −1/ 2 ⎧ ⎫ ⎛ 1 ⎞ ⎛1⎞ M tan −1 ⎜ ⎟ + N tan −1 ⎜ ⎟ − ( M 2 + N 2 ) tan −1 ( M 2 + N 2 ) ⎪ ⎪ ⎝M ⎠ ⎝N⎠ ⎪ ⎪ 1 ⎪ ⎪ M2 N2 ⎤ Fws = ⎡⎛ ⎨ 1 + M 2 )(1 + N 2 ) ⎞⎛ M 2 (1 + M 2 + N 2 ) ⎞ ⎛ N 2 (1 + M 2 + N 2 ) ⎞ ⎥ ⎬ πM ⎪ 1 ⎢ ( ⎟⎜ ⎟ ⎜ ⎟ ⎪ + ln ⎜ ⎪ 4 ⎢⎜ 1 + M 2 + N 2 ⎟⎜ (1 + M 2 )( M 2 + N 2 ) ⎟ ⎜ (1 + N 2 )( M 2 + N 2 ) ⎟ ⎥ ⎪ ⎠⎝ ⎠ ⎝ ⎠ ⎥⎪ ⎢⎝ ⎪ ⎣ ⎦⎭ ⎩ ( ) where M = H /W N = L /W Outside Shelf Heat Balance The heat balance simulation does not do a sky and ground integration. sky f skymult ρ sol Fws and Isun = solar beam intensity of the sun Ih. sky where ′′ qshelf −inc . Consequently. sun = ( I sun cos θ ) f sunlit ρ sol Fws ′′ qshelf −inc .

The resulting transmitted diffuse radiation from the sky.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE ρsol = shelf surface reflectivity in the solar spectrum Fws = view factor from window to shelf The view factor Fws is the same as described above for daylighting. Window Calculation Module In EnergyPlus a window is considered to be composed of the following four components. sol gnd where τdiff. With the incident radiation determined. 1993]. Shading device. The absorbed solar radiation enters the glazing heat balance calculation that determines the inside surface temperature and. which consists of one or more plane/parallel glass layers. Table 24. Frame. the description of glazing algorithms is based on material from Finlayson et al. the remaining window heat balance is calculated normally. Divider. which is a separate layer. F. [Finlayson et al. The frame and divider thermal model. If there are two or more glass layers. Optical Properties of Glazing The solar radiation transmitted by a system of glass layers and the solar radiation absorbed in each layer depends on the solar transmittance. the heat gain to the zone from the glazing (see “Window Heat Balance Calculation”). sol = diffuse window transmittance in the solar spectrum References Mills. Heat and Mass Transfer. ground. 1993. reduce heat loss (movable insulation) or control daylight glare. therefore. such as drapery. roller shade or blind. which consists of horizontal and/or vertical elements that divide the glazing into individual lites. In the following. and the shading device optical and thermal models. on the inside or outside of the glazing. whose purpose is to reduce solar gain. are new to EnergyPlus. the visible transmittance of the glazing is an important factor in the calculation of interior daylight illuminance from the glazing. p.. The glazing optical and thermal calculations are based on algorithms from the WINDOW 4 and WINDOW 5 programs [Arasteh et al. Variables in Window Calculations Mathematical variable T 4/23/05 Description Transmittance Units - FORTRAN variable 142 .. In addition. reflectance and absorptance properties of the individual layers. therefore. which surrounds the glazing on four sides. only the first of which. glazing.. The transmitted solar radiation is absorbed by interior zone surfaces and. For reporting purposes the shelf radiation is included in the Surface Ext Solar Ground Diffuse Incident output variable. and shelf is: ′′ ′′ ′′ qtrans = ( qsky −inc + q′′ −inc + qshelf −inc )τ diff . 1995. the layers are separated by gaps filled with air or another gas. 499. 1989]. contributes to the zone heat balance. A. is required to be present: Glazing. The total diffuse incident radiation due to the shelf is internally added to the ground diffuse incident radiation on the window.

ρbl.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE R R.R Ti.j Tdirgl R i. A N λ Es(λ) R(λ) φ f b i microns W/m micron 2 nlayer wle e y30 Phi τ ρ α d n κ β P. back absorptance of layer i Number of glass layers Wavelength Solar spectral irradiance function Photopic response function of the eye Angle of incidence (angle between surface normal and direction of incident beam radiation) Transmittivity or transmittance Reflectivity or reflectance Spectral absorption coefficient Glass thickness Index of refraction Extinction coefficient Intermediate variable A general property. p τsh ρsh αsh τbl. such as transmittance Shade transmittance Shade reflectance Shade absorptance Blind transmittance. R f b i.j. back reflectance Transmittance through glass layers i to j Direct transmittance of glazing Front reflectance. G.f. αbl Q. rb0 Material%Thickness ngf. reflectance. betab Material%Trans Material%AbsorpSolar - m W/m2 Fij fswitch 4/23/05 - SwitchFac 143 . back reflectance from glass layers i to j Direct front and back reflectance of glazing Front absorptance.b A i. Rdirgl. ngb betaf. irradiance and radiosity for blind optical properties calculation View factor between segments i and j Switching factor - - Rdirgl.j f b Reflectance Front reflectance. J m -1 tf0 rf0. absorptance Source.

such as transmittance Shade transmittance Shade reflectance Shade absorptance Switching factor - - A i. and “back” refers to radiant incident on the side of the glass closest to the 4/23/05 144 b .R Ti.j f b Transmittance Reflectance Front reflectance. R f f b i. R f Back reflectance. back reflectance from glass layers i to j Front absorptance.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE T R R.j. ngb betaf. the optical properties of individual glass layers are given by the following quantities at normal incidence as a function of wavelength: Transmittance. R Here “front” refers to radiation incident on the side of the glass closest to the outside environment. back reflectance Transmittance through glass layers i to j Front reflectance. back absorptance of layer i Number of glass layers Wavelength Solar spectral irradiance function Photopic response function of the eye Angle of incidence (angle between surface normal and direction of incident beam radiation) Transmittivity or transmittance Reflectivity or reflectance Spectral absorption coefficient Glass thickness Index of refraction Extinction coefficient Intermediate variable A general property. A N λ b i microns W/m micron 2 nlayer wle e y30 Phi Es(λ) R(λ) φ τ ρ Α D N Κ Β P. betab Material%Trans Material%AbsorpSolar SwitchFac m - Glass Layer Properties In EnergyPlus.j R i. T Front reflectance. rb0 Material%Thickness ngf. p τsh ρsh αsh fswitch m -1 tf0 rf0.

Glass Optical Properties Conversion Conversion from Glass Optical Properties Specified as Index of Refraction and Transmittance at Normal Incidence The optical properties of uncoated glass are sometimes specified by index of refraction. The following equations show how to convert from this set of values to the transmittance and reflectance values required by Material:WindowGlass.j .526 1/ 2 − (1 − r ) 2 ⎛ 1. Note that since the glass is uncoated.07846 2 Glazing System Properties The optical properties of a glazing system consisting of N glass layers separated by nonabsorbing gas layers (Figure 63. reflection and absorption of solar radiation within a multi-layer glazing system.j. the absorption in layer j..86156 n = 1.) are determined by solving the following f b recursion relations for Ti. interzone) walls. Each of the variables is a function of wavelength. For glazing in interior (i.e. n. T. and transmittance at normal incidence.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE inside environment. n and T are different for the solar and visible).526 − 1 ⎞ r =⎜ ⎟ ⎝ 1. the front and back reflectances are the same and equal to the R that is solved for in the following equations. the front and back reflectance.j and R i. “back” is the side closest to the zone in which the wall is defined in and “front” is the side closest to the adjacent zone. “front” is therefore the side closest to the outside air and “back” is the side closest to the zone air. find R: ⎛ n −1 ⎞ r =⎜ ⎟ ⎝ n +1 ⎠ 2 ⎡(1 − r ) 4 + 4r 2T 2 ⎤ ⎦ τ=⎣ 2 2r T 2 (1 − r ) rτ 2 R=r+ 1 − r 2τ 2 Example: T = 0. from layers i to j. For glazing in exterior walls. R i. and can be used to convert either spectral-average solar properties or spectral-average visible properties (in general.526 + 1 ⎠ τ = 0. These equations apply only to uncoated glass.93974 R = 0. and Aj . Here layer 1 is the outermost layer and layer N is the innermost layer. These relations account for multiple internal reflections within the glazing system. the transmittance through layers i to j. Schematic of transmission. Given n and T. respectively. 4/23/05 145 .

i j Ti . Schematic of transmission.1 2 f T1.1 Outside A1 Rf1.1 2 b T2.1 R =R b 2.2 f 1. j R jf+1.2 R1. for double glazing (N=2) these equations reduce to T1.1T2.1 R2.N RbN.2 R1.1 b f 1 − R1. j −1 (1 − T j .2 R =R + f 1. j ) j + f b f b 1 − R j .i R jf. j R b−1.2j −1 R jf. N Figure 63. N R j −1.i b j. j j (149) Ri . j = Ti .2 Inside . reflection and absorption of solar radiation within a multi-layer glazing system.j = 0 if i<0 or j>N.1 (152) In Eq. T1. As an example.2 = T1..DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Ti .j = 1 and Ri.i j 1 − R b−1.1 Rf1.1 R2. j R b−1. j ) T1.f j −1 + (150) R =R b j . j R b−1. j 1 − R jf.i j T j2. j −1T j . j + (151) Ajf = T1.2 + 4/23/05 146 .2 f b 1 − R2.N T1. N (1 − T j . j − R b.1 1 − R j .2 f b 1 − R2. N R j −1.2 R1.. j − R jf.f j = Ri . j 1 − R jf.1 Layer 1 2 Rb1.1 A2 T2. (152) Ti.1 b 2.

9.6.0.3100.1130.0.4100.7.0. 60.4. 3.3250.3050.5200.8. 460.0.0.9.1 − R ) + f 1 f 1. the f b wavelength-dependent values of Ti.0. 82. 21. These functions are shown in Table 25 and Table 26.0.1. 1.0.0. 666. 390.3.3800.0.5.5400.1485. 99.0.4.2.3. EnergyPlus calculates “spectral average” values of the above glazing system properties by integrating over wavelength: The spectral-average solar property is Ps = ∫ P (λ ) E (λ ) d λ ∫ E (λ ) d λ s s The spectral-average visible property is Pv = ∫ P ( λ ) E (λ ) R (λ ) d λ ∫ E (λ ) R (λ ) d λ s s where Es (λ ) is the solar spectral irradiance function and R (λ ) is the photopic response function of the eye. 646.2. 978.5.7.6. 748.4800.7.3.1.2.3500.1.5300.1 − R1.0.5. Table 25: Solar spectral irradiance function.0. 1043.2.3. 483.0.5.1484. 2. 228.4.2 ) f b 1 − R2.1030.4000.1 WINDOW CALCULATION MODULE A = (1 − T1.0.1. derived from Optics5 data file ISO-9845GlobalNorm. 42.5.0.1.9.1211. 71.1184.5. 0.0.9.1081.2. 1526. 395. 438. 132.1 A2f = f T1.1586.6.4. 313.1395.1.DAYLIGHTING AND WINDOW CALCULATIONS f b T1. 246.8.7.5. 520. 250.8.0.4300.0.0. 339. 667. R i.0.5.1539. 10-14-99.3.6.1175.0.3450. 182. 690.3200. 480.4200.0.2. 0.1131.2 R1. 712. 720.1071. 107.0.2. 70.1628. 91. 690. 1.0. 0. 244. Air mass 1.5100.1392. 435.0.3.4400. They are expressed as a set of values followed by the corresponding wavelengths for values.5.3.1434.9.6.0. If a glazing layer has optical properties that are roughly constant with wavelength.i and R i.5.1.5. 234.1013. 413. 220.0.8.1599.2. 933.1. 423. 89. 105.2. 849.2.1. 258.2.1.9.2.0.3. 785.3150.5.5.7.1 ) f b 1 − R2.0. Based on ISO 9845-1 and ASTM E 892. 1561. 275.1572.3. 30.4500. 274. 643.4.std.6.1581.5.1419.1550.7.1 R2. 4/23/05 147 . 55.2 − R2. 403.2.7. 916. 247.7.3400.0.0.5. 412.6.0.9.6.3350. 189.0.0.2.3300.1.5. 171.5 terrestrial solar global spectral irradiance values (W/m2-micron) on a 37o tilted surface.3.0.4900. 26. 244. 9.1193.9.4600. 526.0.3600.2 Wavelengths (microns) corresponding to above data block 0. 181. (149) to (152) can be replaced with constant values for that layer.2.1158.5.1501.0.1010.1302.5. 18.5.3700.1 If the above transmittance and reflectance properties are input as a function of wavelength. 108.1 (1 − T2.2.i . 637.7. 32.2 (1 − T1.3000.1548.3900.0. 62.2 R1. Corresponds to wavelengths in following data block.0. 262.0.8. 959.4700. 746.5.i in Eqs. 21.1316.5000. 105.

1. .3230..575 microns is used in the calculations.0232.0040.2.705.0004.775.0.0320.455.5370 Table 26: Photopic response function.9300.0..1126.630..1.0.1.0..2080.9480..0.6460.9950.9520.1.1.0000.2.0.0.8620.3200.0.0380.0007.0.445.7525.730.1370.1980.4970.0816. which is the same as WINDOW4 file Cie31t.0...0.0.0.405.0. 0.0.0.0015.0.0.410.0.0004.0.6700.1.2650..0.610.2586.675..0..0.700..0116.6310.0002.555.7675.0002.dat.9154.7100. .0.6500.1.2700.2900.0.0001.0006.1.1.8163.1.0.7400.1000...0000. 0.3810.685.1.0. 0.3210.1.8237.8700.2. For uncoated glass the reflectance is the same for the front and back surfaces...1480.0001.0158.605..650.690. and the reflectance.8800..9149.465.1.1.0.0029.585.9800.0022..0..0073.898 microns is used.0.485.2. .750. The spectral data include the transmittance..9850.380...9150.0.0..0.1300..515.4500.1. ISO/CIE 10527..0.0.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE 0.0600.6100. . 0.530.595.5780.. R.765. T..0.2.1693.5030.0.440.0.1. absorptivity and index of refraction are a function of wavelength.0.495..0119.0.715.420.480..1000.0.550.720.1. CIE Standard Calorimetric Observers..740. 1.620.0000.0005.0.680.5668.0001.7400.0700. 0.0. For angle of incidence.755.0446...3500.9370.660. λ.txt".7570.0650.4425.0. ρ..600. Photopic response function values corresponding to wavelengths in following data block.2000.0002.0.0001.8000.0.0.0.0.0.0.0.0...0. φ ..1.2350.390.7100.0021.770. 1.0.580.8400.1.0.. 0.0..0.5580..0.0.0610. by the following relationships: 4/23/05 148 . 0.395.1.665.1750.565..0.460.0350.. .0.425.710.4073.0.7180. .3600.0.520.745.2.0.430.7800.5500.0.640.0.4940.735.0480.8600. In the visible range an average wavelength of 0.7575.0.0.6100.0.1.1.0.1..470..5200. τ.1.0400.400.385...0.760.9050.9786.0.510.. 0.475.0.. 2.9803..545.0168.0.4625..1070.0170.525.5700.560.635...0. 0. Angular Properties for Uncoated Glass The following discussion assumes that optical quantities such transmittivity.1...2. Based on CIE 1931 observer.8160.0.0.540.9200.5390.615.2..570. 0.0. 0.6780.3950.505.8600.1.575. .0..1.0000.1.695.0.. the transmittance and reflectance are related to the transmissivity.1200..5900..2.435.5030.9540.0012.0.535.6300.1.0910.0.625.780 Calculation of Angular Properties Calculation of optical properties is divided into two categories: uncoated glass and coated glass..645.4412.725.9600.0000 / Wavelengths (microns) corresponding to above data block .9250.0.1382.0.0.9935.1.0.1800.9650.6300.0.0. and reflectivity. 1.. If there are no spectral data the angular dependence is calculated based on the single values for transmittance and reflectance in the visible and solar range.0739.8315..0001.0. derived from Optics5 data file "CIE 1931 Color Match from E308..4770.0230.0..0.0000. 0.8000.0.9950.0.490..500.1390.1.450.6082.655..590.0.6900.. In the solar range an average wavelength of 0.7932.7244. .0082.0010.1610.0298...7625.2170.0041.670.415.0...6949.2.0050.0.0...0.0.5920. reflectvity.

(155) which is then used to calculate the transmittivity in Eq. from Eq. (159) is used to calculate the absorption coefficient in Eq. ρ(0). The index of refraction is used to calculate the reflectivity in Eq. The reflectivity. (153) and (154) to obtain the angular values of the reflectance and transmittance. (153) and (154) gives the reflectivity at normal incidence: ρ (0) = where β − β 2 − 4(2 − R (0)) R(0) 2(2 − R(0)) (160) β = T (0) 2 − R(0) 2 + 2 R(0) + 1 (161) The value for the reflectivity. transmissivity and absorption coefficient are then substituted into Eqs. Solving Eq.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE τ (φ ) 2 e −α d / cosφ ′ T (φ ) = 1 − ρ (φ ) 2 e −2α d / cosφ ′ R (φ ) = ρ (φ ) (1 + T (φ )e −α d / cosφ ′ ) (153) (154) The spectral reflectivity is calculated from Fresnel’s equation assuming unpolarized incident radiation: 1 ⎛ ⎛ n cos φ − cos φ ′ ⎞ ⎛ n cos φ ′ − cos φ ⎞ ρ (φ ) = ⎜ ⎜ ⎟ +⎜ ⎟ 2 ⎜ ⎝ n cos φ + cos φ ′ ⎠ ⎝ n cos φ ′ + cos φ ⎠ ⎝ 2 2 ⎞ ⎟ ⎟ ⎠ (155) The spectral transmittivity is given by τ (φ ) = 1 − ρ (φ ) The spectral absorption coefficient is defined as (156) α = 4πκ / λ (157) where κ is the dimensionless spectrally-dependent extinction coefficient and λ is the wavelength expressed in the same units as the sample thickness. 4/23/05 149 . (154) at normal incidence gives the following expression for κ κ =− λ R(0) − ρ (0) ln 4π d ρ (0)T (0) (159) Eliminating the exponential in Eqs. The result from Eq. (160) is substituted into Eqs. (155) at normal incidence gives n= 1 + ρ (0) 1 − ρ (0) (158) Evaluating Eq. (157). (156). (158) and (159).

355 -0.460 3 4 0.0288 1.645. the angular dependence of uncoated clear glass is used.532 -0.645: T (φ ) = T (0)τ bnz (φ ) R (φ ) = R(0)(1 − ρbnz (φ )) + ρbnz (φ ) Calculation of Hemispherical Values The hemispherical value of a property is determined from the following integral: Phemispherical = 2∫ 2 P(φ ) cos(φ ) sin(φ )dφ 0 π The integral is evaluated by Simpson’s rule for property values at angles of incidence from 0 to 90 degrees in 10-degree increments.599 3. ρ clr ) and bronze glass (τ bnz .563 2.043 -2.002 0. the angular dependence of uncoated bronze glass is used.997 -2. 0 1 3.225 τ clr ρclr τ bnz ρbnz -0.999 -0.05725 -7. If the transmittance of the coated glass is > 0.513 1.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Angular Properties for Coated Glass A regression fit is used to calculate the angular properties of coated glass from properties at normal incidence. ρbnz ) are determined from a fourth-order polynomial regression: τ (φ ) = τ 0 + τ 1 cos(φ ) + τ 2 cos 2 (φ ) + τ 3 cos3 (φ ) + τ 4 cos 4 (φ ) and ρ (φ ) = ρ 0 + ρ1 cos(φ ) + ρ 2 cos 2 (φ ) + ρ3 cos3 (φ ) + ρ 4 cos 4 (φ ) − τ (φ ) The polynomial coefficients are given in Table 27. If the transmittance of the coated glass is ≤ 0.0015 0.645. The values for the angular functions for the transmittance and reflectance of both clear glass (τ clr .868 2 -3.341 6.840 2. 4/23/05 150 .862 These factors are used as follows to calculate the angular transmittance and reflectance: For T(0) > 0.645: T (φ ) = T (0)τ clr (φ ) R (φ ) = R(0)(1 − ρ clr (φ )) + ρ clr (φ ) For T(0) ≤ 0. Table 27: Polynomial coefficients used to determine angular properties of coated glass.054 0.813 -1.

In EnergyPlus. With “Switchable glazing” shading is achieved making the glazing more absorbing or more reflecting. such as solar radiation incident on the window. This means that direct radiation incident on the shade is reflected and transmitted as hemispherically uniform diffuse radiation: there is no direct component of transmitted radiation.. An example is electrochromic glazing where the application of an electrical voltage or current causes the glazing to switch from light to dark. Shades Shade/Glazing System Properties for Short-Wave Radiation Short-wave radiation includes 1) Beam solar radiation from the sun and diffuse solar radiation from the sky and ground incident on the outside of the window.” “Blinds” in EnergyPlus are slat-type devices such as venetian blinds. Tsys (φ ) = T1. and absorptance. τsh. Many types of drapery and pull-down roller devices are close to being perfect diffusers and can be categorized as “shades. the optical properties of blinds are strongly dependent on angle of incidence. reflectance. its transmittance. and 5) Short-wave radiation from electric lights incident as diffuse radiation on the inside of thewindow. Also of interest is the amount of radiation absorbed by the shading device.dif N 1 − R dif ρ sh f τ sh 4/23/05 151 . Shades and blinds can be either fixed or moveable.e. Here.” ‘blinds. “shades. are the same for the front and back of the shade and are independent of angle of incidence. Exterior Shade For an exterior shade we have the following expressions for the system transmittance. 4) Beam solar radiation from one exterior window incident on the inside of another window in the same zone. and the amount of inter-reflection between the shading device and the glazing. giving a direct component of transmitted radiation.” and “switchable glazing. they can be deployed according to a schedule or according to a trigger variable. The effect depends on the shade position (interior. Also. glazing properties in the absence of the shade). Unlike shades. The system properties are given in terms of the isolated shade properties (i. 3) Solar radiation reflected from the inside zone surfaces and incident as diffuse radiation on the inside of the window. “system” refers to the combination of glazing and shade.. taking interreflection between shade and glazing into account. usually by an electrical or chemical mechanism.e. the effective system glass layer absorptance. exterior or between-glass).” “Shades” are assumed to be perfect diffusers. αsh.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Optical Properties of Window Shading Devices Shading devices affect the system transmittance and glass layer absorptance for short-wave radiation and for long-wave (thermal) radiation. some of the direct radiation may pass between the slats. It is also assumed that the transmittance. and the system shade absorptance. shading devices are divided into three categories. 2) Beam and/or diffuse radiation reflected from exterior obstructions or the building itself. shade properties in the absence of the glazing) and the isolated glazing properties (i. depending on slat angle and the profile angle of incident direct radiation. ρsh. If moveable.

f Adiff . j = 1 to N sys α sh = α sh ⎜1 + ⎛ ⎜ ⎝ τ sh R f ⎞ ⎟ 1 − R f ρ sh ⎟ ⎠ Interior Shade The system properties when an interior shade is in place are the following.b j .b j = 1 to N 1 − R ρ sh dif b τ sh j = 1 to N sys α sh (φ ) = T1. j = 1 to N Adif . 1 − R f ρ sh T1.dif ρ sh N 1 − R f ρ sh τ sh . f j. j .b .dif j.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE dif Tsys = T1. j . j. sys = Adiff j.dif N 1 − R dif ρ sh f 1 − R f ρ sh τ sh Asys (φ ) = Adiff j.b j = 1 to N 1 − Rbdif ρ sh Adif . N Adif . j = 1 to N Adiff .b ρ sh Adif . sys = Adif j . N (φ ) j.sys = j . N (φ ) 1 − Rbdif ρ sh α sh dif α sh . sys = Adiff + T1. j. f (φ ) + T1. N (φ ) dif Tsys = T1.dif N 1 − Rbdif ρ sh τ sh 1 − Rbdif ρ sh 1 − R ρ sh dif b τ sh Asys (φ ) = Aj . τ sh . sys = T1.b ρ sh Adif . Tsys (φ ) = T1. j .dif N 1 − Rbdif ρ sh α sh Long-Wave Radiation Properties of Window Shades Long-wave radiation includes 4/23/05 152 .

or dark state. Blinds Window blinds in EnergyPlus are defined as a series of equidistant slats that are oriented horizontally or vertically. incident on the inside of the window. such as transmittance or glass layer absorptance. which can be 0 or 1. or light state.” If “schedule.” “trigger.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Thermal radiation from the sky. The switching factor. sys = ε sh ⎜ 1 + ⎛ ⎜ ⎝ lw lw τ sh ρ gl ⎞ lw lw ⎟ 1 − ρ sh ρ gl ⎟ ⎠ where lw ρ gl is the long-wave reflectance of the outermost glass surface for an exterior shade or the innermost glass surface for an interior shade.” fswitch = schedule value. The system emissivity (thermal absorptance) for an interior or exterior shade. is given by lw lw ε sh . The innermost (for interior shade) or outermost (for exterior shade) glass surface emissivity when the shade is present is lw lw ε gl . such as electrochromics. p. 1996. The following assumptions are made in calculating the blind optical properties: 4/23/05 153 . An optical property. for this state is given by p = (1 − f switch ) plight + f switch pdark where plight is the property value for the unswitched. such as equipment and electric lights. determines what state the glazing is in. All of the slats are assumed to have the same optical properties. sys = ε gl lw τ sh lw lw 1 − ρ sh ρ gl Switchable Glazing For switchable glazing. Fischer and Winkelmann. and Thermal radiation from internal sources. taking into account reflection of long-wave radiation between the glass and shade. and pdark is the property value for the fully switched. The blind optical model in EnergyPlus is based on Simmler.” or “daylighting. The program calculates how much long-wave radiation is absorbed by the shade and by the adjacent glass surface. however.” which is the angle of incidence in a plane that is perpendicular to the window plane and to the direction of the slats. fswitch. Thermal radiation from other room surfaces incident on the inside of the window. and it is assumed that the long-wave transmittance of the glass is zero. ground and exterior obstructions incident on the outside of the window. the solar and visible optical properties of the glazing can switch from a light state to a dark state. Blind properties for direct radiation are also sensitive to the “profile angle. separation and angle) and the slat optical properties (front-side and back-side transmittance and reflectance). The overall optical properties of the blind are determined by the slat geometry (width. The value of the switching factor in a particular time step depends on what type of switching control has been specified: “schedule. that document has numerous typographical errors and should be used with caution.

Any other structures that support or move the slats are ignored. Direct Transmittance of Blind The direct-to-direct and direct-to-diffuse transmittance of a blind is calculated using the slat geometry shown in Figure 64 (a). denoted by s1 through s6 (note in the following that si refers to both segment i and the length of segment i). the reflection has no specular component. This also means that absorption by the slats is hemispherically uniform with no incidence angle dependence. The goal of the blind direct transmission calculation is to determine the direct and diffuse radiation leaving the cell through s2 for unit direct radiation entering the cell through s1. In the case shown in Figure 64 (a) the cell receives radiation by reflection of the direct radiation incident on s4 and. by transmission through s3. • • Slat Optical Properties The slat optical properties used by EnergyPlus are shown in the following table.dif . which shows the side view of one of the cells of the blind. Slat Optical Properties τ dir .dif = ρ dif . ρ dir . The spectral dependence of inter-reflections between slats and glazing is ignored.dif . and b b ρ dir .dir = 0 . For the case shown. s3 and s4 are the segments illuminated by direct radiation. ρ dif . spectral-average slat optical properties are used.dif .dif f b ρ dir .. so that f b τ dir .dif f b ρ dif . ρ dir .The lengths of s1 and s2 are equal to the slat separation. f b ρ dir . ρ dir .dif . each slat is divided into two segments. The slats are perfect diffusers. It is further assumed that the slats are perfect diffusers. so that the cell is bounded by a total of six segments. Until the EnergyPlus model is improved to take into account the angle-of-incidence dependence of slat transmission and reflection. h.dif are independent of angle of incidence.dif .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE • • • The slats are flat. if the slats have non-zero transmittance.dif . Table 28.dif = τ dif . which is the distance between adjacent slat faces.e. They have a perfectly matte finish so that reflection from a slat is isotropic (hemispherically uniform) and independent of angle of incidence.dir = 0 . If the slats have holes through which support strings pass. the holes and strings are ignored.dif τ dif . Inter-reflection between the blind and wall elements near the periphery of the blind is ignored. i.dif = ρ dif . so that τ dir . the transmitted radiation is isotropic and the transmittance is independent of angle of incidence. 4/23/05 154 . which is illuminated from above.dir = 0 .dif Direct-to-diffuse transmittance (same for front and back of slat) Diffuse-to-diffuse transmittance (same for front and back of slat) Front and back direct-to-diffuse reflectance Front and back diffuse-to-diffuse reflectance It is assumed that there is no direct-to-direct transmission or reflection.dif and ρ dir . If the transmittance of a slat is non-zero. it is assumed that f f τ dir . and ρ dir .

Reflectance and Absorptance The direct-to-diffuse and transmittance and reflectance of the blind are calculated using a radiosity method that involves the following three vector quantities: Ji = the radiosity of segment si. A correction for non-zero slat thickness is described later. the total radiant flux into the cell from si Gi = the irradiance on the cell side of si Qi = the source flux from the cell side of si Based on these definitions we have the following equations that relate J.e.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Figure 64. (b) side view of a cell showing case where some of the direct solar passes between adjacent slats without touching either of them. h | w| ≤ h where w=s cos(ϕb − ϕ s ) cos ϕ s Note that we are assuming that the slat thickness is zero.. si. Direct-to-Diffuse Blind Transmittance. In this figure φs is the profile angle and φb is the slat angle. G and Q for the different segments: 4/23/05 155 . i.dir = 1 − | w| . From the geometry in this figure we see that dir τ bl . f. for the calculation of direct solar transmittance. (a) Side view of a cell formed by adjacent slats showing how the cell is divided into segments. Direct-to-Direct Blind Transmittance Figure 64 (b) shows the case where some of the direct radiation passes through the cell without hitting the slats.

dif G6 f J 6 = Q6 + τ dif .dif G5 + ρ dif .dif ∑ J j Fj 6 = Q5 j =3 j =3 f J 6 − τ dif .dif ∑ J j Fj 6 = Q6 j =3 j =3 6 6 This can be written in the form (162) where X is a 4x4 matrix and ⎡ J3 ⎤ ⎢J ⎥ J′ = ⎢ 4⎥ ⎢ J5 ⎥ ⎢ ⎥ ⎣ J6 ⎦ ⎡Q3 ⎤ ⎢Q ⎥ Q′ = ⎢ 4 ⎥ ⎢Q5 ⎥ ⎢ ⎥ ⎣Q6 ⎦ We then obtain J ′ from 4/23/05 156 . j =1 6 i = 1.dif ∑ J j Fj 3 − ρ dif .dif G4 b J 5 = Q5 + ρ dif . Using J1 = Q1 = 0 and J 2 = Q2 = 0 and combining the above equations gives the following equation set relating J and Q: b J 3 − ρ dif . Fji is the fraction of radiation leaving s j that is intercepted by si .dif ∑ J j Fj 3 − ρ dif .dif ∑ J j Fj 5 − τ dif . 6 where Fji is the view factor between s j and si .dif G3 + ρ dif .dif ∑ J j Fj 3 − τ dif .dif G5 + τ dif . i.dif G4 f J 4 = Q4 + τ dif ..dif G6 In addition we have the following equation relating G and J: Gi = ∑ J j Fji .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE J1 = Q1 J 2 = Q2 b J 3 = Q3 + ρ dif .dif ∑ J j Fj 4 = Q3 j =3 6 j =3 6 6 6 f J 4 − τ dif .e.dif G3 + τ dif .dif ∑ J j Fj 4 = Q4 j =3 6 j =3 6 b J 5 − ρ dif .

i = 1. n These constraints lead to simple equations for the view factors for n = 3 and 4. n si Fij = s j Fji . are obtained as follows. i = 1. 2h 4/23/05 157 . we have the following geometry and view factor expression: s2 s1 For n = 4 we have: s2 s3 d2 s1 Applying these to the slat cell shown in Figure 65 we have the following: d1 s4 s3 F12 = s1 + s2 − s3 2 s1 F12 = d1 + d 2 − ( s3 + s4 ) 2s1 F12 = F13 = d1 + d 2 − 2s 2h h + s3 − d3 . In such a polygon the view factors must satisfy the following constraints: ∑F j =1 n ij = 1. j = 1. n Fii = 0. Fij . i = 1. For n = 3. n. The cell we are dealing with is a convex polygon with n sides. etc.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE J ′ = X −1Q′ The view factors.

f.dif The direct-to-diffuse calculations are performed separately for solar and visible slat properties to get the corresponding solar and visible blind properties.dif − τ bl .dir − ρbl . f = 1 − τ bl . (162). The sources for the direct-to-diffuse transmittance calculation are: Q1 = Q2 = Q5 = Q6 = 0 (and therefore J1 = J 2 = 0) Q3 = τ dir . Slat cell showing geometry for calculation of view factors between the segments of the cell. f. The front direct absorptance of the blind is then dir dir dir dir α bl . Dependence on Profile Angle The direct-to-direct and direct-to-diffuse blind properties are calculated for direct radiation profile angles (see Figure 64) ranging from –90O to +90O in 5O increments.dif ⎪ ⎭ b Q3 = ρ dir . the front direct-to-diffuse transmittance and reflectance of the blind are: dir τ bl . f.dif = G2 dir ρbl .dif ⎫ π ⎪ (beam hits front of slats) ⎬ ϕb ≤ ϕ s + f 2 Q4 = ρ dir .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Figure 65. f. f.dif ⎫ π ⎪ (beam hits back of slats) ⎬ ϕb > ϕ s + 2 Q4 = τ dir .dif ⎪ ⎭ For unit incident direct flux.dif = G1 where G2 = ∑ J j Fj 2 j =3 6 6 G1 = ∑ J j Fj1 j =3 and J 3 to J 6 are given by Eq. (The “profile angle” is the angle of incidence in a plane that is perpendicular to the window and 4/23/05 158 .

e. Slat cell showing arrangement of segments and location of source for calculation of diffuse-todiffuse optical properties. (2) slat angle is determined by a schedule. f − dif . f = 1 − τ bl . We then have dif τ bl . Blind properties for sky and ground diffuse radiation For horizontal slats on a vertical window (the most common configuration) the blind diffuseto-diffuse properties will be sensitve to whether the radiation is incident upward from the ground or downward from the sky (Figure 67). visible and IR slat properties to get the corresponding solar. For front-side properties we have a unit source. All the other Qi are zero.dif = front transmittance for ground diffuse solar front transmittance for sky diffuse solar 4/23/05 159 .dif = sky τ bl . Diffuse-to-Diffuse Transmittance and Reflectance of Blind To calculate the diffuse-to-diffuse properties. f.) In the time step loop the blind properties for a particular profile angle are obtained by interpolation. f. Using this source value. i. For this reason we also calculate the following solar properties for a blind consisting of horizontal slats in a vertical plane: gnd τ bl . visible and IR blind properties. assuming uniformly distributed incident diffuse radiation.dif The back-side properties are calculated in a similar way by setting Q2 = 1 with the other Qi equal to zero.dif − ρbl . f−dif . we apply the methodology described above to obtain G2 and G1. In the time-step loop the slat angle is determined by the slat-angle control mechanism and then the blind properties at that slat angle are determined by interpolation. s3 = s4 and s5 = s6 . Three slat-angle controls are available: (1) slat angle is adjusted to just block beam solar incident on the window.dif = G2 dif ρbl . and (3) slat angle is fixed.dif = G1 dif dif dif α bl .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE perpendicular to the slat direction. Figure 66. f. Dependence on Slat Angle All blind properties are calculated for slat angles ranging from –90O to +90O in 10O increments. each slat bounding the cell is divided into two segments of equal length (Figure 66). Q1 = 1 . The diffuse-to-diffuse calculations are performed separately for solar. f..

f− dif = ∫α 0 dir bl . f− dif . This gives: π /2 sky τ bl . f − dif . f. as shown in Figure 67. treating each element as a source of direct radiation of intensity I (φs ) incident on the blind at profile angle φs .dif = WINDOW CALCULATION MODULE front reflectance for sky diffuse solar front absorptance for ground diffuse solar front absorptance for sky diffuse solar These are obtained by integrating over sky and ground elements. f.dif I sky (φs ) cos φs dφs π /2 ∫ 0 I sky (φs ) cos φs dφs π /2 sky α bl . dir bl . f− dif .dif = ∫ ⎡τ ⎣ 0 dir . f I sky (φs ) cos φs dφs π /2 ∫ 0 I sky (φs ) cos φs dφs 4/23/05 160 .dif = sky α bl .dif = ∫ 0 dir ρbl . f− dif .dif = gnd α bl . f−dif .dif (φs ) ⎤ I sky (φs ) cos φs dφs ⎦ π /2 ∫ 0 I sky (φs ) cos φs dφs π /2 sky ρbl . f − dif .dif = front reflectance for ground diffuse solar sky ρbl .DAYLIGHTING AND WINDOW CALCULATIONS gnd ρbl . f dir (φs ) + τ bl .

f. f. Side view of horizontal slats in a vertical blind showing geometry for calculating blind transmission. f.dif = π /2 ∫ 0 dir ρbl . f = ∫ 0 dir dir ⎡τ bl .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Sky diffuse radiation Sky element dφs Blind φs φs Ground diffuse radiation dφs Ground element Figure 67. f− dif . reflectance and absorptance properties to account for the amount of radiation incident on a blind that is reflected and 4/23/05 161 . This means that I sky is independent of giving: π /2 φs . τ sky − dif .dif ⎤ cos φs dφs ⎣ ⎦ sky ρbl . f− dif = π /2 ∫α 0 dir bl . dif bl . An improvement to this calculation would be to allow the sky radiance distribution to be nonuniform. We assume that the sky radiance is uniform.e.dir + τ bl . to depend on sun position and sky conditions.dif cos φs dφs sky α bl . reflection and absorption properties for sky and ground diffuse radiation. Correction Factor for Slat Thickness A correction has to be made to the blind transmittance. i. as is done in the detailed daylighting calculation (see “Sky Luminance Distributions” under “Daylight Factor Calculation”). f cos φs dφs The corresponding ground diffuse quantities are obtained by integrating φs from −π / 2 to 0..

f. The quantity of interest is the fraction. This is illustrated in Figure 68 for the case of direct radiation incident on the blind. Based on the geometry shown in Figure 68 we see that f edge = t cos(ϕ s − ξ ) t sin(ϕb − ϕ s ) t cos γ = = ⎛ ⎛ ⎛ t ⎞ t ⎞ t ⎞ ⎜ h + cos ξ ⎟ cos ϕ s ⎜ h + cos ξ ⎟ cos ϕ s ⎜ h + ⎟ cos ϕ s sin ϕb ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ The edge correction factor for diffuse incident radiation is calculated by averaging this value of fedge over profile angles. fedge. f (1 − f edge ) + f edge ρ f 4/23/05 162 . the following two equations show how blind front diffuse transmittance and reflectance calculated assuming zero slat thickness are modified by the edge correction factor. of direct radiation incident on the blind that strikes the slat edges.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE absorbed by the slat edges (the slats are assumed to be opaque to radiation striking the slat edges). φs. The slat cross-section is assumed to be rectangular. f. from –90O to +90O.dif (1 − f edge ) dif dif ρbl .dif → τ bl . ρf. f → ρbl . It is assumed that the edge transmittance is zero and that the edge reflectance is the same as the slat front reflectance. dif dif τ bl . As an example of how the edge correction factor is applied.

4/23/05 163 . except for the solar transmittances for the 10-degree slat angle case. The ISO 15099 calculation method is similar to that used in EnergyPlus. Comparison with ISO 15099 Calculation of Blind Optical Properties Table 29 compares EnergyPlus and ISO 15099 [2001] calculations of blind optical properties for a variety of profile angles. slat angles and slat optical properties.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Figure 68. except that the slats are divided into five equal segments. The ISO 15099 and EnergyPlus results agree to within 12%. This indicates that the slats should be divided into more than two segments at small slat angles. Here the transmittances are small (from 1% to about 5%) but differ by about a factor of up to two between ISO 15099 and EnergyPlus. Side view of slats showing geometry for calculation of slat edge correction factor for incident direct radiation.

890 0.636 0.729 0.0 0.216 0.294 0.019 0.155 0.012 0.216 0.418 0.053 0.273 0.310 0.416 0.298 0.567 0.0 0.090 0.053 0.309 0.375 0.057 0.057 0. front side Solar reflectance.227 0.0 0.0 0.0 0.729 0.141 0.0 0.430 0. direct to diffuse Back solar reflectance.338 0.294 0.227 0.0 0.304 0.9 0.495 0. direct to direct Back solar transmittance.730 0. second row (in italics) = EnergyPlus calculation) Front solar transmittance. hemispherical diffuse to Back solar transmittance.55 0. direct to direct Front solar transmittance.057 0.70 0.051 0.375 0. back side 0.227 0.027 0.272 0.087 0.038 0.227 0. back side Solar transmittance Solar reflectance.895 ’ 0.730 0.275 0.0 0.057 0.070 0.0 0.025 0.227 0.057 0.227 0.40 0.431 0.0 0.306 0.9 0.005 0.100 0.077 0.502 0.0 0. direct to diffuse Front solar reflectance.066 0.214 0.057 0.141 0.0 0.9 0.55 0.057 0.385 0.534 4/23/05 164 .057 0.057 0.55 0.679 0.371 0.227 0.310 0.227 0.227 0.558 0.544 0.291 0.295 0.418 0.057 0.0 0.038 0.895 0.0 0.394 0.298 0.730 0.55 0.295 0.536 0.0 0.389 0.012 0.678 0.70 0.729 0.0 0.729 0.502 0.051 0.495 0.373 0.0 0.9 0.0 0.057 0.073 0.291 0.534 0.090 0.012 0.012 0.012 0.50 Solar Profile angle (deg) 0 60 0 60 0 60 0 60 0 60 Calculated blind properties (first row = ISO 15099 calculation.0 0.227 0.363 0.295 0.332 0. front side IR emissivity.016 10 0.0245 0.50 0.006 0.048 0.016 45 0.085 0.70 0.293 0.9 0.100 0.057 0. direct to diffuse Back solar transmittance.338 0.057 0.012 0.9 0.0245 0.70 0.0 0.309 0.373 0.568 0.057 0.074 0.309 0.019 0.057 0.057 0.096 0.076 0.104 0.273 0.016 45 0.416 0.011 0.310 0.155 0.284 0.558 0.016 45 0. EnergyPlus values that differ by more than 12% from ISO 15099 values are shown in bold italics. Slat properties Separation (m) Width (m) Angle (deg) IR transmittance IR emissivity. hemispherical diffuse to diffuse Front hemispherical IR transmittance Back hemispherical IR transmittance Front hemispherical IR emissivity Back hemispherical IR emissivity 0.269 0.368 0.227 0.4 0.356 0.288 0.074 0.389 0.227 0.387 0.275 0.016 45 0.293 0.622 0. Comparison of blind optical properties calculated with the EnergyPlus and ISO 15099 methods.9 0.890 0.729 0.271 0.309 0.0 0.103 0. direct to diffuse Front solar transmittance.295 0.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Table 29.047 0.0 0.4 0.730 0.385 0.115 0.387 0.730 0.536 0.332 0.40 0.214 0.394 0.052 0.729 0.546 0.025 0.730 0.088 0.9 0.310 0.277 0.

f.b 165 . φs ) = Tgl (φ ) ⎜τ bl . f − dif Agl .b ρbl . dif f ..dif (φs ) Rgl . f Rgl .b ⎞ ⎟ ⎟ ⎠ A dir . j .dif gnd dif 1 − ρbl . f Rgl .b dif gnd Tgl τ bl . j . f − dif Rgl .bα bl .b ⎝ dir dif dir Tgl (φ )α gl .b gnd dif 1 − ρbl . f (φ . the system glass layer absorptance.b ⎝ dir gl dif dif Tgl τ bl . dif . j . sys bl . f Rgl .dif sky dif 1 − ρbl . f Agl . sys = dif sky Tgl τ bl . j−. f dif dif 1 − ρbl .dir (φs ) + τ bl . f − dif .dif dif dif 1 − ρbl . f− dif Agl . blind properties in the absence of the glazing)— indicated by “bl” —and the isolated glazing properties (i. f (φs ) dir dif 1 − ρbl .b dif dif 1 − ρbl . j . f Rgl . j . f− dif . f− dif Rgl . sys = Agl . j = 1. j = 1. f Rgl . dif f . φs ) = A dir gl . f (φs ) Rgl . j . N sky dif Agl .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE Blind/Glazing System Properties for Short-Wave Radiation When a blind is in place we have the following expressions for the system transmittance. sys bl .b dir dir T fdir . f = dif dif Tgl α bl . f− dif Rgl . glazing properties in the absence of the blind)—indicated by “gl.” Interior Blind The system properties when an interior blind is in place are the following: dif dir dif ⎛ dir τ bl . j . f + . f =A dif gl . j . f ρbl . f =A dif gl . f (φ ) + .e. f. f. N α 4/23/05 dif . N A gnd − dif . sys gl .b dif dif dif Tgl ρbl ..b T gnd − dif . sys gl . sys gl .b ⎞ ⎟ ⎟ ⎠ T dif .” are given in terms of the isolated blind properties (i. indicated by “sys. j . j = 1. taking inter-reflection between blind and glazing into account. f + dif gnd dif Tgl ρbl . j . f (φ . The system properties.e. dif f . f + f dif sky dif Tgl ρbl . f − dif Rgl .dif (φs ) + .b . and the system blind absorptance. f (φs ) + dif dif ⎜ 1 − ρbl . sys dif dif ⎜ 1 − ρbl . f (φs ) Rgl . f.b sky dif 1 − ρbl . φs ) = T (φ ) ⎜ α bl .b . sys = A dif . N α dir .all (φ . sys = T sky − dif . j = 1. f dir dif dif ⎛ dir ρbl . j .

f.b WINDOW CALCULATION MODULE α sky − dif .b T dif . N A gnd − dif . f (φ )τ bl . sys gl . dif f .dif (φs ) Agl . f ρ dif bl .dif Agl .dif (φs ) ⎜ Tgl (φ ) + dif dif ⎜ 1 − Rgl . j. f. f ρbl . f − dif .b ⎠ dir dir dir Agl . f ρbl .dir (φs ) ρbl . f = .dir (φs ) Agl . j . φs ) = τ dir . j = 1. sys gl . sys T gnd − dif . f−dif . f ρbl .b. sys T sky − dif .b sky dif τ bl .b . f − dif . f. f. f dir dir dir dif (φs ) Rgl (φ ) ρbl . j .b j = 1.dif Agl .dif Agl . f dif dif τ bl .DAYLIGHTING AND WINDOW CALCULATIONS dif sky Tgl α bl . f ρbl . dif f . f ρbl . f = α gnd − dif . f 1 − ρbl . dif dif 1 − Rgl . f− dif Rgl .b ⎝ dir dif ⎞ τ bl .b gnd dif τ bl . dir bl .dif (φs ) ) dif dif ( gl . f = . dif dif 1 − Rgl . f dir dir dir R dir (φ )τ bl . f ρbl .b (φs ) Rgl . f (φs ) + α bl . j . f. f = .b (φs ) + τ bl . f ) 1− R dif gl . sys bl . f (φ ) + . f. sys gl . f. φs ) = α bl .dif (φs )Tgl ⎟+ dif dif ⎟ 1 − Rgl . sys A dif .dif Tgl = dif dif 1 − Rgl . sys (φ . N dir dir dir dir dir α bl . dif f . N A sky − dif . f = Exterior Blind The system properties when an exterior blind is in place are the following: dir .b (φs ) + τ bl . sys T (φ . j .b Rgl . f ρbl . sys bl . f ρbl . sys (φ .dif Tgl = dif dif 1 − Rgl .b Rgl . f− dif . f− dif sky dif 1 − ρbl . φs ) = τ bl . N 4/23/05 166 . all f . f − dif Rgl . j .b dif gnd Tgl α bl .f (τ dir . dif dif 1 − Rgl . f dif dif τ bl . f sky dif τ bl . f − dif gnd dif 1 − ρbl . j . f. f dif dir dir ⎛ dir Tgl Rgl . j . f ρbl .b j = 1.dif Tgl = dif dif 1 − Rgl . dir bl .b gnd dif τ bl .dir (φs ) + dif dif α bl . f j = 1. j .

f dif α bl .eff = ε bl . The effective emissivity (long-wave absorptance) of an interior or exterior blind. f dif + α bl .eff T Solar Radiation Transmitted and Absorbed by a Window/Blind System Let the direct solar incident on the window be I dir .eff The effective temperature of the blind/glazing combination that is used to calculate the window’s contribution to the zone’s mean radiant temperature (MRT) is given by eff lw lw ε bl . f =α dif bl .b dif dif 1 − Rgl .DAYLIGHTING AND WINDOW CALCULATIONS dif dif τ bl . f.b sky dif τ bl . is given by ε lw . f− dif + α gnd − dif .dif Rgl . f ρbl .norm cos φ 4/23/05 (W / m 2 ) 167 .b Blind/Glazing System Properties for Long-Wave Radiation The program calculates how much long-wave radiation is absorbed by the blind and by the adjacent glass surface. taking into account reflection of long-wave radiation between the glass and blind.b dif dif 1 − Rgl .inc = f sunlit I dir . f =α gnd − dif bl . sys bl . f ρbl . sys bl .eff Tbl + ε gl . f− dif .eff + ε gl . and it is assumed that the long-wave transmittance of the glass is zero. sys = α bl . f − dif . f gnd dif τ bl .eff Tgl = lw lw ε bl .eff bl lw lw ⎛ τ bl ρ gl ⎞ = ε ⎜1 + lw lw ⎟ ⎜ 1 − ρbl ρ gl ⎟ ⎝ ⎠ lw bl where lw ρ gl is the long-wave reflectance of the outermost glass surface for an exterior blind or the innermost glass surface for an interior blind. f− dif . The effective innermost (for interior blind) or outermost (for exterior blind) glass surface emissivity when the blind is present is lw lw ε gl . f sky sky α bl .dif Rgl . f ρbl .eff = ε gl lw τ bl lw lw 1 − ρbl ρ gl The effective inside surface emissivity is the sum of the effective blind and effective glass emissivities: lw lw lw ε ins.b WINDOW CALCULATION MODULE α dif . f dif + α bl .b dif dif 1 − Rgl .eff + ε gl .dif Rgl .

I dir . sys Diffuse solar absorbed by blind: dif ( I sky .sys . f.dir (φ . N f For windows of tilt angle γ whose blinds have horizontal slats: (vertical windows have tilt = 90O. j.f j = 1.inc be the irradiance on the window due to diffuse solar radiation from the sky (W/m2) and let I gnd .inc be the irradiance on the window due to diffuse solar radiation from the ground (W/m2). j = 1. all in W/m2: Direct solar entering zone from incident direct solar: I dir . . j..incT fdir . and φ is the angle of incidence. φs ) Direct solar absorbed by glass layers: dir I dir . Then we have the following expressions for different classes of transmitted and absorbed solar radiation for the window/blind system (where φs is the direct solar profile angle). sys (φ . sys (φ . φs ) . norm is the direct normal solar irradiance.inc + I gnd .inc + I gnd .inc )α bl . where Let I sky .inc + I gnd .inc ) Agl . φs ).dif .incT fdir . φs ) .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW CALCULATION MODULE f sunlit is the fraction of the window that is sunlit (determined by the shadowing calculations). N For windows whose blinds have vertical slats: Diffuse solar entering zone from incident diffuse solar: ( I sky .inc )T fdif . horizontal windows have tilt = 0O) Diffuse solar entering zone from incident diffuse solar: 4/23/05 168 .dif (φ . sys Diffuse solar absorbed by glass layers: dif ( I sky . sys Diffuse solar entering zone from incident direct solar: I dir .incα bl . f. sys Direct solar absorbed by blind: dir I dir .inc Agl .

inside air film convective conductance Conductance of gap j Outside and inside air temperature Exterior. Ti Eo. tin outir. sys . Fortran Variables used in Window Heat Balance Calculations Mathematical variable N σ εi ki h o.inc + ⎜1 − ⎟ I sky .inc + 2 ⎠ ⎦ j = 1.inc + ⎟ I gnd .inc + ⎜1 − . interior long-wave radiation incident on window Temperature of face i Radiation (short-wave. rmir thetas AbsRadGlassFace I bm dif W/m2 W/m2 W/m2 BeamSolarRad QS I ext int I sw 4/23/05 169 . sys ⎟ I sky .inc ⎥ 2 ⎠ 2 2 ⎠ ⎝ ⎣⎝ ⎦ ⎣ 2 ⎦ Diffuse solar absorbed by blind: sky α bl .inc ⎥ + ⎟ I sky .inc + ⎜ 1 − ⎟ I gnd . Ei θi Si ext Description Number of glass layers Stefan-Boltzmann constant Emissivity of face i Conductance of glass layer i Outside. hcout hgap tout. f− dif .inc ⎥ + T fgnd − dif . and long-wave from zone internal sources) absorbed by face i Exterior beam normal solar intensity Exterior diffuse solar intensity on glazing Interior short-wave radiation (from lights and from reflected diffuse solar) incident on glazing from inside Units W/m -K W/m -K W/m2-K K W/m2 K W/m 2 2 2 FORTRAN variable nlayer sigma emis scon hcout. sys ⎢⎜1 − ⎡⎛ ⎣⎝ ⎤ ⎡ | cos γ | ⎤ | cos γ | ⎞ | cos γ | ⎛ | cos γ | ⎞ gnd I gnd . f −dif . dif .inc ⎥ .inc ⎥ 2 ⎠ 2 2 ⎠ ⎝ ⎦ ⎣ 2 ⎦ ⎡⎛ | cos γ sky Agl .−fdif . 2 ⎠ ⎝ ⎣ 2 ⎦ Window Heat Balance Calculation Table 30.dif ⎢ I sky . j .dif ⎢⎜1 − I gnd . j−. sys ⎢⎜1 − f 2 ⎣⎝ ⎤ |⎞ | cos γ | I gnd . sys ⎢ I sky .inc ⎥ + α bl .inc + ⎟ I gnd . h i hj To. sys ⎢ I sky .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION ⎡⎛ | cos γ | ⎞ ⎤ ⎡ | cos γ | ⎤ | cos γ | ⎛ | cos γ | ⎞ T fsky − dif . N Diffuse solar absorbed by glass layers: ⎡ | cos γ | ⎤ ⎛ | cos γ | ⎞ gnd Agl .

dif W/m2 QL φ Aj Aj f. W/m2-K W/m2-K K AbsDiff.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION I int lw Long-wave radiation from lights and equipment incident on glazing from inside Angle of incidence Front beam solar absorptance of glass layer j b. See “Edge of Glass Corrections. 2) The heat flow is perpendicular to the glass faces and is one dimensional. for adjustments to the gap conduction in multi-pane glazing to 4/23/05 170 . Glass layer #1 Outside To Eo ho S1 ε1 k1 S2 ε2 h1 S3 ε3 θ1 θ2 θ3 Glass layer #2 Inside θ4 Ti Ei k2 hi S4 ε4 Figure 69. AbsDiffBack Aface. Aj Front and back diffuse solar absorptance of glass layer j Matrices used to solve glazing heat balance equations Radiative conductance for face i Difference in temperature of face i between successive iterations A.dif f radians W/m2. there are no heat capacity terms in the equations. B hr. Glazing system with two glass layers showing variables used in heat balance equations. therefore.” below. For a window with N glass layers there are 2N faces and therefore 2N equations to solve. Figure 69 shows the variables used for double glazing (N=2). Bface . The following assumptions are made in deriving the heat balance equations: 1) The glass layers are thin enough (a few millimeters) that heat storage in the glass can be neglected.i ∆θi hr(i) - The Glazing Heat Balance Equations The window glass face temperatures are determined by solving the heat balance equations on each face every time step.

For N glass layers Si is given by S 2 j −1 = S2 j = 1 ext ext int I bm cos φ Ajf (φ ) + I dif Ajf . 5) The short wave radiation absorbed in a glass layer can be apportioned equally to the two faces of the layer. so for the purposes of the heat balance calculation it is split equally between the two faces of a layer. Glass layers are assumed to be opaque to IR so that the thermal radiation from lights and equipment is assigned only to the inside (room-side) face of the inside glass layer. 3) The glass layers are opaque to IR. so the heat balance equations would have to be modified to handle this case. Short-wave radiation (solar and short-wave from lights) is assumed to be absorbed uniformly along a glass layer. 4) The glass faces are isothermal. Eoε1 − ε1σθ14 + k1 (θ 2 − θ1 ) + ho (To − θ1 ) + S1 = 0 k1 (θ1 − θ 2 ) + h1 (θ3 − θ 2 ) + σ h1 (θ 2 − θ 3 ) + k2 (θ 4 − θ3 ) + σ (163) ε 2ε 3 (θ34 − θ 24 ) + S2 = 0 1 − (1 − ε 2 )(1 − ε 3 ) ε 2ε 3 (θ 24 − θ34 ) + S3 = 0 1 − (1 − ε 2 )(1 − ε 3 ) (164) (165) Eiε 4 − ε 4σθ 44 + k2 (θ3 − θ 4 ) + hi (Ti − θ 4 ) + S4 = 0 Absorbed Radiation (166) Si in Equations (163) to (166) is the radiation (short-wave and long-wave from zone lights and th equipment) absorbed on the i face.dif + I sw Ab . dif . This is generally a good assumption since glass conductivity is very high. The equations for single glazing (N=1) and for N=3 and N=4 are analogous and are not shown. The four equations for double-glazing are as follows. This is true for most glass products. j 2 ( ) j = 1 to N int S 2 N = S2 N + ε 2 N I lw Here ext I bm = exterior beam normal solar intensity ext I dif = exterior diffuse solar incident on glazing from outside int I sw = interior short-wave radiation (from lights and from reflected diffuse solar) incident on glazing from inside int I lw = long-wave radiation from lights and equipment incident on glazing from inside ε 2 N = emissivity (thermal absorptance) of the room-side face of the inside glass layer 4/23/05 171 .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION account for 2-D conduction effects across the pane separators at the boundaries of the glazing. For thin plastic suspended films this is not a good assumption.

The value of the inside face temperature.i 6) Find the solution θ = A−1 B by LU decomposition 7) Perform relaxation on the the new 8) Go to step 4 θi : θi → (θi + θ prev . ∆θi .i = θi evaluate the radiative conductances hr .2K. For the first time step of a design day or run period the initial values are estimated by treating the layers as a simple RC network. θ 2 N . Edge-Of-Glass Effects Table 31. Equation (163) becomes 3 Eoε1 − hr .1θ1 + k1 (θ 2 − θ1 ) + ho (To − θ1 ) + S1 = 0 2) Write the equations in the matrix form Aθ = B 3) Use previous time step’s values of θi as intitial values for the current time step. If the test still fails the program stops and an error message is issued. the tolerance is increased to 0. 4) Save the 5) Using the θi θi for use next iteration: θ prev . the test is 1 2N ∑ | ∆θ i =1 2N i | < 0.02 K If this test does not pass after 100 iterations.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Solving the Glazing Heat Balance Equations The equations are solved as follows: 1) Linearize the equations by defining hr . Currently.i = ε iσθ i . Fortran Variables used in Edge of Glass calculations Mathematical variable Description Area-weighted net conductance of glazing including edge-ofglass effects Area of center-of-glass region Area of frame edge region Area of divider edge region Total glazing area Conductance of center-of-glass region (without air films) Units W/m2-K FORTRAN variable h Acg Afe Ade Atot hcg m2 m 2 2 CenterGlArea FrameEdgeArea DividerEdgeArea Surface%Area - m m2 W/m -K 2 4/23/05 172 .i ) / 2 θi in successive Repeat steps 4 to 8 until the difference. between values of the iterations is less than some tolerance value. determined in this way participates in the zone heat balance solution (see External Convection) and thermal comfort calculation (see Thermal Comfort (Heat Balance)). For example.

is higher than it is in the center of the glass.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION hfe hde hck hrk η α Conductance of frame edge region (without air films) Conductance of divider edge region (without air films) Convective conductance of gap k Radiative conductance of gap k Area ratio Conductance ratio W/m2-K W/m2-K W/m2-K W/m2-K - FrEdgeToCenterGlCondRatio. where the spacers are located. the conductance of the glazing near the frame and divider. DivEdgeToCenterGlCondRatio Because of thermal bridging across the spacer separating the glass layers in multi-pane glazing. The area-weighted net conductance (without inside and outside air films) of the glazing in this case can be written h = ( Acg hcg + Afe h fe + Ade hde ) / Atot where (167) hcg = conductance of center-of-glass region (without air films) hfe = conductance of frame edge region (without air films) hde = conductance of divider edge region (without air films) Acg = area of center-of-glass region Afe = area of frame edge region Ade = area of divider edge region Atot = total glazing area = Acg + Afe + Ade The different regions are shown in Figure 70: 4/23/05 173 .

In the EnergyPlus glazing heat balance calculation effective gap convective conductances are used to account for the edge-of-glass effects.2 + hc .2 ) −1 ( −1 −1 ) where 4/23/05 174 . gap width. the center-of-glass conductance (without inside and outside air films) can be written as hcg = ( hr . They depend on the glazing construction as well as the spacer type. Neglecting the very small resistance of the glass layers. The approach for other numbers of gaps is analogous. Equation (167) can be rewritten as h = hcg (ηcg + α feη fe + α deη de ) where (168) ηcg = Acg / Atot η fe = Afe / Atot η de = Ade / Atot α fe = h fe / hcg α de = hde / hcg The conductance ratios α fe and α de are user inputs obtained from Window 5.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Center-of-glass region Frame-edge region Divider-edge region Frame Divider Figure 70: Different types of glass regions. and frame and divider type.1 ) + ( hr . These effective conductances are determined as follows for the case with two gaps (triple glazing).1 + hc .

(169) and (170) we obtain hr .k = convective conductance of the kth gap hr . ε j = emissivity of the faces bounding the gap θi . f 2 = 0 if τ sh = 0 Otherwise 4/23/05 175 . The following algorithm is used to apportion the absorbed short-wave radiation to the two faces of the layer.1 + h c .k = radiative conductance of the kth gap ε iε j 3 1 = σ (θi + θ j ) 2 1 − (1 − ε i ) (1 − ε j ) ε i .1 (( ) ( −1 + hr . f1 = 1. k = (ηcg + α feη fe + α deη de ) ( hr .2 + hc .2 ) −1 ( −1 −1 ) (169) We can also write h in terms of effective convective conductances of the gaps as h = hr .θ j = temperature of faces bounding the gap (K) Equation (168) then becomes h = (ηcg + α feη fe + α deη de ) ( hr .2 ) −1 −1 ) (170) Comparing Eqs.1 + hc .k η fe (α fe − 1) + η de (α de − 1) + hc .k + hc .k ) Using ηcg = 1 − η fe − ηde gives hc . Here f1 is the fraction assigned to the face closest to the incident radiation and f2 is the fraction assigned to the face furthest from the incident radiation.2 + h c .k 1 + η fe (α fe − 1) + η de (α de − 1) ( ) ( ) This is the expression used by EnergyPlus for the gap convective conductance when a frame or divider is present. k = hr .1 ) + ( hr .k + hc . Apportioning of Absorbed Short-Wave Radiation in Shading Device Layers If a shading device has a non-zero short-wave transmittance then absorption takes place throughout the shading device layer.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION hc .

abs Description IR from the exterior surround absorbed by outside frame surfaces IR emitted by outside frame surfaces Convection from outside air to outside frame surfaces Conduction through frame from inside frame surfaces to outside frame surfaces Solar radiation plus outside glass IR absorbed by outside of frame Diffuse solar absorbed by outside frame surfaces. f 2 = 0 if α sh > 0. f 2 = 0 if α sh ≤ 0.bm cos(βface) CosIncAng 4/23/05 176 .emitted Qconv Qcond W W W - Qabs dif W - Q abs.01 f1 = 1.sol W/ m2 - I I dif ext ext W/ m2 W/ m2 - FrameSolAbsorp bm fr sol f. per unit frame face area Beam solar absorbed by outside frame surfaces.sol W/ m2 - Q bm abs.01 < α sh ≤ 0. Fortran Variables used in Window/Frame and Divider calculations Mathematical variable QExtIR. These temperatures are determined by solving the heat balance equations on the inside and outside surfaces of the frame and divider.999 ⎪ ⎪ ⎭ Window Frame and Divider Calculation For the zone heat balance calculation the inside surface temperature of the frame and that of the divider are needed.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION f1 = 0. per unit frame face area Diffuse solar incident on window Direct normal solar irradiance Solar absorptance of frame Front diffuse solar reflectance of glazing Front beam solar reflectance of glazing Cosine of angle of incidence of beam solar on frame outside face Units W FORTRAN variable QIR.dif Rgl α Rgl f.999 f1 = 1− e2 1 ln(1−α sh ) α sh f 2 = 1 − f1 ⎫ ⎪ ⎪ ⎬ 0. Table 32.

form factor of frame inside projection for interior IR Exterior IR incident on window plane Interior IR incident on window plane Outside.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Cos(βh) Cosine of angle of incidence of beam solar on frame projection parallel to window x-axis Cosine of angle of incidence of beam solar on frame projection parallel to window y-axis Fraction of window that is sunlit Area of frame’s outside face (same as area of frame’s inside face) Area of frame’s outside and inside projection faces Form factor of frame’s outside or inside face for IR Form factor of frame outside projection for exterior IR. hi. ε2 θ1.c. Fp2 m2 - - Eo Ei ε1. FrameCon S1 S2 W/m2-K W/m -K 2 FrameQRadOutAbs FrameQRadInAbs η1. Ap2 Ff Fp1. DividerWidth 177 . θ2 To. Ti ho.c k W/m2 W/m2 K K W/m2-K W/m2-K outir SurroundIRfromParentZone FrameEmis FrameTempSurfOut. divider width - CosIncAngHorProj Cos(βv) - CosIncAngVertProj fsunlit Af m2 SunlitFrac - Ap1. Ap2/Af Height of glazed portion of window Width of glazed portion of window Frame width. w d 4/23/05 m m m Surface%Height Surface%Width FrameWidth. FrameTempSurfIn tout. inside surface temperature Outside and inside air temperature Frame outside and inside air film convectinve conductance Effective inside-surface to outside-surface conductance of frame per unit area of frame projected onto window plane Qabs/Af Interior short-wave radiation plus interior IR from internal sources absorbed by inside of frame divided by Af Ap1/Af. inside frame surface emissivity Frame outside. tin HOutConv. HInConv FrameConductance. η2 H W w f.

The frame and divider profiles are approximated as rectangular since this simplifies calculating heat gains and losses (see “Error Due to Assuming a Rectangular Profile.” below). The outside and inside frame and divider surfaces are assumed to be isothermal.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION pf1. Cross section through a window showing frame and divider (exaggerated horizontally).r.r A C B. pf2 N h. vertical dividers Frame outside. N v To. inside surface radiative conductance Intermediate variable in frame heat balance solution Intermediate variable in frame heat balance solution Intermediate variables in frame heat balance solution m K W/m2-K K - FrameProjectionOut. D Frame outside. HInRad Afac Efac Bfac. TInRadFr HOutRad. hi.r. Wall Frame Frame outside projection Frame width Frame inside projection Double-pane glazing Divider Divider outside projection Divider width Divider inside projection Spacer Frame Wall Figure 71. inside radiative temperature Frame outside. Dfac Frame Temperature Calculation Figure 71 shows a cross section through a window showing frame and divider. VertDividers TOutRadFr. inside projection Number of horizontal.r ho. Frame Outside Surface Heat Balance The outside surface heat balance equation is 4/23/05 178 . Ti. FrameProjectionIn HorDividers.

abs − QIR . per unit area of frame projected onto the plane of the window. which gives QExtIR .abs = ε1 Eo Af Ff + ε1 Eo Ap1 Fp1 where ε1 is the outside surface emissivity. is the sum of the IR from the sky.” below). Eo.emitted = IR emitted by outside frame surfaces Qconv = convection from outside air to outside frame surfaces Qcond = conduction through frame from inside frame surfaces to outside frame surfaces Qabs = solar radiation (from sun.0 and Fp1 = 0.5.c (1 + η1 )(T0 − θ1 ) + k (θ 2 − θ1 ) + S1 = 0 ⎝ 2 ⎠ where S1 = Qabs/Af and (171) 4/23/05 179 .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION QExtIR .emitted + Qconv + Qcond + Qabs = 0 where QExtIR . ground and obstructions.abs = ε1 Eo ( Af + 1 Ap1 ) 2 The IR emitted by the outside frame surfaces is QExtIR . The exterior IR incident on the plane of the window. The first term can be written as the sum of the exterior IR absorbed by the outside face of the frame and the exterior IR absorbed by the frame’s outside projection surfaces. abs = IR from the exterior surround (sky and ground) absorbed by outside frame surfaces QIR . Ff = 1. by definition. For the purposes of the frame heat balance calculation it is assumed to be isotropic.emitted = ε1σ ( Af + Ap1 )θ14 The convective heat flow from the outside air to the outside frame surfaces is Qconv = ho .c ( Af + Ap1 )(To − θ1 ) The conduction through the frame from inside to outside is Qcond = kAf (θ 2 − θ1 ) Note that Af is used here since the conductance. For isotropic incident IR. sky and ground) plus IR from outside window surface absorbed by outside frame surfaces (see “Calculation of Absorbed Solar Radiation. Adding these expressions for the Q terms and dividing by Af gives ⎛ 1 ⎞ E0ε1 ⎜ 1 + η1 ⎟ − ε1 (1 + η1 ) θ14 + ho . QExtIR . is. k.

(171) as follows.r + θ1 )3 2 (To . within a few percent.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION η1 = Ap1 ⎛ p ⎞ H + W − ( N h + N v ) wd = ⎜ f .1 ⎟ ⎜ Af ⎝ w f ⎟ H + W + 2w f ⎠ We linearize Eq.r + θ1 )3 2 ho .r − θ1 ) Defining an outside surface radiative conductance as follows ho . r (To .r = [ Eo (1 + η1 ) /(1 + η1 )]1/ 4 2 This gives ε1 (1 + η1 )[To4.c (1 + η2 )(Ti − θ 2 ) + k (θ1 − θ 2 ) + S 2 = 0 where (173) 1 Ti .r = ε1 (1 + η1 ) then gives (To .r − θ 2 ) + hi .c (1 + η1 )(To − θ1 ) + k (θ 2 − θ1 ) + S1 = 0 Frame Inside Surface Heat Balance (172) A similar approach can be used to obtain the following linearized inside surface heat balance equation: hi .r − θ1 ) + ho .r = [ Ei (1 + η 2 ) /(1 + η2 )]1/ 4 2 4/23/05 180 .r − θ14 ] which. r − θ1 ) The final outside surface heat balance equation in linearized form is then ho . r (To . r (Ti . Write the first two terms as ε1 (1 + η1 )[ Eo (1 + η1 ) /(1 + η1 ) − θ14 ] and define a radiative temperature 1 2 1 To . equals ε1 (1 + η1 ) (To .

DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION η2 = Ap 2 ⎛ p ⎞ H + W − ( N h + N v ) wd = ⎜ f .cTi + S 2 hi .rTi . sol ⎜ 1 + 0. per unit frame face area is bm face bm Qabs. dif dif f I ext → I ext 1 + Rgl ..r + ho .dif dif ( ) The beam solar absorbed by the outside face of the frame. rTo .2 ⎟ Af ⎜ w f ⎟ H + W + 2w f ⎝ ⎠ and Ei is the interior IR intensity incident on the plane of the window. the absorbed diffuse solar per unit frame face area is ⎛ Ap1 ⎞ dif dif dif Qabs ..r + k + hi . r + k + ho .c B= C= D= hi ..c Calculation of Solar Radiation Absorbed by Frame The frame outside face and outside projections and inside projections absorb beam solar radiation (if sunlight is striking the window) and diffuse solar radiation from the sky and ground.c k hi . sol cos β face f sunlit The beam solar absorbed by the frame outside projection parallel to the window x-axis is bm bm Qabs.cTo + S1 ho . sol = I ext α fr . For the outside surfaces of the frame. sol cos β h p f 1 (W − N v wd ) f sunlit / Af sol 4/23/05 181 . sol ( Af + Fp1 Ap1 ) / Af = I ext α fr .r + k + ho . i.r + hi .e. sol = I ext α fr .h = I ext α fr .5 ⎟ ⎜ Af ⎟ ⎝ ⎠ If there is no exterior window shade.c k ho .r + k + hi . (172) and (173) simultaneously gives θ2 = with D + CA 1 − CB A= ho . I ext includes the effect of diffuse solar reflecting off of the glazing onto the outside frame projection. Solving Eqs.

for a circular profile vs. Note that at any given time beam solar can strike only one of the two projection bm surfaces that are parallel to the window x-axis. The error in the calculation of absorbed beam radiation is close to zero since the beam radiation intercepted by the profile is insensitive to the shape of the profile. And the right reveal surfaces shadow the bottom reveal surfaces. The result is that the left/outside. bm bm f I ext → I ext 1 + Rgl . the beam and diffuse solar absorbed by the inside frame projections is calculated. The percent error in the calculation of convection and emitted IR is approximately 100 L profile . Reflection from outside reveal surfaces—which are associated with the setback of the glazing from the outside surface of the window’s parent wall—increases the solar gain through the glazing.bm ( ) The beam solar absorbed by the frame outside projection parallel to the window y-axis is bm bm Qabs. 4/23/05 182 .rect is the profile length for a rectangular profile (wf + pf1 for outside of frame or wf + pf2 for inside of frame) and Lprofile. I ext includes the effect of beam solar reflecting off of the glazing onto the outside frame projection. An example of this shadowing is shown in Figure 72. Reflection from inside reveal surfaces—which are associated with the setback of the glazing from the inside surface of the window’s parent wall—decreases the solar gain to the zone because some of this radiation is reflected back out of the window. Beam Solar Reflection from Window Reveal Surfaces This section describes how beam solar radiation that is reflected from window reveal surfaces is calculated. Note that the top and right reveal surfaces are facing away from the sun in this example so their sunlit areas are zero. For example. i. rect . In this case the sun is positioned such that the top reveal surfaces shadow the left and bottom reveal surfaces. where Lprofile. sol cos β v p f 1 ( H − N h wd ) f sunlit / Af Using a similar approach. The amount of beam solar reflected from reveal surfaces depends. left/inside and bottom/inside reveal surfaces each have sunlit areas. actual / L profile . For example. Error Due to Assuming a Rectangular Profile Assuming that the inside and outside frame profile is rectangular introduces an error in the surface heat transfer calculation if the profile is non-rectangular. a square profile the error is about 15%.. for the window can be applied to the window frame. for a circular profile vs a square profile the error is about 22%.actual is the actual profile length. on the extent to which reveal surfaces are shadowed by other reveal surfaces.e. among other things.. fsunlit. Divider Temperature Calculation The divider inside and outside surface temperatures are determined by a heat balance calculation that is analogous to the frame heat balance calculation described above. If there is no exterior window shade.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Here it is assumed that the sunlit fraction. rect − L profile . taking the transmittance of the glazing into account.vsol = I ext α fr . bottom/outside. The error in the absorbed diffuse radiation and absorbed IR depends on details of the shape of the profile.

which is determined from the vertex locations of the window and its parent wall. The case with a frame has to be considered separately because the frame can cast an additional shadow on the inside reveal surfaces. and the size of the corresponding illuminated areas. the top reveal surfaces will never be illuminated by beam solar if the reveal surfaces are perpendicular to the glazing. for a window of arbitrary tilt. Representative shadow patterns are shown in Figure 73 for a window with no shading device. it is assumed that an interior or between-glass shading device blocks beam solar before it reaches inside reveal surfaces. The patterns shown apply to both vertical and horizontal reveal surfaces. This calculation is done in subroutine CalcBeamSolarReflectedFromWinRevealSurface in the SolarShading module. 4/23/05 183 . Correspondingly. (Of course. The size of the shadowed areas. It is important to keep in mind that. Then the above factors determine a unique shadow pattern. Example of shadowing of reveal surfaces by other reveal surfaces. And if the bottom reveal surfaces are illuminated the top surfaces will not be. depends on the following factors: • • The sun position relative to the window The height and width of the window • The depth of the outside and inside reveal surfaces We will assume that the reveal surfaces are perpendicular to the window plane and that the window is rectangular. and without and with a frame. and vice versa. If an exterior shading device (shade or blind) is in place it is assumed that it blocks beam solar before it reaches outside or inside reveal surfaces. if the left reveal surfaces are illuminated the right surfaces will not be. From the geometry of the pattern the shadowed areas and corresponding illuminated areas can be determined. The window reveal input data is specified in the WindowFrameAndDivider object expect for the depth of the outside reveal.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Inside reveal surfaces Glazing Shadow Outside reveal surfaces Figure 72. as is being assumed. and vice versa. for a vertical window.

DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION For each shadow pattern in Figure 73. plus one half of the glazing thickness. = window height for vertical reveal surfaces or window width for horizontal reveal surfaces = vertical solar profile angle for shadowing on vertical reveal surfaces or horizontal solar profile angle for shadowing on horizontal reveal surfaces. = depth of inside reveal (or. d1 and d 2 are measured from the midplane of the glazing. d 2′ d 2′′ = depth of shadow cast by top reveal on bottom reveal. or by left reveal on right reveal. the shadowed and illuminated areas extend into the glazing region. respectively. α p1 ( p2 ) = distance from outside (inside) surface of frame to glazing midplane. the beam solar absorbed by the surfaces formed by the frame outside and inside projections perpendicular to the glazing is calculated as described in “Window Frame and Divider Calculation: Calculation of Solar Radiation Absorbed by Frame. = depth of shadow cast by frame.sh and A2. inside sill depth). measured from the outside plane of the glazing to the edge of the reveal.” 4/23/05 184 . measured from the inside plane of the glazing to the edge of the reveal or the sill. For the case with a frame.sh of the outside and inside reveal surfaces. For this reason. equations are given for the shadowed areas A1. For simplicity it is assumed that. plus one half of the glazing thickness. for the case without a frame. The variables in these equations are the following (see also Figure 74): d1 d2 L = depth of outside reveal. or by right reveal on left reveal. for illumination on bottom reveal surfaces.

Expression for area of shaded regions for different shadow patterns: (a) window without frame.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Figure 73. (b) window with frame 4/23/05 185 .

DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION 4/23/05 186 .

The following logic gives expressions for the shadowed areas for all possible shadow patterns. sh L= L d12 = d1 + d 2 − d 2′ 4/23/05 187 .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION α Frame Wall Inside reveal depth Glazing H Inside sill depth Outside reveal depth d 2′′ = wf tan α wf d 2′ = d 1 + d 2 − H tan α H tan α d1 d2 Figure 74. Vertical section through a vertical window with outside and inside reveal showing calculation of the shadows cast by the top reveal onto the inside sill and by the frame onto the inside sill. Here: d1 = d1 d2 = d 2 P1 = p1 P2 = p2 f1 = d1 − p1 f2 = d 2 − p2 d2prime = d 2′ d2prime2 = d 2′′ TanAlpha = tan α A1sh = A1. sh A2sh = A2.

5/TanAlpha)*L**2 A2sh = d2*L END IF END IF ELSE ! d2prime > d2 A2sh = d2*L IF(d2prime < d1+d2) THEN IF(d12*TanAlpha <= L) THEN A1sh = L*(d2prime-d2) + 0.5*GlazingThickness <= P2) d2 = P2 + 0.5*(d2-(d2prime+d2prime2+P2)) ELSE ! d2-(d2prime+d2prime2+P2) < 0. IF(window does not have a frame) THEN IF(d2prime <= d2) THEN IF(d12*TanAlpha <= L) THEN A1sh = 0.5*L**2/TanAlpha END IF ELSE ! d2prime >= d1+d2 A1sh = d1*L END IF END IF ELSE ! Window has a frame f1 = d1-P1 f2 = d2-P2 d2prime2 = FrameWidth/TanGamma IF(vertical reveal) THEN ! Vertical reveal IF(InsReveal+0.5*TanAlpha*d1**2 A2sh = d2prime*L + 0. L2 = average distance to frame of illuminated area of inside reveal (used to calculate view factor to frame).5*TanAlpha*(L/TanAlpha .001 ELSE ! Horizontal IF(bottom reveal surfaces may be illuminated) THEN ! Bottom reveal surfaces may be illuminated IF(InsSillDepth+0.d1)**2 ELSE ! d1*TanAlpha > L A1sh = d1*L .0.) THEN A2sh = (d2prime+d2prime2)*L + & 0.5) IF(d2-(d2prime+d2prime2+P2) >= 0. ! Inside reveal is fully shadowed by frame and/or !opposing reveal A2sh = f2*L L2 = f2 4/23/05 188 .5*GlazingThickness<=P2) d2= P2 + 0.5*TanAlpha*f1**2 L1 = f1*(f1*TanAlpha/(6*L)+0.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION L1 = average distance to frame of illuminated area of outside reveal (used to calculate view factor to frame).(0.0.A1sh ELSE ! d12*TanAlpha > L IF(d1*TanAlpha <= L) THEN A1sh = 0.001 ELSE ! Top reveal surfaces may be illuminated IF(InsReveal+0.001 END IF END IF IF(d2prime <= f2) THEN ! Shadow from opposing reveal does not go beyond inside ! surface of frame IF(d12*TanAlpha <= L) THEN A1sh = 0.5*TanAlpha*d1**2 A2sh = d2*L .5*TanAlpha*d12**2 .5*TanAlpha*d12**2 ELSE ! d12*TanAlpha > L A1sh = d1*L .5*GlazingThickness <= P2) d2 = P2 + 0.5*TanAlpha*((d1+d2-d2prime)**2-d1+p2+d2prime2)**2) L2 = d2prime2 + 0.

5*L**2/TanAlpha L1 = f1 .5*L**2/TanAlpha L1 = f1-(L/TanAlpha)/3 END IF END IF END IF ELSE ! d2prime > f2 -.5*L**2/TanAlpha L1 = f1 .0.5*TanAlpha*f1**2 L1 = f1*((f1*TanAlpha)/(6*L) + 0.ATAN(FrameWidth/L2)/PiOvr2) END IF ! End of check if window has frame The beam solar reflected from a sunlit region of area A is given by R = I B A cos β (1 − a ) 4/23/05 189 .0 L1 = f1 ELSE ! d2prime < d1+d2 IF(d2prime <= d2+P1) THEN IF(f1*TanAlpha <= L) THEN A1sh = 0.5*(1.5*(1.5) IF((d1+P2+d2prime2)*TanAlpha >= L) THEN A2sh = f2*L L2 = f2 ELSE ! (d1+P2+d2prime2)*TanAlpha < L A2sh = f2*L .5) ELSE ! f1*TanAlpha > L A1sh = f1*L .0 .5) ELSE ! f1*TanAlpha > L A1sh = f1*L .5*TanAlpha*f1**2 L1 = f1*((f1*TanAlpha)/(6*L) + 0.(d1+P2+d2prime2))/3 END IF ELSE ! (d1+P2)*TanAlpha > L L2 = f2 A2sh = f2*L IF(f1*TanAlpha <= L) THEN A1sh = 0.0 .5*TanAlpha*d12**2 L1 = (L*(f1-d12/2)-d12*TanAlpha* & (f1/2-d12/3))/(L-d12*TanAlpha/2) ELSE ! d12*TanAlpha > L A1sh = f1*L .0.5*TanAlpha*f1**2 L1 = f1*((f1*TanAlpha)/(6*L) + 0.0.5*(L-(d1+P2)*TanAlpha)**2/TanAlpha & + d2prime2*(L-(d1+P2+d2prime2/2)*TanAlpha) L2 = d2prime2 + (L/TanAlpha .(L/TanAlpha)/3 END IF ELSE ! d2prime > d2+P1 IF(d12*TanAlpha <= L) THEN A1sh = L*(d2prime-(d2+P1)) + 0.ATAN(FrameWidth/L1)/PiOvr2) FracToGlassIns = 0.0.(L/TanAlpha)/3 END IF END IF END IF END IF FracToGlassOuts = 0.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION END IF ELSE ! d12*TanAlpha >= L IF((d1+P2)*TanAlpha <= L) THEN A1sh = 0.Shadow from opposing reveal goes beyond ! inside of frame A2sh = f2*L L2 = f2 IF(d2prime >= d1+d2) THEN A1sh = 0.

and ( R / 2)(1 − F2 ) goes toward the glazing ( F2 is calculated using a method analogous to that used for F1 ). and layer front α f l . of the portion. The beam solar absorbed by an outside reveal surface is added to the other solar radiation absorbed by the outside of the window’s parent wall. ( R / 2)(1 − F1 ) . is assumed to be reflected back into the zone. ( R / 2)(1 − F2 )τ diff is transmitted 4/23/05 190 . of the source to the frame is calculated from the shape of the illuminated area (see above psuedo-code). If the glazing has diffuse back reflectance b ρ diff . then.diff . and layer back absorptance α lb. that goes toward the glazing. ( R / 2)(1 − F1 )τ diff is transmitted to the zone. Then F1 is related as follows to the average angle subtended by the frame of width w f : F1 = tan −1 ( w f / L1 ) π /2 For the portion going towards the frame. The view factor F1 to the frame calculated by assuming that the illuminated area can be considered to be a line source. ( R / 2) F2 a f is absorbed by the frame and contributes to the frame heat conduction calculation. Of the portion that goes toward the window a fraction ( R / 2) F2 goes toward the frame. L1 . if present. The rest. and 1 − F1 goes toward the glazing. ( R / 2) F1a f is absorbed by the frame (where a f is the solar absorptance of the frame) and contributes to the frame heat conduction calculation. Then the area-weighted average distance. The rest.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION where R = reflected solar radiation [W] I B = beam normal irradiance [W/m2] A = sunlit area [m2] β = beam solar angle of incidence on reveal surface a = solar absorptance of reveal surface All reflected radiation is assumed to be isotropic diffuse. ( R / 2) F2 (1 − a f ) . diffuse front reflectance f ρ diff . of the portion ( R / 2)(1 − F2 ) that goes toward the glazing. and ( R / 2)(1 − F1 ) ρ diff is reflected to the exterior environment. Of the portion that goes toward the window a fraction F1 goes toward the frame. diff . For the portion going towards the frame. For outside reveal surfaces it is assumed that R / 2 goes toward the window and R / 2 goes to the exterior environment.diff is absorbed in glass layer f l and contributes to the glazing heat balance calculation. is assumed to be reflected to the exterior environment. If the glazing has diffuse transmittance absorptance τ diff . ( R / 2)(1 − F1 )α l f. For inside reveal surfaces it is assumed that R / 2 goes towards the window and R / 2 goes into the zone. ( R / 2) F1 (1 − a f ) . then. if present.

diff is absorbed in glass layer l and contributes to the glazing heat balance calculation.”) The following effects are considered by the shading device thermal model: • For interior and exterior shading device: Long-wave radiation (IR) from the surround absorbed by shading device. when in place. Figure 75 illustrates the case of double glazing with an interior shading device. for between-glass shading. and two new equations. b The beam solar absorbed by an inside reveal surface is added to the other solar radiation absorbed by the inside of the window’s parent wall. covers the glazed part of the window (and dividers. Inter-reflection of solar radiation between shading layer and glass layers. and the effect of this flow on the shading-to-gap and glass-to-gap convection coefficients. The window shading device thermal model accounts for the thermal interactions between the shading layer (shade or blind) and the adjacent glass. Heat Balance Equations for Shading Device and Adjacent Glass If a window shading device is deployed the heat balance equations for the glass surfaces facing the shading layer are modified.DAYLIGHTING AND WINDOW CALCULATIONS b WINDOW HEAT BALANCE CALCULATION back out the glazing. “Thermal Performance of Windows. and convection from interior shading layer to zone air or from exterior shading layer to outside air. For interior shading the surround consists of the other zone surfaces. Shading Device Thermal Model Shading devices in EnergyPlus can be on the exterior or interior sides of the window or between glass layers. In the following it is assumed that the shading device. For interior and exterior shading devices it is assumed that the shading layer is separated from the glazing by an air gap. 4/23/05 191 . and ( R / 2)(1 − F2 ) ρ diff is reflected into the zone. ( R / 2)(1 − F2 )α l . if present) and is parallel to the glazing. Natural convection airflow in the gap (or. Direct and diffuse solar radiation absorbed by the shading device. gaps) between the shading layer and adjacent glass. Doors and Shading Devices—Detailed Calculations” [ISO15099. and between the shading layer and the room (for interior shading) or the shading layer and the outside surround (for exterior shading). If the window has a frame. convective gain (or loss) to zone air from gap airflow. one for each face of the shading layer. A between-glass shading layer is assumed to be centered between two glass layers and separated from the adjacent glass layers by gaps that is filled with the same gas. for between-glass shading. are added. 2001]. see “Airflow Windows. The airflow model is based on one described in the ISO Standard 15099. Inter-reflection of IR between the shading device and adjacent glass. for interior shading. An important feature of the shading device thermal model is calculating the natural convection airflow between the shading device and glass. Convection from shading layer and glass to the air in the gap (or. For exterior shading the surround is the sky and ground plus exterior shadowing surfaces and exterior building surfaces “seen” by the window. is a determinant of the convective heat gain from the shading layer and glazing to the zone air. gaps) between shading layer and adjacent glass induced by buoyancy effects. or transmitted by the shading device and absorbed by the adjacent glass. it is assumed that the shading device does not cover the frame. • • • • • • For interior shading. (Between-glass forced airflow is also modeled. This flow affects the temperature of the shading device and glazing and.

The corresponding heat transfer from shading layer to gap air is qc . gl = hcv (θ 4 − Tgap ) where Tgap = effective mean temperature of the gap air (K).DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Glass layer #1 Outside To Eo ho S1 ε1 k1 S2 ε2 h1 S3 ε3 θ1 θ2 θ3 Glass layer #2 Shade layer Inside θ4 θ5 θ6 Ei k2 S4 ε4 hcv Tgap hcv Ssh. “gap”) is Eiε 4τ sh σε 4 ⎡θ54ε sh − θ 44 (1 − ρ sh ) ⎤ + k2 (θ3 − θ 4 ) + hcv (Tgap − θ 4 ) + S4 = 0 + ⎦ 1 − ρ 4 ρ sh 1 − ρ 4 ρ sh ⎣ where τsh = IR diffuse transmittance of shading device εsh = diffuse emissivity of shading device ρsh = IR diffuse reflectance of shading device ( = 1 . The term 1 – ρ4 ρsh accounts for the inter-reflection of IR radiation between glass and shading layer. sh = hcv (θ5 − Tgap ) The convective heat transfer coefficient is given by hcv = 2hc + 4v 4/23/05 (174) 192 . The convective heat transfer from glass layer #2 to the air in the gap is qc . Glazing system with two glass layers and an interior shading layer showing variables used in heat balance equations. The heat balance equation for the glass surface facing the gap between glass and shading layer (called in the following. hcv = convective heat transfer coefficient from glass or shading layer to gap air (W/m2K).( τsh + εsh)) θ5 = temperature of the surface of the shading layer that faces the gap (K).1 εsh ksh hi Ssh.2 εsh Figure 75.

The variables used in the following analysis of the interior shading case are shown in Figure 76. zone air temperature for interior shading or outside air temperature for exterior shading. and shading layer and glass face temperatures. θ6 = temperature of shading layer surface facing the zone air (K). gap air mean equivalent temperature and gap outlet air temperature given values of zone air temperature (or outside temperature for exterior shading). 4/23/05 193 . including height of shading layer.2 = solar radiation plus short-wave radiation from lights plus IR radiation from lights and zone equipment absorbed by the zone-side face of the shading layer (W/m2K).1 = solar radiation plus short-wave radiation from lights plus IR radiation from lights and Eiε sh − ε shσθ 64 + ksh (θ5 − θ 6 ) + hi (Ti − θ 6 ) + S sh . Ssh. The calculation of hcv and Tgap is described in the following sections. which in turn depends on several factors.2 = 0 where Ssh. Solving for Gap Airflow and Temperature For interior and exterior shading devices a pressure-balance equation is used to determine gap air velocity.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION where hc = surface-to-surface heat transfer coefficient for non-vented (closed) cavities (W/m2K) v = mean air velocity in the gap (m/s). glass/shading layer separation (gap depth).1 = 0 where zone equipment absorbed by the gap-side face of the shading layer (W/m2K). shading layer face temperatures and gap geometry. 2001]. The quantities hcv and Tgap depend on the airflow velocity in the gap. The heat balance equation for the shading layer surface facing the zone air is ksh = shading layer conductance (W/m2K). The heat balance equation for the shading layer surface facing the gap is Eiτ sh ρ 4ε sh σε sh ⎡ε 4θ 44 − θ54 (1 − ρ 4 (ε sh + ρ sh )) ⎤ + + ⎦ 1 − ρ 4 ρ sh 1 − ρ 4 ρ sh ⎣ k sh (θ 6 − θ5 ) + hcv (Tgap − θ 5 ) + S sh . The pressure balance equates the buoyancy pressure acting on the gap air to the pressure losses associated with gap airflow between gap inlet and outlet [ISO15099.

in ρ0 = density of air at temperature T0 (kg/m3) T0 = reference temperature (283K) g = acceleration due to gravity (m/s2) H = height of shading layer (m) φ = tilt angle of window (vertical = 90o) Tgap = effective mean temperature of the gap air (K) Tgap. Pressure Balance Equation The pressure balance equation for airflow through the gap is ∆pT = ∆pB + ∆pHP + ∆pZ (175) Here. Ar and Ah are shown schematically.in Tsh s W Abot H Inside Glass Figure 76. ∆pT is the driving pressure difference between room air and gap air. Atop. The opening areas Abot. Al.in = gap inlet temperature ( = zone air temperature for interior shading) (K) 4/23/05 194 .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Air flow (a) Tgap.out (b) Atop Shade Outside Tgap Ar v H Al Ah Tgl s Glass Shade Tgap.in TgapTgap . It is given by ∆pT = ρ0T0 gH sin φ where Tgap − Tgap . Vertical section (a) and perspective view (b) of glass layer and interior shading layer showing variables used in the gap airflow analysis.

Assuming steady laminar flow.in ⎟ ⎝ ⎠ Z out where 2 ⎛ Agap ⎞ =⎜ − 1⎟ ⎜ 0. It is given by ∆pB = ρ 2 v2 (Pa) where ρ is the gap air density evaluated at Tgap (kg/m3). the inlet pressure drop factor.in = equivalent inlet opening area (m2) Aeq. Zin. The ∆pZ term is the sum of the pressure drops at the inlet and outlet openings: ∆pZ = ρ v2 2 ( Zin + Z out ) (Pa) Here.in Abot ( Al + Ar + Ah ) 2( Abot + Atop ) 4/23/05 195 .in Aeq .66 Aeq . The ∆pHP term represents the pressure drop due to friction with the shading layer and glass surfaces as the air moves through the gap.out ⎟ ⎝ ⎠ 2 Aeq.in = Abot + Atop 2( Abot + Atop ) ( Al + Ar + Ah ) Aeq . 1998]: ∆pHP = 12µ H v s2 (Pa) where µ is the viscosity of air at temperature Tgap (Pa-s).out = equivalent outlet opening area (m2) Agap = cross-sectional area of the gap = sW (m2) If Tgap > Tgap.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION The ∆pB term is due to the acceleration of air to velocity v (Bernoulli’s law).60 Aeq . Zout. it is given by the Hagen-Poiseuille law for flow between parallel plates [Munson et al.out = Atop + If Tgap ≤ Tgap. are given by ⎛ Agap ⎞ − 1⎟ Z in = ⎜ ⎜ 0. and the outlet pressure drop factor.

the area of the openings through which airflow occurs (see Figure 76 and Figure 77) are defined as follows: Abot = area of the bottom opening (m2) Atop = area of the top opening (m2) Al = area of the left-side opening (m2) Ar = area of the right-side opening (m2) Ah = air permeability of the shading device expressed as the total area of openings (“holes”) in the shade surface (these openings are assumed to be uniformly distributed over the shade) (m2) Figure 77 shows examples of Abot.out = Abot + 2( Abot + Atop ) ( Al + Ar + Ah ) Here.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Aeq . (a) Horizontal section through shading layer with openings on the left and right sides (top view). Al and Ar for different shading device configurations. These areas range from zero to a maximum value equal to the associated shade/blind-toglass cross-sectional area. Expression for the Gap Air Velocity Expressing Equation (174) in terms of v yields the following quadratic equation: 4/23/05 196 . Examples of openings for an interior shading layer covering glass of height H and width W. Al and Ar ≤ sH. Abot and Atop ≤ sW..in = Atop + Abot ( Al + Ar + Ah ) 2( Abot + Atop ) Atop Aeq . Atop. Atop = tW t (a) r (b) Ar = min(rH. Not to scale. (b) Vertical section through shading layer with openings at the top and bottom (side view).e.sH) H s Shade Glass l b Abot = bW Figure 77. i.sH) W s Shade Glass Al = min(lH.

out = Tave − (Tave − Tgap .in )e − h / H 0 where Tave = Tgl + Tsh 2 (177) is the average temperature of the glass and shading layer surfaces facing the gap (K). from the gap inlet (Figure 78) is Tgap (h) = Tave − (Tave − Tgap . given by H0 = ρC p s 2hcv v where Cp is the heat capacity of air. ⎤ ⎥ ⎥ ⎦ 1/ 2 − 12µ H s2 (176) The choice of the root of the quadratic equation is dictated by the requirement that v = 0 if Gap Outlet Temperature and Equivalent Mean Air Temperature The temperature of air in the gap as a function of distance.in )e − H / H 0 The thermal equivalent mean temperature of the gap air is (178) Tgap = H 1 ∫ Tgap (h)dh = Tave − H0 (Tgap. h.in − Tgap 12µ H v − ρ0T0 gH sin φ =0 2 s Tgap .in = Tgap.in − Tgap ⎢⎜ 2 ⎟ + Tgap . The gap outlet temperature is given by Tgap . H0 = characteristic height (m).inTgap Solving this gives ⎡⎛ 12µ H ⎞2 2 ρ 2 (1 + Z in + Z out ) ρ0T0 gH sin φ Tgap .in ) H 0 H (179) 4/23/05 197 .out − Tgap.inTgap ⎢⎝ s ⎠ v= ⎣ ρ (1 + Z in + Z out ) Tgap.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION ρ v2 2 (1 + Zin + Z out ) + Tgap .

in Mean gap air temperature. The solution sequence in WinShadeGapFlow is as follows: At start of iteration. temperature.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Inlet air temperature.in)/2. Tgap Outlet air Average surface temperature. by calling WindowGasConductance and NusseltNumber. Thereafter use value from previous iteration. h. Tgap.out Tave h axis Air flow Gap air temperature vs. Variation of gap air temperature with distance from the inlet for upward flow. Convective Heat Gain to Zone from Gap Airflow The heat added (or removed) from the air as it passes through the gap produces a convective gain (or loss) to the zone air given by 4/23/05 198 . Tgap. Tgap(h) H H0 Temperature axis Figure 78. Get v from Equation (176) Get hcv from Equation (174) Get Tave from Equation (177) Get Tgap.out. Solution Sequence for Gap Air Velocity and Outlet Temperature The routine WinShadeGapFlow is called within the glazing heat balance iterative loop in SolveForWindowTemperatures to determine v and Tgap. guess Tgap as ((Tgl + Tsh)/2 + Tgap. hc.out from Equation (178) Get new value of Tgap from Equation (179) The values of hcv and Tgap so determined are then used in the window heat balance equations to find new values of the face temperatures of the glass and shading layers. These temperatures are used in turn to get new values of hcv and Tgap until the whole iterative process converges. Get still-air conductance.

2 Figure 79. Glass layer #1 Outside To Eo ho S1 ε1 k1 θ1 θ2 Gap #1 s v θ5 Shading layer Gap #2 s θ6 v θ3 Glass layer #2 Inside θ4 Ti Ei hcv.in ) (W) where the air mass flow rate in the gap is given by m = ρ Agap v (kg/s) Heat Balance Equations for Between-Glass Shading Device In EnergyPlus shading devices are allowed between the two glass panes of double glazing and between the two inner glass panes of triple glazing.1 Tgap.2 S3 ε3 k2 hi S4 ε4 Tgap.2 S6 ε6 hcv.out − Tgap .1 − θ 2 ) + k1 (θ1 − θ 2 ) + ⎤ σε 2 ⎡ τ sh (ε 3θ34 + ε 6θ64 ρ3 ) + ε 5θ54 + ε 2θ 24 R1 ⎥ − σε 2θ 24 + S2 = 0 ⎢ 1 − ρ 2 R1 ⎣1 − ρ6 ρ3 ⎦ where 4/23/05 199 .1 S5 ε5 ksh hcv. Figure 79 shows the case of a between-glass shading device in double glazing. For face #2: hcv .1 (Tgap .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION qv = LW ⎡ hcv (Tgl − Tgap ) + hcv (Tsh − Tgap ) ⎤ = 2hcv LW (Tave − Tgap ) ⎣ ⎦ This can also be expressed as (W) qv = mC p (Tgap . The heat balance equations for the two glass surfaces facing the shading device are the following.1 S2 ε2 hcv. Glazing system with two glass layers and a between-glass shading device showing variables used in the heat balance equations.

DAYLIGHTING AND WINDOW CALCULATIONS 2 τ sh ρ3 R1 = ρ5 + 1 − ρ 6 ρ3 WINDOW HEAT BALANCE CALCULATION Tgap .1 − θ5 ) + k sh (θ 6 − θ5 ) + ⎤ σε 5 ⎡ τ sh ρ 2 (ε 3θ34 + ε 6θ64 ρ3 ) + ε 2θ 24 + ε 5θ54 ρ2 ⎥ − σε 5θ54 + S5 = 0 ⎢ 1 − ρ 2 R1 ⎣1 − ρ5 ρ3 ⎦ For face #6: hcv .1 .2 = effective mean air temperature in gap 2 (K) hcv .2 + 4v where (180) hc . when these gaps are non-vented (closed).2 (Tgap .1 = 2hc .1 (Tgap .2 (Tgap .2 = 4/23/05 surface-to-surface heat transfer coefficients for gap #1 and #2.1 = convective heat transfer coefficient from glass or shading layer to gas in gap 1 (W/m2K) For face #3: hcv .2 = convective heat transfer coefficient from glass or shading layer to gas in gap 2 (W/m2K) The heat balance equations for the shading layer faces are: For face #5: hcv . hc . respectively.2 − θ3 ) + k2 (θ 4 − θ 3 ) + σε 3 1 − ρ3 R2 where 2 τ sh ρ 2 1 − ρ5 ρ 2 ⎡ τ sh ⎤ ε 2θ 24 + ε 5θ54 ρ 2 + ε 6θ 64 + ε 7θ 34 R2 ⎥ − σε 3θ 34 + S3 = 0 ⎢ ⎣1 − ρ5 ρ 2 ⎦ ( ) R2 = ρ6 + Tgap . 200 .2 = 2hc .1 = effective mean air temperature in gap 1 (K) hcv .2 − θ3 ) + k2 (θ 4 − θ 3 ) + σε 3 1 − ρ3 R2 ⎡ τ sh ⎤ ε 2θ 24 + ε 5θ54 ρ 2 + ε 6θ 64 + ε 7θ 34 R2 ⎥ − σε 3θ 34 + S3 = 0 ⎢ ⎣1 − ρ5 ρ 2 ⎦ ( ) The convective heat transfer coefficients are given by hcv .1 + 4v hcv .

Then. when the airflow is upward in gap #1 it is downward in gap #2 and vice-versa.1.i + Z out .2 ) H ⎞ 2 ⎤ ⎢⎜ + 2∆pT .2 (Pa) The pressure drops on the right-hand side of this equation are: ∆pB .in )ξ1 where Tave .1 = ρ gap .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION air velocity in the gaps (m/s).1.i ∆pZ . we obtain the following expression for the airflow velocity: ⎡⎛ 12( µ gap .out this gives: Tgap .1C p sv /(2hcv .in = Tgap .1 with H 0.1.2 )(1 + Z in + Z out ) ⎥ ⎟ s2 ⎢⎝ ⎥ ⎠ ⎣ ⎦ v= ( ρ gap .i = ρ gap .1. Since Tgap .1 − Tgap .1 − (Tave.2 = ∆pB .2 ) H s2 (182) The choice of the sign of the square root term is dictated by the requirement that v = 0 if ∆pT .i 2 v2 H s2 ∆pHP .1.2 Tgap .out = Tave .2.1.1 + ∆pZ .1 = Tgap . which gives ∆pT .e.2 + ∆pZ . The procedure is as follows.out )ξ1 4/23/05 201 .1 − Tgap .1 + ∆pHP .2 = ρ0T0 gH sin φ Tgap .1. the air velocity is determined by solving the following pressure balance equation: v= ∆pT . i.1 − (Tave.i ) where i = gap number (1 or 2).2.1 + Z out .out = Tave .1 + µ gap .2 .i v 2 2 ( Z in .1 + ρ gap .i = ρ gap ..1.2 . We have Tgap .1 + ∆pB .1.2 .2 where the driving pressure difference between gap #1 and #2 is (181) ∆pT .2 = 0 . It is assumed that the gap widths are equal.2 )(1 + Z in + Z out ) 1/ 2 − 12( µ gap . inserting these pressure drop expressions into (181).1 − Tgap .i = 12µ gap .1 + µ gap . i. It can be shown that Z in .2 .2 + Z out .1 and Tgap .2 ( ρ gap .. so that the velocity of flow in the gaps is equal and opposite.e.1 ) . In analogy to the interior or exterior shading device case. Given v we can now calculate Tgap .1Tgap .1 = Z in .2 + ∆pHP .1 + ρ gap .1 = (θ 2 + θ5 ) / 2 and ξ1 = e − H H 0 . Tgap .

At start of iteration guess Tgap .1 .2 H (T gap .2 .1 .1 − Tave.1 = Tave . from Tgap .2 − Tave.2.1 .2 − Tgap .2.2 .1 and Tgap . hc .2 by calling WindowGasConductance and NusseltNumber.1 H ξ (Tave .1.2 from Equation (180) 4/23/05 202 .out )ξ 2 Solving these simultaneous equations gives: Tgap .1 .1 = Tave.1 H (T gap .2 H ξ (Tave . Then 1) Get 2) Get ρ gap . Tgap .out − Tgap .2.out = Tave .out ) Tgap .1 − gives H 0.1 = Tave .2 using Tgap . hcv .1.1 − with H 0.2.2 (1 − ξ 2 ) + ξ 2Tave .1 .2 − H 0. µ gap . Tgap .out = Using these in Tgap . ρ gap .2 (1 − ξ 2 ) 1 − ξ1ξ 2 Tave .out ) Tgap .2 − (Tave.1.1 (1 − ξ1 ) + ξ1Tave .1 (1 − ξ1 ) 1 − ξ1ξ 2 Tgap . µ gap .2 = Tave .2 − we get H 0.1 ) (184) The overall solution sequence is as follows.2 . 3) Get v from Equation (182) 4) Get hcv . still-air conductances hc .2 ) (183) ξ= (1 − ξ1 )(1 − ξ 2 ) 1 − ξ1ξ 2 Similarly.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Similarly.1.out − Tgap .out = Tave .2 = Tave .2 = Tave .

1 . The allowed combinations of Airflow Source and Airflow Destination are: InsideAir OutsideAir InsideAir InsideAir InsideAir ReturnAir OutsideAir InsideAir OutsideAir OutsideAir 4/23/05 203 .1 . Tgap .2 . hcv . Ho.2 from Equations (183) and (184) The values hcv . Such windows are also known as “heat-extract windows” and “climate windows.1 and Tgap .2 until the whole iterative process converges.1. Tave . θ3 .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION 5) Get Tave .2.” Five configurations of airflow windows are modeled (Figure 80) that depend on the source and destination of forced air. ξ1 and ξ 2 .2 6) Get Ho.1 . θ5 and θ 6 .2 are then used in the face heat balance equations to find new values of the face temperatures θ 2 .2 . These are used in turn to get new values of hcv .1 . Tgap . 7) Get Tgap . Airflow Windows In airflow windows forced air flows in the gap between adjacent layers of glass. hcv .1 and Tgap . Tgap .

Gap airflow configurations for airflow windows. thereby picking up and rejecting heat from the glazing (Figure 80). 4/23/05 204 . 1. A common application of airflow windows is to reduce the zone cooling load by exhausting indoor air through the window. From “Active facades. Belgian Building Research Institute.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION (a) (b) (c) (d) Outside Inside Outside Inside Outside Inside Outside Inside “Air exhaust” “Indoor air curtain” “Air supply” “Outdoor air curtain” To zone return air (e) Outside Inside Airflow to return air Figure 80. June 2002. Figure 81 shows the variables used in the heat balance equations for forced airflow in a double-glazed window.” Version no.

Glazing system with forced airflow between two glass layers showing variables used in the heat balance equations. The heat balance equation for the left-hand glass surface facing the gap in Figure 81 is: k1 (θ1 − θ 2 ) + hcv (Tgap − θ 2 ) + σ ε 2ε 3 (θ34 − θ 24 ) + S2 = 0 1 − (1 − ε 2 )(1 − ε 3 ) The corresponding equation for the right-hand glass surface facing the gap is: k2 (θ 4 − θ 3 ) + hcv (Tgap − θ3 ) + σ Here.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Glass layer #1 Outside To Eo ? ho S1 ε1 k1 S2 ε2 θ1 θ2 Tgap.out Glass layer #2 Inside θ3 θ4 Ti ? Ei v hcv Tgap hcv k2 hi S4 ε4 s S3 ε3 Tgap. which is an input value that is constant or can vary according to a user-specified schedule: v= 4/23/05 F Agap (m/s) 205 .in Figure 81. The convective heat transfer coefficient is given by hcv = 2hc + 4v where hc = glass-to-glass heat transfer coefficient for non-vented (closed) cavity (W/m2K) v = mean air velocity in the gap (m/s). ε 2ε 3 (θ24 − θ34 ) + S3 = 0 1 − (1 − ε 2 )(1 − ε 3 ) Tgap = effective mean temperature of the gap air (K) hcv = convective heat transfer coefficient from glass to gap air (W/m2K). The air velocity is determined by the gap cross-sectional area in the flow direction and the air flow rate.

The mean temperature of the gap air is given by the following expression.out − C p .out = Tave − (Tave − Tgap . whose derivation follows that for (179) for the case of an interior shading device: Tgap = Tave − where H (Tgap .i = heat capacity of the inside air (J/kg-K) Cp.out = heat capacity of the gap outlet air (J/kg-K) and where the air mass flow rate in the gap is m = ρF Fan Energy (kg/s) The fan energy used to move air through the gap is very small and is ignored. To if the airflow source is outside air) (K) The outlet air temperature is given by Tgap .in ) H0 Tave = θ 2 + θ3 2 H = glazing height (m) ρC p s H0 = v 2hcv Tgap.iTi ) where (W) Cp.in )e − H / H 0 The equations for glass face #1 and #4 are the same as those for no airflow in the gap (Equations (163) and (166)).in = gap air inlet temperature (Ti if the airflow source is inside air. The convective heat gain to the zone air due to the gap airflow when the airflow destination is inside air is qv = m(C p .DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION where F = airflow rate (m3/s) Agap = gap cross-sectional area in direction of flow (m2) It is assumed that the airflow is uniform across the width of the window.out − Tgap . 4/23/05 206 .outTgap .

2.out Glass layer #2 Inside Ti Tgap.ave.out ) / 2 and Cp. Airflow window with between-glass shading device showing variables used in the heat balance equations.in Figure 82. 4/23/05 207 . The heat balance equations in this case are the same as those for the between-glass shading device with natural convection (Figure 79 and following equations) except that now v= F /2 Agap (m/s) where Agap = sW is the cross-sectional area of the gap on either side of the shading device.1.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION Airflow Window with Between-Glass Shading Device Figure 82 shows the case of a double-glazed airflow window with a between glass shading device.ave.outTgap .2.out = (Tgap .out = heat capacity of the outlet air evaluated at Tgap.out + Tgap . It is assumed that the shading device is centered between the two panes of glass so that the airflow.in Tgap.2 s θ1 θ2 v θ5 θ6 θ3 v Ei θ4 Tgap.ave . is divided equally between the two gaps.iTi ) (W) where the average temperature of the two outlet air streams is Tgap .1.out − C p .out Shading layer Gap #2 Tgap.ave.ave.1.1 s Tgap.2.1 S2 ε2 hcv.2 S3 ε3 k2 hi S4 ε4 Tgap.2 S6 ε6 hcv. The convective heat gain to the zone air due to the airflow through the two gaps when the airflow destination is inside air is qv = m(C p .1 S5 ε5 ksh hcv.out (J/kg-K) Gap #1 Glass layer #1 Outside To Eo ho S1 ε1 k1 hcv. F.

Fischer and F. Solar-Thermal Window Blind Model for DOE-2.” Third Edition Update.S. C. Rubin. Pt. U.D. Huizenga. WINDOW 4. ISO15099. 1998.DAYLIGHTING AND WINDOW CALCULATIONS WINDOW HEAT BALANCE CALCULATION References Arasteh. doors and shading devices—detailed calculations. B. Young and T.K. Draft. 1996 (unpublished). D.Air Heat Balance Manager / Processes 4/23/05 208 . A versatile procedure for calculating heat transfer through windows. E. Reilly and M. LBL-33943.. Thermal performance of windows. Arasteh.K. D. Vol.D.H.R.S. “Fundamentals of Fluid Mechanics.F. Reilly. Simmler. 1989.U.. M. 2. Inc.. July 18. Finlayson.. 95. M. Lawrence Berkeley National Laboratory. 2001. Winkelmann. Lawrence Berkeley National Laboratory report no. Munson. 1993. ASHRAE Trans. Okiishi. Rubin and M..0: documentation of calculation procedures. John Wiley & Sons. H. Simulation Research Group internal report. D.

It is added directly into the air heat balance.i ) (185) The inside heat transfer coefficient is modeled from a choice of correlations. The EnergyPlus defaults are 1. In the most common procedure.35 m/s (7.0 which gives a constant volume flow of infiltration under all conditions.0.4 mph) windspeed. and even in very small amounts through building elements. These coefficients produce a value of 1. 0.606.AIR HEAT BALANCE MANAGER / PROCESSES CONVECTION FROM SURFACES Air Heat Balance Manager / Processes Convection from Surfaces This contribution is expressed using the convective heat transfer coefficient as follows: qconv = nsurfaces ∑ i =1 hc . 0. BLAST (one of the EnergyPlus predecessors) used the following values as defaults: 0. However. Ideally. which corresponds to a typical summer condition. one should do a detailed analysis of the infiltration situation and then determine a custom set of coefficients using methods such as those laid out in Chapter 26 of the ASHRAE Handbook of Fundamentals. Infiltration described by the equation shown above is entered into EnergyPlus using the following syntax.i Ai (Ta − Ts . the details required to include this component in the heat balance are generally not available.75. 0.1177. cracks around windows. Such a treatment also violates the tenets of the heat balance since the surface temperature of the surfaces producing the internal loads exchange heat with the zone air through normal convective processes. Infiltration Infiltration (Ref Object: Infiltration) is the unintended flow of air from the outdoor environment directly into a thermal zone. Exfiltration (the leakage of zone air to the outside) is generally handled better as zone exhaust air in the zone equipment description.5 mph) windspeed. The basic equation used to calculate infiltration in EnergyPlus is: Infiltration = ( I design ) ( Fschedule ) ⎡ A + B (Tzone − Todb ) + C (WindSpeed ) + D (Windspeed 2 ) ⎤ ⎣ ⎦ More advanced infiltration calculations are possible using the EnergyPlus link to the COMIS program for natural infiltration and/or the Air Distribution System model for times when the HVAC system operates. the infiltration quantity is converted from a number of air changes per hour (ACH) and included in the zone air heat balance using the outside temperature at the current hour.0 at 0C deltaT and 3. these coefficients would increase the infiltration rate by a factor of 2. Infiltration is generally caused by the opening and closing of exterior doors.03636. Convection from Internal Sources This component is the companion part of the radiant contribution from internal gains described previously. The determination of the amount of infiltration air is quite complicated and subject to significant uncertainty. Infiltration/Ventilation Any air that enters by way of infiltration is assumed to be immediately mixed with the zone air. 4/23/05 209 . and its direct inclusion into the air heat balance is a reasonable approach.0. At a winter condition of 40C deltaT and 6 m/s (13. The question of typical values for these coefficients is subject to debate. once again..

5ZoneAirCooled). With these coefficients. Calculation of Zone Air Temperature The zone air heat balance is the primary mechanism for linking the loads calculation to the system simulation. Ventilation Ventilation (Ref Object: Ventilation) is the purposeful flow of air from the outdoor environment directly into a thermal zone in order to provide some amount of non-mechanical cooling. COMIS COMIS was originally developed in 1994 as a stand-alone multi-zone air flow program with its own input and output processors.. GeometryTest). and the winter conditions would give 1. the zone air temperature becomes the main interface variable. 0.0. the DOE-2 defaults in some (e. The COMIS air flow building description shares many similarities to the building description used for modeling heat transfer. the summer conditions above would give a factor of 0.g. the Heat Balance Air Manager module generates an internal COMIS input file (comistest. as well as any changes in window and door opening conditions. For each time step. Ventilation as specified by this input syntax is intended to model “simple” ventilation as opposed to the more detailed ventilation investigations that can be performed with the link to COMIS. Air Exchange Air exchange and interchange between zones is treated as a convective gain.34.. COMIS. the air change method defaults are (adjusted to SI units) 0. Its role in the integration process was described previously (“Basis for the Zone and System Integration”). in turn. Both treat the building as a collection of zones that are linked 4/23/05 210 . the actual flow rate of ventilation can be modified by the temperature difference between the inside and outside environment and the wind speed. As such.224 (windspeed). The basic equation used to calculate ventilation in EnergyPlus is: Ventilation = (Vdesign )(Fschedule ) A + B Tzone − Todb + C (WindSpeed ) + D WindSpeed 2 [ ( )] More advanced ventilation calculations are possible using the EnergyPlus link to the COMIS program. the COMIS source code has been modified into ten modules that are called by the EnergyPlus program during each time step. and the exterior environmental conditions and internal zone temperatures. The EnergyPlus example files use all of the above.47 m/s (10 mph) gives a factor of 1. returns to the Heat Balance Air Manager the calculated air flows from the outside as infiltration. In this implementation into EnergyPlus.AIR HEAT BALANCE MANAGER / PROCESSES AIR EXCHANGE In DOE-2 (the other EnergyPlus predecessor). The source of the BLAST defaults is noted in the BLAST documentation as: "Empirical equation and the coefficient default were determined from ASHRAE journal articles and other data on the effects of outdoor weather conditions. and the EnergyPlus defaults in some (e. and from zone to zone as cross-mixing air flows.cif) based on the building air flow description described in this section. Comistest.75. Simple ventilation in EnergyPlus can be controlled by a schedule and through the specification of minimum and delta temperatures as described below. A windspeed of 4. the exterior and interior temperatures are updated. As with infiltration. the BLAST defaults in some (e. At the beginning of the simulation.g. 0.cif can be used as input to the original COMIS program for testing and debugging purposes..g. 0." The source of the DOE-2 defaults is based on examining the infiltration relationships decsribed in the ASHRAE Handbook of Fundamentals. LgOffVAVDetCoil).

pdf 4/23/05 211 . wind speed and direction.pdf COMIS – An International Multizone Air-Flow and Contaminant Transport Model.lbl.) 1998. The COMIS model. humidity. windows. Environment COMIS External Nodes. etc. etc. the orientation and location of the building surfaces. Window or Door Construction Surface: HeatTransfer Environmental Data (temperature. does require substantially more information about the external environment. however. while a COMIS surface needs only references to the air flow type and the corresponding heat transfer surface. radiation. COMIS Air Flow : Opening Surfaces COMIS Surface External Environmental Data. however.lbl. Lawrence Berkeley National Laboratory. References More detailed discussions of their use. the nodes are the air flow properties rather than the thermal properties of the various building components such as walls.gov/Publications/lbnl-42182. B. defaults and suggested values can be found in Feustel. Air Flow representations in Buildings Thermal inputs Zone Building Component Thermal Zone Wall. H.AIR HEAT BALANCE MANAGER / PROCESSES COMIS to the external environment and other zones by a network of nodes.0 User’s Guide. Thermal vs. Roof or Floor Construction. such as the volume and neutral height of the zones. (ed. http://epb1. The following table shows the general parallels between the thermal and air flow representations of a building : Table 33.gov/comis/ Comis 3. COMIS Site Wind Conditions. In the air flow representation. pressure.) Air flow inputs COMIS Zone COMIS Air Flow : Crack. the air flow modeling requires differing external nodes where the pressure coefficients (Cp) by wind direction may differ.gov/comis/composite. roofs. and Smith.0 User's Guide". "COMIS 3. no additional inputs are needed to define an air flow zone. COMIS Cp Values A substantial part of the information needed for the air flow modeling already exists and can be extracted or referenced from the building description for the thermal modeling. http://epb1. In fact. Berkeley CA.lbl. etc. This document and other information is available on the LBNL web site: http://epb1. Whereas the thermal modeling assumes the same external environment for all surfaces (except solar and wind). doors.

The primary air system includes supply and return fans. actuator at a component inlet) then iteration can be confined to the components and the overall air system equations can be solved directly. Consequently EnergyPlus uses iteration over the entire primary air system in order to converge the system controllers. outside air economizer. central heating and cooling coils. instead iterative methods are needed. The air loop is just the “air side” of a full HVAC system. As a result the usual algebraic loops arising in steady state air system simulations are eliminated from the EnergyPlus system simulation. Supply and return plenums are also included on the zone equipment side of the loop. For each component several choices of inputs and outputs are possible. They are embodied as Fortran90 subroutines within software modules. EnergyPlus uses algebraic energy and mass balance equations combined with steady state component models. The term “loop” is used because in most cases some air is recirculated so that the air system forms a fluid loop. During each system simulation time step new zone temperatures and humidities are predicted using past values. The zone temperatures and humidities are then held constant during the simulation of the air system (and the plant). System set points are held fixed during a system time step. Component Models EnergyPlus component models are algorithmic with fixed inputs and outputs. Air Loops Definition of Air Loop In EnergyPlus an air loop is a central forced air HVAC system. When the zone air and the air system are modeled with algebraic equations (steady state) then – in cases with recirculated air – there will be one or more algebraic loops. baseboards. or lagged and control is local (sensor located at a component outlet. The zone air is assigned heat and moisture storage capacities and the capacity effects are modeled by 1st order ordinary differential equations. The need for iteration can be reintroduced by the need for system control. and so forth. and any other central conditioning equipment and controls. There is no one choice that will be most usable and efficient for all system configurations and control schemes. In EnergyPlus the zone humidity ratios and temperatures are decoupled from the solution of the air system equations.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES AIR LOOPS Building System Simulation System Manager / Processes EnergyPlus uses a loop based HVAC system formulation. However these requirements are too restrictive to simulate real systems. But controller sensors are allowed to be remote from the location of the actuator. and heating and cooling demands are known inputs to the system simulation. The zone equipment side of the loop contains the air terminal units as well as fan coils. For reasons of consistency and comprehensibility the requirement was imposed that all EnergyPlus models be forward 4/23/05 212 . externally determined. For simulation purposes the air loop is divided into 2 parts: the primary air system (representing the supply side of the loop) and the zone equipment (representing the demand side of the loop). Simulation Method Simulating a forced air system and its associated zones can be done in a number of ways. If system set points are fixed. window air conditioners. Then the zone temperatures and humidity ratios are corrected using results from the system simulation. Typically a humidity ratio and a mass flow rate will be variables involved in the iterative solution. humidity ratios. An example of the dual duct VAV system is shown below. The zone temperatures. In other words it is not possible to solve the equations directly.

This method was chosen (rather than for instance Newton-Raphson) for its robustness. just call ManageAirLoops (simulates the primary air systems) and ManageZoneEquipment (simulates the zone equipment sets) once and quit. the component inputs correspond to the inlet conditions and the outputs to the outlet conditions. The controllers are converged by the method of interval halving.SimZoneEquipment) SimPlantDemandLoops = . enthalpies.LE. If another choice of inputs and outputs is needed it is obtained by numerical inversion of the forward model. The system controllers are simulated in sequence. flow rates. SimElecCircuits =.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES AIR LOOP SIMULATION models. Otherwise loop over primary air systems and zone equipment sets until the temperatures. END IF IF (SimZoneEquipment) THEN CALL ResolveAirLoopFlowLimits(IterAir+1) CALL ManageZoneEquipment(FirstHVACIteration. Iteration Scheme The primary air system simulation uses successive substitution starting at the return air inlet node and proceeding in the flow direction to the supply air outlet nodes.SimAirLoops.OR. These flow rates are passed back upstream to the primary air system. DO WHILE ((SimAirLoops . This initial pass is simulated with full design air flow rates and allows the zone equipment to set the flow rates for each zone that will meet the zone loads.SimAirLoops) SimPlantDemandLoops = . The air terminal unit components with their built in dampers and controllers respond to the zone heating and cooling loads by setting air flow rates at their inlet nodes. That is. Determination of Air Mass Flow Rates In most cases the air mass flow rate of the central air system is set by zone equipment downstream of the primary air system.MaxAir) ) IF (SimAirLoops) THEN CALL ManageAirLoops(FirstHVACIteration. These flow rates are held fixed while the primary air system is simulated.AND. agree to within tolerance at each primary air system – zone equipment gap. These flags are set by the subroutine UpdateHVACInterface which is called from within ManageAirLoops and 4/23/05 213 . During this sequence of iterative solutions the air mass flow rates are held constant. Air Loop Simulation A complete simulation of each primary air system – zone equipment air loop is done in the following manner.SimZoneEquipment. The iteration proceeds until an individual actuator-controller has converged (the sensed value matches the set point within the specified tolerance).TRUE.TRUE. humidity ratios etc. END IF IterAir = IterAir + 1 END DO CALL ResolveLockoutFlags(SimAirLoops) The logical flags SimAirLoops and SimZoneEquipment are used to signal whether the primary air systems or the zone equipment sets need to be resimulated.TRUE. (IterAir.TRUE SimElecCircuits =. If this is the first simulation this system time-step. SimZoneEquipment) . establishing the flow rates in the primary air system branches.

Like the other manager routines in EnergyPlus. node at the entering node of each ATU in the system). This subroutine checks if any air system component has requested that the economizer be locked out. sometimes it means just the supply side – the primary air system.01 0. END IF 4/23/05 214 . Primary Air System Simulation When the EnergyPlus HVAC simulation manager needs to simulate the primary air system side of the air loop it calls ManageAirLoops.false. The values checked by UpdateHVACInterface and their tolerances are as follows. ResolveLockoutFlags sets SimAirLoops to true and the EconoLockout flag to true to ensure that the air loop will be resimulated with the economizer forced off. the primary air system simulation manager subroutine. If such a request has been made and if the economizer is active.0 ResolveAirLoopFlowLimits is invoked to deal with zone equipment – primary air system flow mismatches. In this case ResolveAirLoopFlowLimits takes the air flow that the fan can supply and apportions it among the ATUs in proportion to their design maximum air flow rates (ResolveAirLoopFlowLimits sets the mmax avail . Note that “air loop” is used inconsistently in the program: sometimes it means the full loop consisting of both supply & demand sides – primary air system and zone equipment.01 10.0 0. If they are not it sets the simulate flag of the downstream half-loop to true. Each time UpdateHVACInterface is called it also checks whether the values at the half-loop outlet nodes are in agreement with the values at the downstream half-loop inlet nodes.01 10. UpdateHVACInterface (when called from ManageAirLoops) passes the values at the outlet nodes of a primary air system on to the inlet nodes of the corresponding zone equipment half-loop and similarly (when called from ManageZoneEquipment) passes the values of the outlet nodes of a zone equipment half-loop on to the inlet nodes of its corresponding primary air system.0 0. At the end of the air loop simulation ResolveLockoutFlags is called. ManageAirLoop Code IF (GetInputFlag) THEN !First time subroutine has been entered CALL GetAirPathData ! Get air loop descriptions from input file GetInputFlag=. ManageAirLoops has a very simple structure: Table 34.0001 0.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PRIMARY AIR SYSTEM SIMULATION ManageZoneEquipment at the end of each half-loop simulation. Quantities specific enthalpy [J/kg} mass flow rate [kg/s] humidity ratio [kg H2O / kg dry air] quality air pressure [Pa] temperature [C] energy [J] Tolerances 10. For instance the zone air terminal units (ATUs) may be asking for more air than the central fan can supply.

the air loop design supply air flow rate is obtained by summing the design heating or cooling air flow rates of the zones connected to the air loop. These objects and their data are described in the EnergyPlus Input Output Reference document. Mixers. 3) If automatic sizing of the loop flow rates is called for.SimZoneEquipment) END IF ! No Update ! Report information at the Manage Air Loop Level CALL ReportAirLoops 1) If the user input data has not been input. and Branches. Instead that knowledge is embedded implicitly in the overall air loop – zone equipment network description.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PRIMARY AIR SYSTEM SIMULATION ! Initialize air loop related parameters CALL InitAirLoops(FirstHVACIteration) ! Call the AirLoop Simulation IF (SysSizingCalc) THEN CALL SizeAirLoops ELSE CALL SimAirLoops(FirstHVACIteration. For instance. For sizing calculations it is important to have a data structure that explicitly shows which zones each air loop serves. Initialization Calculations One Time Calculations Zones Served by System The main one time calculation involves figuring out what zones are served by each air loop. Input data The input data specifying an air loop consists of: 1) the loop configuration. The EnergyPlus input does not explicitly describe which zones receive supply air from a given air loop. 4/23/05 215 . 4) design flow rate. 2) Perform air loop initialization calculations: a) at the beginning of the simulation (one time initializations). do it. 4) Otherwise perform a simulation of the air loop. For each air loop. b) at the start of each environment (design day or simulation run period). The utility routines used to get and check the data are described in the EnergyPlus Guide for Module Developers. (b) Components on the Branches 2) loop control. (a) Controllers. c) before each air loop simulation. (a) Splitters. get the data and store it in the air loop data structures. the following calculation is performed. section Input Services. (b) System Availability Managers. 3) connection to zone equipment.

(5) The case where there is no Zone Splitter for a Zone Supply Air Path must be handled. (6) The list of cooled and heated zones are saved in the air loop data structure AirToZoneNodeInfo. node = mbr .0. humidity ratio = 0. Branch Sizing If this not an air loop sizing calculation. A match indicates that this zone is served by heating supply air from the air loop. loop over all the nodes on the branch and set the node data to the following values: Tnode = 20o C Wnode = Woa hnode = PsyHFnTdbW (20. The actual calculation is described in the Sizing section of this document.0 4/23/05 216 .0 msetpt . For each branch.min = ρ std ⋅ Vbr . Initialize each branch mass flow rate: mbr . loop over all the branches in all air loops and trigger the branch design air flow auto-sizing calculation. In this case the program looks for a match between the zone equipment inlet node and an Air Distribution Unit heating or cooling inlet node. standard pressure.min where ρ std is the density of air at 20 degrees C. (2) This node number is compared to the inlet node of all Zone Supply Air Paths. When a match is found this zone is identified as being served by cooling supply air from the air loop. When a match is found.node = 0.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PRIMARY AIR SYSTEM SIMULATION (1) For each air loop outlet branch. the corresponding zone equipment inlet node is identified.node = 0. the Zone Splitter for this ZoneSupplyAir Path is identified.max mmax.node = mbr . When a match is found that zone is identified as being served with heating or cooling supply air from the air loop. (4) Similarly the outlet nodes of the Zone Splitter are compared with the heating inlet nodes of all zone Air distribution Units.max mmax avail . but is the first pass through the HVAC code in a normal simulation.max = ρ std ⋅ Vbr . loop over all the branches in the loop.Woa ) mnode = mbr .max mbr . Begin Environment Initializations For each air loop. (3) The outlet nodes of the Zone Splitter are compared to the cooling inlet nodes of all the zone Air Distribution Units.max mmin.

node = mmax.baro Qunode = 0. Thus a single call to SimAirLoops results in a simulation of all the central air system equipment. at the start of each HVAC solution sequence.node = 0. At the end of the first pass of loop 2. The lack of a system mass balance requires a resimulation of the central air system: i.node = 0.outletnode zone requirement. PsyHFnTdbW is the EnergyPlus psychrometric function for enthalpy h.node .0 . a second pass in loop 2. simulate in sequence each component During and at the end of each loop some tests are performed. given temperature and humidity ratio. For each air system. Loop over all of the central air systems (Air Primary Loops). 1) On each air loop. loop over the outlet branches and find the loop outlet nodes. Then set the mass flow rate setpoints for the air loop nodes. zone eq inletnode . This is usually the case. Central air system simulation The subroutine SimAirLoops does the actual simulation the central air system equipment for all the air loops.e. repeat the simulation of all the Air Primary Loop components until the controller has converged Loop over each branch in the Air Primary Loop On each branch. loop over all the nodes on the branch and set msetpt . make 1 or 2 simulation passes Loop over each controller in the Air Primary Loop For each controller. This sets the air loop flow rate to the total Otherwise. if one exists. If it is the outletbr . loop over the inlet nodes and. System Time Step Initializations For each branch in each air loop.0 pnode = pstd . 4/23/05 217 . SimAirLoops accomplishes its task using a set of nested loops. set the entering node mass flow rate equal to the primary air system design mass flow rate (subject to it not being larger than the entering node mass flow rate setpoint). This will insure that start of an HVAC solution sequence..max . and Qu is quality. 2) Pass the mass flow rate setpoint upstream to the start of the outlet branches.0 where Woa is the humidity ratio of the outside air. A call to ResolveSysFlow checks the mass balance and imposes a mass balance if there is not a balance. if it is the start of an HVAC solution sequence set mmax avail . 3) Sum the total return air mass flow rate and save it in the AirLoopFlow data structure. through the splitter.outletnode during the first pass through the full loop that the mass flow rate will be at the maximum. The simulation of a full air loop (central equipment plus zone terminal units and equipment) requires the interaction of 2 managers: ManageAirLoops and ManageZoneEquipment. and upstream to the beginning node of the splitter inlet branch. set setpt . set setpt . a decision is made on whether a second pass is needed. m =m m =m 4) For each air loop.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PRIMARY AIR SYSTEM SIMULATION mmin avail . but this is one part of a larger iterative simulation of the full air loops involving the zone equipment as well. The first pass has been performed assuming that there is a mass flow balance in the central air system simulation.

Input data The input specifying a set of zone equipment consists of: 1) the Controlled Zone Equip Configuration object data. This flag is used at a higher level to decide whether the primary air system needs to be resimulated if an HVAC component is calling for economizer lockout. SimAir) END IF CALL RecordZoneEquipment(SimAir) CALL ReportZoneEquipment SimZone = . At the end of the primary air system simulation a call to UpdateHVACInterface passes the primary air system outlet node data to the zone equipment inlet nodes. steps 5 & 6 are repeated one more time to ensure all the components on the loop have final values. get the data and store it in the zone equipment data structures 2) Perform zone equipment initialization calculations. exhaust nodes. Like the other managers. If convergence is achieved loop 3 is exited. 4/23/05 218 . 5) Transfer the zone equipment outlet node data to the inlet nodes of the primary air systems and check for convergence (done in RecordZoneEquipment by calling UpdateHVACInterface). RETURN 1) If the user input data has not been input. END IF CALL InitZoneEquipment(FirstHVACIteration) IF (ZoneSizingCalc) THEN CALL SizeZoneEquipment ELSE CALL SimZoneEquipment(FirstHVACIteration. zone node. 3) If calculation of the design zone air flow rates and loads needs to be done. the zone equipment simulation manager subroutine. If the data across the supply side – demand side gap doesn’t match to within a preset tolerance. 4) Otherwise simulate all of the zone equipment. outlet node. (a) the zone connection to the air loop – air inlet nodes.. do it. After all the controllers on a loop are converged. Finally a flag indicating whether the economizer is active is set.FALSE. the flag SimZoneEquipment is set to true to ensure that the zone equipment side gets resimulated.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES ZONE EQUIPMENT SIMULATION In loop 3 a call to ManageControllers simulates controller action and checks for controller convergence. Zone Equipment Simulation When the EnergyPlus HVAC simulation manager needs to simulate the zone equipment side of the air loop it calls ManageZoneEquipment. ManageZoneEquipment has a very simple structure: IF (GetInputFlag) THEN CALL GetZoneEquipment GetInputFlag = . The results of this calculation are stored in the zone sizing data structures and used by the component automatic sizing algorithms and the central system sizing calculations. fan coils etc. (b) the components serving each zone – air terminal units.False.

BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES ZONE EQUIPMENT SIMULATION 2) zone supply air path data. These mass flow rates are then passed back upstream to the air loop demand-side inlet nodes at the end of each zone equipment simulation sequence. Simulation The subroutine SimZoneEquipment does the actual simulation of all the zone equipment. the components are simulated in the order specified by the user assigned priority given in the Zone Equipment List object.0. First the supply air plenums and zone splitters are simulated in their input order. poa is the pressure of the outside air.Woa ) mnode = 0 Qunode = 1. (a) zone splitters and supply plenums. These demand-side inlet mass flow rates are then used as mass flow rate setpoints in the air loop supply-side simulation. Note that this includes components that are part of the demand side of an air loop as well as components that are independent of any air loop. If multiple air-conditioning components are attached to a zone. Each air terminal unit sets its inlet node to the air mass flow rate required to satisfy its zone load. given temperature and humidity ratio. (a) zone mixer and return plenums. Initialization Calculations One Time Calculations There are no one time calculations for zone equipment Begin Environment Initializations For each controlled zone initialize the zone inlet. except set the exhaust nodes mass flow rate and min and max available mass flow rates to zero. set the exhaust nodes conditions equal to the zone node conditions. exhaust and zone nodes to standard conditions.0 pnode = poa where Woa is the humidity of the outside air. and Qu is quality. System Time Step Initializations For each controlled zone. Then the air terminal units are simulated followed by the zone return plenums and zone mixer. 4/23/05 219 . PsyHFnTdbW is the EnergyPlus psychrometric function for enthalpy h. For zone equipment components that are part of the demand side of an air loop the simulation sequence is effectively performed in the direction of the air flow. 3) zone return air path data. Namely: Tnode = 20o C Wnode = Woa hnode = PsyHFnTdbW (20.

there may be multiple return plenums. (a) Loop over each component (supply plenum or zone splitter) on the supply air path and simulate each component. This reverse order simulation passes the air terminal units inlet mass flows back upstream to the return air path inlet node. (2) Loop over all the controlled zones. The Direct Air unit allows the program to know what zone this branch of the air system is attached to and a place to input the maximum air flow rate. 2 for dual duct systems). For the Direct Air objects to work correctly. (ii) Increment the required system output. Sometimes. The supply air paths consist of any combination of zone splitters and zone supply air plenums as long as it forms a tree structure with the supply air path inlet node the root and the air terminal unit inlet nodes as the leaves. (3) Loop over all the supply air paths (a) Loop over the components on each supply air path in reverse input order. The components are simulated in input order. 1 for single duct systems. (i) Simulate each component. (1) Loop over all the supply air paths. For each full air loop there should be one return air path. it is desirable to supply central system air directly to a zone without any zone level control or tempering. (a) Set the required system output. Direct Air Central system air is usually supplied to a zone through a terminal unit such as a single duct VAV reheat box. Here energy not included in the zone energy balance such as light-heat-to-return-air is added to the return nodes of the controlled zones. however. (a) Loop over each component (return plenum or zone mixer) on the return air path and simulate each component. it is important in any systems including Direct Air objects for the sum of the maximum zone air flow rates to be equal to the maximum central system flow rate. This signals the HVAC manager that the supply-side of the air loop needs to be resimulated. This inlet flow rate can be adjusted by a schedule. The return air path configuration is limited to a single mixer. An example would be Furnace or Central DX equipment. (b) Check to see if the supply air path inlet node mass flow rate has changed. The Direct Air object creates the capability of supplying central system air directly to a zone and only contains the zone inlet node. (4) Calculate the zone air flow mass balance – the zone inlet and exhaust mass flow rates are summed and the zone node and return air node mass flow rates are determined by a mass balance for each zone. The zone maximum flow rates are specified in the direct air inputs. (b) Loop over the components serving the zone in the user prioritized order. This completes a single simulation sequence of all the zone equipment. It can be thought of as a balancing damper in the duct branch going to the zone. This can be thought of as a seasonal adjustment of the balancing damper. 4/23/05 220 . (6) Loop over all of the return air paths. (5) Calculate the conditions at each zone return air node. If it has set the SimAir flag to true. Direct air is the component used to pass supply air directly into a zone without any thermostatic control. This node is both the zone inlet node and the outlet node of the Zone Splitter.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES DIRECT AIR For each full air loop there should be 1 supply air path for each primary air system outlet (i. The central air system flow rate is specified in the Air Primary Loop input and also in the air loop branch and central fan inputs.e.

and the chilled water flow rate must satisfy mass continuity. methods were developed to model the system air loop and its interactions with the zones due to temperature controls and the relative difference between the zone and supply air temperatures. a plant loop. The plant demand loop contains equipment that places a load on the primary equipment. 4/23/05 221 . In EnergyPlus the performance of the systems and plant are interdependent because the simulations are combined. There are no explicit limits on the number of loops within the program—the user is only limited by computer hardware. A user may have any number of each type of loop in a particular input file. Current Primary System Modeling Methodology There are three main loops within the HVAC simulation in the new program: an air loop. called either heating or cooling coils. The load is met by primary equipment such as chillers or boilers on the plant supply loop. Execution speed will naturally vary with the complexity of the input file. Each plant supply loop must be connected to a plant demand loop and vice versa.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS Plant/Condenser Loops Integration of System and Plant In order to integrate the air handling system simulation with the zones simulation. radiant systems. In addition. That is. This might include coils. baseboards. and a condenser loop. Plant and condenser loops are broken into supply and demand sides. coil controls are usually necessary to ensure that the values of chilled water flow variables entering and leaving the coil remain in a reasonable range. etc. A similar situation is encountered when integrating the central plant simulation. the central plant interacts with the systems via a fluid loop between the plant components and heat exchangers. Typically. the temperature of the chilled water leaving the plant must equal the temperature of the water entering the coils. A similar breakdown is present on condenser loops where the demand side includes the water side of condensers while the supply side includes condenser equipment such as cooling towers. Main loops are further divided into “sub-loops” or “semi-loops” for organizational clarity and simulation logistics (see Figure “Connections between the Main HVAC Simulation Loops and Sub-Loops”). The specific controls vary from application to application but two common possibilities are: maintaining a constant coil leaving air temperature or limiting the water temperature rise across the coil. The plant outputs must match the system inputs and vice versa. These sub-loops are matched pairs that consist of half of a main loop. The air loop is assumed to use air as the transport medium as part of an air handling system while the plant and condenser loops may use a fluid of the user’s choosing (typically water).

Note that since plant supply and demand are broken up into two separate sub-loops chillers or boilers may be in parallel to each other in the supply side and coils may be in parallel to each other on the demand side. within any branch. Condenser Loop Demand Supply Cond. For example. The fluid side of the coil is handled within the plant demand side. The breakdown into sub-loops allows for better handling and control of information and simulation flow throughout the program. in parallel. Plant Equip. the restriction of only a single splitter and mixer on a particular sub-loop does not limit the normal configurations. the segment between point A and point B is defined as a branch. Branches further divide the sub-loops into groups as they would appear within any HVAC system. Overall iterations ensure that the results for the current time step are balanced and updated information has been passed to both sides of the sub-loops as well as across to the other side of indirect connections such as coils. Thus. as is the section between points E and F. a sub-loop does not require a splitter or mixer if all equipment on the sub-loop are simply in series—this would correspond to a single branch that would define the entire sub-loop. Connections between the Main HVAC Simulation Loops and Sub-Loops.. or both with some restrictions. coils (heating or cooling) are in reality heat exchangers with an air and a water or refrigerant side. and condenser loops are enhanced by indirect connections between the various main loop types. though sub-iteration loops are maintained between the two sides of any loop to speed convergence. Branches are defined as individual legs within the loop structure. Elements can be defined in series. there can only be elements in series and not in parallel. plant. etc. Zones & Equip. Towers. Coils. Also. Thus. The air side of the coil is handled within the air loop where the control of the device is also maintained. Baseb. All loops are simulated simultaneously. etc. etc. Each sub-loop may only have one splitter and one mixer. future link Figure 83. Figure “Branch Layout for Individual HVAC Sub-Loops” provides an overview of a generic sub-loop representation. Wells. equipment may be in parallel between the mixer and splitter. The topology rules for individual subloops allow a reasonable amount of flexibility without requiring a complicated solver routine to determine the actual flow and temperature conditions. Thus. however. 4/23/05 222 .BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS Building Systems Simulation Manager Zone Conditions Predictor Simulate Building Systems Zone Conditions Corrector Air Loop Main Air Handler Plant Loop Demand Supply Coils. which passes the energy requirements of the coil on to the plant supply side.. There may be multiple sections (C1 to D1 through Cn to Dn) in between the splitter and mixer. Direct connections between the sub-loops of the air.

This information is then used either by the next component on the branch or establishes the outlet conditions for the branch. flow rate. 1 to j Components Dn Figure 84. maximum allowed (design) flow rate. enthalpy.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS A 1 to m Components B n Splitter C1 1 to i Components D1 n Mixer E 1 to k Components F Cn . and minimum available flow rate. flow rate. the loop manager would call the appropriate module to simulate (in order) all of the components on each branch of the loop except for splitters and mixers. This would be based purely on the component’s own control scheme and thus each component would be free to request as much (or as little) flow as desired. This involves making the supply side meet a particular load based on the simulation of the demand side loops. and the algorithms described are important to how the HVAC simulation functions. maximum available flow rate. In this step. each branch is made up of one or more components linked together in series. In the first step. Components on the branch take the conditions of the node at their inlet and use that information as well as overall control information to simulate the component and write the outlet data to the node following the component. These issues are discussed in the next several subsections. each component would set the conditions at the outlet node including temperature. Load distribution is an issue that must be addressed as well as how flow rates are adjusted and temperatures are updated. etc. Branch Layout for Individual HVAC Sub-Loops Essentially. The branch has an information node containing properties of the loop (temperature.. 4/23/05 223 .) at the beginning and end of the branch as well as between components. minimum allowed (design) flow rate. Plant Flow Resolver Overview of the Plant Flow Resolver Concept One of the most important aspects of the solution procedure within the plant and condenser loops of the new program is the method used to solve the various sub-loops..

chillers. For this iteration.g. If the user designates a constant speed pump that is operating continuously. two of which are defined by user input. This information is important not only to the pump itself but also to other pieces of equipment which may control their flow rates and thus require knowledge of the limits within which they may work. whether the pump operation is continuous or intermittent. the loop manager would resolve the flow at all nodes and through all branches of the local loop. For constant speed pumps. The operation of a constant speed pump is fairly straightforward.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS In the second step. because there is the possibility that the coils or some other component might request more flow in future iterations during the same time step. the pump will operate and select a flow somewhere between the minimum and maximum limits. In order to allow specification of realistic configurations. In these cases where the pump is running.g. and whether or not there is a load on the loop. These three deciding factors are whether the pump is constant or variable speed. In total. As a result. the first estimate of flow requested by the demand side tends to be fairly accurate and the flow rate does not vary in subsequent iterations. then the pump can shutdown. a rules-based “flow resolver” was developed for the 4/23/05 224 . However. In many cases. If there is no load on the loop and the pump is operating intermittently. This setpoint temperature is the fluid temperature that one is attempting to maintain at the outlet of the supply side and can be scheduled to different values on an hourly basis. This may have the net effect of adding heat to the loop if no equipment is turned on. In addition. loop managers were designed to support both semi-deterministic models (e. the flow resolver sets the flow rate through each loop component. the pump will run regardless of whether or not there is a load. How it reacts depends on several different conditions. the maximum and minimum flow rate values are the same and thus if the flow requested does not match this. the parameter estimation models of the ASHRAE Primary Toolkit [Pedersen 2001]) and “demand based” models (e. the performance map models of BLAST and DOE2. the maximum and minimum flow rates are set by the user-defined limits. The pump is simulated first on the supply side loop after the demand side loop has determined what the demand on the loop will be. it will try to meet the flow request made by demand side components. For any other condition such as the loop having a load and the pump is operating intermittently or the pump is continuously operating (regardless of the loading condition). Plant/Condenser Supply Side Component models. A variable speed pump is defined with maximum and minimum flow rates that are the physical limits of the device.1E). The components are then simulated with the corrected flows. condensers and cooling towers are simulated on the supply side of the plant and condenser loops. In general. The load is simply calculated by multiplying the requested flow rate from the demand side by the difference between the enthalpy at the supply side inlet and the enthalpy that corresponds to the current loop setpoint temperature. such as boilers. the decisions on what to set the maximum and minimum flow rates is directly related to the type of pump (constant or variable speed). If the pump is constant speed and operates intermittently. For variable speed pumps. In order to achieve these design criteria without resorting to a pressure based flow network solver in the HVAC portion of the code. the program must not only set flow rates but also maintain a record of the current maximum and minimum flow rate limits. the loop manager must be able to simulate models that require the mass flow rate as an input and models that calculate the mass flow rate as an output—sometimes in the context of a single loop configuration. the pump will run at its capacity if a load is sensed and will shut off if there is no load on the loop. there are three different decision variables. Pump Control for Plant and Condenser Loops. The pump is quite simply the component that drives the flow (also see Pumps). the plant and condenser supply side loop managers were designed to support parallel-serial connection of component models on the loop. the other components must either deal with the flow or a bypass branch must be available to handle the excess flow.

An ACTIVE type describes a demand based plant model that calculates mass flow rate as an output. The BYPASS type designates a loop bypass. i. Each component is tagged in the user input file as an ACTIVE..e. The flow resolver is based on the following assumptions and limitations: Each loop is only allowed to have a single splitter and a single mixer Due to the fact that there can only be one splitter and one mixer on a given loop. maximum allowed (design) flow rate. As previously discussed. 1 Pump sets branch 1 mass flow rate 2 Predictor estimates mass flow rate of each parallel branch 4 3 Correcter adjusts mass flow rate of each parallel branch to enforce loop mass balance Figure 85. flow rate. These predicted values are based purely on the component’s own control scheme and thus each component is free to request as much (or as little) flow as desired. maximum available flow rate. minimum allowed (design) flow rate. the pump establishes the total loop mass flow rate by setting the flow in the first supply side branch. The loop manager calls the appropriate module to simulate (in order) all of the components on each branch of the loop except for splitters and mixers. In the second step. A PASSIVE type describes a semi-deterministic model that is simulated with the mass flow rate as an input. a branch that contains a bypass may have no other equipment on that branch Equipment may be in parallel only between the splitter and mixer component of a loop or between one of those types of equipment and the loop inlet/outlet nodes Equipment may be hooked together in series in each branch of the loop Flow rates on individual branches will be controlled using maximum and minimum available flow rate limits The flow resolver employs a simple predictor-corrector algorithm to enforce mass continuity across the plant loop splitter as shown in the following figure. 4/23/05 225 . and minimum available flow rate. Plant/Condenser Supply Side Solution Scheme. a predictor algorithm polls each piece of equipment on the loop and “predicts” branch mass flow rates based on the requested flow rate for each. each component sets the conditions at its outlet node including temperature. PASSIVE or BYPASS type of model. In this step.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS EnergyPlus plant and condenser supply side managers. it follows logically that there can be at most one bypass on each loop No other components may be in series with a bypass.

the BYPASS. then it is assumed that the first active component on the branch is the highest priority and dictates the flow request. then the remaining flow is divided between the PASSIVE branches and as a last resort. individual branch flow rates must be adjusted to provide for the extra flow or the “flow deficit”. In the demand side scheme. If there is insufficient flow to meet the branch demand.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS The predictor algorithm first establishes the desired flow rate of each branch by searching for ACTIVE components on the branch. On the demand sides. The most restrictive flow constraints are assumed to be valid for the entire branch regardless of component type. also in reverse order. The user may specify individual components as either active or passive. When all of these branches have been exhausted. the zone equipment. Thus. 4/23/05 226 . the branch flow rates and constraints between the splitter and mixer can be adjusted. If there is more than one active component on a particular branch. the mixer and any branch following the mixer is completely passive. When there is a mismatch between the outlet constraints and the inlet constraints. ACTIVE branch requests are met first in the order that the branches appear in the branch list in the input file. The flow rate is resolved first for each individual branch. the loop manager makes any necessary “corrections” to the requested branch flows in order to enforce overall continuity on the loop. If mass conservation allows all ACTIVE branches to be satisfied. the splitter and mixer must resolve the various flow requests. all of the control happens at the splitter. If there is not enough flow for all of the requests. The splitter first attempts to sum the maximum and minimum constraints from all of the branches coming out of the device and compares those to the constraints that are valid for the branch leading into the splitter. The first ACTIVE component in simulation order sets the desired branch flow. Branches with a BYPASS component have a branch flow only when all other branches combined cannot handle the entire loop flow. Once the flow has been resolved at the splitter. this will be mandatory to maintain a mass balance at the splitter. For every branch. On the supply sides. all of the components have already been simulated and controlled by the air loop. the simulation will defer to the inlet constraints due to the fact that the pump is in reality controlling flow on the loop. Thus. there is a load management scheme and other constraints that must be resolved. When the flow rate coming out of the splitter does not match the branch requests. flow rates will be decreased in a similar order: passive branches first in reverse order followed by active branches in reverse order. The reverse order guarantees that the branch appearing first has the highest priority to receive the flow rate is has requested. Since the constraints of the pump would be passed across to the demand side from the supply side. the demand side management module only needs to resolve the actual flow rate through each section or branch of the subloop and also monitor the maximum and minimum flow rates that are available. Plant/Condenser Demand Side The plant and condenser demand side are simulated in a different manner than the supply sides because in reality there are no components to simulate or control. When there is extra flow. flow will be increased to the active branches. In the third step. an assumption is made that the coils or other demand side components must live within the bounds of the pump. the program cycles through each node on the branch and determines what the flow requests and flow limits are. if necessary. Flow rates are increased or decreased until a mass balance at the splitter exists. Once all of the branches have set their flow rates and constraints. there is also a defined scheme for assigning priority to components that will have the ability to control the flow. Branches with only PASSIVE components require a flow rate between the minimum and maximum allowable branch flow. Active components are given highest priority for requesting a particular flow rate. Since there may be several components in series on a particular branch. In some cases. flow is sent through any bypass branch first and then is sent to passive branches in reverse order of their appearance in the splitter outlet list. or the plant supply side.

(Tdi. this node becomes an energy storage location and its temperature reflects the current capability of the supply side. The user must estimate a fluid volume based on the size of the pipes in the loop. This means there are more degrees of freedom that must be controlled. 4/23/05 227 . Once setpoint temperature is reached. it was decided to add thermal capacitance to the plant loop and benefit from the added stability. Based on that experience. each loop is assigned a fluid volume as user input. The controllers for these components must know if the constraints have been modified so that the simulation does not toggle between a component requesting a flow that the pump cannot meet and the pump then resetting the flow to what it can provide. If the loop setpoint cannot be maintained. The flow resolver concept discussed previously controls the flow rates through the components and maintains an overall mass flow balance through the loop. (Tso) and the demand inlet temperature from the previous timestep. the components must be resimulated by the controlling loop (air loop. Note that once a flow rate for any component has changed that this signals the need to resimulate any sub-loop to which it might have an indirect connection. However. The size of the thermal capacitance affects the speed of recovery from situations where the setpoint was not maintained. No energy is lost or gained because of storage in the loop capacitance. This is used to determine a capacitance concentrated in the supply side outlet node. To implement the capacitance. Similarly. if the condenser demand side simulation results in a change in flow rate through a chiller condenser. the calculated supply side outlet.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS It is also necessary to monitor the flow constraints at the branches and components since once the flow rates are changed. Loop Instability Induced by the Capacitance Calculation The demand inlet temperature. then the plant supply side must be triggered to perform its calculations again. the simulation must recalculate the conditions on both the air loop and zone equipment sub-loops since coils and other equipment could be on either side of the main air loop. the storage effects are not involved. as shown in Figure “Demand and Supply Side Loops”) is calculated at the end of each Iteration of the HVAC simulation. The supply side outlet node temperature and the demand side inlet temperature proceed through smooth paths from one time step to the next. it was necessary to use a third order numerical solution there. In that case. or plant supply side). a simple first order solution has been found to be satisfactory. Care has been taken to avoid cases where the various sub-loops might simply keep triggering the resimulation of their indirect connections in an infinite loop. based on the loop mass. Since the plant loop thermal capacitance is higher. This situation is somewhat analogous to that existing in the link between the zone and the air system. zone equipment. the convergence and stability of the iterative solution was greatly improved by adding the thermal capacitance of the zone air and other fast responding mass within the zone. Currently. the temperatures still need to be controlled. Temperature Resolution The transition from load or energy based plant models to a loop based arrangement makes variables of both the flow rate and the fluid temperature. However rough estimates seem to be sufficient. this means that if a flow rate on the plant demand side changes. A purely iterative procedure can be expected to converge to the appropriate loop temperatures. but the procedure can become slow to converge under conditions where the demand changes rapidly or the supply components may not have enough capacity to meet the system demand. Because the thermal capacitance in the zone/system interaction is relatively small.

(m * t s sys *3600 ) M ≤1 A mass ratio greater than one implies that the water will be pumped around the loop more than once during a single system timestep. This formulation is stable as long as the mass ratio is less than one. and redefining the time step has vast implications throughout the simulation. The formulation assumes that the water in the loop is “well mixed”. ⎛ ( ms * tsys *3600 ) ⎞ Tdi − new = Tdi −old + ⎜ * (Tso − Tdi ) ⎟ ⎜ ⎟ M ⎝ ⎠ There is the option of shortening the time step or increasing the mass to maintain stability. Demand and Supply Side Loops The capacitance of the water in the loop is accounted for by balancing the energy added to the total mass of the water in the loop. M [kg]. and then converts this to a mass quantity.8. Upon entering the Loop Volume with the “autosize” option the program calculates the stable volume with the stability criteria equal to 0. Since the mass quantity is put in to stabilize the loop as its 4/23/05 228 . with the energy added during the current timestep. The resulting gain on the temperature calculation would lead to instability in the calculated loop water temperature. M ( Tdi − new − Tdi −old ) = ( ms * t sys *3600 ) * ( Tso − Tdi ) Where: Tdi −old Previous system time-step demand inlet temperature [°C] Tdi − new Current demand side inlet temperature [°C] ms Maximum expected supply side mass flow rate [Kg/s] tsys System time step [hr] Tso Supply side outlet temperature [°C] M Mass of the water in the loop [Kg] The new inlet temperature is calculated as shown in previous equation. The program is driven top down by time step.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS Demand Outlet Supply Inlet Tdo Demand Side Components Tdi Demand Inlet Tsi Supply Side Components Tso Supply Outlet Figure 86.

The BRANCH definition is inputted in simulation and connection order for all of the components on the branch. The Plant Flow Resolver Algorithm The flow resolver is called from the plant and condenser loop managers as shown below. The flow control algorithm described in the next section determines the flow through the loop uses the branch control types: ACTIVE. Adjust branch flows if necessary to achieve a mass balance on the loop Resimulate Components with adjusted flows. the lower limit will be checked and larger loop volumes can still be entered to simulate large. The BYPASS control type designates a loop bypass. massive systems. An ACTIVE control type describes a demand based plant model that calculates mass flow rate as an output of the model.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS primary purpose this quantity is modified to retain loop stability. The Plant Flow Resolver Input The input specifically related to the flow resolver consists of the Plant BRANCH list and the Plant CONNECTOR LIST as shown in the Input Output Reference. The SPLITTER and MIXER syntax in turn define the relative connection of the branches to each other on the loop. Examples of all the input syntax is shown in the Input/Output Reference for the appropriate object. PASSIVE. A PASSIVE control type describes a model that is simulated with the mass flow rate as a model input. DO All Branches CALL Sim Plant Equipment IF (Branch is Splitter inlet ) CALL Update Splitter END DO CALL Solve Flow Network DO All Branches CALL Sim Plant Equipment IF (Branch is Splitter inlet ) CALL Update Splitter END DO CALL Update Mixer CALL Sim Plant Equipment on Mixer Outlet Branch 4/23/05 229 . If the user needs to do a study and simulating a small loop is the major criteria. then the overall simulation time step can be shortened to facilitate. and BYPASS. users should be particularly mindful of both the maximum loop volumetric flow rate and the plant loop volume parameters in their input file. The user still has the capability to enter the Loop Volume. Any inconsistencies between these two parameters may lead the program to override the loop volume and lead to unexpected results. then the program will recalculate the loop mass. User defined names link the plant loop to its branches (contained in the BRANCH list) and define the loop splitters and mixers contained in the CONNECTOR LIST. Check Branch flows (downstream of splitter) to ensure that continuity is not being violated. The simulation assumes that the inlet node of the first component listed on the branch is the branch inlet node and the outlet node of the last component listed on the branch is the branch outlet node. Finally the loop mixer is updated and the mixer outlet branch is simulated. • • • • Simulate Components with the flow rate “requested” to meet loop demand. then the flow resolver sets the branch flows and the loops are simulated again. Thus. Note that if a user specifies a maximum loop volumetric flow rate and loop volume that may lead to instabilities based on the above noted convergence criteria. All branches are simulated with desired flow rates.

If there is insufficient flow to meet the branch demand. the BYPASS. Load Distribution Scheme 1 (“Optimal”) 1) Fill the component to a minimum PLR level.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS The flow resolver algorithm first establishes the desired flow rate of each branch by searching for ACTIVE components on the branch. If we have remaining loop demand (RLD). Load Distribution Scheme Load distribution consists of two steps: 1. 4/23/05 230 . Fill components up to optimum PRL level sequentially. then remaining flow is divided between the remaining PASSIVE branches and as a last resort. if possible. In order to avoid this problem. Step 2 may result in unmet demand since the remaining loop demand is less than the minimum capacity of a component. 2. If mass conservation allows all ACTIVE branches to be satisfied. Plant Components calls CalcCompCapacity for the calculation of component loads. Once DistributeLoad computes components’ loads. Branches with only PASSIVE components require a flow rate between the minimum and maximum allowable branch flow. Branches with a BYPASS component have a branch flow only when all other branches combined cannot handle the entire loop flow. The figure “Load Distribution Scheme” shows plant loop operation procedure with load distribution. For the component that has mininimum capacity > RLD. each plant component calculates its mass flow rate based on its own load. ACTIVE branch requests are met first in the order that the branches appear in the branch list in input. The first ACTIVE component in simulation order sets the desired branch flow. Figure 87. Summary of Load Distribution Schemes Two load distribution schemes are employed in EnergyPlus. and maximum part load ratio. two schemes are devised. DistributeLoad calls Plant Components in order to figure out component’s minimum. fill components to the maximum PRL level sequentially. optimal.

Since scheme 1 is trying to reach component’s optimal PRL level. Look in the Input Output Reference for the correct usage of these Set Point Managers. then each component can request a desired flow rate. When this value is determined then the Load Distribution scheme explained in the previous section takes this value and distributes the load to the appropriate equipment. 3) Re-compute RLD. If FlowLock = 0. Thus. Loop Demand Calculation Scheme SingleSetPoint The SingleSetPoint scheme for the Plant Loop takes the value that is placed on the Node%TempSetPoint and calculates the heating or cooling load necessary to obtain that setpoint. 4/23/05 231 . if in the DeadBand then no load is calculated. Load Distribution Scheme 2 (“Overloading”) 1) Fill the component to a minimum PRL level. 2) This component has FlowLock =2. There is a SingleSetPoint and a DualSetPointDeadband. The keys “Optimal” or “Overloading” designate the field. The demand calculation scheme determines how the load is calculated. 3) Disregard RLD.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS 2) Reset all other component loads to zero.1. then the flow resolver has locked down the branch flow rates at specified levels to maintain continuity. or 2. DeltaTemp = LoopSetPoint . If FlowLock =1. The designation for the load distribution scheme is added to the ‘Plant Loop’ object (ref Input Output Reference). In the SingleSetPoint scheme the Plant Loop requires that a Set Point Manager set a single setpoint value that sets Node%TempSetPoint. Examples of this set Point manager would be: Set Point Manager:Scheduled. If FlowLock=2. The Plant Loop Demand Calculation Scheme determines the amount of heating or cooling necessary to bring the temperature of the Plant Loop to its setpoint(s). 4) Redistribute remaining load to the reset components. Loop Demand Calculation Scheme DualSetPointDeadband The DualSetPointDeadband scheme for the Plant Loop takes the value that is placed on the Node%TempSetPointHi and Node%TempSetPointLo calculates the heating or cooling load necessary to obtain that setpoint. etc. and since scheme 2 enforces a component to have minimum capacity by adding a certain amount of load. Set Point Manager:Outside Air. it will be called ‘Optimal Operation’. IDF input file should have an appropriate key representing the operation scheme both under ‘Chilled Water Loop’ and ‘Hot Water Loop’. the SingleSetPoint is the default if that field is left blank in the Plant Loop object. Examples of this Set Point manager would be: SET POINT MANAGER:SCHEDULED:DUALSETPOINT. FlowLock is an integer variable with values of either 0. the flowrate for the component is set by the the load distribution scheme under scheme 2 above. For the DualSetPointDeadband scheme the Plant Loop requires that a Set Point Manager that sets the high and low setpoint values for Node%TempSetPointHi and Node%TempSetPointLo. Then the Load Distribution scheme distributes this calculated load to the appropriate equipment. Plant loop will overcool. The pseudo code below shows the basis of the algorithm. it will be called ‘Overloading Operation’. Summary of Plant Loop Demand Calculation Schemes There are two plant loop demand calculations schemes in EnergyPlus.LoopTempIn LoopDemand = mdot * Cp * DeltaTemp The sign of the Loop Demand determines if the loop has a cooling or heating load. In the next section is a summary of the 2 algorithms and how they are used.

The first scheme appearing in the list is given the highest priority.LoopTempIn) ! Possible combinations: ! 1 LoadToHeatingSetPoint > 0 & LoadToCoolingSetPoint > 0 --> Heating required ! 2 LoadToHeatingSetPoint < 0 & LoadToCoolingSetPoint < 0 --> Cooling Required ! 3 LoadToHeatingSetPoint < 0 & LoadToCoolingSetPoint > 0 --> Dead Band Operation ! 4 LoadToHeatingSetPoint > 0 & LoadToCoolingSetPoint < 0 --> Not Feasible IF (LoadToHeatingSetPoint . In other words. See the Input Output Reference for input field details.LT. 0. Load Range operation is used when the loop load is calculated and then the equipment is 4/23/05 232 .AND. Operation Schemes (Plant and Condenser) Plants and condenser loops must have some mechanism for controlling the operation of the loop and which equipment is available under different operating conditions.0 . this load needs to be allocated to the supply equipment according to the users input. its identifying name. the first operation scheme is available. there is a lower limit for the load range. then it is used by the simulation to define how the plant or condenser loop operates.LT. Plant Operation Schemes See the Input Output Reference for input field details. If it is not available.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS !Calculate the demand on the loop IF (mdot > 0. if according to its schedule. The options for plant control schemes are: Uncontrolled Loop Operation The Uncontrolled Scheme takes the full capacity of the supply equipment and cools or heats the loop accordingly. etc. the second scheme has second highest priority. LoadToCoolingSetPoint .0 . This is mainly done by the operation schemes.GT. 0. Cooling Load Range Based Operation or Heating Load Range Based Operation The “COOLING (or HEATING) LOAD RANGE BASED OPERATION” statement defines the different ranges and which equipment list is valid for each range. LoadToCoolingSetPoint . unless turned off by the loop flow resolver to maintain continuity in the fluid loop.AND. an upper limit for the load range.0 .0 END IF IF(ABS(LoopDemand) < LoopDemandTol) LoopDemand = 0. 0. and the schedule that defines its availability.0) THEN LoopDemand = LoadToHeatingSetPoint ELSE IF (LoadToHeatingSetPoint . An example would be a cooling tower where the cooling tower would cool the condenser loop with all of its available capacity and not be limioted by a capacity range or setpoint.GT. and a name that links to an equipment availability list (the “CONDENSER (or ‘LOAD RANGE’) EQUIPMENT LIST”). LoadToCoolingSetPoint .AND. If the loop runs.0) THEN LoopDemand = LoadToCoolingSetPoint ELSE IF (LoadToHeatingSetPoint . Uncontrolled loop operation simply specifies a group of equipment that runs ‘uncontrolled’. 0. this equipment will run also.GT. 0. 0.LT.0) THEN LoopDemand = 0. Each operation scheme must have the type of operation scheme.0 The sign of the Loop Demand determines if the loop has a cooling or heating load. Once the Loop load is calculated by the return conditions from the demand side and using the loop setpoint. Then the Load Distribution scheme distributes this calculated load to the appropriate equipment.0) THEN LoadtoHeatingSetPoint = mdot*Cp*(LoopSetPointLo . In each trio.LoopTempIn) LoadtoCoolingSetPoint = mdot*Cp*(LoopSetPointHi .0 ELSE CALL ShowSevereError END IF ELSE LoopDemand = 0. the second operation scheme in the list is checked to see if it is available until the highest priority scheme that is also available is found. if there is any.

Cooling Load Range Based Operation or Heating Load Range Based Operation The “COOLING (or HEATING) LOAD RANGE BASED OPERATION” statement defines the different ranges and which equipment list is valid for each range. ‘Outdoor…Range Based’ schemes compare the current value of an environmental parameter with user specified ranges of that parameter. This allows for the most efficient operation of the plant equipment or for the user to determine the most efficient plant configuration. Uncontrolled loop operation simply specifies a group of equipment that runs ‘uncontrolled’. Outdoor Dewpoint Range Based Operation. a series of data trios is expected. unless turned off by the loop flow resolver to maintain continuity in the fluid loop. an upper limit for the load range. and a name that links to an equipment availability list (the “CONDENSER (or ‘LOAD RANGE’) EQUIPMENT LIST”). The ‘Range Based Operation’ schemes select a user specified set of equipment for each user specified range of a particular simulation variable. there is a lower limit for the load range. ‘Load Range Based’ schemes compare the demand on the condenser supply side with specified load ranges and associated equipment lists. Uncontrolled Loop Operation The Uncontrolled Scheme takes the full capacity of the supply equipment and cools or heats the loop accordingly. The condenser operation schemes apply to the equipment on the ‘supply side’ of the condenser loop—pumps. 4/23/05 233 . cooling towers. Condenser Operation Schemes This is very similar to the Plant Operation Schemes. In each trio. ground coupled heat exchangers. etc. When the equipment list has been deteremined then the load is allocated to the equipment in a manner selected by the user with “Optimal or Sequential” Load Distribution Scheme. Outdoor RHPercent Range Based Operation The “OUTDOOR…BASED OPERATION” statements define the different ranges of the various environmental parameters and which equipment list is valid for each range. Load Range operation is used when the loop load is calculated and then the equipment is selected in the proper range. Outdoor Drybulb Range Based Operation. The load range based operation scheme has two statements associated with it: a main statement that defines the ranges that individual priority settings are valid and the lists of equipment that may be used for each range. but there are several more options avaible with the Condenser or Environmental Loop connection. Outdoor Wetbulb Range Based Operation. The load range based operation scheme has two statements associated with it: a main statement that defines the ranges that individual priority settings are valid and the lists of equipment that may be used for each range. After the keyword and the identifying name. If the loop runs. When the equipment list has been deteremined then the load is allocated to the equipment in a manner selected by the user with “Optimal or Sequential” Load Distribution Scheme. this equipment will run also.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS selected in the proper range. an upper limit for the load range. The keywords select the algorithm that will be used to determine which equipment is available for each time step. there is a lower limit for the load range. An example would be a cooling tower where the cooling tower would cool the condenser loop with all of its available capacity and not be limioted by a capacity range or setpoint. See the Input Output Reference for input field details. In each trio. This allows for the most efficient operation of the plant equipment or for the user to determine the most efficient plant configuration. and a name that links to an equipment availability list (the “CONDENSER EQUIPMENT LIST”).

Its main function is to resolve mass flow between the two nested plant loops based upon the control modes entered by the user. Outdoor Dewpoint Temperature Difference Based Operation. A difference range is specified for each equipment list. The hydraulic coupling is provided between the primary and secondary loop by the Plant Loop Connection Component. with a variable volume flow system to provide flow to multiple cooling coils. which looks at the difference between the tower inlet temperature and wet bulb temperature. The Plant Loop Connection Component is a control volume around some piping and bypass valve. The connection component does the mass flow resolution between the nested loops and does the proper mixing with the bypass as necessary. a diagram is shown below and a sample input file is included in the test suite. The initial implementation has a constant flow primary loop (Loop with Chillers) and a variable flow secondary loop (Loop with the Coils). Primary-Secondary Loop Systems The Plant Loop Connection Component is used to connect 2 plant loops together. Example of a Primary-Secondary Nested Loop Simulation The Plant Loop Connection Component is simulated on the Supply Side of the Secondary Loop and the results are provided to the Demand Side of the Primary Loop. This keeps the flow through the chillers at a fixed level and allows for variable flow in the secondary systems. This should facilitate the majority of the primary secondary systems being used today. The Delta Temperature based control strategy helps to control any condenser equipment based on the difference between a reference node temperature and any environmental temperature.. This system normally has multiple chillers on the primary constant volume loop. For example a cooling tower can be controlled by a strategy.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS Outdoor Drybulb Temperature Difference Based Operation. Outdoor Wetbulb Temperature Difference Based Operation. typically to create a Primary-Secondary plant loop system. Since the demand side of the primary loop is not simulated until the supply side of the secondary loop 4/23/05 234 . Figure 88. The Plant Loop connection component was designed to use the existing component/loop/solution structure to facilitate the simulation with the existing demand side manager and the supply side manager.

there is an additional iteration through the demand side that is necessary. either by no demand or the secondary pump scheduled off. . The component that can store the excess heat and allow it to be used elsewhere in the system or for domestic hot water is the Water Heater:Simple and is defined in the Input/Output Reference. i. There are some assumptions that are necessary: • • When the secondary side shuts down. These assumptions are essential due to the fact that control information is difficult to pass between the nested loops. the same scheduled needs to be applied to both of the pumps. and the ability to store or use that heat elsewhere in the system. only two loops can be simulated together.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS is simulated.e. Only the Plant Loop Connection Component triggers this additional demand-side iteration. only one level of nesting. and internal combustion and combustion turbine generators is designed to use the existing component/loop/solution structure to facilitate the simulation with the existing demand side manager and the supply side manager. Heat recovery normally contains components that produce heat that can be recovered. Example of a Heat Recovery Loop Simulation In the example above there is a chilled water Loop with chilled water supplied by a diesel engine driven chiller. the primary side will also shut down for that time step. Heat Recovery Loop Systems Heat Recovery is accomplished by specifying another set of supply and demand loops. If either the secondary or primary pump is scheduled. There is a hot water Loop that is being supplied by the water heater: 4/23/05 235 . engine driven and combustion turbine chillers. • Currently. This was necessary to keep the iterated information current and to ensure that an energy balance is achieved. i.e. Each of the heat recovery components. Figure 89.

4/23/05 236 . The Plant water heater simple can also be used to just meet scheduled domestic hot water use. or provide a hot water storage tank for heat recovery as a single function. internal combustion. Then on the demand side of the heat recovery Loop there is the engine driven chiller. but using the Plant supply and demand side configurations this can be extended to meet most user configurations. Example files of some of these configurations are provided in the internal test suite.BUILDING SYSTEM SIMULATION SYSTEM MANAGER / PROCESSES PLANT/CONDENSER LOOPS simple. Or any combination of the above can be configured. There is also scheduled domestic hot water usage on the water heater which excess demand can be met by a number of user-specified heating sources. provide a hot water source for Plant Loop equipment. This hot water is pump on the supply side by the heat recovery pump and provides the heat to the water heater to meet the water heater setpoint. This is probably one of the more complex configurations and interactions that would take place in heat recovery. and combustion turbine electric generators with specified mass flows to recover the heat.

(2) Set the ZoneSizingCalc flag equal to true. GetSystemSizingInput and GetPlantSizingInput reads in all the user sizing input contained in objects Sizing Parameters. Call ManageHeatBalance to do a full heat balance calculation for each zone. ZoneSizingCalc = true signals the HVACManager to ignore the real HVAC system and instead run the ideal zonal system (described below) used to calculate design loads and flow rates. system and plant level calculations of design heating and cooling capacities and fluid flow rates. ManageSizing performs the following tasks. The call to ManageHeatBalance also brings about an HVAC simulation. (1) By calling GetSizingParams. GetZoneSizingInput. System Sizing and Plant Sizing. 2. (c) Call UpdateZoneSizing(EndDay) to calculate peaks and moving averages from the zone design sequences for each design day. This section describes the sizing methodology implemented in EnergyPlus. Sizing Manager The sizing calculations in EnergyPlus are managed by a sizing manager contained in the software module SizingManager. (b) Loop over hours in the day (i) Loop over zone time steps in each hour 1. Group – Design Objects. (5) Set the SysSizingCalc flag equal to true. (b) Loop over hours in the day 4/23/05 237 . (a) Call UpdateZoneSizing(BeginDay) to initialize zone design load and flow rate sequences. (3) Loop over all the design days. Zone. Modular. The corresponding design load and flow rate sequences are saved for use in the system design calculations. A zone by zone heat balance load and air-flow calculation for multiple design days. (6) Call ManageZoneEquipment and ManageAirLoops to read in the zone and central system inputs needed for the system design calculations. This ends the zone design calculations. This method includes: 1. HVACManager also calls UpdateZoneSizing(DuringDay) to save the results of the ideal zonal system calculation in the design load and flow rate sequences. component-specific sizing algorithms for each HVAC component. efficient and flexible sizing calculations can present a serious barrier to the adoption and acceptance of building simulation programs. This starts the zone design calculations. The main sizing manager routine ManageSizing is called from ManageSimulation before the annual simulation sequence is invoked. The lack of reliable. 4. Significant user control with modest input requirements. 3.LOOP AND EQUIPMENT SIZING SIZING MANAGER Loop and Equipment Sizing The importance of correct equipment sizing is often ignored in discussions of building simulation methods. (4) Call UpdateZoneSizing(EndZoneSizingCalc) to calculate for each zone the peak heating & cooling loads and flow rates over all the design days. These objects and their data are described in the EnergyPlus Input Output Reference. Call ManageWeather to obtain outside conditions for this time-step. This starts the system design calculations. (7) Loop over all the design days. The program needs enough information to be able to figure out the overall air loop connectivity. (a) Call UpdateSysSizing(BeginDay) to initialize system design load and flow rate sequences. Zone Sizing. 2.

(1997). (8) Call UpdateSysSizing(EndSysSizingCalc)) to calculate for each system the peak heating & cooling loads and flow rates over all the design days. HVAC equipment sizing begins with the calculation of space heating and cooling loads. And this ends the tasks of the Sizing Manager. j for design days.LOOP AND EQUIPMENT SIZING ZONE DESIGN LOADS AND AIR FLOW RATES (i) Loop over zone time steps in each hour 1. Table 35. A space cooling (heating) load is defined as the rate at which heat must be removed (added) to a space to maintain a constant temperature. CalcZoneSizing(i. ZoneSizing and FinalZoneSizing includes the following data items. Zone Sizing Data Name All the data from ZoneSizingInput DesHeatMassFlow 4/23/05 Description the zone design heating air mass flow rate in 238 .j) corresponds to CalcZoneSizing but includes the effect of the user specified sizing factor or user specified zone design flow rate. 2. The index i is for the controlled zones. (c) Call UpdateSysSizing(EndDay) to calculate peaks and moving averages from the system design sequences for each design day.1. Different building designs. In the end the designer’s experience and engineering judgement will play an important role in any sizing calculation. The index i is for the controlled zones. The corresponding design load and flow rate sequences are saved for use in the system design calculations. The method used to size an HVAC system in a hot. This ends the system design calculations. The current industry standard method for calculating space loads is the heat balance method [ASHRAE Fundamentals (2001). FinalZoneSizing(i) corresponds to CalcFinalZoneSizing but includes the effect of the user specified sizing factor or user specified zone design flow rate. The type of building is also relevant – a simple watts per square foot loads estimate could be adequate for a building containing a network server farm while a detailed. Zone Design Loads and Air Flow Rates Overview There is no single best way to establish design HVAC flow rates and size HVAC equipment. Since EnergyPlus is a heat balance based simulation program it is straightforward for the program to use this method for calculating zone loads.. The data stored in CalcZoneSizing. Call ManageWeather to obtain outside conditions for this time-step. Call UpdateSysSizing(DuringDay) to save the results of the system design calculations in the system design load and flow rate sequences. Pedersen (2001). CalcFinalZoneSizing.j) stores the results of the zone design calculations for all zones and all design days. page 29. CalcFinalZoneSizing(i) stores the results of the zone design calculations for the peak heating and cooling cases for each zone. dynamic loads simulation would be necessary for a passive solar building. climates. and HVAC systems will impose varying constraints on the designer. Zone Design Data Arrays The zone design data arrays are: ZoneSizingInput(i) stores the input data from the Zone Sizing objects. ZoneSizing(i. moist climate such as Miami will be different than the method used for a building in Albuquerque. Pedersen et al.

LOOP AND EQUIPMENT SIZING ZONE DESIGN LOADS AND AIR FLOW RATES DesCoolMassFlow DesHeatLoad DesCoolLoad DesHeatDens DesCoolDens DesHeatVolFlow DesCoolVolFlow DesHeatCoilInTemp DesCooltCoilInTemp DesHeatCoilInHumRat DesCoolCoilInHumRat HeatMassFlow CoolMassFlow HeatLoad CoolLoad HeatZoneTemp HeatZoneRetTemp CoolZoneTemp CoolZoneRetTemp HeatZoneHumRat CoolZoneHumRat ZoneTempAtHeatPeak ZoneRetTempAtHeatPeak ZoneTempAtCoolPeak ZoneRetTempAtCoolPeak ZoneHumRatAtHeatPeak ZoneHumRatAtCoolPeak TimeStepNumAtHeatMax TimeStepNumAtCoolMax HeatDDNum 4/23/05 [kg/s] the zone design cooling air mass flow rate in [kg/s] the zone design heating load in [W] the zone design cooling load in [W] the zone design heating air density [kg/m3] the zone design cooling air density [kg/m3] the zone design heating air volume flow rate [m3/s] the zone design cooling air volume flow rate [m3/s] zone heating coil design air inlet temperature [C] zone cooling coil design air inlet temperature [C] the zone heating coil design air inlet humidity ratio [kg/kg] the zone cooling coil design air inlet humidity ratio [kg/kg] current zone heating air mass flow rate at the HVAC time step [kg/s] current zone cooling air mass flow rate at the HVAC time step [kg/s] current zone heating load [W] current zone cooling load [W] current zone temperature during heating [C] current zone return temperature during heating [C] current zone temperature during cooling [C] current zone return temperature during cooling [C] current zone humidity ratio during heating [C] current zone humidity ratio during cooling [C] zone temperature at maximum heating [C] zone return temperature at maximum heating [C] zone temperature at maximum cooling [C] zone return temperature at maximum cooling [C] zone humidity ratio at maximum heating [kg/kg] zone humidity ratio at maximum cooling [kg/kg] zone time step number (in the day) at the heating peak zone time step number (in the day) at the cooling peak design day index of design day causing heating peak 239 .

zone time step) [kg/kg] daily sequence of zone humidity ratios (cooling. subroutine ManageHVAC calls SimHVAC. zone time step) [C] daily sequence of zone return temperatures (cooling. This unit supplies heating or cooling air at a fixed. The Heat Balance manager then causes the HVAC Manager to be called in a manner identical to a normal simulation. In turn ManageZoneEquipment checks if ZoneSizingCalc is true. The system output Qsys is simply set equal to the zone demand – it is assumed that the ideal system can always meet the zone load. If it is true. The units have infinite capacity – the flow rate can be any amount. SimHVAC calls ManageZoneEquipment and returns. SimHVAC checks ZoneSizingCalc.LOOP AND EQUIPMENT SIZING ZONE DESIGN LOADS AND AIR FLOW RATES CoolDDNum MinOA HeatFlowSeq(i) CoolFlowSeq(i) HeatLoadSeq(i) CoolLoadSeq(i) HeatZoneTempSeq(i) HeatZoneRetTempSeq(i) CooltZoneTempSeq(i) CoolZoneRetTempSeq(i) HeatZoneHumRatSeq(i) CoolZoneHumRatSeq(i) design day index of design day causing cooling peak design minimum outside air [m3/s] daily sequence of zone heating air mass flow rates (zone time step) [kg/s] daily sequence of zone cooling air mass flow rates (zone time step) [kg/s] daily sequence of zone heating loads (zone time step) [W] daily sequence of zone cooling loads (zone time step) [W] daily sequence of zone temperatures (heating. if it is it calls SizeZoneEquipment rather than SimZoneEquipment. zone time step) [kg/kg] Zone Design Load Calculation As described in the preceding section. 2) If the system is active (zone temperature not in the deadband and zone load greater than 1 watt) the sign of the zone load is used to determine whether heating or cooling is required and Tin and Win are set to the appropriate values from the Zone Sizing input. 1) Loop over all the controlled zones.air ⋅ (Tin − Tz )) 4/23/05 240 . rather than simulating the actual system. The air flow rate corresponding to the load is just msys = Qsys /(C p . zone time step) [C] daily sequence of zone humidity ratios (heating. The ZoneSizingCalc set to true signals the HVAC Manager to ignore the actual HVAC system and instead calculate the design zone loads and air flow rates using an ideal zonal system. the Sizing Manager initiates the zone design calculation by looping over all of the design days and calling the Heat Balance Manager for each zone time-step in each design day. SizeZoneEquipment assumes that each controlled zone is served by an ideal air conditioning unit. In module HVACManager. zone time step) [C] daily sequence of zone return temperatures (heating. The calculation steps are as follows. user input temperature and humidity (specified in the Zone Sizing objects). zone time step) [C] daily sequence of zone temperatures (cooling.

3) The results for each zone are stored in the zone sizing data arrays. and user sizing multipliers or user specified design flows taken into account. a design day) name and number are stored in the zone sizing data structures DuringDay The calculated and stored sequences are summed or averaged over the zone time-step. EndZoneSizingCalc (1) Write out onto a comma-separated file the calculated design sequences for each zone: HeatLoadSeq. fixed-width averaging window to the sequences. zone coil loads calculated. peak values calculated for each design day. The width of the window is user specified in the Sizing Parameters input object. at the end of the design day (CallIndicator = EndDay) and at the end of the zone design calculation (CallIndicator = EndZoneSizingCalc).LOOP AND EQUIPMENT SIZING ZONE DESIGN LOADS AND AIR FLOW RATES If the system is not active. HeatFlowSeq. CoolFlowSeq and the corresponding peaks and volumetric flow peaks. the mass flow rate is set to zero and the system output is left at zero. The user modifications to the calculated sizing will be applied to and stored in ZoneSizing and FinalZoneSizing. BeginDay The environment (in this case. All of this input is treated as a sizing factor. It is called at the start of each design day (CallIndicator = BeginDay). a heating & a cooling load sequence chosen for each zone from all the design day results. possible further smoothing of results done. looking at the results for all of the design days. CoolLoadSeq. EndDay (1) Smooth the design sequences by applying a moving. The sequences that are smoothed are: (a) CoolFlowSeq (b) CoolLoadSeq (c) HeatFlowSeq (d) HeatLoadSeq (e) CoolZoneRetTempSeq (f) HeatZoneRetTempSeq (2) The peak heating and cooling loads and mass & volume flow rates are extracted from each set of design sequences. Updating and Adjusting the Zone Results The results from SizeZoneEquipment are at the system time-step and are for all design days. (3) The user can modify the calculated zone design results by specifying a sizing factor at the global or zone level or by specifying and actual design heating or cooling zone design volumetric flow rate. These results then need to be summed or averaged over the zone time-step. at the zone time-step (CallIndicator = DuringDay). (3) Using the time of the peak and the design outside air fraction the design zone heating and cooling coil inlet temperatures and humidity ratios are calculated. (4) For each zone. (2) The data in CalcZoneSizing and CalcFinalZoneSizing is moved to ZoneSizing and FinalZoneSizing. These tasks are accomplished by the subroutine UpdateZoneSizing. If the user inputs a 4/23/05 241 . the design days that cause the peak heating and peak cooling for that zone are chosen and the corresponding design sequences and peak loads and flow rates are saved in the CalcFinalZoneSizing array. This finishes the calculated – unmodified by the user – portion of the zone design calculation.

At this point the condition that the design flow rates are never allowed to be less than the minimum outside air flow rate is imposed. This is the data that will be used for sizing zone HVAC equipment and in the system sizing calculation. This concludes the calculation of the zone design flow rates and loads. the design flow and load sequences as well as peak loads and flows are multiplied by the appropriate sizing factor and stored in ZoneSizing and FinalZoneSizing. j for design days. System Design Data Arrays The system design data arrays are: SysSizInput(i) stores the input data from the System sizing objects. during. The index i is for the air loops. Table 36. The data stored in SysSizing. There is an update subroutine UpdateSysSizing called at the beginning. CalcSysSizing(i) stores the results of the system design calculations for the peak heating and cooling cases for each air loop. Note that the user can input a zone sizing factor or a zone design flow rate – not both – so there is never a conflict. SysSizing(i. The results of the system design calculation are stored in the system sizing arrays described below. CalcSysSizing and FinalSysSizing includes the following data items. System Sizing Data Name All the data from SysSizInput CoinCoolMassFlow CoinHeatMassFlow NonCoinCoolMassFlow NonCoinHeatMassFlow coincident peak cooling mass flow rate [kg/s] coincident peak heating mass flow rate [kg/s] noncoincident peak cooling mass flow rate [kg/s] noncoincident peak heating mass flow rate [kg/s] 242 Description 4/23/05 . The major difference is that this calculation is done at the zone time-step only.j) stores the results of the system design calculations for all systems and all design days. The calculation sequence for system level design loads and air flow rates resembles the calculation sequence for zone loads and air flow rates. The index i is for air loops. (4) Once the zone heating and cooling sizing factors are established. There is no idealized component calculation triggered at the system time-step as in the zone calculation. System Design Loads and Air Flow Rates Overview The purpose of the system design calculation is to estimate design heating and cooling loads and air flow rates for each air loop in the simulation problem. FinalSysSizing(i) corresponds to CalcSysSizing but includes the effect of the user specified sizing factor or user specified system design flow rate. The system design calculation operates at the zone time step using the design environment weather data and the data stored in the zone sizing arrays. (5) The outside air fractions are recalculated using the new user-modified design flow rates and new design zone coil inlet conditions calculated and stored. and end of a loop in the Sizing Manager over all the design days.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES cooling design volumetric flow rate for a zone it is divided by the calculated cooling design volumetric flow rate for the zone to give a zone cooling sizing factor.

LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES DesMainVolFlow DesHeatVolFlow DesCoolVolFlow SensCoolCap HeatCap PreheatCap CoolMixTemp CoolMixHumRat CoolRetTemp CoolRetHumRat CoolOutTemp CoolOutHumRat HeatMixTemp HeatMixHumRat HeatRetTemp HeatRetHumRat HeatOutTemp HeatOutHumRat HeatFlowSeq(i) CoolFlowSeq(i) SensCoolCapSeq(I) HeatCapSeq(i) PreHeatCapSeq(i) SysCoolRetTempSeq(i) SysCoolRetHumRatSeq(I) SysHeatRetTempSeq(i) SysHeatRetHumRatSeq(I) SysCoolOutTempSeq SysCoolOutHumRatSeq 4/23/05 design main supply duct volume flow [m3/s] design heat supply duct volume flow [m3/s] design cool supply duct volume flow [m3/s] design sensible cooling capacity [W] design heating capacity [W] design preheat capacity [W] design mixed air temperature for cooling [C] design mixed air humidity ratio for cooling [kg water/kg dry air] design return air temperature for cooling [C] design return air humidity ratio for cooling [kg water/kg dry air] design outside air temperature for cooling [C] design outside air humidity ratio for cooling [kg water/kg dry air] design mixed air temperature for heating [C] design mixed air humidity ratio for heating [kg water/kg dry air] design return air temperature for heating [C] design return air humidity ratio for heating [kg water/kg dry air] design outside air temperature for heating [C] design outside air humidity ratio for heating [kg water/kg dry air] daily sequence of system heating air mass flow rate (zone time step) [kg/s] daily sequence of system cooling air mass flow rate (zone time step) [kg/s] daily sequence of system sensible cooling capacity (zone time step) [W] daily sequence of system heating capacity (zone time step) [W] daily sequence of system preheat capacity (zone time step) [W] daily sequence of system cooling return temperatures (zone time step) [C] daily sequence of system cooling return humidity ratios (zone time step) [kg water/kg dry air] daily sequence of system heating return temperatures (zone time step) [C] daily sequence of system heating return humidity ratios (zone time step) [kg water/kg dry air] daily sequence of system cooling outside temperatures (zone time step) [C] daily sequence of system cooling outside 243 .

These activities are all performed within the UpdateSysSizing subroutine in the SimAirServingZones module. It is called at the start of each design day (CallIndicator = BeginDay). There is a logical flag SysSizingCalc corresponding to ZoneSizingCalc. The zone design flow rates for the zones served by an air loop are summed to obtain the system level design flow rates. Instead the system design loads and flow rates are calculated using the zone level results. at the zone time-step (CallIndicator = DuringDay). These air flows are mixed with the system level design minimum outside air flow rate to obtain system design coil loads. BeginDay (1) The environment (in this case. at the end of the design day (CallIndicator = EndDay) and at the end of the zone design calculation (CallIndicator = EndSysSizingCalc). (2) Loop over the zones cooled by this air loop: (a) NonCoinCoolMassFlowsys=∑DesCoolMassFlowzone (3) Loop over the zones heated by this air loop: (a) NonCoinHeatMassFlowsys=∑DesHeatMassFlowzone DuringDay (1) Loop over the zones cooled by this air loop: CoolFlowSeqsys(i) =∑CoolFlowSeq zone(i) SysCoolRetTemp(i)= CoolFlowSeqsys(i) ∑(CoolZoneRetTempSeq(i)• CoolFlowSeqzone(i))/ ∑(CoolZoneHumRatSeq(i)• CoolFlowSeqzone(i))/ SysCoolRetHumRat(i)= CoolFlowSeqsys(i) FracOA=ρair• DesOutAirVolFlowsys/ CoolFlowSeqsys(i) Tmix=Toutside• FracOA + SysCoolRetTemp(i)(1 – FracOA) Wmix=Woutside• FracOA + SysCoolRetHumRat (i)(1 – FracOA) SysCoolOutTempSeq(i)= Toutside SysCoolOutHumRatSeq(i)= Woutside Get the current (zone time-step) system cooling capacity: SysSensCoolCapcur=Cp. It is used to allow the component routines to distinguish a normal simulation call from a being called during a system sizing calculation.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES SysHeatOutTempSeq SysHeatOutHumRatSeq humidity ratios (zone time step) [kg water/kg dry air] daily sequence of system heating outside temperatures (zone time step) [C] daily sequence of system heating outside humidity ratios (zone time step) [kg water/kg dry air] System Design Flow Rate and Load Summation and Adjustment There is no system level subroutine corresponding to SizeZoneEquipment.air• CoolFlowSeqsys(i) •( Tmix-Tsup) SensCoolCapSeq(I)= SysSensCoolCapcur 4/23/05 244 . a design day) name is stored in the system sizing data structures.

Wmix is the current mixed air humidity ratio [kg water / kg dry air]. [J/kg-K]. Toutside is the current outside air temperature [C]. Tsup is the user specified design cooling supply temperature [C]. Tmix is the current mixed air temperature [C]. FracOA is the outside air fraction.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES If SysSensCoolCapcur is the maximum for the day so far then save SysSensCoolCapcur as the design value: SensCoolCap(i )sys= SysSensCoolCapcur And save the corresponding mixed. return and outside conditions: HeatMixTempsys= Tmix HeatMixHumRatsys=Wmix HeatRetTempsys= SysHeatRetTemp(i) HeatRetHumRatsys= SysHeatRetHumRat(I) HeatOutTempsys= Toutside 4/23/05 245 . Woutside is the current outside air humidity ratio [kg water / kg dry air].air• MinFlowRatsys•HeatFlowSeqsys(i) •( Tsup-Tmix) HeatCapSeq(I)= SysHeatCapcur If SysHeatCapcur is the maximum for the day so far then save SysHeatCapcur as the design value: HeatCap(i )sys= SysHeatCapcur And save the corresponding mixed. (2) Loop over the zones heated by this air loop. return and outside conditions: CoolMixTempsys= Tmix CoolMixHumRatsys=Wmix CoolRetTempsys= SysCoolRetTemp(i) CoolRetHumRatsys= SysCoolRetHumRat(I) CoolOutTempsys= Toutside CoolOutHumRatsys= Woutside Here [kg/m3]. ∑(HeatZoneRetTempSeq(i)• HeatFlowSeqzone(i))/ ∑(HeatZoneHumRatSeq(i)• HeatFlowSeqzone(i))/ HeatFlowSeqsys(i) SysHeatRetHumRat(i)= HeatFlowSeqsys(i) FracOA=ρair• DesOutAirVolFlowsys/ HeatFlowSeqsys(i) Tmix=Toutside• FracOA + SysHeatRetTemp(i)(1 – FracOA) Wmix=Woutside• FracOA + SysHeatRetHumRat (i)(1 – FracOA) SysHeatOutTempSeq(i)= Toutside SysHeatOutHumRatSeq(i)= Woutside Get the current (zone time-step) system heating capacity: SysHeatCapcur=Cp. Cp.air is the specific heat of dry air at 20C. HeatFlowSeqsys(i) =∑HeatFlowSeq zone(i) SysHeatRetTemp(i)= ρair is the density of dry air at 20C and standard elevation corrected pressure.

HeatCapSeq(i). CoolRetHumRat. then move DesCoolVolFlow from SysSizing into CalcSysSizing along with CoolDesDay. So there is no obvious value for outside air temperature at the peak. (b) Loop over all of the zones cooled by this air loop (i) In FinalZoneSizing replace the value in DesCoolCoilInTemp with the user specified CoolSupTempsys. Now these results need to be processed to find the heating and cooling design quantities for each system over all the design days. (a) Loop over all of the design days. CoolOutTemp. SysCoolRetTempSeq(i). SensCoolCap. PreHeatCap. return air temperature at the peak and so forth. CoolMixTemp. SysHeatRetTempSeq(i). Do the same for DesCoolCoilInHumRat and 4/23/05 246 . HeatOutHumRat. HeatMixTemp. SysHeatOutTempSeq(i) and SysHeatOutHumRatSeq(i).j): we have results for each design day. DesHeatVolFlowsys) If the user has specified noncoincidentsystem sizing then: DesCoolVolFlowsys=ρair•NonCoinCoolMassFlowsys DesHeatVolFlowsys=ρair•NonCoinHeatMassFlowsys DesMainVolFlowsys=Max(DesCoolVolFlowsys. EndDay If the user has specified coincident system sizing then: DesCoolVolFlowsys=ρair•CoinCoolMassFlowsys DesHeatVolFlowsys=ρair•CoinHeatMassFlowsys DesMainVolFlowsys=Max(DesCoolVolFlowsys. CoolOutHumRat. At the end of each design day loop the peak cooling and the peak heating data will be stored in CalcSysSizing. For noncoincident sizing the task is harder since we don’t have a single time-step during which all the zone peaks occur. HeatRetHumRat. (ii) If the value of DesHeatVolFlow in SysSizing for the current design day is greater than the value stored in CalcSysSizing. SensCoolCapSeq(i).LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES HeatOutHumRatsys= Woutside Here MinFlowRatsys is the user specified minimum supply flow ratio. then move DesHeatVolFlow from SysSizing into CalcSysSizing along with HeatDesDay. We must return to the zone sizing data and calculate average values for return and outside conditions. HeatOutTemp. CoinHeatMassFlow. HeatRetTemp. CoolRetTemp. HeatCap. CoolFlowSeq(i). SysCoolRetHumRatSeq(i). CoolMixHumRat. (i) If the value of DesCoolVolFlow in SysSizing for the current design day is greater than the value stored in CalcSysSizing. At this point we set DesMainVolFlow in CalcSysSizing equal to the maximum of DesCoolVolFlow and DesHeatVolFlow. SysHeatRetHumRatSeq(i). HeatMixHumRat. HeatFlowSeq(i). (1) Loop over all of the air loops. CoinCoolMassFlow. PreHeatCapSeq(i). DesHeatVolFlowsys) EndSysSizingCalc At this point all the calculations have been done in SysSizing(i. SysCoolOutTempSeq(i) and SysCoolOutHumRatSeq(i). For coincident sizing the task is quite easy.

zone peak•DesCoolMassFlowzone)/ NonCoinCoolMassFlowsys At the end of the zone loop calculate mixed air conditions and the system sensible cooling capacity.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES CoolSupHumRat. FracOA=ρair• DesOutAirVolFlowsys/ NonCoinCoolMassFlowsys Tmix =SysCoolOutTemp• FracOA + SysCoolRetTemp• (1 – FracOA) Wmix = SysCoolOutHumRat • FracOA + SysCoolRetHumRat • (1 – FracOA) SysSensCoolCap=Cp. This ensures that zone equipment connected to an air loop will use the system design supply air conditions as coil entering conditions. (ii) NonCoinCoolMassFlowsys=∑DesCoolMassFlowzone SysCoolRetTemp=∑(ZoneRetTempAtCoolPeak•DesCoolMassFlowzone) / NonCoinCoolMassFlowsys SysCoolRetHumRat=∑(ZoneHumRatAtCoolPeak• DesCoolMassFlowzone)/ NonCoinCoolMassFlowsys SysCoolOutTemp=∑(TOA. This ensures that zone equipment connected to an air loop will use the system design supply air conditions as coil entering conditions. (c) Loop over all of the zones heated by this air loop.zone peak•DesHeatMassFlowzone)/ 4/23/05 247 . (ii) NonCoinHeatMassFlowsys=∑DesHeatMassFlowzone SysHeatRetTemp=∑(ZoneRetTempAtHeatPeak•DesHeatMassFlowzone) / NonCoinHeatMassFlowsys SysHeatRetHumRat=∑(ZoneHumRatAtHeatPeak• DesHeatMassFlowzone)/ NonCoinHeatMassFlowsys SysHeatOutTemp=∑(TOA. Do the same for DesHeatCoilInHumRat and HeatSupHumRat.air• NonCoinCoolMassFlow •( Tmix-Tsup) Then (for noncoincident sizing) the variables calculated in section (ii) are moved into the CalcSysSizing Array.zone peak•DesCoolMassFlowzone)/ NonCoinCoolMassFlowsys SysCoolOutHumRat=∑(WOA. (i) In FinalZoneSizing replace the value in DesHeatCoilInTemp with the user specified HeatSupTempsys.

Note that user specified sizing ratios have already been applied to the zone sizing data which have been used in out preceding system sizing calculation. This data needs to be altered to take into account the user input system design flow rates (if any). First we move the calculated system sizing data from CalcSysSizing array into the FinalSysSizing array. CoinCoolMassFlow= SizRatcool• CoinCoolMassFlowcalc NonCoinCoolMassFlow= SizRatcool• NonCoinCoolMassFlowcalc DesCoolVolFlow= SizRatcool• DesCoolVolFlowcalc Since the flow rates have been altered the outside air fraction will change.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES NonCoinHeatMassFlowsys SysHeatOutHumRat=∑(WOA. FracOA=ρair• DesOutAirVolFlowsys/ NonCoinHeatMassFlowsys Tmix =SysHeatOutTemp• FracOA + SysHeatRetTemp• (1 – FracOA) Wmix = SysHeatOutHumRat • FracOA + SysHeatRetHumRat • (1 – FracOA) SysHeatlCap=Cp. This must be done for the time-step sequence and for the peak value. (i) As in the zone case.zone peak•DesHeatMassFlowzone)/ NonCoinHeatMassFlowsys At the end of the zone loop calculate mixed air conditions and the system sensible cooling capacity. Thus the user specified sizing ratios do not have to be explicitly taken into account at the system level. CoolFlowSeqsys(i)= SizRatcool• CoolFlowSeqsys. This will alter the design mixed air conditions and lead to an altered value for the cooling capacity. (2) We now have the calculated system sizing data. For each air loop this gives us a SizRatcool and SizRatheat. (ii) Loop over the zone timesteps (index=i). the user specified system design flow rates are turned into sizing ratios by dividing the user input value by the calculated value.air• NonCoinHeatlMassFlow •( Tsup-Tmix) Then (for noncoincident sizing) the variables calculated in section (ii) are moved into the CalcSysSizing Array. FinalSysSizing will contain the user modified system design data when we are all done.calc(i) FracOA=ρair• DesOutAirVolFlowsys/ CoolFlowSeqsys(i) Tmix= SysCoolOutTempSeq(i)•FracOA + 4/23/05 248 . Loop over the air loops.

air• DesHeatVolFlowsys •( Tsup-Tmix) Tmix and Wmix are saved in FinalSysSizing . FracOA=ρair• DesOutAirVolFlowsys/ DesCoolVolFlow Tmix= CoolOutTempsys•FracOA + CoolRetTempsys•(1-FracOA) Wmix= CoolOutHumRatsys•FracOA + CoolRetHumRatsys• (1-FracOA) SensCoolCapsys= Cp. (iv) Do the same calculation for the heating case.LOOP AND EQUIPMENT SIZING SYSTEM DESIGN LOADS AND AIR FLOW RATES SysCoolRetTempSeq(i)•(1-FracOA) SensCoolCapSeq(i)= Cp.air• DesCoolVolFlowsys •( Tmix-Tsup) Tmix and Wmix are saved in FinalSysSizing .calc(i) FracOA=ρair• DesOutAirVolFlowsys/ HeatFlowSeqsys(i) Tmix= SysHeatOutTempSeq(i)• FracOA + SysHeatRetTempSeq(i)• (1-FracOA) HeatCapSeq(i)= Cp.air• CoolFlowSeqsys(i) •( Tmix-Tsup) (iii) Do the same calculation for peak cooling. HeatFlowSeqsys(i)= SizRatHeat • HeatFlowSeqsys.air• HeatFlowSeqsys(i) •( Tsup-Tmix) (vi) Do the same calculation for peak heating. 4/23/05 249 . FracOA=ρair• DesOutAirVolFlowsys/ DesHeatVolFlow Tmix= HeatOutTempsys•FracOA + HeatRetTempsys• (1-FracOA) Wmix= HeatOutHumRatsys•FracOA + HeatRetHumRatsys• (1-FracOA) HeatCapsys= Cp. CoinHeatMassFlow= SizRatheat• CoinHeatMassFlowcalc NonCoinHeatMassFlow= SizRatheat• NonCoinHeatMassFlowcalc DesHeatVolFlow= SizRatheat• DesHeatVolFlowcalc (v) Loop over the zone timesteps (index=i).

Each component module contains a sizing subroutine. labeled by its inlet water supply node number. and stored in the PlantSizingData array. Volume of the plant loop Since the loop capacitance has a stability requirement of (V i ∆tstep / V ) ≤ 1 the volume is set so that the stability requirement will be 0. For hot water loops – hot water coils. 4/23/05 250 . max i∆tstep .8 Condenser Loop Sizing Maximum Loop Volumetric Flow Rate The loop maximum volumetric flow rate (m3) is just set equal to the value stored in the PlantSizData array for this loop. zonei3600) / 0. When a component is called for the first time in a simulation. max i∆tstep .DesHeatVolFlowsys) This concludes the system design calculation. which is the largest time step encountered at the max flow rate the loop can reach. And for condenser loops – various types of chiller that use condenser water for cooling.8 at the zone time step. Volume of the plant loop Since the loop capacitance has a stability requirement of (V i ∆tstep / V ) ≤ 1 the volume is set so that the stability requirement will be 0. zonei3600) / 0. it reads in its user specified input data and then calls the sizing subroutine. These individual component design water flow rates are then accessed. heating or other.8 Component Sizing Introduction In EnergyPlus each HVAC component sizes itself. Vloop = (Vloop . Vloop = (Vloop . CurDuctType (in DataSizing) contains the information about the current duct type.LOOP AND EQUIPMENT SIZING PLANT LOOP SIZING (vii) DesMainVolFlowsys=MAX(DesCoolVolFlowsys.8 at the zone time step. A number of high-level variables are used in the sizing subroutines. Plant Loop Sizing Introduction The program needs to be able to autosize the fluid flow rate in each plant fluid loop. cooling. For chilled water loops these components will be cooling coils. The design plant loop flow rates are set by the sum of the needs of the demanding components on each loop. This routine checks the autosizable input fields for missing data and calculates the data when needed. Hot and Chilled Water Loop Sizing Maximum Loop Volumetric Flow Rate The loop maximum volumetric flow rate (m3) is just set equal to the value stored in the PlantSizData array for this loop. Each component that uses water for heating or cooling stores its design water flow rate (in its sizing routine) in the array CompDesWaterFlow. This array also contains the user specified design values for each plant loop. summed for each plant loop. which is the largest time step encountered at the max flow rate the loop can reach. The types can be main.

DesCoolVolFlowzone) If the max fan flow rate is less than SmallAirVolFlow the max flow rate is set to zero. Otherwise Vfan .LOOP AND EQUIPMENT SIZING COMPONENT SIZING CurZoneEqNum (in DataSizing) is the current zone equipment set index and indicates that the component is a piece of zone equipment and should size itself using the zone sizing data arrays. Fan Sizing Fan sizing is done in subroutine SizeFan. COIL:Water:SimpleCooling Sizing The sizing is done subroutine SizeWaterCoil. other or default Vfan . des = ρ air i DesHeatVolFlowsys Using the system design mixed and supply air conditions calculate the design coil load. min = DesCoolVolFlowsys i MinFlowRatsys for duct type=heating Vfan . des = ρ air i DesCoolVolFlowsys for duct type=heating mair . Max Flow Rate If the fan is part of the central air system then check the duct type. CurSysNum (in DataSizing) is the current air loop index and indicates that the component is part of the primary air system and should size itself using the system sizing data arrays. other or default mair . max = DesHeatVolFlowzone . For duct type = main. 4/23/05 251 . max = DesCoolVolFlowsys For variable volume fans the minimum flow rate is set to Vfan . max = DesMainVolFlowsys For variable volume fans the minimum flow rate is set to Vfan . des = ρ air i DesMainVolFlowsys for duct type=cooling mair . min = DesHeatVolFlowsys i MinFlowRatsys If the fan is zone equipment then check whether it is part of a component that only does heating. For heating only Vfan . get the coil design air flow rate. Max Water Flow Rate of Coil System Coils Depending on the duct type. max = Max( DesHeatVolFlowzone. For duct type = main. min = DesMainVolFlowsys i MinFlowRatsys for duct type=cooling Vfan . max = DesHeatVolFlowsys For variable volume fans the minimum flow rate is set to Vfan .

air=CoolDesTempzone Wout. des = mair . air ) With the coil load and the user specified (in a Plant Sizing object) design chilled water temperature rise.air. For duct type = main. des = DesCoilMassFlowzonei(hin .air= PsyHFnTdbW(Tout. cw. water .air= PsyHFnTdbW(Tcoil. calculate the max water flow rate: Vcoil .air= PsyHFnTdbW(Tin.water= ExitTempplt. Qcoil.air= CoolMixHumRatsys Tin. Win. water i ρ water i∆Tplt .LOOP AND EQUIPMENT SIZING COMPONENT SIZING Tin.air= PsyHFnTdbW(Tin. des ) Zone Coils Using the zone design coil inlet and supply air conditions calculate the design coil load.air= CoolMixTempsys Win. Then we numerically invert the coil model to solve for the UA that will enable the coil to meet the design coil load given the specified inputs.air= CoolSupHumRatsys Win. air = ρ air i DesCoolVolFlowsys for duct type=heating 4/23/05 252 . air − hout .in. cw.air.air.des min . water . we specify the model inputs (other than the UA) at design conditions and the design coil load that the coil must meet. other or default min . Win.air) Qcoil . des i(hin . air − hout . System Coils The design coil load is the system design sensible cooling capacity. max = Qcoil . water = ρ water iVcoil . des /(Cp .air= DesCoolCoilInHumRatzone hin.air= CoolMixHumRat sys hin.out.air) Depending on the duct type.des= SensCoolCapsys The required inputs for the simple coil model are: Tin. max = Qcoil . Wcoil. air = ρ air i DesMainVolFlowsys for duct type=cooling min . max hin. Tin.air= PsyHFnTdbW(Tcoil. Wout.out) Qcoil .air) hout. water i ρ water i∆Tplt . calculate the max water flow rate: Vcoil . water . air ) With the coil load and the user specified (in a Plant Sizing object) design chilled water temperature rise.air= CoolMixTempsys Tout. des ) UA of the Coil To obtain the UA of the coil.in) hout. des /(Cp . get the coil design air flow rate.air= CoolDesHumRatzone Win. Wcoil.cw.air= CoolSupTempsys Wout.air= DesCoolCoilInTemp zone Tout.

des = ρ air i DesHeatVolFlowsys Dfin = 0.air= DesCoolCoilInTempzone Tout. The residual is calculated in the function SimpleCoolingCoilUAResidual. water .LOOP AND EQUIPMENT SIZING COMPONENT SIZING min . DesHeatMassFlowzone) Tin. Win.water= ExitTempplt. des = ρ air i DesMainVolFlowsys for duct type=cooling mair . other or default mair . air − hout . des = ρ air i DesCoolVolFlowsys for duct type=heating mair .air. air = ρ air i DesHeatVolFlowsys We now have all the data needed to obtain UA.air. air ) We now have all the data needed to obtain UA. Zone Coils The required inputs for the simple coil model are: Tin. This is a general utility routine for finding the zero of a function. The numerical inversion is carried out by calling subroutine SolveRegulaFalsi. des 4/23/05 253 . Number of Tubes per Row Ntube / row = Int(13750iVcoil . des = DesCoolMassFlowzonei(hin . max hin. max ) Ntube/row=Max(Ntube/row.air) The design coil load is then Qcoil . The residual is calculated in the function SimpleCoolingCoilUAResidual. water = ρ water iVcoil . This is a general utility routine for finding the zero of a function.air= PsyHFnTdbW(Tout.air=CoolDesHumRatzone Win. air = Max( DesCoolMassFlowzone.3) Fin Diameter Depending on the duct type. The numerical inversion is carried out by calling subroutine SolveRegulaFalsi.air) hout. In this case it finds the UA that will zero the residual function – the difference between the design coil load and the coil output divided by the design coil load. COIL:Water:DetailedFlatCooling Sizing The sizing is done in subroutine SizeWaterCoil Max Water Flow Rate of Coil The calculation is identical to that done for COIL:Water:SimpleCooling.cw.air= PsyHFnTdbW(Tin. get the coil design air flow rate. Wout. water .des min .335imair . For duct type = main. In this case it finds the UA that will zero the residual function – the difference between the design coil load and the coil output divided by the design coil load.air= CoolDesTempzone Wout.air= DesCoolCoilInHumRatzone min .

4•Dtube. calculate the max water flow rate: Vcoil . des Fin Surface Area Depending on the duct type. get the coil design air flow rate. Tube Outside Surf Area Atube. Tin.44imair .inside is the tube inside diameter. hw. Max Water Flow Rate of Coil System Coils With the coil load from the system design data array and the user specified (in a Plant Sizing object) design hot water temperature fall.outside is the tube outside diameter. max = HeatCapsys /(Cp . If the coil is not part of an induction unit then obtain the coil inlet temperature from the zone design data array.5imair . For duct type = main. water . get the coil design air flow rate.1•Dtube. For duct type = main. des Total Tube Inside Area Atube. other or default mair .air= DesHeatCoilInTempzone If the coil is part of an induction unit take into account the induced air: 4/23/05 254 . Coil Depth Depthcoil=Depthtube spacing• Ntube rows COIL:Water:SimpleHeating Sizing The sizing is done in subroutine SizeWaterCoil. des = ρ air i DesMainVolFlowsys for duct type=cooling mair .LOOP AND EQUIPMENT SIZING COMPONENT SIZING Minimum Air Flow Area Depending on the duct type.inside•Ntube rows•Ntubes/row Where Dtube.outside=4. des = ρ air i DesCoolVolFlowsys for duct type=heating mair . water i ρ water i∆Tplt . des = ρ air i DesMainVolFlowsys for duct type=cooling mair . des = ρ air i DesHeatVolFlowsys AFinSurf = 78.outside•Ntube rows•Ntubes/row Where Dtube. des = ρ air i DesHeatVolFlowsys AMinAirFlow = 0. other or default mair . des = ρ air i DesCoolVolFlowsys for duct type=heating mair . des ) Zone Coils Using the zone design coil inlet and supply air conditions calculate the design coil load.total inside=4.

air = ρ air i DesCoolVolFlowsys for duct type=heating min .des= HeatCapsys The required inputs for the simple coil model are: Tin. get the coil design air flow rate.air is calculated at the outlet humidity and the average of the inlet and outlet temperatures. The numerical inversion is carried out by calling subroutine SolveRegulaFalsi. tu Otherwise the design flow is obtained from the zone design data array: mair .des min . With the coil load and the user specified (in a Plant Sizing object) design hot water temperature decrease. air = ρ air i DesHeatVolFlowsys We now have all the data needed to obtain UA.air= HeatDesHumRatzone If the coil is part of a terminal unit the mass flow rate is determined by the volumetric flow rate of the terminal unit: mair .water= ExitTempplt.air= HeatMixHumRatsys Tin. In this case it finds the UA that will zero the residual function – the difference between the design coil load and the coil output divided by the design coil load. des = ρ air imair . For duct type = main. water = ρ water iVcoil . water . airmair .hw. Then we numerically invert the coil model to solve for the UA that will enable the coil to meet the design coil load given the specified inputs. des = DesHeatMassFlowzone Qcoil . calculate the max water flow rate: Vcoil .LOOP AND EQUIPMENT SIZING COMPONENT SIZING Fracminflow=MinFlowFraczone Tin. other or default min . 4/23/05 255 .air= DesHeatCoilInTempzone• Fracminflow + ZoneTempAtHeatPeakzone•(1− Fracminflow) Tout. des /(Cp . hw. max = Qcoil . des . air = ρ air i DesMainVolFlowsys for duct type=cooling min . air − Tin .air= HeatMixTempsys Win. des = cp . water . max Depending on the duct type. we specify the model inputs (other than the UA) at design conditions and the design coil load that the coil must meet.air=HeatDesTempzone Wout. water i ρ water i∆Tplt . This is a general utility routine for finding the zero of a function. The residual is calculated in the function SimpleHeatingCoilUAResidual. des i(Tout . Qcoil. air ) Here cp. System Coils The design coil load is the system design sensible cooling capacity. des ) UA of the Coil To obtain the UA of the coil.

The residual is calculated in the function SimpleHeatingCoilUAResidual. water = ρ water iVcoil .hw. air imair .air= DesHeatCoilInHumRatzone Tin.water= ExitTempplt. des = DesHeatMassFlowzone Qcoil . tu Otherwise the design flow is obtained from the zone design data array: mair . Nominal Capacity of the Coil System Coils The value is obtained from the system design array. des = cp .des min . des = ρ air imair . air − Tin .air=HeatDesTempzone Wout. This is a general utility routine for finding the zero of a function.air= DesHeatCoilInTempzone• Fracminflow + ZoneTempAtHeatPeakzone•(1− Fracminflow) Tout.air= DesHeatCoilInTempzone If the coil is part of an induction unit take into account the induced air: Fracminflow=MinFlowFraczone Tin.air= HeatDesHumRatzone If the coil is part of a terminal unit the mass flow rate is determined by the volumetric flow rate of the terminal unit: mair . des . Tin. des i(Tout . We now have all the data needed to obtain UA. The numerical inversion is carried out by calling subroutine SolveRegulaFalsi. If the coil is not part of an induction unit then obtain the coil inlet temperature from the zone design data array. air ) Here cp. In this case it finds the UA that will zero the residual function – the difference between the design coil load and the coil output divided by the design coil load.air is calculated at the outlet humidity and the average of the inlet and outlet temperatures. max Tout.air= HeatDesHumRatzone 4/23/05 256 .air=HeatDesTempzone Wout. water . Sizing of Gas and Electric Heating Coils The sizing calculation is done in subroutine SizeHeatingCoil in module HeatingCoils. Tin.air= DesHeatCoilInTempzone• Fracminflow + ZoneTempAtHeatPeakzone•(1− Fracminflow) Win.LOOP AND EQUIPMENT SIZING COMPONENT SIZING Zone Coils If the coil is not part of an induction unit then obtain the coil inlet temperature from the zone design data array.air= DesHeatCoilInTempzone If the coil is part of an induction unit take into account the induced air: Fracminflow=MinFlowFraczone Tin. Capnom= HeatCapsys Zone Coils The capacity is calculated from the design coil inlet and outlet conditions.

Vair . rated = DesMainVolFlowsys Zone Coils The rated air flow rate is the maximum of the heating and cooling design flow rates from the zone design array. COIL:DX:HEATINGEMPIRICAL 3. rated = Max( DesCoolVolFlowzone. COIL:DX:MULTIMODE:COOLINGEMPIRICAL with COILPERFORMANCE:DX:COOLINGBYPASSFACTOREMPIRICAL Rated Air Volume Flow Rate System Coils The rated air flow rate is obtained from the system design array.Wmix) hmix= PsyHFnTdbW(Tmix. rated / FlowCapRatiomin If FlowCapRatio < FlowCapRatiomax then CCaprated = Vair . DX Coil Sizing The sizing calculations are done in subroutine SizeDXCoil in module DXCoils.Wmix) hsup= PsyHFnTdbW(Tsup. Vair . DesHeatVolFlowzone) Rated Total Cooling Capacity System Coils The rated cooling capacity is obtained by dividing the peak cooling capacity by the Cooling Capacity Modifier Curve evaluated at peak mixed wetbulb and outdoor drybulb temperatures.wb.LOOP AND EQUIPMENT SIZING COMPONENT SIZING Qcoil.Tmix. COIL:DX:COOLINGBYPASSFACTOREMPIRICAL 2.air−Tin.Wmix.std.air) Here cp. COIL:DX:MULTISPEED:COOLINGEMPIRICAL 4. pair.air is calculated at the outlet humidity and the average of the inlet and outlet temperatures. rated i(hmix − hsup ) CCaprated=CCappeak/ CapModFac We check that the design volume flow per total capacity is within the prescribed range: FlowCapRatio = Vair .des=Cp.Toutside) CCappeak = ρ air iVair .air• DesHeatMassFlowzone•(Tout. rated / FlowCapRatiomax 4/23/05 257 . This section covers the sizing of the objects 1.wb= PsyTwbFnTdbWPb(Tmix. rated / CCaprated If FlowCapRatio < FlowCapRatiomin then CCaprated = Vair .std) CapModFac=CurveValue(CCapFTemp. Tmix= CoolMixTempsys Wmix=CoolMixHumRatsys Tsup=CoolSupTempsys Wsup=CoolSupHumRatsys Toutside=CoolOutTempsys ρair=PsyRhoAirFnPbTdbW(pair. Tmix.Wsup) Tmix.

00006041 m3/s per watt(450 cfm/ton) Rated Total Heating Capacity For COIL:DX:HEATINGEMPIRICAL the rated heating capacity is set equal to the cooling capacity.desday.wb.Wmix) hsup= PsyHFnTdbW(Tsup.std) CapModFac=CurveValue(CCapFTemp.Wmix) hmix= PsyHFnTdbW(Tmix. Tin.Wsup) Tmix. rated / FlowCapRatiomax where FlowCapRatiomin = 0. Tmix.Wmix. Tmix= DesCoolCoilInTempzone Wmix= DesCoolCoilInHumRatzone Tsup= CoolDesTempzone Wsup= CoolDesHumRatzone Toutside=Toutside.rated= 26.00004699 m3/s per watt (350 cfm/ton) And FlowCapRatiomax= 0.rated.Toutside) CCappeak = ρ air iVair .6667 oC (80 oF) Win.LOOP AND EQUIPMENT SIZING COMPONENT SIZING where FlowCapRatiomin = 0. 67 oF wetbulb) Cp. rated / CCaprated If FlowCapRatio < FlowCapRatiomin then CCaprated = Vair .00004699 m3/s per watt(350 cfm/ton) And FlowCapRatiomax= 0. rated / FlowCapRatiomin If FlowCapRatio < FlowCapRatiomax then CCaprated = Vair .01125 (corresponds to 80 oF drybulb. Rated SHR The rated sensible heat ratio is calculated to be the sensible cooling (from rated inlet conditions to user specified supply conditions) divided by the total cooling (from rated inlet to specified supply). rated i(hmix − hsup ) CCaprated=CCappeak/ CapModFac We check that the design volume flow per total capacity is within the prescribed range: FlowCapRatio = Vair .wb= PsyTwbFnTdbWPb(Tmix.Tmix.peak ρair=PsyRhoAirFnPbTdbW(pair.rated) For system coils Tsup=CoolSupTempsys Wsup=CoolSupHumRatsys 4/23/05 258 .air= PsyCpAirFnWTdb(Win. pair.rated= 0.00006041 m3/s per watt (450 cfm/ton) Zone Coils The rated cooling capacity for zone coils is calculated in the same manner as for system coils.std. Tin.

Low Speed The rated total cooling capacity.000114 m3/s per watt (850 cfm/ton) times the total rated cooling capacity. Evaporative Condenser Pump Rated Power Consumption The evaporative condenser pump rated power consumption is set equal to the total cooling capacity times 0.000114 m3/s per watt (850 cfm/ton) times the total rated cooling capacity.004266 watts pump power per watt capacity (15 W/ton). Resistive Defrost Heater Capacity For the heat pump the resistive defrost heat capacity is set equal to the cooling capacity. Wsup) ∆hrated. Rated SHR. We have Efftot=Effmot•Effimpeller The motor efficiency is an input.sup=hrated−hsup ∆Qsrated.sup Evaporative Condenser Air Volume Flow Rate The evaporative condenser air volume flow rate (m3/s) is set to 0. low speed. Since we need the total efficiency to calculate the nominal power consumption we assume an impeller efficiency of 0. Evaporative Condenser Air Volume Flow Rate. low speed.78 for purposes of sizing.LOOP AND EQUIPMENT SIZING COMPONENT SIZING For zone coils Tsup= CoolDesTempzone Wsup= CoolDesHumRatzone Then hrated= PsyHFnTdbW(Tin. Rated Total Cooling Capacity. Low Speed The evaporative condenser pump rated power consumption.rated−Tsup) SHRrated=∆Qsrated. low speed The rated sensible heat ratio. Pump Sizing The loop pumps’ autosizable inputs are nominal volumetric flow rate and nominal power consumption.sup/∆hrated. 4/23/05 259 .rated) hsup= PsyHFnTdbW(Tsup. Low Speed The evaporative condenser air volume flow rate.004266 watts pump power per watt capacity (15 W/ton). low speed The rated air volume flow rate. Rated Volumetric Flow Rate This is just set equal to the design loop demand obtained from summing the needs of the components on the demand side of the loop. is set equal to 1/3 times the full rated total cooling capacity.rated. low speed. low speed. Evaporative Condenser Pump Rated Power Consumption.air•(Tin. is set equal to 1/3 times the full rated air volume flow rate. Rated Air Volume Flow Rate. Win.sup=Cp. low speed (m3/s) is set to 1/3 times 0. is set equal to the full speed SHR. is set equal to 1/3 times the total cooling capacity times 0.

Design Evaporator Volumetric Water Flow Rate Vevap . des iVloop . nom = Cp . des is the loop design volumetric flow rate. ∆Tloop. If there are multiple boilers on a loop that call for autosizing. des iVloop . ρw is the density of water at standard conditions (5. nomi(1 + 1/ COPchiller .nom is the chiller nominal COP. des = Vloop . nom = Cp . All boilers on a loop are sized to meet the full loop load.05 oC). COPchiller. Nominal Cooling Capacity Qchiller .w is the specific heat of water at design condenser inlet temperature. they will all be assigned the same heating capacity and flow rate. All chillers on a loop are sized to meet the full loop load. is an input. Nominal Capacity Qboiler . ρw is the density of water at standard conditions (5. they will all be assigned the same cooling capacity and evaporator flow rate. des where Cp. the nominal head. des where Cp. All 3 quantities can be straightforwardly obtained using the user specified loop sizing data and the loop design flow rates. des is the loop design volumetric flow rate.w is the specific heat of water at the boiler design outlet temperature. ∆Tloop. wi ρ w) where Cp.des is the hot water loop design temperature decrease. evaporator flow rate and condenser flow rate. ∆Tloop. des Design Condenser Volumetric Water Flow Rate Vcond . 4/23/05 260 . Boiler Sizing Generally boilers will need nominal heating capacity and rate. Electric Chiller Sizing Generally chillers will need nominal cooling capacity. nom) /(∆Tloop . des = Qchiller . Vloop . ρw is the density of water at standard conditions (5. wi ρ wi∆Tloop . wi ρ wi∆Tloop .des is the chilled water loop design temperature rise.des is the chilled water loop design temperature rise. If there are multiple chillers on a loop that call for autosizing.w is the specific heat of water at 5 oC. Vloop . Both quantities can be straightforwardly obtained using the user specified loop sizing data and the loop design flow rates. desiCp .05 oC).05 oC).LOOP AND EQUIPMENT SIZING COMPONENT SIZING Rated Power Consumption Qnom = HnomiVnom / Efftot Hnom.

des i∆Tloop . EnergyPlus provides two input choices: the user can input the design water flow rate and tower UA at each operating point or the tower nominal capacity (and let the program calculate the water flow rate and UA). nom = 0. Fan Power at Design Air Flow Rate The nominal fan power is sized to be 0. wi ρ wiVtower . ∆Tloop. and the tower UA.5 and a fan pressure rise of 190 Pascals. If Tower Performance Input Method = UA and Design Water Flow Rate then Qtower . des = Cp . w.8iQtower .382•10-08 is m3/s per watt corresponds to the rule-of-thumb of sizing the tower flow rate at 3 gallons per minute per ton.LOOP AND EQUIPMENT SIZING COMPONENT SIZING Design Evaporator Volumetric Water Flow Rate Vdes = Vloop . des = 5. Design Water Flow Rate If Tower Performance Input Method = UA and Design Water Flow Rate then Vtower . des where Cp. The design tower load is: for Tower Performance Input Method = UA and Design Water Flow Rate Qtower . desi∆Tloop . Then we numerically invert the tower model to solve for the UA that will enable the tower to meet the design tower load given the specified inputs. des = Vloop .382 E . air . w.5i ρ air /190 where ρair is the density of air at standard conditions. des Cooling Tower Sizing The quantities needed to autosize a cooling tower include the design water flow rate. the nominal fan power and air flow rate. This data may be need to be given at more than one operating point:. ρw is the density of water at standard conditions (5. for instance – high speed fan.05 oC). nom Design Air Flow Rate We assume a fan efficiency of 0. Tower UA Value at Design Air Flow Rate To obtain the UA of the tower. w.0105iQtower . nom = Cp .des is the condenser water loop design temperature rise. des = Qfan. nom where 5. Choice of input method will affect the sizing calculations in ways noted below. wi ρ wiVtower . w. Finally Qfan . des 4/23/05 261 . we specify the model inputs (other than the UA) at design conditions and the design tower load that the tower must meet.w is the specific heat of water at the condenser loop design exit temperature. nomi0. Then Vtower . low speed fan and free convection.0105 times the design load. des If Tower Performance Input Method = Nominal Capacity then Vtower .

The data needed for sizing the units is obtained from the zone design arrays and the user specified plant sizing input. Fan Coil Unit Sizing Fan Coil units are compound components: each unit contains a fan. chilled water coil and outside air mixer. des = 1. nom .25iQtower . nom (to allow for compressor heat) Then we assign the inputs needed for the model. Air Flow Rate in Free Convection Regime The free convection air flow rate is set to 0. and the design outside air flow rate. Fan Power at Low Fan Speed The fan power at low fan speed is set to 0. DesHeatVolFlowzone) 4/23/05 262 . hot water coil.des Nominal Capacity Tin. The inputs that may need to be autosized are the nominal unit air flow rate. Tower UA Value at Low Fan Speed For Tower Performance Input Method = UA and Design Water Flow Rate the low speed UA is set to 0. The numerical inversion is carried out by calling subroutine SolveRegulaFalsi. This is a general utility routine for finding the zero of a function. The inlet water temperature is set slightly differently for the 2 input methods. In this case it finds the UA that will zero the residual function – the difference between the design tower load and the tower output divided by the design tower load.exit.6 oC (78 oF design air inlet wetbulb temperature) Win is calculated from the entering air drybulb and wetbulb.water=35 oC (95 oF design inlet water temperature). For Tower Performance Input Method = Nominal Capacity the low speed UA is calculated in the same manner as the full speed UA using Qtower . Tin.LOOP AND EQUIPMENT SIZING COMPONENT SIZING for Tower Performance Input Method = Nominal Capacity Qtower . max = Max( DesCoolVolFlowzone. Tower UA Value in Free Convection Regime For Tower Performance Input Method = UA and Design Water Flow Rate the low speed UA is set to 0. For UA and Design Water Flow Rate Tin.16 times the fan power at full speed. The inlet water mass flow rate is just the design eater volumetric flow rate times the density of water. freeconv instead of Qtower . nom . For Tower Performance Input Method = Nominal Capacity the low speed UA is calculated in the same manner as the full speed UA using Qtower .des+∆Tloop. The residual is calculated in the function SimpleTowerUAResidual.1 times the full air flow rate.6 times the full speed UA. nom .1 times the full speed UA. lowspeed instead of Qtower . Maximum Air Flow Rate Vair . Air Flow Rate at Low Fan Speed The nominal air flow rate at low fan speed is set to one-half of the full speed air flow rate. the maximum hot and chilled water flow rates. We now have all the data needed to obtain UA. nom .wb=25.water=Tloop.air=35 oC (95 oF design air inlet temperature) Tin.air.

des = DesCoolMassFlowzonei(hin . des ) where cp.air is evaluated at the average of the inlet & outlet temperatures and the coil outlet humidity ratio. if a hot water reheat coil is present. des = cp .out) Qcoil .in=DesColdCoilInTempzone Tcoil. hw = Qcoil . max = Max( DesCoolVolFlowzone. max ) Maximum Hot Water Flow Tcoil. max ) Single Duct Terminal Units These are the Single Duct:Const Volume:Reheat. hw = Qcoil . coil ) Vmax . des /(cp .out= CoolDesHumRatzone Hcoil. des /(cp . max = Min( MinOAzone.in= PsyHFnTdbW(Tcoil.in) Hcoil. Single Duct:VAV:Reheat and Single Duct:VAV:Noreheat objects.out= PsyHFnTdbW(Tcoil. max = Max( DesCoolVolFlowzone.out=ColdDesTempzone Wcoil. coil − hout . wi ρ wi∆Tloop . Window Air Conditioner Sizing Window air conditioners are compound components: each unit contains a fan. Wcoil. The inputs that may need to be autosized are the nominal unit air flow rate and the design outside air flow rate. Vair .LOOP AND EQUIPMENT SIZING COMPONENT SIZING Maximum Outside Air Flow Rate Voutsideair . Vair . DesHeatVolFlowzone) Maximum Outside Air Flow Rate Voutsideair .in. coil − Tin . The inputs that may need to be autosized are the maximum air flow rate through the terminal unit and the maximum hot water flow rate through the reheat coil. coil ) Vmax . max = Min( MinOAzone.out.out=HeatDesTempzone Qcoil . des ) where cp. air i DesHeatMassFlowzonei(Tout . Maximum Air Flow Rate Vair . wi ρ wi∆Tloop .air is evaluated at the average of the inlet & outlet temperatures and the coil outlet humidity ratio.in=DesHeatCoilInTempzone Tcoil. The data needed for sizing the units is obtained from the zone design arrays. DesHeatVolFlowzone) 4/23/05 263 . Wcoil. A DX coil and outside air mixer. Maximum Cold Water Flow Tcoil.in= DesCoolCoilInHumRatzone Wcoil. Maximum Air Flow Rate Vair .

C. ASHRAE Transactions. 459-468. But if the unit is VAV and normal action then: Vair . Fisher. 1997. and R. max . hw = Qcoil . Toolkit for Building Load Calculations. min iVair .. Refrigerating and Air-Conditioning Engineers. coil ) Vmax . heating i(Tout . max . des ) where cp. air i ρ air iVair . Pedersen. Vol.air is evaluated at the average of the inlet & outlet temperatures and the coil outlet humidity ratio. 103(2). 2001 ASHRAE Fundamentals Handbook. D.E. A heat balance based cooling load calculation procedure. coil . Refrigerating and Air-Conditioning Engineers. pp. coil − Tin . Inc. 2001. 4/23/05 264 . Atlanta: American Society of Heating. Atlanta: American Society of Heating. coil . des = cp . heating = Vair . C.J. coil . Liesen. If the terminal unit is VAV and reverse action or if the terminal unit is constant volume then: Vair .LOOP AND EQUIPMENT SIZING COMPONENT SIZING Max Reheat Water Flow Tcoil. des /(cp .O.O. heating = Fracair . Pedersen. The coil load and max hot water flow rate are then: Qcoil .in=DesHeatCoilInTempzone Tcoil. Inc. wi ρ wi∆Tloop .out=HeatDesTempzone The design air flow for heating depends on the terminal unit characteristics. References ASHRAE Fundamentals 2001.

‘UCSD Cross Ventilation Model Controls’ UCSD Displacement Ventilation UCSD Cross Ventilation Mundt model Mundt (1996) points out that a floor air heat balance provides a simple and reasonably accurate method of modeling the temperature near the floor surface. The figure below diagrams the temperature distribution versus height being calculated by the model. and how to apply the room air models available in EnergyPlus. RANS-CFD) are too computationally expensive to use with EnergyPlus for the foreseeable future.g. Room air modeling was added to EnergyPlus starting with Version 1. Summary of room air models available in EnergyPlus Air model name Well-Mixed Mundt Applicability All zones displacement ventilation displacement ventilation cross ventilation Input Objects Required None. Such models (e. Therefore. indoor air quality.2. ‘RoomAir Node’’ ‘RoomAir Model’ . it is up to the user to have a good understanding of when. Mundt’s floor air heat balance is extended to include convection heat gain from equipment and by ventilation or infiltration that may be introduced near the floor in order to maintain all the 4/23/05 265 . default ‘RoomAir Model’. Note that EnergyPlus does not have completely general methods of modeling room air that are applicable to every conceivable type of airflow that might occur in a zone. The rest of this section provides some guidance in the way of examples and further discussion of the models available in EnergyPlus. there is not currently a consensus on how to best model non-uniform air temperatures in buildings. etc) that might benefit from localized modeling of how room air varies across space. The slope of a linear temperature gradient can then be obtained by adding a second upper air temperature value that comes from the usual overall air system cooling load heat balance. Table 37. This allows surface heat transfer and air system heat balance calculations to be made taking into account natural thermal stratification of air and different types of intentional air distribution designs such as under-floor and side-wall displacement ventilation that purport to extract room air at higher-than-mean temperatures. Also note that because the complete mixing model for room air has long been the standard in building energy simulation. The models that are available in EnergyPlus offer only limited modeling capabilities for select room airflow configurations. Although there are many types of analyses (comfort. EnergyPlus offers the different types of air models listed in the table below along with the input objects associated with the use of that model. only the temperature distribution of room air within the zone is currently addressed in EnergyPlus. ‘Mundt Model Controls’.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Alternative Modeling Processes Room Air Models The group of input objects described in this section is used to account for non-uniform room air temperatures that may occur within the interior air volume of a zone. ‘UCSD Displacement Ventilation Model Controls’ ‘RoomAir Model’ . where.

ALTERNATIVE MODELING PROCESSES COMPONENT SIZING terms in the air heat balance of the Heat Balance Model. TLeaving = − Qsys ρ c pV + TSupply (186) where Qsys is the air system heat load with negative values indicating a positive cooling load. such as tower computer cases. The user prescribes values for floor splits as input. 4/23/05 266 . Tleaving. is obtained from. No guidance is known to be available to use in recommending floor splits. The equation above can be solved directly for TAirFloor and is used in the form of the equation below. but the user could for example account for equipment known to be near the floor. ρ c pV (TAirFloor − TSupply ) = hcFloor AFloor (TFloor − TAirFloor ) + QConvSourceFloor + QInfilFloor where ρ is the air density cp is the air specific heat at constant pressure i V is the air system flow rate Tsupply is the air system’s supply air drybulb temperature hcFloor is the convection heat transfer coefficient for the floor Afloor is the surface area of the floor Tfloor is the surface temperature of the floor QconvSourceFloor is the convection from internal sources near the floor (< 0. This yields the following heat balance for a floor air node.2 m) QInfilFloor is the heat gain (or loss) from infiltration or ventilation near the floor “Floor splits” are the fraction of total convective or infiltration loads that are dispersed so as to add heat to the air located near the floor. TAirFloor = ρ c pVTSupply + ∑ hcFloor AFloorTFloor + QConvSourceFloor + QInfilFloor ρ c pV + ∑ hcFloor AFloor The upper air node temperature is obtained by solving the overall air heat balance for the entire thermal zone for the temperature of the air leaving the zone and going into the air system return. Values for Qsys are computed by the load calculation routines and passed to the air model. dT TLeaving − TAirFloor = dz H return where Hreturn is the distance between the air system return and the floor air node assumed to be 0. or supplementary ventilation designed to enter along the floor. dT/dz. The vertical temperature gradient or slope.1 m from the floor and z is the vertical distance.

ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Ta. Tai = Tleaving − dT ( zleaving − zi ) dz So for example the temperatures near the ceiling can easily be determined. However.1 m) is accomplished by returning the temperature for the appropriate height to the appropriate air node used for control.3 Z-direction height Ta. then the air model can provide individual values for each surface based on the height and slope. 4/23/05 267 . If the walls are subdivided in the vertical direction as shown in the figure above. ceil Tleaving Ta.g. Height versus temperature schematic for Mundt model The constant slope allows obtaining temperatures at any vertical location using. Accounting for the location of the thermostat inside the zone (e. 1.1 Ta. floor Temperature Figure 90.2 TstatDB Ta. no additional heat balances are necessarily made (in the air domain) at these points as all the surface convection is passed to the model in the totaled value for Qsys . supply Ta.

These models must be simple enough not to impose an undue 4/23/05 268 . The use of three nodes allows for greatly improved prediction of thermal comfort and overall building energy performance in low energy cooling strategies that make use of unmixed stratified ventilation flows. UCSD Displacement Ventilation Room Air Model Overview The UCSD Displacement Ventilation Room Air Model provides a simple model for heat transfer and vertical temperature profile prediction in displacement ventilation.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING References: Mundt. E. Such techniques include displacement ventilation (DV) and underfloor air distribution (UFAD) systems. in principle. Similarly the fresh air will be used more effectively than with a fully mixed system: the fresh air won’t be “wasted” in the upper. Finally. The fullymixed room air approximation that is currently used in most whole building analysis tools is extended to a three node approach. Often DV systems use 100% outside air.. In order to avoid chilling the occupants the supply air temperature used for DV is considerably higher than that used in conventional forced-air systems. Stockholm. Ph. 1996. D. natural displacement ventilation and also wind-driven cross-ventilation (CV). The inflow temperature advantage is then only useful for the minimum outside air that must always be provided (in most cases this remaining advantage is negligible). The UCSD Displacement Ventilation Model is one of the non-uniform zone models provided through the Room Air Manager in EnergyPlus. (1997)]. Non-uniform zone models Several types of models have been proposed as suitable for inclusion in building energy simulation (BES) programs. In a DV system conditioned air is delivered at floor level and low velocity in order to minimize mixing and to establish a vertical temperature gradient. the vertical temperature gradient means that the average room temperature can be higher for a DV conditioned room than with a conventionally conditioned room: the occupants feel the lower temperature in the lower region of the room and are unaffected by the higher temperature near the ceiling. The performance of displacement ventilation systems-experimental and theoretical studies. Maintaining the lower boundary of this warm layer above the occupied zone is one of the many unique challenges of displacement ventilation design. In DV a noticeable interface occurs between the occupied zone of the room and a mixed hot layer near the ceiling of the room [Dominique & Guitton. operating in closed loop). DV systems have limitations. Royal Institute of Technology. is exhausted through ceiling level vents. The vertical displacement air movement means that convective heat gains introduced near the ceiling will be removed without affecting the occupied region of the room. with the purpose of obtaining a first order precision model for vertical temperature profiles in displacement ventilation systems. in turn. DV systems seem to be best suited to interior spaces with only moderate loads. Also a fraction of the heat gains that occur in the occupied zones rise as plumes into the upper part of the space. Thesis. However. The intent is to provide a selection of useful non-uniform zone air models to enable the evaluation of air-conditioning techniques that use stratified or partially stratified room air. The incoming air “displaces” the air above it which. The methodology can also include. unoccupied region of the room. In a conventional system conditioned air is delivered at ceiling level and the intent is to create a fully mixed space with uniform conditions. Exterior spaces may have conditions that are not conducive to establishing a vertical temperature gradient. thereby reducing the cooling load. This can lead to problems in removing both sensible and latent loads. whenever the outside air temperature is above ≈19°C this advantage is mostly lost: the internal loads must be removed from the space independently of the airflow pattern (during the warmer hours buildings tend to be almost closed to the outside. Displacement Ventilation A DV system is a complete contrast to a conventional forced air system.

are not explicitly modeled. successfully validated against scaled model experiments [Linden et al. (1990)]. suitable for top-hat profiles for density and velocity across the plumes. it is best to classify it in a separate category: plume equation based multi-layer models [Linden et al. However. since it is implemented in EnergyPlus. each node couples convectively to one or more surfaces. yet provide enough predictive capability to produce useful comparisons between conventional and stratified zone operation strategies. (1990). The Rees-Haves model. (1956)]. and the Rees-Haves model [Rees & Haves 2001] since it is a well developed nodal-type model and is implemented in the RP-1222 toolkit.127 is used. Zonal models are coarse–grained finite volume models. This is the main difference between the DV models implemented in the RP-1222 toolkit and the model that is described here. The main assumption of this model. This characterization allows the physics of the heat gains and the ventilation flow to be represented in a realistic manner. Morton et al.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING computational burden on a BES program. ASHRAE RP-1222 provides a short history (and examples) of each type of model. The result is a zone divided vertically into two or more well separated regions – each region characterized by a single temperature or temperature profile. is that the interface between the two layers occurs at the height (h) where the vertical buoyancy driven plume flow rate is the same as the inflow rate. These models assume that the dominant mechanism is plumedriven flow from discrete internal sources and that other effects (such as buoyancy driven flow at walls or windows) may be neglected. Model Description Single Plume Two Layer Model The simplest form of the plume equation based models is the case of a single plume in an adiabatic box with constant supply air flow. a value of 0. these heat sources also produce plumes that can be included in the model. while successful in predicting the flow and temperature field for geometries similar to those used in its development. Alternatively. can suffer from lack of flexibility and clarity in the modeling approximations. ASHRAE RP-1222 [Chen & Griffith 2002] divides the candidate models into two categories: nodal and zonal. without the introduction of ad hoc assumptions. The UCSD DV model is closer to a nodal model than to a zonal model. When dealing with diverse geometries it is not clear that the flow coefficients used in the model are applicable or why they can be used since plumes. For this configuration two layers form in the room: a lower layer with similar density and temperature as the inflow air and a mixed upper layer with the same density / temperature as the outflow air. In terms of nodal models for displacement ventilation we mention the Mundt model [Mundt 1996]. Nodal models describe the zone air as a network of nodes connected by flow paths. the fundamental driving mechanisms of the displacement flow. For a point source of buoyancy in a nonstratified environment (a plume) the airflow rate increases with vertical distance from the source according to: V = C i B1/ 3 z 5/ 3 where (187) V = plume volume flux [m3/s] B = buoyancy flux [m4/s3] z = vertical distance above source [m] 6 9 C = α ( α )1/ 3 π 2 / 3 5 10 α = plume entrainment constant. For an ideal gas 4/23/05 269 .

for any room inflow rate (F. nearby plumes may coalesce.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING ∆ρ/ρ = ∆Τ/Τ resulting in the following relation between heat input rate and buoyancy flux: (188) B= where gQ ρ C pT (189) ρ = density of air [kg/m3] T = air temperature [K] g = acceleration of gravity [m/s2] Q = heat input rate [W] C p =specific heat capacity of air [J/kgK] Since the plume volume flow rate increases with height with exponent 5/3. 2. 4/23/05 270 . This case is much more complex but if we are satisfied with a first order precision model the equal strength model can be used by averaging the plume strengths [Carrilho da Graça. two layer flow can be established whenever the plume buoyancy flux is more than 1/7 of the horizontal flux [Hunt et al. the biggest challenge remains the interaction between wall driven boundary layers (positively and negatively buoyant) and displacement flows. For this reason. For multiple plumes of equal strength a straight-forward extension of the single is possible. (m3/s)) there will always be a height (h. Even in a case where all plumes are of equal strength. it would be rare for a real world case to consist of a single point-source plume originating on the floor.(2004)].1) equal to F and solving for z=h: h = F 3/ 5 /(C 3/ 5 B1/ 5 ) Substituting (189) in (190) and introducing air properties at 20 C gives: (190) h = 24. Although these limitations are significant it is important to note that even in the presence of dominant convection from the floor surface. Upward moving wall or floor generated buoyancy flux in the lower layer is of the same order of magnitude as plume driven flow. (2002)]. as long as the heated area does not exceed 15% of the room floor [Holford et al. (2002)]. This height is obtained by setting (1. a buoyancy. As the complexity of the physical systems modeled increases some limitations must be imposed. the model that is developed below is not applicable when: 1.5 meters apart at their source will coalesce within 2 meters [Kaye & Linden. In particular.. N plumes of unequal strength result in the formation of n vertical layers.55i F 3/ 5 / Q1/ 5 Multiple Plumes and Wall Heat Transfer (191) Of course. A two layer structure can also originate when the only heat source is a heated portion of the room floor. Plumes that are less than 0.(m)) where the plume driven flow rate matches the inflow rate. (2003)]. Downward moving buoyancy driven airflow rate is of the same order of magnitude as plume driven flow (these airflow currents are typically generated on lateral surfaces or in the ceiling whenever these surfaces are much cooler than the room air). unaffected by heat gains from walls and windows.

3. a floor level temperature Tfloor to account for the heat transfer from the floor into the supply air 2.conv Qmxz = Qgl .conv + Qbb . an upper level temperature Tmx representing the temperature of the upper.conv + Qgaseq .conv + Qhw.conv + Qotheq . The heat gain Q is estimated by summing all the convective internal gains located in the occupied subzone – task lights. With these assumptions we can describe the implementation. with equal strength. equipment – and dividing this power equally among the n plumes. an occupied subzone temperature Toc representing the temperature of the occupied region. people.conv + Qeleq . Qocz = Qoc . The calculation is done in subroutine CalcUCSDDV. The UCSD DV model is controlled by the subroutine ManageUCSDDVModel which is called from the RoomAirModelManager. First we calculate the convective heat gain going into the upper and lower regions. The supply air flow rate V is obtained by summing all the air flows entering the zone: supply air.i MCPtot = MCPzone + MCPsys 4/23/05 271 . mixed region and the outflow temperature. and inter-zone flow.conv + Qstmeq .conv + Qtl . ventilation.conv Qtot .conv = Qocz + Qmxz Next we sum up the inlet air flows in the form of MCP (mass flow rate times the air specific heat capacity) and MCPT (mass flow rate times Cp times air temperature). We assume that the model for multiple. equal strength plumes (equations (192) and (193) will be adequate for our calculations. MCPzone = MCPi + MCPvent + MCPmix MCPsys = inlets ∑ mC i p .ALTERNATIVE MODELING PROCESSES COMPONENT SIZING For the case of multiple non-coalescing plumes (n).conv + Qltp + Qhtrad . infiltration. the total vertical airflow for a given height is: V = C in1/ 3 B1/ 3 z 5/ 3 resulting in a mixed layer height of: (192) h = 24.55i F 3/ 5 /(nQ)1/ 5 Implementation (193) The model predicts three temperatures that characterize the three main levels in the stratification of the room: 1. The RoomAirModelManager selects which zone model will be used for each zone.

3. HAToc.i ⋅ Ai for a region and HAT is surfaces ∑ hc . The hc’s calculated in step 1 depend on the subzone temperaures and the boundary layer height. 4/23/05 272 .i is the convective heat transfer coefficient for surface i. Iterative procedure 1.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING MCPTzone = MCPTi + MCPTvent + MCPTmix MCPTsys = inlets ∑ mC i p .000833 = 1/( ρ air ⋅ c p . The sum is over all the surfaces bounding the region.000833 ⋅ MCPtot Frhb = ( 24. Calculate the three subzone temperatutes: Tfloor. Ai is the area of surface i. and Ti is the surface temperature of surface i. Using this we calculate the total power in the plumes and the power per plume.i i T MCPTtot = MCPTzone + MCPTsys The number of plumes per occupant N plumesperpers is a user input. The total number of plumes in the zone is: N plumes = N occ ⋅ N plumesperperson The gains fraction Frgains is a user input via a schedule. Hence the need for iteration Next we describe each steps 1 and 3 in more detail. HATmx. It is the fraction of the convective gains in the occupied subzone that remain in that subzone. hc . air ) converts MCP to a volumetric flow rate.55 H ceil ) ⎜ 1/ 3 ⎜ N ⎝ plumes ⋅ Q perplume ⎞ ⎟ ⎟ ⎠ 3/ 5 (194) where 0. Call subroutine HcUCSDDV to calculate a convective heat transfer coefficient for each surface in the zone. Toc and Tmx. Q plumes = (1 − Frgains ) ⋅ Qtot . Recalculate Frhb using the equation (194). In turn the subzone temperatures depend on the HA and HAT’s calculated in step 1. Next we iterate tot over the following 3 steps. and HATfl. and HAmx. HAoc. 2.conv Q perplume = Q plumes / N plumes We now make an initial estimate of the height fraction Frhb (height of the boundary layer divided by the total zone height). an effective air temperature for each surface.i ⋅ Ai ⋅ Ti for a region. HAfl. Here HA is surfaces ∑ hc . ⎛ 0.

oc .oc + HAoc + MCPtot ) Tmx = (Cair . fl − (3 / 2) ⋅ T−2. zone conditions must be recalculated assuming a well-mixed zone. It loops through all the surfaces in each zone and decides whether the surface is located in the upper. T fl = (Cair .oc + (1/ 3) ⋅ T−3. This routine uses the “detailed” natural convection coefficient calculation that depends on surface tilt and ∆T .mx = Rair . Otherwise the subzone temperatures are obtained straightforwardly by solving an energy balance equation for each subzone. Cair .mx ⋅ Mulcap /(∆t z ⋅ 3600) Cair . mx − (3 / 2) ⋅ T−2. the appropriate subzone HA and HAT sums are incremented.mx ) + Qocz ⋅ (1 − Frgains ) + Qmxz + HATmx + Toc ⋅ MCPtot ) /((11/ 6) ⋅ Cair .e. mx ⋅ (3 ⋅ T−1.mx + HAmx + MCPtot ) Here Cair . If entirely in one subzone the subzone temperature is stored in the surface effective temperature variable TempEffBulkAir(SurfNum) and hc for the surface is calculated by a call to subroutine CalcDetailedHcInForDVModel. Now that this calculation is complete and the subzone temperatures and depths calculated. fl . occupied subzone. we check to see if this assumption was justified.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Step 1 Subroutine HcUCSDDV is quite straightforward. fl + (1/ 3) ⋅ T−3. During the surface loop. or if the surface is in both subzones. This calculation is appropriate for situations with low air velocity. the subroutine calculates hc for each subzone and then averages them.oc ) + Qocz ⋅ Frgains + HAToc + T fl ⋅ MCPtot ) /((11/ 6) ⋅ Cair . mixed subzone or the lower.. Namely a third order finite difference expansion of the temperature time derivative is used in updating the subzone temperatures. Cair . Mixed calculation The above iterative procedure assumed that displacement ventilation was taking place: i. 1/ 3 Step 3 The calculation of subzone temperatures follows the method used in the ZoneTempPredictorCorrector module and described in the section Basis for the System and Zone Integration. If not. 4/23/05 273 . fl ) + HAT fl + MCPTtot ) /((11/ 6) ⋅ Cair . fl + HAfl + MCPtot ) Toc = (Cair .mx is calculated by Rair .mx ⋅ ∆t z / ∆tsys The other subzone air heat capacities are calculated in the same manner.air .oc ⋅ (3 ⋅ T−1. once the hc for a surface is calculated. mx are the heat capacities of the air volume in each subzone. fl ⋅ (3 ⋅ T−1.mx ⋅ c p .oc − (3 / 2) ⋅ T−2. and Cair .mx + (1/ 3) ⋅ T−3. For surfaces that bound 2 subzones. If a surface is in 2 subzones the HA and HAT for each subzone are incremented based on the area of the surface in each subzone. conditions were favorable temperature stratification in the zone. weighting by the amount of surface in each subzone.mx = Vmx ⋅ (∆zmx / zceil ) ⋅ ρ air .

First.2 meters. setting the boundary height to 1.000833 ⋅ MCPtot / Vmin We define heights: H trans = Frhb ⋅ H ceil H mxavg = ( H ceil + H trans ) / 2 H ocavg = ( H fltop + H trans ) / 2 4/23/05 274 . we calculate a minimum flow fraction: Vmin = (1. z + (1/ 3) ⋅ T−3.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING If Tmx < Toc or MCP ≤ 0 or H fr ⋅ H ceil < H fl . z − (3 / 2) ⋅ T−2. the floor subzone is always 0. Repeat step 1 Final calculations The displacement ventilation calculation finishes by calculating some report variables.top is the height of the top of the floor subzone. H fl . that is.5 / 24. The mixed calculation iteratively calculates surface convection coefficients and room temperature just like the displacement ventilation calculation described above. Call HcUCSDDV to calculate the hc’s.min then the following mixed tot calculation will replace the displacement ventilation calculation.conv + HAToc + HATmx + HAT fl + MCPTtot ) /((11/ 6) ⋅ Cair . z + HAoc + HAmx + HAfl + MCPtot ) Tmx = Tavg Toc = Tavg T fl = Tavg 2. Using equation (194).5 meters and solving for the flow.55)5/3 ⋅ N plumes ⋅ Q1/ 3 perplume Frmin flow = . Note: ∆zocc . It is defined to be 0. z ) + Qtot .min is the minimum thickness of occupied subzone. The 3 subzone temperatures are then set equal to Tavg. It is set to 0. Then the code iterates over these steps.top + ∆zocc .1 meter above the floor surface. Calculate Tavg using Tavg = (Cair . 3. only one zone temperature Tavg is calculated.2 meters thick and T fl is the temperature at 0. z ⋅ (3 ⋅ T−1. 1. In the mixed case however.2 meters. Frhb is set equal to zero.

ALTERNATIVE MODELING PROCESSES COMPONENT SIZING H flavg = H fltop / 2 Using the user defined comfort height we calculate the comfort temperature. If mixing: Tstat = Tavg If displacement ventilation: If Hstat < Hflavg Tstat = T fl Else if H stat ≥ H flavg and H stat < H ocavg Tstat = (T fl ( H ocavg − H stat ) + Tmx ( H stat − H flavg ) /( H ocavg − H flavg ) Else if H stat ≥ H ocavg and H stat < H mxavg Tstat = (Toc ( H mxavg − H stat ) + Tmx ( H stat − H ocavg ) /( H mxavg − H ocavg ) Else if H stat ≥ H mxavg and H stat < H ceil 4/23/05 275 . If mixing: Tcomf = Tavg If displacement ventilation: If Hcomf < Hflavg Tcomf = T fl Else if H comf ≥ H flavg and H comf < H ocavg Tcomf = (T fl ( H ocavg − H comf ) + Tmx ( H comf − H flavg ) /( H ocavg − H flavg ) Else if H comf ≥ H ocavg and H comf < H mxavg Tcomf = (Toc ( H mxavg − H comf ) + Tmx ( H comf − H ocavg ) /( H mxavg − H ocavg ) Else if H comf ≥ H mxavg and H comf < H ceil Tcomf = Tmx Using the user defined thermostat height we calculate the temperature at the thermostat.

the zone node temperature is set to Tmx.999 The maximum temperature gradient is: If H ocavg − H flavg > 0.0 H trans = −9. G. the displacement ventilation report variables are set to flag values.999 If H mxavg − H ocavg > 0. and B. Ph. Thesis.999 and GradTmax = max(GradTmax. University of California..1 = −9. Incorporating Nodal Room Air Models into Building Energy Calculation Procedures. Simplified models for heat transfer in rooms. San Diego. ASHRAE RP-1222 Final Report.999 Finally.999 GradTmax = −9. Q.1 GradTavg = (Tmx − T fl ) /( H mxavg − H flavg ) else GradTavg = −9. 2002.1 GradTmax.1 GradTmax. flow = −1.min or Tmx − Toc < ∆TCrit Re p tot GradTavg = −9.999 FRmin. GradTmax. If Tmx < Toc or MCP ≤ 0 or H fr ⋅ H ceil < H fl . Chen.1 = (Toc − T fl ) /( H ocavg − H flavg ) else GradTmax. D. 4/23/05 276 .2 = −9. 2003.2 = (Tmx − Toc ) /( H mxavg − H ocavg ) else GradTmax. Griffith.1 . References Carrilho da Graca.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Tstat = Tmx The average temperature gradient is: If H mxavg − H flavg > 0. if the zone is deemed to be mixed.top + ∆zocc .2 ) For reporting purposes.

the main jet region and the recirculations. 577-580. 2002. 2001. Haves. Holford and P. UCSD Cross Ventilation Room Air Model Overview The UCSD Cross Ventilation Room Air Model provides a simple model for heat transfer and vertical temperature profile prediction in cross ventilated rooms. S. pp. J. A nodal model for displacement ventilation and chilled ceiling systems in office spaces. Steady-state flows in an enclosure ventilated by buoyancy forces assisted by wind. Linden. Proceedings of RoomVent 2002. and P. pp. Multiple sources of buoyancy in a naturally ventilated enclosure. 581-584. G.F. G. Coalescing axisymmetric turbulent plumes. Emptying filling boxes: the fluid mechanics of natural ventilation. Proc. pp. G. Competition between heat sources in a ventilated space.R. Because of the high momentum conservation. Kaye. pp. with air flowing through windows. Bulletin N38. 2004. and approximate solutions of the Navier-Stokes equations. Thesis. 4/23/05 277 . E.. Turner.F. and P. P. and P. Vol. Cross-Ventilation Cross Ventilation (CV) is common in many naturally ventilated buildings. Morton. Vol.F. 1-23. 355-386. Linden. The Performance of Displacement Ventilation Systems – Experimental and Theoretical Studies.S. B...F. pp. Lane-Serff and D.F. 1996. 2002. Vol. Journal of Fluid Mechanics. pp. 41--63. 753-762. 502. Natural ventilation of an enclosure containing two buoyancy sources. Hunt. Vol. and P. CV strategies are often used when there is need for high ventilation airflow rates. CV flows often occur in wind-driven ventilation systems with inflow through operable windows. Building and Environment. Smeed. 1996. Proceedings of RoomVent 2002. and P. Turbulent gravitational convection from maintained and instantaneous sources.I.F. Journal of Fluid Mechanics. Vol A234. 426. Cooper. G.A. open doorways and large internal apertures across rooms and corridors in the building. Rees. This model was developed using scaling analysis. Linden. K. P. Linden. of Building Simulation. . Validation of displacement ventilation simplified models. Taylor andJ. 1996. In these cases significant conservation of inflow momentum occurs with the inflow traveling freely across the room. 1956.M. The model distinguishes two regions in the room.N. 311. Linden. experimental correlations. Holford. Building Services Engineering KTH. 153-176. and P. D.F.J. Hunt. Guitton. Journal of Fluid Mechanics.F. CV flows may be caused by wind or a mechanical ventilation system. Linden. computational fluid dynamics. Vol. Journal of Fluid Mechanics.. 212. pp. 177-192. Proceedings of the Royal Society of London.M. Journal of Fluid Mechanics.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Cooper.R. G. pp.. 2001. pp. and predicts characteristic airflow velocities and average air temperatures. P. Vol. 311. 1990. 1997.R. 309-335. M. Hunt and P. Ph. Dominique. 26. Characterization of the flow driven by a finite area heat source in a ventilated enclosure.R. Stockholm. Mundt. J. The model consists of a set of analytical expressions that clearly display the first order effects of room geometry and inflow characteristics on Cross Ventilation airflow and heat transfer. Linden.

ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Figure 91. Left: basic compartment geometry considered in the model. most of the contact between ventilation flow and the internal surfaces occurs in the recirculation regions that occupy the majority of the room volume. Neiswanger et al. In these cases. when the inlet is located close to the center of the inlet surface. (2003). Baturin & Billington. reflecting energy conservation. The simplest flow configuration. Several studies have shown that surface heat transfer has a large impact on energy consumption predictions in mechanical systems [Carrilho da Graça. (1987). (2001)]. The left hand side of Figure 91 shows a simple room geometry that can lead to crossventilation. (1977). result of a CFD simulation using the geometry on the left (taking advantage of the system symmetry to simulate only one half of the room volume). The most relevant to the present problem are [Aynsley et al. commonly occurs in corridors and long spaces whose inlet aperture area is similar to the room cross-sectional area. These recirculating flow regions have been observed in many experiments. with no recirculations. A more complex airflow pattern occurs when the inlet aperture area is an order of magnitude smaller than the cross sectional area of the room AR=W. AR=4. heat transfer between airflow. As ventilation air flows across the room. (1995)]. Right: top view of one half of the velocity field (at height H/2).H (for the case shown in Figure 91. (1972).H2). or a combination of the two with both recirculation and inlet flow attaching to a lateral surface or the ceiling. Ohba et al. room surfaces and internal heat sources occurs and the airflow temperature changes between inlet and outlet. 4/23/05 278 . Kalema & Haapala. In this case.. before the introduction of the present model. similar to turbulent flow in a channel. Its magnitude has considerable influence in the effective indoor temperature as well as in determining the success of the night cooling system. Recirculating flow Figure 92 shows a schematic representation of the two basic airflow patterns that can occur in CV. the required level of modeling detail was not available in a whole building thermal simulation tool. However. with air moving in the opposite direction to the core jet flow. Any cross-ventilated room will have an airflow pattern that is either similar to one of the two base cases shown in Figure 92 (with or without recirculations). In these room geometries. The flow velocity profile across the channel is approximately flat as a result of the high degree of mixing that is characteristic of turbulent flows. the main CV region in the core of the room entrains air from the adjacent regions and forms recirculations that ensure mass conservation... Heat transfer between air and room surfaces is important in both of these passive-cooling strategies. The average airflow velocity in the cross section can be obtained approximately by dividing the flow rate by the cross sectional area of the space. the flow occupies the full cross section of the room and the transport of pollutants and momentum is unidirectional.

furniture etc). but. all room ventilation situations will display variable temperature gradients as the air temperature changes along the flow (see TR1 and TR2 in the figure b). In the figure b) and c). In the figure a). 4/23/05 279 . these surfaces are flat (walls. as a consequence of heat transfer into a flow stream with limited heat capacity (that depends on the local ventilation rate in different parts of the room volume). schematically. c). nevertheless.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING WALL WALL Main Jet Flow WALL WALL AIN≈AR => NO Recirculations WALL WALL Recirculation Flow Main Jet Flow Recirculation Flow WALL WALL AIN<<AR => flow WITH Recirculations Figure 92. while flat plate heat transfer occurs at constant free stream temperature (TIN=TOUT in a)). a heat sink. Top view of the possible airflow patterns in cross-ventilation. the differences between flat plate heat transfer (a)). with infinite thermal capacity. simple use of flat plate correlations is not possible as the forced air flow velocity or free stream air temperature are not known since the confined flow occurring in most rooms does not resemble a free stream. the effects of flow confinement are clear. Figure 93 shows. floors. Often. d)). the boundary layer that develops exchanges heat with the environment. and two types of cross ventilation flows (no recirculations in b) and with recirculations in c)). Effects of recirculating flow in heat transfer Ventilation air changes temperature as it is exposed to different parts of the internal room surfaces.

4/23/05 280 . Figure 93. fully-mixed model the room temperature is constant (TRM in d)). variable TR2 case R TW T E M P E R A T U R E TR2 ? TRM TR1 TIN A B d) Temperature variations across the room length. fixed stream temperature TIN INTERNAL SURFACE TW hS INLET TIN hS hS TR1 …. (2003)]. Differences between flat plate heat transfer and cross ventilation flow In a one-node. variable TR1 case C INTERNAL SURFACE TW hS ∆T2 TR2 ? hS T2 OUTLET TOUT INLET TIN T1 hS ∆T2 hS INTERNAL SURFACE TW c) Two dimensional room heat transfer. Whenever mixing is not perfect. one node model leads to errors of up to 100%. The differences between CFD simulations of recirculating flows with a fixed heat source in the recirculation region and mixed model predictions can reach 1. the temperature gradients that determine the heat transfer with the room surfaces will be incorrectly modeled.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING INLET TIN TIN hS FLAT PLATE TW OUTLET TOUT=TIN a) Flat plate heat transfer.… hS TR1 OUTLET TOUT INTERNAL SURFACE TW b) One dimensional room heat transfer.5°C [Carrilho da Graça. Confinement effects due to recirculation flow are also very influential in internal surface heat transfer predictions where using a perfectly mixed flow.

Difficulties in determining the local forced convection heat transfer coefficient lead modelers to consider only natural convection. In particular. while simple to integrate in energy analysis software. also available in EnergyPlus). By definition. the use of CFD is impractical and simpler heat transfer models are needed. Fully mixed room air heat transfer models use a single modeling point to characterize indoor air temperature in the room (one example is the mixed flow model. These features make these models complex to use and often imprecise. due to the nonlinear interactions that exist in building heat transfer processes. In all other cases these models fail to predict the magnitude of the heat transfer and also most room geometry induced effects. ignoring airflow pattern effects. have the need to simulate using local measured weather spanning several days or months. in some regions of the room air can move in other directions. Airflow pattern The main challenge when modeling the airflow pattern is to predict the flow characteristics in the recirculation regions. Each of the two parts of the problem poses considerable challenges. (1991). and. 1992]. zonal models. Application of these correlations to most geometries found in real design is difficult as the adequate correlation must be identified (by matching the design geometry to the geometries used to develop the correlations) and is often not available. (1988). designers have to analyze multi-room ventilation geometries. two interrelating components of the CV system must be considered: the airflow pattern and the magnitude of the local heat transfer between room air and internal surfaces (partition elements. This inaccuracy is a consequence of the use of two major approximations: • Room air is considered fully mixed. occupants. Altmayer et al. With this approximation. In these cases. etc. 4/23/05 281 . are. For the present problem. Spitler et al. (1984). In the present model this characterization is done using scaling analysis principles. plumes. precise only when the flow is mixed.. boundary layers. fully mixed room air heat transfer models and experimental correlations. but.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Existing approaches In order to predict heat transfer in CV airflows there are currently four available options: computational fluid dynamics (CFD. etc. Transfers between airflow and the internal sources depend on the local concentration gradient and transfer coefficient. Experimental correlations can be a good solution to model complex physical systems. This approximation neglects the effects of the temperature variations that occur as air moves across a room and comes into contact with internal surfaces and heat sources resulting in inaccurate calculation of the thermal capacity of the airflow. in their present form. Because the momentum equation is not solved in the iteration procedure. correlations are only available for particular cases [Chandra & Kerestecioglu. These models generally require user identification of the dominant room airflow components (jets. and not exhausted by the ventilation air. • Model Description In order to model heat transfer and evaluate thermal comfort. In many situations.). when modeling the effects of internal thermal mass. Temperature variations inside the room are not modeled.) that are “contained” in particular zones [Allard. an artificial flow resistance is imposed between room zones. These models. even if the correct local heat transfer coefficient is used the result can still be imprecise. it is relevant to determine how much energy from the internal gains is transferred to the internal surfaces. any CV airflow pattern has an element of direct air movement between inlet and outlet. using Reynolds averaged turbulence models).. Awbi & Hatton. (2000)]. Zonal models simulate room heat transfer by numerically solving for mass and energy conservation in a set of fully mixed zones (often less then twenty).

0m) and inflow aperture areas (1-4m2). Width: 4. dominated by turbulence).ALTERNATIVE MODELING PROCESSES COMPONENT SIZING The CV flows to which the model applies are bounded by a stationary geometry and are dominated by horizontal momentum flux. AIN. A* = R AIN AIN (196) The velocity scale (UIN) of the inlet jet. all the velocities in the room are expected to scale linearly with the characteristic velocity of the inlet flow [Carrilho da Graça.4m. approximately self similar and suitable for the application of scaling analysis principles. as opposed to buoyancy. Room length in the CV direction. In these flows.W. f ( L. The simulations spanned rooms with variable dimensions (height: 2. By multiplying the velocities predicted using expression (195) by the area over which the recirculation flow occurs correlations for the magnitude of the recirculation flow rates can be obtained.25-13.5m. 2003]: U R = Cn . W.. as in the present case. Square root of the inlet area. and inlet location. This momentum flows into the room in the form of a jet whose characteristic dimension is typically not more than an order of magnitude smaller than the room length (or the characteristic room dimension ((L. AIN . Two clear trends were visible in these flows: recirculations are present whenever A* is bigger than two and the behavior of the recirculations depends on the length to width ratio of the room. if the flow regime is stable (either laminar.H)1/3)). H . Square root of the room cross section (perpendicular to the CV flow AR=W.. U MAX g (r ) dA .59.. U IN = M ρ AIN ( ) 2 (197) In order to correlate the velocities in different regions of the room the corresponding correlation constants Cn and scaling laws f (see right hand side of Figure 91) must be obtained. 2. This correlation results in the following expression for the velocity in the recirculating portion of the flow: 4/23/05 282 . 2003]. CL = W − WIN AR 2 (198) The scaling principle behind the correlations is: the momentum flux through the inlet aperture is the dominant feature in the CV flow. W . Sets of CFD simulations [Carrilho da Graça.) . Using dimensional analysis principles it is possible to compose two non-dimensional parameters with the three independent length scales in this problem: 1. based on geometries similar to Figure 91 were used to obtain the correlation constants and to validate the simplifying assumptions used when developing the correlation expressions (the “f” functions). H. or. is defined using inflow momentum flux: M =∫ AIN ρ . Length: 2. We propose that the two independent parameters (formed using the length scales) that the function f will depend on are: L* = A L . The criterion that most successfully distinguishes between different “groups” of rooms is are: A* > 2 ⇔ 2L AIN 1 < . resulting in a jet flow that is never fully developed and strongly confined. 3.25-3.U IN (195) The function f is expected to depend on L. For ventilation systems with these characteristics.H).. the flow pattern will also be stable.

4/23/05 283 . Internal forced convection models are not available in EnergyPlus (for natural convection the ASHRAE Detailed model is available). As expected the slopes in these correlations are smaller than for the standard correlations. as a consequence of the higher dissipation that occurs in long rooms. The modeling principle described above is used to predict four parameters that will then be used to model thermal comfort and heat transfer in cross ventilated rooms (described below).U IN AR (199) This scaling relation can also be obtained from a simplified solution of the Navier-Stokes equations. Long rooms (4<CL<11): dashed line. 1) Recirculation Region Average Velocity The average recirculating flow velocity in the room cross section with maximum flow rate is predicted using the following expression (1/3<CL<11): U R = CR L AR . Correlation for standard rooms (CL<4): gray line. In most cross ventilation airflows heat transfer occurs though mixed convection and the forced convection component is relevant and must be modeled. by evaluating the momentum flux through the shear layer that develops in the perimeter of the inflow jet. AIN 3 2 FIN . 4 < C L ≤ 11 (200) Figure 94 shows the results of the correlation for the standard (gray line) and long rooms (dashed line). 1 3 ≤ C L ≤ 4 CR = ⎨ ⎩0. the maximum airflow rate in the recirculation regions (defining the heat capacity of the recirculating airflow) and the average airflow velocity near the internal partitions (for forced convection heat transfer). ⎧ 0. The horizontal axis is the value of the non-dimensional factor on the right hand side of each correlation expression (see equation (200)).ALTERNATIVE MODELING PROCESSES COMPONENT SIZING U R = CM 2 AIN . These parameters are the characteristic velocities in the jet and recirculation regions (for thermal comfort). L 2 . Figure 94. The correlation expression labeled with a (*) refers to the cases with CL > 4 (long rooms).298 . Average velocity in the recirculation region.162 .

ALTERNATIVE MODELING PROCESSES COMPONENT SIZING 2) Jet Region Average Velocity Average volumetric velocity in the main jet region is given by (1/3<CL<11): U J = 1. AIN . 1 3 ≤ C L ≤ 4 CF = ⎨ ⎩0.077 . Correlation for standard rooms (CL<4): gray line. The horizontal axis is the value of the non-dimensional factor on the right hand side of each correlation expression (see right hand side of expression (201)).147 . FIN . Average velocity in the main jet region. 3) Total airflow rate in the recirculation regions The total volumetric flow rate of the recirculation region of the room airflow is given by (1/3<CL<11): FR = C F L AR . 1 3 ≤ CL ≤ 11 (201) Figure 95. Long rooms (4<CL<11): dashed line.56 FIN AR . 3/ 2 AIN ⎧ 0. 4 < C L ≤ 11 (202) 4/23/05 284 .

(1983)] and has been subsequently used in two other studies that are relevant to the present work: with vertical surfaces and horizontal forced flow [Neiswanger et al.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Figure 96. Long rooms (4<CL<11): dashed line. Colburn’s analogy between heat and momentum transfers [Bejan. In order to predict the local mixed convection heat transfer coefficient that results from these two mechanisms it is possible to use the following formula [Siebers et al.. Air enters the room in a developing jet composed of two shear layers that quickly become turbulent. it is a reasonable approximation to use a correlation for forced convection turbulent boundary layers in all the internal surfaces.0296 ⎜ ⎟ ⎝ν ⎠ − 1 5 (204) 285 . For this reason. Maximum flow rate in the recirculation region. (1983)]: hMX = (hF ε + hN ε )1 ε (203) This formula was verified experimentally [Seibers et al. Since the air is already turbulent. resulting in: 4 5 hF = ρ c p U ∞ P 4/23/05 −2 1 3 r 2 C fx = ρ c p U ∞ Pr −2 3 ⎛ x⎞ 0. (2000)]. Correlation for standard rooms (CL<4): gray line.. that exist in these regions. 4) Forced convection In CV flow patterns that conform to the conditions mentioned in the previous section.. Equation (203) was found to agree better with experimental measurements when ε was set to 3. Since natural convection always occurs when there is a temperature difference between the air and a surface. The horizontal axis is the value of the non-dimensional factor on the right hand side of each correlation expression (see right hand side of expression (202)).2. Part of the air in the shear layers goes into the recirculation regions forming the wall jets. (1987)] and horizontal surfaces with horizontal and vertically impinging forced flow [Awbi&Hatton. room heat transfer always occurs through mixed convection. there is no significant laminar region in these boundary layers. The boundary layers that form close to the thermally active room surfaces are then driven by buoyancy and momentum transferred from the dominant CV flow. (1994)] to estimate the forced component of the convection heat transfer. airflow is predominantly turbulent and there is forced approximately horizontal air movement close to the room surfaces. or wall currents. In all of these studies.

Therefore. the other being the need to have a flow that is dominated by forced convection. The same reasoning applies to the other extreme case. One clear reason why this limit is needed is the fact that dissipation in the recirculation region is neglected. an upper limit is needed.8 ⎛ 121 + L AIN = CS ⎜ ⎜ AR AIN ⎝ ⎞ 0. The upper limit imposed on CL already indicates that the model should not be applied when the length of expansion for the shear layer is very large. depending on room geometry. 4 < CL ≤ 11 (206) The correlations presented above are expected to be very imprecise for cases with CL <1/3 and CL >11. it is not feasible to integrate the x dependent coefficient in equation (204) for every room surface.3U S 0. This limit ensures that one of the conditions to apply the model is present. scaled by inflow aperture length scale.8 The proposed correlation is: (205) US 0. a limit should be imposed on A* since it can be expected that for very large values of this parameter the momentum flux in the return flow becomes infinitesimal and therefore the present analysis is not applicable. 1 3 ≤ CL ≤ 4 CS = ⎨ ⎩0. the effects of different lengths.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING The heat transfer coefficient has a weak dependence on x through an exponent of -0. Performing the integration using standard properties results in: hF = 4. lateral walls and windows). In the cases with small CL the model is not applicable because the aspect ratio of the room is too extreme (large cross section to room length ratio) and the dominance of the shear layer momentum transfer in recirculation flow may not occur. half way between short L=2m and long rooms L=10m). Forced convection boundary layer heat transfer is not influenced by buoyancy driven transport (the turbulent eddy formation process is driven by momentum transfer) and for this reason there is no temperature or orientation dependence in the result. Since the x dependence is small. In a simple room heat transfer model. 4/23/05 286 . a further approximation is introduced: an integration is performed only once using a typical room surface length of 6 meters (a characteristic value. This is a consequence of boundary layer growth and the resulting reduction in the temperature gradient. in this case due to the large room length. This result will be used for all room surfaces (floor. ceiling. large CL. are smaller than the model’s overall accuracy. After analyzing the cases in the CFD library as well as few exploratory runs for very large rooms it was decided to use the following criterion as the upper limit of application of the model: L* A* < 20 (207) This expression places an upper limit on room dimensions.082 .8 . while the energy source remains constant (the inflow jet).115. ⎟ U IN ⎟ ⎠ 0.4 ⎧ 0.2. clearly as the room becomes larger dissipation effects increase. In addition to this length limitation.

1 0.25 0.2 2.3 0.4 1. 4/23/05 287 .4 Figure 97. Results of the correlation for CV induced airflow velocity near the room surfaces. Long rooms (4<CL<11): dashed line.1145x R2 = 0.5923 1 1.4 0. The wall boundary layer part of the recirculation flow (label W). with coordinate r varying between 0 and L. Figure 98 shows a schematic representation of the heat transfer process (showing a top view of one half of a symmetrical room).8 (the left hand side of expression (203)). Top view of the flow structure in case R. On the horizontal axis the values of the coefficient with exponent 0.2 0.4 on the right hand side of expression (203).6408 y = 0.05 0 y = 0. Recirculation path W TR+∆TR/2 r=0 I N TR R TR+∆TR O U T r=L J TIN Shear layer “interface” TIN+∆TJ Figure 98.8 2.35 (m/s)0. Heat transfer in recirculating flows In most cross ventilation flows with recirculations the mixing between recirculation and inflow jet is only partial. The recirculation region coordinate system is shown in the figure.8 0. The units of the vertical axis are: (m/s)0. The dark gray arrows show heat transfer in the shear layer.0818x R2 = 0.15 0. The part of the recirculation flow that exchanges heat with the jet (label R). The light gray arrows show flow direction. The flow is divided into three distinct streams with connected temperature variations: The main jet (labeled J in the figure). Standard rooms (CL<4): gray line.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING 0.6 3 3.

The temperature in the recirculation region is given by: ρ CP R F L hS AS r − TW (r ) = TS − (TS − TIN ) e e (e −e ASL hSL (1+ R ) ρ CP R F − 1) A h (1+ R ) − SL SL ρ CP R F hS AS ρ CP R F (211) (1 + R) + R 4/23/05 288 . temperature difference: ρ CP F P (TJ ( x) − TR ( x) ) ∂ TJ ( x) A h = − SL SL (TJ ( x) − TR ( x) ) = L ∂x CD AIN 4 σ π (209) Predicting air temperature in the recirculation region of the flow An accurate representation of system behavior must consider both the local heat transfer in the shear layer (varying along the x direction. ASL = L P .ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Heat transfer in a shear layer It is possible to use a formulation with a heat transfer area. local. driven by the temperature profiles TJ(x) and TR(x)) and the heat transfer in the wall boundary layers (with temperature TW(r)). TJ (0) = TIN The first equation in (210) models heat transfer in the wall layer. resulting in the following system of equations: ∂TW (r ) AS hS ⎧ ⎪ ρ CP F R ∂r = L (TS − TW (r ) ) ⎪ ⎪ ⎪ ⎪ ρ CP F R ∂TR ( x) = ASL hSL (TJ ( x) − TR ( x) ) ⎪ ⎨ L ∂x ⎪ ⎪ ASL hSL ∂TJ ( x) ⎪ ⎪ ρ CP F ∂x = − L (TJ ( x) − TR ( x) ) ⎪ ⎪ ⎩ The relevant boundary conditions are: (210) TW ( L ) = TR (0) . a convection heat transfer coefficient and an average shear layer temperature difference to model the heat exchange between jet and recirculation flows: Qy =0 = ASL hSL ∆TSL . TW (0) = TR ( L) . the heat transfer process is driven by the variable. The second models the temperature variation of the portion of the recirculation flow that is in contact with the jet. hSL = ρ CP F CD AIN 4 σ π (208) Locally. it differs from the third equation only in a sign (the temperature variation in TR is symmetric to the variation in TJ) and the flow rate (F. Both transfers can be modeled using differential equations.R for the recirculation and F for the main jet).

Due to this limited ability. In the case of a flow in a room with adiabatic surfaces and heat gains in the recirculation region the indoor air temperatures can be predicted using the solution to the following system of equations: ⎧ ρ CP F R (TR (0) − TR ( L)) = GR ⎪ ⎪ ⎪ ∂TR ( x) ASL hSL = (TJ ( x) − TR ( x) ) ⎪ ρ CP F R ⎪ L ∂x ⎨ ⎪ ⎪ A h ∂TJ ( x) = − SL SL (TJ ( x) − TR ( x) ) ⎪ ρ CP F L ∂x ⎪ ⎪ ⎩ (213) This system of equations differs from system F16 in several ways. The concept of a global heat transfer coefficient is not applicable. heat gains in the jet region. close to the outlet can be ignored in a first order accuracy model. In the model implementation in EnergyPlus the user will not define the exact location of the heat gains within each region of the flow therefore. Referring to the previously used subdivision of the room into jet and recirculation regions we can see that gains can occur exclusively in the jet region. can be characterized as follows: For gains occurring in the jet region (between the inlet and the room outlet) a change in inflow temperature is an adequate. are inserted in two equal parts in two separate points of the jet flow (in the room entrance and before the exit. In all other situations. the recirculation is partially isolated from the main jet flow and higher temperatures are generated in this region whenever heat gains are present. The temperature increase in the recirculation flow is simply defined by the internal gains (first 4/23/05 289 . conservative approximation. Because there is no wall heat transfer there is no need to split the recirculation flow in two parts (R and W). within the first order accuracy goal. Heat gains in the recirculation region When heat gains occur in the recirculation region significant heat accumulation occurs due to the limited heat transfer ability of the shear layer. the first approach provides a conservative approximation. internal gains often dominate the room temperature field. or whenever the gains occur very close to the outlet (in the second case). see points r=0 and r=LR in Figure 98). exclusively in the recirculation region or in both regions. Clearly the two approaches described above are only exact when: the gains occur at the inlet and perfectly mix with the inflow jet (for the first case). From energy conservation we conclude that the altered inflow temperature is given by: TING = TIN + GJ ρ CP F (212) Gains occurring in the jet region.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Heat transfer in recirculating flows with internal gains Whereas surface heat transfer is an important component of room heat transfer. Heat gains in the jet region The effect of gains in the jet region on room air temperature distribution is simple to model and.

The model should be used only in rectangular rooms with flat surfaces.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING equation in (212)). The recirculation flow temperature at x=L is given by: TR ( L) = TR (0) + ∆TR The temperature variations in the recirculation flow is given by: (215) ∆TR = GR ρ CP F R (216) Combined effects of surface heat transfer and internal gains in the recirculation region In this case. the following system of equations must be solved: ∂TW (r ) AS hS GR ⎧ ⎪ ρ CP F R ∂r = − L (TS − TW (r ) ) − L R ⎪ ⎪ ⎪ ⎪ ρ CP F R ∂TR ( x) = ASL hSL (TJ ( x) − TR ( x) ) ⎪ ⎨ L ∂x ⎪ ⎪ ASL hSL ∂TJ ( x) ⎪ ⎪ ρ CP F ∂x = − L (TJ ( x) − TR ( x) ) ⎪ ⎪ ⎩ With the boundary conditions: TW ( L) = TR (0) . The recirculation flow temperature at x=0 is given by: ⎛ ⎞ 1 G (1 + R) ⎜ TR (0) = TIN + ASL hSL (1+ R ) + 1⎟ R ⎜ ⎟ ρ CP F R ⎜ ρ CP F R ⎟ −1 ⎠ ⎝e (214) The temperature in the recirculation is inversely proportional to the recirculation flow ratio (R) and the shear layer area (measured by the product: ASL hSL). in this case the variable cross section means that the flow may 4/23/05 290 . TW (0) = TR ( L ) . Let us consider the example of a triangular room. The second and third equations in (213) model shear layer heat transfer are similar in the two cases ((210) and (213)). As both of these parameters increase this expression becomes similar to (212). TJ (0) = TIN (217) For simplicity the heat gains are considered to be evenly distributed along the recirculation path. The temperature variation in the recirculation region is given by: ρ CP R F LR hS AS r − ASL hSL (1+ R ) ρ CP R F TW (r ) = TS + e GR − AS hS (e e A h (1+ R ) − SL SL ρ CP R F − 1) (GR + AS hS (TS + TIN )) hS AS ρ CP R F (218) + R−e (1 + R ) Limitations of the model A.

4/23/05 291 . The upper limit proposed for this temperature increase is 1.Air must enter the room through one “dominant” aperture. the amount of buoyancy that can be inserted in the flow with out affecting the CV dominated flow pattern is proportional to the inflow momentum flux (and consequently airflow rate). C. In this case the room length perpendicular to the CV direction is not representative of the length of development for the shear layer driving the flow (see expression (199)). whenever the heat gains are very high the flow structure can change and the modeling approach/assumption of a flow dominated by horizontal movement can become inadequate.The temperature variation between inlet and outlet must be smaller than 1. the model will be more precise for rooms where the outlet is in the opposing surface of the inlet.5 °C. the effects of variations in outlet geometry are neglected. Figure 99 shows a schematic depiction of the flow pattern that forms the base of the model. An aperture is considered dominant whenever it introduces in the room more than 90% of the air inflow. Clearly. Heat gains can affect the CV airflow pattern and. In addition the average inflow velocity through that aperture must be higher than 0. Still.5°C.The room and inlet dimensions must be such that the value of CL in the right hand side of Equation (198) must be bigger than 1/3 and smaller than 11 and criterion (207) is met. As buoyancy effects become stronger the flow starts to resemble case shown in Figure 100. see below). The recirculation flow follows an horizontal path exchanging heat and momentum with the main jet flow. Although the model presented above could be extended to more than one inflow aperture a single dominant inflow aperture restriction must be imposed for now.2m/s. One straightforward way to scale the influence of buoyancy is to check the airflow temperature increase as the air flows across the room. whenever this criterion is exceeded it is more appropriate to model the flow as mixed. so as to ensure a clear CV airflow (still additional criteria must be met.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING begin with no recirculations and then become attached (or the contrary). In light of the first order precision of the model variations of ±10% in the room cross section are acceptable. D. B. In this case the recirculation not only exchanges with the main jet flow but is also able to reject heat directly by composing part of the outflow. As discussed above.

In addition a set of flow pattern related parameters are also predicted such as type of flow pattern (with or without recirculations).Room airflow dominated by jet Figure 100. and. check criteria (207). 5) If recirculations are present and all CV flow criteria are met. width and inflow velocity. Schematic View . 4) Predict relevant flow characteristics using expressions (200). when buoyancy effects are important a more complex flow pattern develops.Room airflow -. (201). 3) Check whether recirculations are present in the flow. 4/23/05 292 . Implementation The model predicts representative air temperatures and airflow velocities in two regions of the flow: jet and recirculations. Schematic . evaluate parameter CL. room length along the cross ventilation direction. Structure of the implementation The calculation routine follows the following steps sequentially in every time step: 1) Identify the dominant aperture and obtain its area. The recirculation flow moves vertically along its path and the outflow is a combination of jet and recirculation flow.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING Figure 99. In Figure 100. 6) Calculate jet region temperatures. calculate recirculation flow temperature and surface mixed convection heat transfer coefficients. 2) Calculate room length and cross sectional area.a combination of jet and recirculation flow In Figure 99. (202) and (206). inflow aperture area. the recirculation flow develops horizontally and the room airflow is dominated by the jet. for flow with recirculations the ratio between recirculation and inflow rate.

Correlation function. Nomenclature for Cross-Ventilation RoomAir Model Variable A*: AIN: AR: AS: ASL: CL: Cn: Cp: f: F: FR: H: h F: hFS: h N: hS: hSL: L R: L*: Description room non dimensional cross-section area ratio. The outputs CV Jet Region Average Velocity[m/s] and CV Recirculation Region Average Velocity[m/s] are predicted using expressions (200) and (201). Ventilation airflow rate (m3/s). Room length along the cross ventilation direction (m). chapter 1). Inlet area (m2). Free stream heat transfer coefficient (W/(m2K)). room non-dimensional length ratio. flow rate in the recirculation region (m3/s). correlation constant for correlation index “n”.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING 7) If the air temperature increase is above 1. Shear layer heat transfer coefficient (W/(m2K)). Heat capacity of the air at constant pressure (J/(Kg.5°C or one of the CV flow criteria failed. Outputs Representative airflow temperatures for the jet and recirculation regions are obtained by taking the average between predicted temperatures at the beginning and end of each stream: CV Jet Region Temperature[C] = TJ (0) + TJ ( L) 2 TR (0) + TR ( L) 2 (219) CV Recirculation Region Temperature[C] = (220) In both cases TJ and TR are the solutions of the system of differential equations shown in Equation(s) (217). Local surface heat transfer coefficient (W/(m2K)).K)).18. Nomenclature for Cross-Ventilation RoomAir Model Table 38. Surface averaged natural convection heat transfer coefficient. Room height (m). Shear layer heat transfer area (m2) Non-dimensional room aspect ratio (expression 1. perform a mixed flow calculation. Room cross section area (m2) Total room surface area (m2). 4/23/05 293 . Surface averaged forced convection heat transfer coefficient.

Vickery. 446-453. Simplified models for heat transfer in rooms. pp. Franklin Book Company 1972.. Temperature variation in the jet region (K). Perfectly mixed room air temperature (°C). W. 2003. D. University of California. N. Baturin. Aynsley. Effect of interior heat transfer coefficients on thermal dynamics and energy consumption. V. averaged in the YZ plane (°C). Temperature in the section of the recirculation that flows along the room walls. averaged velocity in a given region of the room that is being modeled (m/s). Energy and Buildings 22 (2) (1995) pp. G. Air density (Kg/m3). Transactions of ASME. 174-179. width of the inlet aperture (m). Thesis.. S. Vol.. G. Average inlet airflow temperature (°C). 1987. 101-113. 4/23/05 294 . Room width (m). averaged in the YZ plane (°C). Melbourne.ALTERNATIVE MODELING PROCESSES COMPONENT SIZING M: QY=0: R: T(x): TIN: TJ(x): TOUT: TR: TR(x): TW(r): UJ: U R: U ∞: W: WIN: x: xOutlet: Momentum flux (N or J/m).M. averaged in the YZ plane (°C).A. average velocity in the room volume occupied by inlet jet flow (m/s). Temperature variation in the recirculation region (K). Kalema. V. Johnson. experimentally determined constant that scales mixing in the shear layer. Room air temperature. Ph. 109. L. T.. Neiswanger. Heat transfer in the shear layer mid plane (W). Architectural Aerodynamics. averaged in the YZ plane (°C).. Average temperature of the air leaving the room (K). B. Applied Science London (1977).. An experimental study of high Raleigh number mixed convection in a rectangular enclosure with restricted inlet and outlet openings. Haapala. San Diego. Billington. Temperature in the section of the recirculation flow that is parallel to the jet region.J. Non-dimensional recirculation flow ratio.P. ∆TJ: ∆TR: ∆TSL: ρ: σ: References Carrilho da Graca. . T.. Location of the outlet along the room length (m). Fundamentals of Industrial Ventilation. R. Characteristic velocity of the forced flow that “drives” the forced convection boundary layer. Temperature in the jet region.. V. Carey. Temperature variation in the shear layer region (K).. Coordinate along the length of the boundary layer.

.. 97. ASHRAE Transactions... Pt. D. Irie. Unfortunately. The impact of thermal capacity will be addressed in future upgrades to this model.. Energy and Buildings. Model Description The present ADS model is applied to a single heating and cooling system that uses a single air distribution system. 505-515.B. SAND 83-8225. and Walton and Dols (2003). 1983. A detailed description of the airflow network model may be found in the work of Walton (1989). R. A. Mixed convection from heated room surfaces. Gadgil. pp. it is expensive to develop an in-depth understanding of these complicated interactions through field research. (1999) 233-244. Plans are to expand the model in the future to include more complex systems (e. A. R. Siebers. Dols and Walton (2002). Kerestecioglu. Awbi.E. and ventilation flow rate using wind tunnel experiments.B. C.A. H. Therefore. it is a simplified airflow model.. Interior Convective Heat Transfer in Buildings with Large Ventilative Flow Rates.. D. in press.. M.. Bejan.. ASHRAE Transactions.C. No... A. part 1b (1984) 211-224. (2000) 153-166. 30. Chandra. Livermore CA. Study on airflow characteristics inside and outside a CVmodel. L. Volume 90. ASHRAE Transactions. A. J. due to interactions between air distribution systems and buildings. Kammerud.. Moffat. through air leakage and heat conduction. A. Detailed air distribution system (ADS) models can be used to predict the detailed behavior of forced-air distribution systems coupled with buildings. E. Convection Heat Transfer 2nd ed. 1994. H. The present ADS model consists of three sections: Pressure and airflow calculations Node temperature calculations Node humidity ratio calculations Pressure and airflow calculations The EnergyPlus airflow network basically consists of a set of nodes linked by airflow elements. & Hatton. USA. D. J.J. Wiley.1(1991). Natural convection from heated room surfaces.F. R. S. 20% to 40% of the energy used to condition building spaces. In addition. The node variable is pressure and the element variable is airflow rate. 4/23/05 295 . Sandia National Laboratories. A brief description is presented below.g. The model excludes the impact of thermal capacities of the air and duct systems. such as a computational fluid dynamics (CFD) model.. Pedersen.. 2001. variable air volume systems in large commercial buildings). T. Fisher. Spitler.. Vol. Bauman. Awbi. Experimental Mixed Convection Heat Transfer From a Large Vertical Surface in a Horizontal Flow.. Kurabuchi. air leakage often causes health and safety concerns through the creation of uncontrolled airflows and differential pressures in buildings.ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL Ohba. G. Air Distribution System (ADS) Model Overview Forced-air distribution systems often waste. compared to detailed models. Energy and Buildings. 2764. The present ADS model in EnergyPlus is one of the detailed models and is used to simulate thermal conduction and air leakage losses for air distribution systems in residential or light commercial buildings. Schwind. Heat transfer in naturally ventilated rooms data from fullscale measurements. & Hatton. Correlations for convective heat transfer from room surfaces.O. K.. F. 32.. Journal of Wind Engineering and Industrial Aerodynamics.S. and affects the performance of air conditioners and efficiencies of heating and cooling equipment. Altmayer.

248 − 0. z2 = Entry and exit elevations [m] By rearranging terms and adding wind pressure impacts.703sin(α / 2) − 1.n ⎡1. The normalized surface pressure coefficient may be written as (Swami and Chandra. and is linked by an element type which determines the relationship between airflow and pressure. Wind pressure calculations The wind pressure is determined by Bernoulli’s equation. P2 = Entry and exit static pressures [Pa] V1.81 m/s2] z1.131sin 3 (2α G ) ⎤ = ln ⎢ ⎥ 2 2 2 ⎣ +0. inlet and outlet.175sin 2 (α ) + 0.717 cos (α / 2) ⎦ (224) 4/23/05 296 . 1988): C p .07G sin (α / 2) + 0.ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL Element models An element used in the ADS model has two nodes. V2 = Entry and exit airflow velocities [m/s] ρ = Air density [kg/m3] g = Acceleration of gravity [9. assuming no height change or pressure losses: V2 pw = C p ρ 2 where (223) pw ρ V Cp = Wind surface pressure relative to static pressure in undisturbed flow [Pa] = Air density [kg/m3] = Wind speed [m/s] = Wind surface pressure coefficient [dimensionless] Cp is a function of location on the building envelope and wind direction. the above equation may be rewritten in the format used by the airflow network model: ∆P = Pn − Pm + PS + P W (222) where Pn.769 cos(α / 2) + 0. The airflow within each element is assumed to be governed by Bernoulli’s equation: ⎛ ρV 2 ⎞ ⎛ ρV 2 ⎞ ∆P = ⎜ P + 1 ⎟ − ⎜ P2 + 2 ⎟ + ρ g ( z1 − z2 ) 1 2 ⎠ ⎝ 2 ⎠ ⎝ (221) where ∆P = Total pressure difference between points 1 and 2 [Pa] P1. Pm = Total pressures at nodes n and m [Pa] PS = Pressure difference due to density and height differences [Pa] PW = Pressure difference due to wind [Pa] Section “Air Distribution System Module” in the Input Output Reference provides the relationship between airflow and pressure for each given element type.

(1992). n * C n * ρ V2 2 (225) In order to reduce user input requirements. A detailed description can be found in the work of Swami et al. G=1.ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL where Cp. The following equation is used to calculate temperature distribution across a duct element at the given airflow rate and inlet temperature: 4/23/05 297 .m ∂m =∑ i ∂P m • (229) Node Temperature Calculations A brief description of node temperature calculation is given below. Newton’s method is used to iteratively solve for the pressure at each node. {P}*. A new estimated vector of all node pressures. is computed from the current estimated vector of node pressures. and [J] is the square Jacobian matrix whose elements are given by: J n. the user is not required to input Cp values for different angles. a system of equations with all the elements can be assembled together in an n x n square matrix.0 and calculates the wind pressure automatically. {C}.n α G = Normalized Cp [dimensionless] = Angle between wind direction and outward normal of wall under consideration [degrees] = Natural log of ratio of width of wall under consideration to width of adjacent wall The wind surface pressure can be calculated by combining the above two equations together: pw = C p . where n is the number of nodes. In this way. by: {P} = {P} − {C} * (226) where the correction vector.6. is computed by the matrix relationship: [ J ]{C} = {B} {B} is a column vector with each element given by: • (227) Bn = ∑ mi i (228) where n is the node number and i indicates all flow paths connecting node n to other nodes. the ADS module assumes Cn=0. Solution method Based on the relationship between airflow rate and pressure drop for each element. {P}.

K] = Airflow rate [kg/s] = Perimeter of a duct element [m] = Temperature as a field variable [oC] = Ambient temperature surrounding the duct element [oC] = Overall heat transfer coefficient [W/m2.K] (231) hi ho tj kj The outlet temperature at x=L is ⎛ UA To = T∞ + (Ti − T∞ ) *exp ⎜ − • ⎜ ⎜ mC p ⎝ where Ti To A ⎞ ⎟ ⎟ ⎟ ⎠ (232) = Inlet air temperature [oC] = Outlet air temperature [oC] = Surface area = PL [m2] The heat transfer by convection to ambient. is: ⎡ ⎛ ⎢1 − exp ⎜ − UA Q = m C p (T∞ − Ti ) • ⎢ ⎜ ⎜ mC p ⎢ ⎝ ⎣ • ⎞⎤ ⎟⎥ ⎟⎥ ⎟ ⎠⎥ ⎦ (233) The outlet node temperature can be solved using the above equation at the given inlet temperature. Since the inlet temperature at one linkage can be the outlet temperature for the connected linkage.ALTERNATIVE MODELING PROCESSES • AIR DISTRIBUTION SYSTEM (ADS) MODEL m Cp where dT = UP(T ∞ −T ) dx (230) Cp • = Specific heat of duct wall [J/kg.K] = Thickness at j-th layer [m] = Thermal conductivity at j-th layer [W/m. A square linear system assembled by the ADS module is expressed below: 4/23/05 298 . the outlet temperatures at all nodes are solved simultaneously.K] m P T T∞ U U= 1 t 1 1 + +∑ j hi ho kj = Inside heat transfer coefficient [W/m2.K] = Outside heat transfer coefficient [W/m2. Q.

s] = Thickness at j-th layer [m] = Moisture diffusivity at j-th layer [kg/m. A detailed description can found in the work of Swami et al. Node Humidity Ratio Calculations A brief description of node humidity ratio calculation is given below.ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL {M }[T ] = [ B] (234) Where {M} = Airflow matrix [T] = Temperature vector [B] = Given boundary conditions The zone temperatures and system component conditions are used as prescribed conditions.o Dj = Inside moisture transfer coefficient [kg/m2.i hm. (1992).s] 1 t 1 1 + +∑ j hm.s] (236) hm. For example. The following equation is used to calculate humidity ratio distribution across a duct element at the given airflow rate and inlet humidity ratio: • m where • dW = U m P(W ∞ −W ) dx (235) m P W W∞ Um Um = = Airflow rate [kg/s] = Perimeter of a duct element = Humidity ratio [kg/kg] = Ambient humidity ratio [kg/kg] = Overall moisture transfer coefficient [kg/m2.i hm.s] = Outside moisture transfer coefficient [kg/m2. thermal zone temperatures calculated in the module ZoneTempPredictorCorrector are used as given temperatures (prescribed conditions) to calculate the node temperatures in the ADS module.o tj Dj The outlet humidity ratio at x=L is: ⎛ U A⎞ Wo = W∞ + (Wi − W∞ ) *exp ⎜ − m ⎟ • ⎜ ⎟ ⎝ m ⎠ where (237) Wi Wo 4/23/05 = Inlet air humidity ratio [kg/kg] = Outlet air humidity ratio [kg/kg] 299 .

A square linear system assembled by the ADS module is expressed below: {M m }[W ] = [ Bm ] where {Mm} [W] [Bm] = Airflow matrix = Humidity ratio vector = Given boundary conditions (239) The zone humidity ratios and system component conditions are used as prescribed conditions. For example..m (i . ducts) are located.i = Total sensible load in the i-th zone due to ADS losses [W] Qcond(ij) = Duct wall conduction loss at the j-th duct located in the i-th zone [W] Qleak(ij) = Sensible supply leak loss at the j-th linkage located in the i-th zone [W] 4/23/05 300 . The ADS model is activated by the ADS Simulation object in the input data file.i = ∑ Qcond . j ) + ∑ Qmulti . j ) j j j (241) where QADS. the outlet humidity ratio at all nodes are solved simultaneously. Qm.i = ∑ Qcond (i . j ) + ∑ Qleak . and model output. m.f90) is true. Model Activation The ADS model is activated when TurnFansOn (set in ZoneEquipmentManager. the ADS simulation is not performed. Integration of the ADS Model Integration of the ADS model consists of three parts: model activation. load assignment. Load assignment The loads from the ADS model consist of sensible and latent loads for each zone where air distribution system elements (e.g. Since the inlet humidity ratio at one linkage can be the outlet humidity ratio for the connected linkage. The sensible and latent loads are calculated as follows: QADS .m ( i . When TurnFansOn is false. j ) j j j (240) QADS . j ) + ∑ Qmulti ( i . is ⎡ • ⎛ U A ⎞⎤ Qm = m(W∞ − Wi ) ⎢1 − exp ⎜ − m ⎟ ⎥ • ⎜ ⎟ ⎢ ⎝ m ⎠⎥ ⎣ ⎦ (238) The outlet node humidity ratio can be solved using the above equation at the given inlet humidity ratio. thermal zone humidity ratios calculated in the module ZoneTempPredictorCorrector are used as given humidity ratios (prescribed conditions) to calculate the node humidity ratios in the ADS module. j ) + ∑ Qleak (i .m ( i .ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL A = Surface area = PL [m2] The moisture transfer by convection to ambient.

the output values for the Infiltration and Ventilation objects are equal to zero. the airflow model is not simulated for the zones defined as “Thermal Zone” in the ADS Node Data object(s).loads + minf W∞ + m • • sysin Wsys + ∑ Ai hmi ρ airWsurfsi + QADS .m is the added total latent load in the i=th zone due to ADS losses from the ADS module. The airflow model can be either the simplified infiltration model and user-specified interzone airflows (SIMPLE). Group – Simulation Parameters. The revised zone temperature update equation (11) used in the subroutine CorrectZoneAirTemp in the module ZoneTempPredictorCorrector described in the section Integrated Solution Manager becomes: Tzt = ∑ Qi + i =1 N sl • N surfaces ∑ i =1 zones • • • ⎛C hi ATsi + ∑ m C pTzi + minf C pT∞ + m sys C pTsup ply + QADS . ventilation and zone-to-zone air flows (MIXING and CROSS MIXING).i = Total latent load in the i-th zone due to ADS losses [kg/s] Qcond. and the latent load in the ith zone is added to the non-mechanical ventilation and cross ventilation latent load variable (B) in subroutine CorrectZoneHumRat in the same module. The revised zone humidity ratio update equation used in the subroutine CorrectZoneHumRat in the module ZoneTempPredictorCorrector described in the section Moisture PredictorCorrector in the Integrated Solution Manager becomes: B = ∑ kg masssched .ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL Qmulti(ij) = Sensible airflow losses caused by unbalance forced airflow from the other connected zones and outdoors at the j-th linkage located in the i-th zone [W] QADS.z is the added total sensible load in the i-th zone due to ADS losses from the ADS module.m(ij) = Latent supply leak loss at the j-th linkage located in the i-th zone [kg/s] Qmulti.m(ij) = Duct wall vapor diffusion loss at the j-th duct located in the i-th zone [kg/s] Qleak. In addition. Using Airflow Models with the ADS Model The airflow models (Ref.m. z − ⎜ z i ⎝ δt i =1 N 3 t − 2δ t 1 t −3δ t ⎞ ⎞⎛ t −δ t − Tz ⎟ ⎜ −3Tz + Tz ⎟ 2 3 ⎠ ⎠⎝ ⎛ 6 ⎞ Cz ⎜ ⎟ + ⎝ 11 ⎠ δ t N surfaces ∑ i =1 hi Ai + ∑ m C p + minf C p + msys C p i =1 N zones • • • where QADS. The program assumes that when the ADS is modeled during a period of HVAC system operation. Airflow Model in the IO Reference) and the ADS model can be used in the same input data file. the ADS model is not simulated and the air flow model (SIMPLE or COMIS) calculates the infiltration.m(ij) = Latent airflow losses caused by unbalance forced airflow from the other connected zones and outdoors at the j-th linkage located in the i-th zone [kg/s] The change in the total sensible load due to the air distribution systems in the i-th zone is added to the temperature independent coefficient (TempIndCoef) in subroutine CorrectZoneAirTemp in the module ZoneTempPredictorCorrector. except for the ventilation inlet air temperature which is set to the outdoor dry-bulb temperature. when the COMIS model is used air flows through COMIS Link are also equal to zero for all “thermal zones” specified in the ADS input. When the HVAC system fans are not operating. or the COMIS multi-zone airflow model. 4/23/05 301 . Therefore.m i =1 surfs (242) where QADS.

1 Supplemental User Guide and Program Documentation. N. Vasanth. “Integration of Radon and Energy Models for Building. V. Dols. Gu. Walton. N. W. References Dols. Chandra. National Institute of Standards and Technology. Florida Solar Energy Center. ASHRAE Transactions 81(1):202 Swami. National Institute of Standards and Technology. 1989. “AIRNET – A Computer Program for Building Airflow Network Modeling. 1992. Maryland Swami. 2003.” FSEC-CR-553-92. & W. Maryland 4/23/05 302 .” NISTIR 6921. “CONTAMW 2. Maryland Walton.” NISTIR 7049. Gathersburg. S. Florida Walton.ALTERNATIVE MODELING PROCESSES AIR DISTRIBUTION SYSTEM (ADS) MODEL Model Output The available outputs from the ADS model are described in the EnergyPlus Input Output Reference manual. N. 2002. S. V. and S. National Institute of Standards and Technology. Correlation for pressure distribution on buildings and calculation of natural-ventilation airflow.” NISTIR 89-4072. Cocoa. G. G. M. M.. L. Gathersburg. 1988. Gathersburg. “CONTAM 2. & G.0 User Manual. & V.

ductwork. This equation is typically coupled with Fourier’s law of conduction that relates the heat flux at any position and time to temperature as follows: q′′ ( x. Energy is thus either added to or removed from the space. (243) α= k is the thermal diffusivity of the layer material. Despite the relative simplicity of the low temperature radiant systems. one must integrate this formulation within an energy analysis program like EnergyPlus. t ) = −k ∂T ( x. ceiling. and roofs. and zone occupants are conditioned by both radiation exchange with the system and convection from the surrounding air that is also affected by the system. One Dimensional Heat Transfer Through Multilayered Slabs One of the most important forms of heat transfer in energy analysis is heat conduction through building elements such as walls. fans. Finally. the conduction transfer function method for modeling transient conduction must be extended to include embedded heat sources or sinks. The system circulates hot or cold fluid through tubes embedded in a wall. Second. for systems with significant thermal mass. ρ is its density. to be relatively simple systems. dampers. are not needed. the integration of such a system within an energy analysis program requires one to overcome several challenges. Transient one dimensional heat conduction through a homogeneous layer with constant thermal properties such as the one shown in Figure 101 is governed by the following equation: ∂ 2T 1 ∂T = ∂x 2 α ∂t where: T is the temperature as a function of position and time. t is the time. floors. and cp is its specific heat. a method that applies to all structures must account for the presence of thermal mass within the building elements. etc. or floor or runs current through electric resistance wires embedded in a surface or a panel. x is the position. on the surface. one must overcome the fact that the radiant system is both a zone heat balance element and a conditioning system. While some thermally lightweight structures can be approximated by steady state heat conduction. ρcp k is its thermal conductivity. Each of these issues will be addressed in the next several subsections. First. t ) ∂x (244) 4/23/05 303 . Unless specifically required for indoor air quality considerations.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Radiant System Models Low Temperature Radiant System Model Low temperature radiant heating and cooling systems appear.

t) Tn (x n-1.t) n+1 . each of these methods have drawbacks which render them inappropriate for use within an energy analysis program which requires both accuracy and efficiency from the simulation. As mentioned in the previous section.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL T1(0. Multilayered Building Element Time Series Solutions: Conduction Transfer Functions Equations (243) and (244) can be solved numerically in a variety of ways.. x .t) q" (x 3.t) q"(0.t) q" (x1. 4/23/05 304 . the solution becomes extremely tedious for the multiple layered slab shown in Figure 102.t) 1 T2 (x1.t) q" (x1.t) 4 T5 (x 4. T1(0.t) q" (x n-1...t) q"(0.t) 5 n Tn+1(x n.. x=0 x = x1 x = x2 x = x3 x = x4 x = xn-1 x = xn Figure 102.t) q" (x 2.t) 1 T2(x1.t) q" (x n.t) q" (x 4.t) 2 T3 (x 2. However.t) 3 T4 (x 3. other models have used control theory and numerical methods such as finite difference and finite element.t) 2 x x=0 x = x1 Figure 101. Single Layered Building Element While analytical solutions exist for the single homogeneous layer shown in Figure 101.

As the next several sections will detail.t − m +1 − ∑ YmTo . s ) dx (247) (248) 4/23/05 305 .t − m +1 . The basic form of a conduction transfer function solution is shown by the following equation: qi′′t = ∑ X mTi . Y. the infinite number of terms needed for an exact response factor solution makes it less than desirable. d 2T ( x. s ) = −k dT ( x. A similar equation holds for the heat flux at the exterior surface. there are two main methods for calculating conduction transfer functions: the Laplace Transform method and the State Space method. the conduction heat transfer through an element can be calculated. s ) dx 2 α q′′ ( x. Fortunately. The most basic time series solution is the response factor equation which relates the flux at one surface of an element to an infinite series of temperature histories at both sides as shown by: qi′′t = ∑ X mTi . Beginning with the transient one dimensional heat conduction equation {Equation (243)} and Fourier's law of conduction {Equation (244)}. With a single. Both methods are well suited for the main focus of this research. i signifies the inside of the building element. and F are the conduction transfer functions. Laplace Transform Formulation The traditional method for calculating conduction transfer functions is described in detail by Hittle (1981). The coefficients (CTFs) in the equation are constants that only need to be determined once. . m =1 m =1 ∞ ∞ (245) where q” is heat flux. The new solution contains elements that are called conduction transfer functions (CTFs).t − m +1 − ∑ YmTo . s ) s = T ( x. the Laplace transform method is used to convert the governing equations into the s-domain for a single layer such as the one shown in Figure 101. T is temperature. M is a finite number defined by the order of the conduction transfer functions.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Another possible modeling method is a time series solution. This equation states that the heat flux at the interior surface of any generic building element for which the assumption of one dimensional conduction heat transfer is valid is linearly related to the current and some of the previous temperatures at both the interior and exterior surface as well as some of the previous flux values at the interior surface. m =1 m =1 m =1 M M k (246) where k is the order of the conduction transfer functions. The final CTF solution form reveals why it is so elegant and powerful. The only storage of data required is the CTFs themselves and a limited number of temperature and flux terms. the similarity of higher order terms can be used to replace them with flux history terms. the extension of conduction transfer functions to include heat sources or sinks. relatively simple equation. The formulation is valid for any surface type and does not require the calculation or storage of element interior temperatures. and t represents the current time step.t − m +1 + ∑ Fm qi′′t − m . o signifies the outside of the building element. Several of the detailed energy analysis programs such as EnergyPlus use a time series solution to transient heat conduction. While in most cases the terms in the series decay fairly rapidly. and X.

) B1 ( s ) = 1 k1 s α1 sinh C1 ( s ) = k1 s α1 sinh D1 ( s ) = cosh ( ) ( 1 s α1 . α1 is the thermal diffusivity of the layer. A1 ( s ) = cosh ( 1 s α1 . The form of the matrix equation for the multilayered element is the same as the equation for a single layer: ⎡T1 ( s ) ⎤ ⎡ A ( s ) B ( s ) ⎤ ⎡Tn +1 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ ⎣ q1 ( s ) ⎦ ⎣C ( s ) D ( s ) ⎦ ⎣ qn +1 ( s ) ⎦ but the transmission matrix is replaced by: (252) 4/23/05 306 . q1(s). The 2 x 2 matrix consisting of A1(s). It can easily be shown that a second layer could be characterized in a similar way as: ⎡T2 ( s ) ⎤ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎡T3 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ ⎣ q2 ( s ) ⎦ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ q3 ( s ) ⎦ (250) where A2(s). for a multilayered element as shown in Figure 102. and 1 is the thickness of the layer. ) ( 1 s α1 .RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL The transformed equations are solved and then put in matrix form as shown below: ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡T2 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎣ q2 ( s ) ⎦ (249) where: T1(s). Ci(s). ) ( 1 s α1 . This can be substituted into Equation (249) to provide insight how the extension to multilayered slabs is achieved. ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎡T3 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥⎢ ⎥ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ q3 ( s ) ⎦ (251) Thus. Bi(s). and Di(s) associated with it. and q2(s) are the temperature and flux terms in the Laplace domain. each separate layer has a transmission matrix of Ai(s). and D2(s) are calculated using the properties of the second layer. C2(s). ) k1 is the thermal conductivity of the layer. B1(s). B2(s). C1(s). T2(s). and D1(s) is called the transmission matrix and contains all of the thermophysical properties of the layer necessary to calculate transient conduction heat transfer through it.

When the temperature histories are formulated as triangular pulses made up of simple ramp functions. B. the state variables are the nodal temperatures. State Space Formulation Recently. Through the use of matrix algebra. the roots of this equation can be found and result in response factors. Ouyang and Haghighat 1991). and the output vector (y) can be related directly to the input vector (u) and time histories of the input and output vectors. The response factors can be simplified as described above through the introduction of flux history terms to form conduction transfer functions. the environmental temperatures (interior and exterior) are the inputs. another method of finding conduction transfer functions starting from a state space representation has begun receiving increased attention (Ceylan and Myers 1980. This formulation can be used to solve the transient heat conduction equation by enforcing a finite difference grid over the various layers in the building element being analyzed. Thus. In this case.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL ⎡ A ( s ) B ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ C ( s ) D ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ Equation (252) is typically rearranged as follows: ⎡ An ( s ) Bn ( s ) ⎤ ⎢ ⎥ (253) Cn ( s ) Dn ( s ) ⎦ ⎣ ⎡ D (s) ⎢ ⎡ q1 ( s ) ⎤ ⎢ B ( s ) ⎢ ⎥=⎢ ⎣ qn +1 ( s ) ⎦ ⎢ 1 ⎢ B (s) ⎣ ⎤ ⎥ ⎥ ⎡ T1 ( s ) ⎤ ⎢ ⎥ − A ( s ) ⎥ ⎣Tn +1 ( s ) ⎦ ⎥ B (s) ⎥ ⎦ −1 B (s) (254) which relates the flux at either surface of the element to the temperature histories at both surfaces. y is the output vector. A simplified method of finding the roots of the Laplace domain equations is described by Hittle and Bishop (1983) and is used by the current version of BLAST. and the resulting heat fluxes at both surfaces are the outputs. and A. C. u is a vector of inputs. the vector of state variables (x) can be eliminated from the system of equations. t is time. the state space representation with finite difference variables would take the following form: ⎡ T1 ⎤ d⎢ ⎥ ⎢ ⎥ ⎡ T1 ⎤ ⎢ ⎥ ⎣Tn ⎦ = A ⎢ ⎥ + B ⎡ Ti ⎤ [ ] ⎢ ⎥ [ ] ⎢T ⎥ dt ⎣ o⎦ ⎢Tn ⎥ ⎣ ⎦ 4/23/05 (257) 307 . Seem 1987. The basic state space system is defined by the following linear matrix equations: d [ x] = [ A][ x ] + [ B ][u ] dt (255) [ y ] = [C ][ x ] + [ D ][u ] (256) where x is a vector of state variables. and D are coefficient matrices.

Tn-1.. C= ρcp A 2 . In matrix format: ⎡ dT1 ⎤ ⎡ −1 hA ⎢ dt ⎥ ⎢ RC − C ⎢ ⎥=⎢ 1 ⎢ dT2 ⎥ ⎢ ⎢ dt ⎥ ⎢ RC ⎣ ⎦ ⎣ 1 ⎤ ⎡ hA RC ⎥ ⎡ T1 ⎤ ⎢ C +⎢ ⎥ −1 hA ⎥ ⎢T2 ⎥ ⎢ ⎣ ⎦ 0 − ⎢ RC C ⎥ ⎦ ⎣ ⎤ 0 ⎥ ⎡To ⎤ ⎥⎢ ⎥ hA ⎥ ⎣ Ti ⎦ C⎥ ⎦ (263) ′′ ⎡ q1 ⎤ ⎡ 0 − h ⎤ ⎡ T1 ⎤ ⎡ 0 h ⎤ ⎡To ⎤ ⎢ q′′ ⎥ = ⎢ h 0 ⎥ ⎢T ⎥ + ⎢ − h 0 ⎥ ⎢ T ⎥ ⎦⎣ 2⎦ ⎣ ⎦⎣ i ⎦ ⎣ 2⎦ ⎣ (264) 4/23/05 308 .. Tn are the finite difference nodal temperatures. T2. .RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL ⎡ T1 ⎤ ⎡ qi′′⎤ ⎡ Ti ⎤ ⎢ ⎥ ⎢ q′′ ⎥ = [C ] ⎢ ⎥ + [ D ] ⎢T ⎥ ⎣ o⎦ ⎣ o⎦ ⎢Tn ⎥ ⎣ ⎦ (258) where T1. and qi′′ and qo are the heat fluxes (desired output).. and A is the area of the surface exposed to the environmental temperatures. n is the number of ′′ nodes. Seem (1987) shows that for a simple one layer slab with two interior nodes as in Figure 103 and convection at both sides the resulting finite difference equations are given by: ′′ qo = A (T1 − To ) C dT1 T −T = hA (To − T1 ) + 2 1 dt R (259) C dT2 T −T = hA (Ti − T2 ) + 1 2 dt R (260) qi′′ = A (Ti − T2 ) (261) (262) where: R= kA . Ti and To are the interior and exterior environmental temperatures.

For a wall composed of insulation between two layers of concrete. Two Node State Space Example The important aspect of the state space technique is that through the use of relatively simple matrix algebra the state space variables (nodal temperatures) can be eliminated to arrive at a matrix equation that gives the outputs (heat fluxes) as a function of the inputs (environmental temperatures) only. The resulting equations involve some fairly complicated terms including spatial derivatives. 4/23/05 309 . they found almost no difference in the response factors calculated by each method. The derivation of the necessary equations is begun by analyzing the simple two layer element shown in Figure 104. The accuracy of the state space method of calculating CTFs has been addressed in the literature. In addition. This eliminates the need to solve for roots in the Laplace domain. Extension of Time Series Solutions to Include Heat Sources and Obtain Internal Temperatures Laplace Transform Formulation Degiovanni (1988) proposed two methodologies for including sources or sinks in the Laplace Transform Formulation. The first method shows how a source that varies as a function of time and location can be incorporated. The current version of EnergyPlus uses the state space method for computing CTFs. Their results showed that for an adequate number of nodes the state space method computed a heat flux at the surface of a simple one layer slab within 1% of the analytical solution. The second method that will be analyzed in more detail involves the addition of a source or sink at the interface between two layers. Ceylan and Myers (1980) compared the response predicted by the state space method to various other solution techniques including an analytical solution. the resulting matrix form has more physical meaning than complex functions required by the Laplace transform method. Ouyang and Haghighat (1991) made a direct comparison between the Laplace and state space methods.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL To 1 hA T1 R C C T2 1 hA Ti Figure 103.

t) q" (x 2.t) T3 (x 2.t) 2 T3 (x 2. Two Layer Example for Deriving the Laplace Transform Extension to Include Sources and Sinks For the first layer.t) 2 2 3 x heat source x x=0 x = x1 (a) x = x2 x=0 x = x1 (b) x = x1+ x = x2 ⎡T2 ( s ) ⎤ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎡T3 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ ⎣ q2 ( s ) ⎦ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ q3 ( s ) ⎦ Figure 104.t) 1 T2 (x1.t) 1 T2 (x1.t) q" (x1.t) q" (x1. the following substitution is made: 0 ⎡T2 ( s ) ⎤ ⎡T2+ ( s ) ⎤ ⎡ ⎤ ⎢ ⎥=⎢ ⎥ + ⎢q ⎥ ⎣ q2 ( s ) ⎦ ⎣ q2+ ( s ) ⎦ ⎣ source ( s ) ⎦ which results in: (267) 0 ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎧ ⎡T2+ ( s ) ⎤ ⎡ ⎤⎫ ⎪ ⎪ ⎢ ⎥=⎢ ⎥ ⎨⎢ ⎥ + ⎢q ⎥⎬ ⎪ ⎪ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎩ ⎣ q2+ ( s ) ⎦ ⎣ source ( s ) ⎦ ⎭ 0 ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎧ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎡T3 ( s ) ⎤ ⎡ ⎤⎫ ⎪ ⎪ ⎬ ⎢ ⎥=⎢ ⎥ ⎨⎢ ⎥⎢ ⎥ + ⎢q s )⎥ ⎭ ⎪ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎩ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ q3 ( s ) ⎦ ⎣ source ( ⎦ ⎪ (268) (269) 4/23/05 310 .t) 3 T1(0.t) q" (0.t) q" (x 2.t) T2 + (x1+.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL T1(0.t) q" (0.t) q" +(x1+. it was determined that in the Laplace domain ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡T2 ( s ) ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎣ q2 ( s ) ⎦ For the second layer: (265) (266) To link the two layers and include the heat source between them.

RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL 0 ⎡T1 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡ A2 ( s ) B2 ( s ) ⎤ ⎡T3 ( s ) ⎤ ⎡ A1 ( s ) B1 ( s ) ⎤ ⎡ ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥⎢ ⎥+⎢ ⎥ ⎢q ⎥ ⎣ q1 ( s ) ⎦ ⎣C1 ( s ) D1 ( s ) ⎦ ⎣C2 ( s ) D2 ( s ) ⎦ ⎣ q3 ( s ) ⎦ ⎣C1 ( s ) D1 ( s )⎦ ⎣ source ( s ) ⎦ (270) While Degiovanni concludes with this formula. This result conforms with the superposition principle which was used to develop the conduction transfer functions from the summation of a series of triangular pulses or ramp sets. If a layer is added to the left of the first layer. The general equation for a building element with n layers and m layers between the left hand surface and the heat source can be derived as: 0 ⎡T1 ( s ) ⎤ ⎛ n ⎡ Ai ( s ) Bi ( s ) ⎤ ⎞ ⎡Tn +1 ( s ) ⎤ ⎛ m ⎡ Ai ( s ) Bi ( s ) ⎤ ⎞ ⎡ ⎤ ⎜ ⎟ ⎜ ⎟ ⎢ ⎥ = ⎜∏ ⎢ ⎥⎟⎢ ⎥ + ⎜∏⎢ ⎥ ⎟ ⎢q ⎥ ⎣ q1 ( s ) ⎦ ⎝ i =1 ⎣Ci ( s ) Di ( s ) ⎦ ⎠ ⎣ qn +1 ( s ) ⎦ ⎝ i =1 ⎣Ci ( s ) Di ( s ) ⎦ ⎠ ⎣ source ( s ) ⎦ (271) or in more compact form: 0 ⎡T1 ( s ) ⎤ ⎡ A ( s ) B ( s ) ⎤ ⎡Tn +1 ( s ) ⎤ ⎡ a ( s ) b ( s ) ⎤ ⎡ ⎤ ⎢ ⎥=⎢ ⎥⎢ ⎥+⎢ ⎥ ⎢q ⎥ ⎣ q1 ( s ) ⎦ ⎣C ( s ) D ( s ) ⎦ ⎣ qn +1 ( s ) ⎦ ⎣ c ( s ) d ( s ) ⎦ ⎣ source ( s ) ⎦ where: (272) ⎡a ( s ) ⎢ ⎣c ( s) ⎡ A ( s ) B ( s ) ⎤ n ⎡ Ai ( s ) Bi ( s ) ⎤ ⎢ ⎥ = ∏⎢ ⎥ ⎣C ( s ) D ( s ) ⎦ i =1 ⎣Ci ( s ) Di ( s ) ⎦ b ( s ) ⎤ m ⎡ Ai ( s ) Bi ( s ) ⎤ ⎥ = ∏⎢ ⎥ . While Equation (273) is correct for any single or multilayered element. Now. Conversely. which relates the heat flux at both sides of the element to the temperature at each side. the first term in the heat source transmission matrix does not appear to match the compactness of the other terms in the matrix equation. the entire right hand side of Equation (270) is multiplied by the transmission matrix of the new layer. The matrix equation that is obtained shows that: ⎡ D (s) ⎢ ⎡ q1 ( s ) ⎤ ⎢ B ( s ) ⎢ ⎥=⎢ ⎣ qn +1 ( s ) ⎦ ⎢ 1 ⎢ B (s) ⎣ D (s)b(s) ⎤ ⎤ ⎡ ⎥ ⎢d ( s ) − ⎥ B (s) ⎥ ⎥ ⎡ T1 ( s ) ⎤ + ⎢ s ⎤ (273) ⎡q ⎢ ⎥ ⎥ ⎣ source ( ) ⎦ − A ( s ) ⎥ ⎣Tn +1 ( s ) ⎦ ⎢ b (s) ⎥ ⎢ ⎥ B (s) ⎥ B (s) ⎢ ⎥ ⎦ ⎣ ⎦ −1 B (s) This equation bears a striking resemblance to Equation (254). the effect of the heat source is simply added to the response to the temperature inputs. If the source term in Equation (273) is dropped. the vector containing the Laplace transform of the temperature and heat flux at interface 3 is replaced by the product of the transmission matrix of the new layer and the vector for temperature and heat flux at the next interface. d ( s ) ⎦ i =1 ⎣Ci ( s ) Di ( s ) ⎦ and Next. Equation (272) must be rearranged to match the form of Equation (254). It can be shown (see Strand 1995: equations 32 through 42 4/23/05 311 . if a layer is added to the right of the second layer in Figure 104. and the term dealing with the heat source is not affected. some insight into what the generic equation for an element that has n layers might look like is gained by working with Equation (270). then the equation is identical to Equation (254).

as the terms in the temperature transmission matrix.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL which detail this derivation) that the heat source transmission term for a two-layer problem reduces to ⎡ D (s) ⎢ ⎡ q1 ( s ) ⎤ ⎢ B ( s ) ⎢ ⎥=⎢ ⎣ q3 ( s ) ⎦ ⎢ 1 ⎢ B (s) ⎣ ⎡ B2 ( s ) ⎤ −1 ⎤ ⎥ ⎢ ⎥ B ( s ) ⎥ ⎡T1 ( s ) ⎤ ⎢ B ( s ) ⎥ ⎡ qsource ( s ) ⎤ ⎢ ⎥+ ⎦ − A ( s ) ⎥ ⎣T3 ( s ) ⎦ ⎢ B1 ( s ) ⎥ ⎣ ⎥ ⎢ ⎥ B (s) ⎥ ⎢ B (s) ⎥ ⎦ ⎣ ⎦ (274) If this is extended to a slab with n layers and a source between the m and m+1 layers. In essence. extremely convenient because the terms in the heat source transmission matrix have the same denominators. The form displayed by Equation (275) is. saving a considerable amount of computer time during the calculation of the transfer functions. It is expected that only the layers to the left of the source will affect q1(s). ⎣C ( s ) D ( s ) ⎦ i =1 ⎣Ci ( s ) Di ( s ) ⎦ ⎡ a ( s ) b ( s ) ⎤ m ⎡ Ai ( s ) Bi ( s ) ⎤ ⎢ ⎥ = ∏⎢ ⎥ .t − m +1 + ∑ Fm qi′′t − m + ∑ Wm qsource . . the combined CTF–QTF solution takes the following form: qi′′t = ∑ X mTi . the general matrix equation for obtaining heat source transfer functions using the Laplace transform method is: ⎡ D (s) ⎢ ⎡ q1 ( s ) ⎤ ⎢ B ( s ) ⎢ ⎥=⎢ ⎣ qn +1 ( s ) ⎦ ⎢ 1 ⎢ B (s) ⎣ where: ⎢ ⎡ b (s) ⎤ ⎤ ⎢ ⎥ ⎥ ⎥ ⎡ T1 ( s ) ⎤ + ⎢ B ( s ) ⎥ ⎡ q ⎢ ⎥ source ( s ) ⎤ ⎦ − A ( s ) ⎥ ⎣Tn +1 ( s ) ⎦ ⎢ b ( s ) ⎥ ⎣ ⎢ ⎥ ⎥ B (s) ⎥ ⎢ B (s) ⎥ ⎦ ⎣ ⎦ −1 B (s) (275) ⎡ A ( s ) B ( s ) ⎤ n ⎡ Ai ( s ) Bi ( s ) ⎤ ⎥ = ∏⎢ ⎥ . ⎣ c ( s ) d ( s ) ⎦ i = m +1 ⎣Ci ( s ) Di ( s ) ⎦ At first glance.t − m +1 − ∑ YmTo . Once Equation (275) is inverted from the Laplace domain back into the time domain. b ( s ) B ( s ) . In fact. and c ( s ) d ( s ) ⎦ i =1 ⎣Ci ( s ) Di ( s ) ⎦ ⎣ n ⎡A s ⎡a ( s ) b ( s ) ⎤ ( ) Bi ( s ) ⎤ =∏⎢ i ⎢ ⎥ ⎥ . however. the appearance of b ( s ) removes the effects of the layers to the right of the source from B(s) leaving only the influence of the layers to the left of the source. the entire term. and thus roots. m =1 m =1 m =1 m =1 M M k M (276) 4/23/05 312 . must be analyzed to determine the effect of qsource(s) on q1(s). the terms in the heat source transmission matrix may appear to be reversed. the same roots that are calculated for the CTFs can be used for the QTFs.t − m+1 . but the presence of b ( s ) in the element multiplied by qsource(s) to obtain q1(s) seems to be contradictory. Thus.

Two Node State Space Example with a Heat Source In obtaining the matrix equivalent for this set of equations. State Space Formulation The two-node example introduced by Seem (1987) can be utilized to examine the extension of the state space method to include heat sources or sinks. The resulting matrix equations take the following form: ⎡ dT1 ⎤ ⎡ −1 hA ⎢ dt ⎥ ⎢ RC − C ⎢ ⎥=⎢ 1 ⎢ dT2 ⎥ ⎢ ⎢ dt ⎥ ⎢ RC ⎣ ⎦ ⎣ 4/23/05 1 ⎤ ⎡ hA ⎥ ⎡ T1 ⎤ ⎢ C RC +⎢ ⎥ −1 hA ⎥ ⎢T2 ⎥ ⎢ ⎣ ⎦ − 0 ⎢ RC C ⎥ ⎦ ⎣ 0 hA C A⎤ T ⎡ o ⎤ C⎥⎢ ⎥ ⎥ ⎢ Ti ⎥ 0 ⎥ ⎢ qsource ⎥ ⎦ ⎥⎣ ⎦ (281) 313 . Thus. it must be grouped with the environmental temperatures as inputs. The nodal equations for the finite difference network shown in Figure 105 are: C dT1 T −T = hA (To − T1 ) + 2 1 + qsource A dt R dT2 T −T = hA (Ti − T2 ) + 1 2 dt R (277) C (278) qi′′ = A (Ti − T2 ) (279) (280) ′′ qo = A (T1 − To ) q source To 1 hA C T1 R C T2 1 hA Ti Figure 105.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL This relation is identical to Equation (246) except for the presence of the QTF series that takes the heat source or sink into account. Figure 105 shows the simple two node network with a heat source added at node 1. it should be noted that the source term is not a constant but rather an input that varies with time.

Determination of Internal Temperatures One aspect of low temperature radiant systems that has not been addressed to this point is the appropriateness of specifying the effect of the system on slab response via a heat source term. The use of a temperature to simulate the presence of a heating or cooling system presents one major obstacle. The If this analysis is extended to a finite difference network with n nodes. the introduction of the source term did not substantially increase the computing effort required to calculate the additional transfer functions. the known quantity is not heat but rather the temperature of the water being sent to the building element. So. the corresponding matrix equations can be shown to be: ⎡ T1 ⎤ d⎢ ⎥ ⎢ ⎥ ⎡ T1 ⎤ ⎡ To ⎤ ⎢Tn ⎥ ⎣ ⎦ = A ⎢ ⎥+ B ⎢ T ⎥ [ ]⎢ ⎥ [ ]⎢ i ⎥ dt ⎢Tn ⎥ ⎢ qsource ⎥ ⎣ ⎦ ⎣ ⎦ ⎡ T1 ⎤ ⎡ To ⎤ ⎡ qi′′⎤ ⎢ ⎥+ D ⎢ T ⎥ ⎢ q′′ ⎥ = [C ] ⎢ ⎥ [ ] ⎢ i ⎥ ⎣ o⎦ ⎢Tn ⎥ ⎢ qsource ⎥ ⎣ ⎦ ⎣ ⎦ (283) (284) The influence of the heat source is also confirmed by the final solution form. In the state space method. 4/23/05 314 . Since the state space method requires the inversion and the exponentiation of the A matrix only. this was shown by the common roots. while Ti has no same would hold for the influence of qsource. the additional QTF terms will not require a substantial amount of additional computing time for their calculation. ′′ direct influence on qo . the principle of superposition is still valid. However. It is interesting to note that as with the Laplace method there is no alteration of the CTFs calculated by the state space method. Thus. there is no readily available temperature value available for use as an input variable. As with the Laplace solution method. The heat produced by such a system can easily be related to the current passing through the heating wire. for a hydronic heating or cooling system. the use of a heat source as the input variable is logical. For a heating system that employs electrical resistance heating. B(s). In the Laplace method. Equation (282) only relates variables that have a direct influence on heat flux. This is not the case. it does have an indirect influence through the nodal network. When fluid is not being circulated. shared by both the CTFs and the QTFs. Furthermore.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL ⎡ To ⎤ ′′ ⎡ q1 ⎤ ⎡ 0 − h ⎤ ⎡ T1 ⎤ ⎡ 0 h 0 ⎤ ⎢ ⎥ ⎢ q′′ ⎥ = ⎢ h 0 ⎥ ⎢T ⎥ + ⎢ − h 0 0 ⎥ ⎢ Ti ⎥ ⎦⎣ 2⎦ ⎣ ⎦⎢ ⎣ 2⎦ ⎣ ⎣ qsource ⎥ ⎦ (282) Equation (282) appears to suggest that the source term has no direct effect on the heat flux at either side of the element because its coefficients are zero. Other similarities between the two solution methods are evident. which is identical to the Laplace transform result shown in Equation (276). the state space method results in a set of QTFs that relate the heat source at the current time step and several previous time steps to the current heat flux at the surface of the element. it can be noted that the A matrices in Equations (263) and (281) are identical.

m = xk .t − m m =1 m =1 m =1 M M k (286) While the current temperature at the interface is not known. This leads to an interesting question: what is the solid temperature? By definition. For a hydronic system. Rearranging Equation (286) provides an equation from which the temperature at the source location may be calculated: ′′ T2. it can be shown that the final solution form in the time domain for the slab with a source at the interface between the two layers is: ′′ ′′ q1.m y1 yk . Typically. and heat source histories.t = ∑ X k . the solid temperature is the temperature of the building element at the depth where the hydronic loop is located.M ) . Most heat exchangers are used to thermally link two fluids. ( m = 1. the only remaining unknown in this equation is the current temperature at surface 2. mT1.t − m+1 m =1 m =1 m =1 m =1 M M k M (285) A similar equation could be written for the response of the first layer in absence of any source term and is given by: ′′ ′′ q1.t − m + ∑ Fm q1.t − m +1 − ∑ Yk .RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL In a hydronic system. The most effective way of relating these two variables is to consider the slab to be a heat exchanger.t = ∑ X k . m +1 y1 ( m = 1. if the inlet fluid temperature. The goal of both methods of calculating CTFs was the elimination of internal temperatures that were not needed for the simulation. Two methods of accomplishing this are described below. the system geometry. there is only one fluid and a stationary solid. an equation could then be formulated to obtain the heat delivered to the slab based on the inlet fluid temperature. However.m = 4/23/05 (289) 315 . presumably the previous values of this parameter will be known.mT2. heat flux. In the case of a hydronic radiant system. mT3.t − m +1 m =1 m =1 m =1 M M −1 k +1 (287) where the new coefficients are obtained from the standard conduction transfer functions for the first layer via the following equations: X k .mT2.t − m +1 + ∑ Fm q1. the surface where the heat source or sink is present. a link between the fluid temperature being sent to the slab and the heat delivered to the slab exist.t − m +1 − ∑ yk . then the outlet temperature and thus the heat transfer to the building element can be computed.t − m +1 − ∑ Yk .mT1. for one dimensional conduction heat transfer. if this value is used in Equation (286). Presumably. In addition. Thus. and the solid temperature are known.t − m + ∑ Wm qsource . mT1. this temperature is not known because it is not needed. . it is necessary to extract this information to solve for the heat source term. Equation (285) does define the current value of the heat flux at surface 1 based on temperature.t − m +1 + ∑ f m q1. M − 1) (288) Yk .t = ∑ xk . Using heat exchanger relationships. Returning to the two layer example shown in Figure 104. the temperatures and the flux histories at surface 1 are also know. The unknowns in Equation (286) are the current heat flux at surface 1 and the temperature at surface 2.

. Consider again the state space example shown in Figure 105. k + 1) (291) This system for backing out an internal temperature through the use of a second. For an element with n layers. The dual role of T1 as a state variable and an output parameter may seem to contradict the goal of the state space method of eliminating the state variables. The validity of these internal temperature 4/23/05 316 . the resulting output matrix equation for the heat flux at both surfaces and the internal temperature is: ⎡ qi′′⎤ ⎡ 0 − h ⎤ ⎡ 0 h 0 ⎤ ⎡ Ti ⎤ ⎢ q′′ ⎥ = ⎢ h 0 ⎥ ⎡ T1 ⎤ + ⎢ − h 0 0 ⎥ ⎢ T ⎥ ⎢ o⎥ ⎢ ⎥ ⎢T ⎥ ⎢ ⎥⎢ o ⎥ ⎣ 2 ⎦ ⎢ 0 0 0⎥ ⎢ q ⎢ T1 ⎥ ⎢ 1 0 ⎥ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ source ⎥ ⎦ (293) The only difference between this relation and Equation (282) is the presence of T1 on both the right and left hand side of the equation. The state space method. mTo . Equation (293) becomes: ⎡ qi′′⎤ ⎡ T1 ⎤ ⎡ Ti ⎤ ⎢ q′′ ⎥ = C ⎢ ⎥ + D ⎢ T ⎥ ⎢ o ⎥ [ ]⎢ ⎥ [ ]⎢ o ⎥ ⎢ Ts ⎥ ⎢Tn ⎥ ⎢ qsource ⎥ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ (294) where Ts is the temperature of the node where the heat source or sink is present. flux. due to the flexibility of the formulation. The transfer function equation for the calculation of Ts that results from Equation (294) is identical in form to Equation (276): Ts . mTi .t = ∑ xk . nodal temperatures can be extracted in the same manner that any other output quantity would be obtained. the temperature at the source location is related to source and temperature histories including histories of Ts. Two output variables were ′′ defined for this example: qi′′ and qo . rearranged CTF equation is valid regardless of whether the Laplace transform or state space method is utilized to calculate the CTFs and QTFs.t − m +1 + ∑ f mTs .t − m + ∑ wm qsource .t − m+1 m =1 m =1 m =1 m =1 M M k M (295) Instead of the flux at either side of the element characterized as a function of temperature. offers a more direct method of obtaining an internal temperature through its definition as an additional output variable.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL F1 = Fm = 1 y1 (290) f m −1 y1 ( m = 2.t − m +1 − ∑ yk . The temperature of the node where the source is present can also be defined as an output variable through the identity equation: T1 = T1 (292) When this equation for T1 is added to Equation (282). However. however. and source history terms.

This offers greater flexibility for defining how the radiant system operates though it may not model every situation. the user does have the ability to specify on an hourly basis through a schedule the temperature of the water that would be supplied to the radiant system. the radiant system model is fully integrated into the heat balance. The model includes a local loop that attempts to meet this demand temperature through mixing and recirculation. The program user is allowed to define a setpoint temperature as well as a throttling range through which the system varies the flow rate of water (or current) to the system from zero to the user defined maximum flow rate. Figure 107 shows how the “desired” inlet water temperature is controlled based on user schedules. and any improvements that are made in areas such as convection coefficients. there is no explicit control on the inlet water temperature or mixing to achieve some inlet water temperature in a hydronic system. the mean radiant temperature. The controlling temperature can be the mean air temperature. More details of the controls are described below. are immediately available to the radiant system as part of the overall heat balance solution. but in this case the user selects both the control temperatures and the water temperatures via schedules. or the operative temperature of the zone. The user has the ability to specify the high and low water and control temperature schedules for heating and cooling (separately. Controls are problematic for almost any simulation program. In a system that uses electric resistance heating. Setpoint temperatures can be varied on an hourly basis throughout the year if desired. However. Rather. Graphical descriptions of the controls for the low temperature radiant system model in EnergyPlus are shown in Figure 106 for a hydronic system.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL calculation methods as well as heat source transfer functions in general will be discussed in the next chapter. The problem is not whether something can be simulated because typically a simulation program offers the ability to experiment with many different control strategies. the problem is typically the diversity of controls that are implemented and keeping the controls that can be simulated up to date. one must then turn to the next difficult issue: controls. the controls are also considered piecewise linear functions. The flow rate is varied linearly with the flow reaching 50% of the maximum when the controlling temperature reaches the setpoint temperature. shading models. a total of eight temperature schedules). In the constant flow-variable temperature systems. Note that this inlet temperature is a “desired” inlet temperature in that there is no guarantee that the system will provide water to the system at that temperature. 4/23/05 317 . Once the transient nature of the system is accounted for. The control strategies are different enough that they were developed as separate system types. Low Temperature Radiant System Controls The use of this equation allows the low temperature radiant system to be handled like any other surface within the heat balance framework. the power or heat addition to the system varies in a manner similar to mass flow rate variation shown in Figure 106. (Operative temperature for radiant system controls is the average of MAT and MRT. Heat balances at the inside and outside surfaces take on the same form as other surfaces. and this choice is also left to the user’s discretion. EnergyPlus offers two different control schemes: variable flow (Low Temp Radiant System:Hydronic) and variable temperature (Low Temp Radiant System:Constant Flow). The controls for variable flow low temperature radiant systems within EnergyPlus are fairly simple though there is some flexibility through the use of schedules. and the participation of the radiant system in the radiation balance within the space and thermal comfort models is automatically included. etc.) Since flow rate is varied. Thus.

etc. MRT.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Mass Flow Rate Heating Setpoint Temperature Controlling Temperature Throttling Range Mass Flow Rate Setpoint Temperature Cooling Controlling Temperature Throttling Range Figure 106.) High Control Temperature Schedule Value Inlet Water Temperature High Water Temperature Schedule Value Low Water Temperature Schedule Value Cooling System Off Control Temperature (MAT. Variable Temperature Low Temperature Radiant System Controls 4/23/05 318 . etc. Variable Flow Low Temperature Radiant System Controls Inlet Water Temperature High Water Temperature Schedule Value Low Water Temperature Schedule Value Low Control Temperature Schedule Value Heating System Off Control Temperature (MAT. MRT.) High Control Temperature Schedule Value Low Control Temperature Schedule Value Figure 107.

This is the best condition and recirculation and bypass amounts are adjusted accordingly based on radiant system outlet temperatures. the space conditioning systems. The system controls determine the desired inlet temperature and system flow rate while loop controls determine the flow rate and temperature of the loop. it was reported as an “unmet load” and does not affect the temperatures experienced within the building. In this second case. and the central plant equipment in three separate steps. First. a pump (constant speed. or group of surfaces/panels). Finally. Variable Temperature Low Temperature Radiant System Component Details One remaining challenge is the merging of the low temperature radiant system model with an integrated building simulation program. Note that pump heat also factors into the model through a simple constant speed pump model and user input. if the loop flow is greater than the radiant flow but the temperature of the loop is not adequate.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL The constant flow (variable temperature) low temperature radiant system model is actually a combination of mixing valves. 4/23/05 319 . then the model simply solves for the actual radiant system inlet temperature and does not try to meet the controls (merely tries to get as close as physically possible given the loop conditions). if the loop has adequate temperature and flow to meet system requests. In the past. IBLAST. and the radiant system (surface. This is connected to the main loop through the standard inlet connections as shown in Figure 108. All of the radiant flow comes from the loop and the loop temperature (after pump heat addition) becomes the radiant system inlet regardless of the temperature controls. if both the temperature and the flow of the loop are inadequate. if the system or plant was undersized. we may or may not be able to meet the desired inlet temperature since recirculation might lower the temperature below the desired temperature. Thus. Third. then there is no amount of mixing that will solve this problem. the model first simulates the radiant system with the desired conditions and then resimulates it to solve for the actual inlet temperature (see later in this section) if it cannot achieve the desired inlet temperature. if the loop temperature is adequate but the loop flow rate is less than the radiant system flow rate. then the model sets the radiant system inlet temperature and controls to the desired values based on the controls and simulates. the large drawback for this arrangement is that there is no feedback from the space conditioning system or central plant response to the building conditions. Loop Inlet Node Loop Bypass Mixing Valve pump Recirculation Rad Surf Rad Surf Rad Surf Loop Outlet Node Figure 108. most simulation programs have simulated the building envelope. Second. but the maximum flow can be modified by a schedule). panel. While this had some advantages and was partly due to a lack of computing capacity. There are four possible conditions (separate for heating and cooling). resolved this issue by integrating all three major components of a building simulation and thus allowing feedback between the equipment and the building envelope. a predecessor (Taylor 1991) to EnergyPlus.

This maintains the conservation of energy within the heat balance simulation over the zone time steps and defines more appropriate temperature and flux histories at each surface that are critical to the success of a conduction transfer function based solution. On the other hand. In EnergyPlus. which can be less than the building time step and can also vary within EnergyPlus. The radiant system has either a direct or an indirect impact on the surfaces within a building. as the system and plant are simulated at multiple shorter time steps. but this could lead to instability within the heat balance due to limits on the precision of the conduction transfer function coefficients. other approaches to solving the mismatch between the system and the zone response of radiant systems are not feasible. but this could result in instabilities in other types of systems that might be present in the simulation. and the radiant system model is simulated at each time step as if the current flow rate was being used throughout the entire zone time step. the radiant system is temporarily shut-off to find how the building would respond if there was no heat source/sink. Then. Yet. Despite the fact that the simulation algorithm described above may either over. the radiant system will rerun the heat balance using the average heat source/sink over all of the system time steps during the past zone time step. When this happens. This means that each time the heat source/sink in the radiant system is varied during the system simulation the zone heat balance must be recomputed to see what the reaction of the rest of the zone is to this change in the conditions of one (or more) of the surfaces.or underpredict system response dependent on how the system has been controlled in previous system time steps. it is reasonable to expect that the effect of these variations will balance out over time even though it might lead to slightly inaccurate results at any particular system time step. After the system has simulated through enough system time steps to equal a zone time step. Flow rate is allowed to vary at each system time step. the system simulation will shorten its time step whenever it senses that conditions are changing too rapidly. the heat balance is always simulated first. it is also a space conditioning system that must act on the space like any other system and thus must also be simulated at the system time step. 4/23/05 320 . it must be simulated with the building envelope. The long-term approach is also in view in the final simulation step at each zone time step. This issue was handled using a multi-step approach. it can present problems for a radiant system. A graphical picture of this somewhat complex multiple step simulation is shown in the figure below. the radiant system is allowed to operate per the controls specified by the user. One could force the system to run at the same time step as the zone. one could try to force the zone to run at the shorter time steps of the system. So. In essence. Yet. this is not physically correct because each change in the flow rate throughout the system simulation will have an impact on the system time steps remaining before the heat balance is simulated during the next zone time step. In reality. While this is effective in maintaining solution stability.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL This integration was not a trivial task and required that the systems be simulated at shorter time steps in some cases to maintain solution stability.

321 . Recalc Heat Bal Sim Radiant System. Resolution of Radiant System Response at Varying Time Steps Heat Exchanger Formulation for Hydronic Systems As has been mentioned previously. it is assumed that the water tubing itself has no appreciable effect on the heat transfer process being modeled. the temperature at the inside surface of the water tubing is assumed to be equal to the temperature at the source location. In either case. It is assumed that the building element that contains the hydronic loop is stationary and that its temperature along the length of the tubing is constant. Recalc Heat Bal Sim Radiant System.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Heat Balance (Zero Radiant System Source/Sink) Simulate Systems and Plants (feedback to zone) Simulate Radiant System/Controls Recalculate Heat Balance System Time Steps System Time Steps Zone Time Step 4/23/05 Sim Radiant System. the actual heat transferred between the building element and the hydronic loop is related to the temperature of the building element at the source location as well as the water system inlet and outlet temperatures. this assumption must be made again in the heat exchanger analysis. Recalc Heat Bal (Note: System time steps can vary but add up to zone time step) Heat Balance (Average Radiant System Source/Sink) Go to next Zone Time Step Figure 109. Recalc Heat Bal Sim Radiant System. the heat balance equations require the heat transferred to the building element from the water loop in order to calculate the heat transferred from the element to the building environment. it is more convenient to use the effectiveness-NTU method in this case. Several assumptions will be incorporated into the heat exchanger analysis. Recalc Heat Bal Sim Radiant System. the source was added at a single node that was characterized by a single temperature. the same characteristic link between the system variables exist. the overall water/slab system can be thought of as a heat exchanger. For consistency. The latter part of this assumption stems from assumptions made in both the one and two dimensional heat source transfer function derivations. In EnergyPlus. Another assumption for the current EnergyPlus model is that the fluid in the tubing is water. In other words. Even though systems defined by this model can vary somewhat. For modeling purposes. While in principle there are two alternative heat exchanger methodologies. Recalc Heat Bal Sim Radiant System. However. Recalc Heat Bal Sim Radiant System. it is assumed that the inlet temperature to the slab (defined by a user schedule and the plant simulation) and the mass flow rate (determined by the control algorithm) are known and that the remaining parameters must be calculated. Finally. it is assumed that the thermal properties of the water do not vary significantly over the length of the tubing and that the water flows at a constant flow rate. Additionally.

by the following equation (Incropera and DeWitt 1985): ε = 1 − e − NTU where NTU is defined by: (299) NTU ≡ UA ( mc p ) (300) water Since the water tubes were assumed to have no effect on the heat transfer process. 4/23/05 322 . is a convection term. First. The convection coefficient can be obtained from internal flow correlations that relate the Nusselt dimensionless number to other flow properties. and L is the total length of the tube. The effectiveness of the heat exchanger. a heat balance on the water loop results in: q = ( mc p ) water (Twi − Two ) (296) where q is the energy transferred between the water loop and the building element. is defined as the ratio of the actual energy transfer to the maximum possible. cp is the specific heat of the water. the number of transfer units. The maximum amount of heat transfer that is possible according to the Second Law of Thermodynamics is: qmax = ( mc p ) water (Twi − Ts ) (297) where qmax is the maximum amount of energy transfer that is possible and Ts is the temperature at the source location. D is the interior tube diameter. or: ε≡ q qmax (298) For a heat exchanger where one fluid is stationary. m is the mass flow rate of the water. ε. UA. the equation for UA is: UA = h (π DL ) (301) where h is the convection coefficient. several equations can be defined which establish the relationship between the heat source and the water temperatures. the effectiveness can be related to NTU. the only term present in the overall heat transfer coefficient. and Two is the outlet water temperature. Thus.66 k (302) where k is the thermal conductivity of the water. Twi is the inlet water temperature.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Using these assumptions and the effectiveness-NTU heat exchanger algorithm. For laminar flow in a tube of constant surface temperature. the Nusselt number is defined by: NuD = hD = 3.

This essentially eliminates ε as an unknown in Equation (297). On the right hand side of Equation (305).t − m +1 − ∑ Yk . assumed to be turbulent for ReD ≥ 2300. outside surface temperature (To). the outside surface temperature is defined as ground temperature and does not require an outside surface heat 4/23/05 323 . the inside surface temperature at the current time step.t − m +1 (305) ⎢ ⎥ ∑ m =1 m =1 m =1 m =1 ⎢ Balance ⎥ ⎣ ⎦ The surface heat balance includes terms for incident solar energy. and source histories). the terms of the heat balance on the left hand side of the equation have been set with the only unknown quantity being Ti. if the surface is a slab on grade floor. radiation heat transfer from internal heat sources such as lights and electrical equipment.mT1. the calculation procedure is somewhat involved and requires. Equation (276) is the standard conduction transfer function formula for a building element with an embedded source/sink of heat. Thus. and convection to the surrounding air. flux. Knowing the inlet water temperature and water mass flow rate.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL For turbulent internal flow. radiation between surfaces using Hottel’s Gray Interchange concept. The controls and the plant define the water mass flow rate and the inlet water temperature. the flow is Knowledge of the flow conditions allows Equations (298) through (304) to be calculated. Since the radiation heat balance is dependent on conditions at the other surfaces.t − m + ∑ Wm qsource. In EnergyPlus. which is the CTF/QTF equation for the temperature at the source location. leaving two equations (Equations (296) and (297)) and three unknowns. An assumption of the heat balance is that the mean temperature of the surrounding air is equal to the final air temperature of the previous time step. The only terms which are not defined are the inside surface temperature (Ti). and (297). the surface flux on the left hand side of the equation is replaced with a surface heat balance: ⎡ Surface ⎤ M M k M ⎢ Heat ⎥ = ′′ X k . (296). This is not the case with the mean air temperature. The presence of the surface temperature in the heat balance does not pose any problems since Equation (305) will be rearranged to solve for this temperature. an iteration loop is required to provide a more accurate estimate of the radiative exchange within the building.mT3.t − m +1 + ∑ Fm q1. and internal heat source/sink (qsource) of the current time step. For example.023Re4 5 Pr1 3 D k (303) where Pr is the Prandtl number of water and ReD is the Reynolds number which is defined by: Re D = 4m πµ D (304) The parameter µ is the absolute viscosity of water. The third equation that can be used in conjunction with Equations (296) and (297) is Equation (295). the use of a modified form of Equation (276). For internal pipe flow. the Colburn equation can be used to define the Nusselt number: NuD = hD = 0. The outside surface temperature will depend on the type of environment to which it is exposed. Using this estimate in the heat balance avoids a second iterative loop around the radiative iteration loop. in addition to Equations (295). most of the terms are already defined since they depend on known values from previous time steps (temperature.

to again avoid iteration. C4 is the heat source transfer function for the current time step.t − m +1 − ∑ yk . In cases where the outside surface temperature is not simply defined such as a surface exposed to the exterior environment.t − m +1 + ∑ f mTs . Equation (305) can be rewritten in a simpler form: Ti . This approximate inside balance uses mean air and radiant temperatures from the previous time step. which results from a known water inlet temperature.t = C1 + C2 qsource. In most cases. the original equation Ts . Substituting Equation (306) into Equation (307) and noting that qsource.t (307) where C3 includes all of the history terms and the effect of the current outside temperature. this is a valid assumption. which was the CTF/QTF equation for the temperature at the source location. Thus. Grouping the temperature and source history terms which are known quantities together with the effect of the outside surface temperature which is defined as described above.t. mTi . the heat balance equation for the outside surface assumes that conditions at the inside surface were the same as the previous time step. can be simplified in a similar manner. The term C2 will depend on the heat source transfer function term and the coefficients of terms linked directly to Ti. mTo . all of the terms in Equation (305) have been defined except the value at the current time step of the inside surface temperature and the heat source/sink. can be shown to be: q= Twi − C3 − C1C5 1 + C4 + C2C5 ε ( mc p ) water (309) 4/23/05 324 .t is the same quantity as q in Equations (296) and (297) results in: Ts = C3 + C4 q + C5 ( C1 + C2 q ) (308) When this equation is combined with Equation (297).t − m +1 m =1 m =1 m =1 m =1 M M k M (295) can be reduced to: Ts = C3 + C4 qsource. Equation (295). and C5 is the conduction transfer function for the inside surface temperature at the current time step.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL balance. the outside surface temperature is redefined to be equal to the inside surface temperature. since the influence of the current inside surface temperature on the outside surface temperature is very small.t − m + ∑ wm qsource .t (306) where the variable C1 includes surface heat balance and past history terms as well as the influence of the current outside temperature.t = ∑ xk . a heat balance similar to Equation (305) is required to define the outside surface temperature. At this point in the simulation algorithm then.t + C5Ti . However. the heat source. In cases were the inside surface temperature has a significant effect. an approximate inside surface heat balance which defines the inside surface temperature is used. If the element is an interior surface which has both surfaces exposed to the same air space.

This. The development shown above follows the heat balance conventions that assume previous values of the inside temperature to calculate the outside temperature. conduction history terms. etc. in reality.) Cb is the current cross CTF term Cc is the QTF inside term for the current heat source/sink Cd is all of the other terms in the outside heat balance (solar.) Ce is the current cross CTF term (should be equal to Cb) Cf is the QTF outside term for the current heat source/sink Cg is the summation of all temperature and source history terms at the source/sink location Ch is the QTF term at the source/sink location for the current heat source/sink Ci is the CTF inside term for the current inside surface temperature Cj is the CTF outside term for the current outside surface temperature Equations (310) and (311) above can be solved to remove the other surface temperature. Even though the coefficients in Equation (309) are fairly complex. the inside surface heat balance can be performed to determine the surface temperatures. we can establish three basic equations for the temperature at the inside and outside surface as well as at the location of the heat source/sink: Tinside = Ca + CbToutside + CC q′′ Toutside = Cd + CeTinside + C f q′′ Tsource = Cg + Ch q′′ + CiTinside + C jToutside (310) (311) (312) where: Tinside is the temperature at the inside surface Toutside is the temperature at the outside surface Tsource is the temperature within the radiant system at the location of the source/sink Ca is all of the other terms in the inside heat balance (solar. once the heat source/sink is defined. etc. is not necessary and since the radiant system can be significantly influenced by the delay that such an assumption might cause. In effect.RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL With both q and Twi known. It should be noted that Equations (306) through (309) are a slight simplification of the actual implementation in EnergyPlus. conduction history terms. respectively. Substituting the new equations for Tinside and Toutside as a function of C and q" into the equation for Tsource and simplifying results in the following equation: Tsource = Ck + Cl q′′ where: (313) Ck = C g + Ci ( Ca + CbCd ) + C j ( Cd + CeCa ) 1 − CeCb 1 − CeCb 325 Cl = Ch + 4/23/05 Ci ( Cc + Cb C f ) + C j ( C f + CeCc ) . LW exchange. LW exchange. the final equation relating the heat source directly to inlet water temperature is compact and does not require any iteration. As with flux control. it is a trivial matter to calculate Two and Ts from Equations (306) and (307). the initial implementation of radiant systems in EnergyPlus used a development (shown below) that does not lag either the inside or the outside surface temperature.

j Ck . Note that the above development is valid for both the hydronic (variable flow) low temperature radiant system and for constant flow (variable temperature) low temperature radiant systems where the inlet temperature is known (based on controls).RADIANT SYSTEM MODELS LOW TEMPERATURE RADIANT SYSTEM MODEL Combining this with heat exchanger analysis as shown above.out. and Twater. j 1+ A If for each surface in the radiant system. it is not possible to know the inlet temperature to the radiant system without simulating it. j ( Twater. we must either iterate or perform more mathematics to arrive at the inlet temperature.Twater.j ) = ε j ( Twater. Combining the previous two equations results in: (Twater. j ) ε j m j c p Cl .in .out. Once the flux to the slab is known.out. The previous equation combines with the following equation which is valid for an surface in the current radiant system: q j = m jc p (Twater.j ) where qj is the heat transfer to the jth surface in the radiant system. The implementation in EnergyPlus chose to avoid iteration and solved for the inlet temperature as shown in the next paragraphs.in .j = (1 − Cm. j ) Twater.) can be calculated using the relatively simpler equations shown above. the remaining terms of interest (outlet water temperature. mj is the mass flow rate only to this surface. etc. when due to loop conditions and the presence of recirculation. j A εj then the previous equations because the slightly less complex: (Twater.out.in . inside and outside surface temperatures.j is the outlet temperature for the jth surface.in + Cm. we eventually arrive at the following equation to relate the flux to the slab to the water inlet temperature and mass flow rate: q′′ = Twater .in − Ck .in − Ck Cl 1 + A ε ( mc p ) (314) water which includes all of the inside and outside heat balance terms (“hidden” in the Ck and Cl coefficients). j = 1+ ε j m j c p Cl . j The overall outlet temperature from the radiant system is just a simple mixing of all of the surface outlet temperatures based on flow fraction and results in Twater.Twater.out. we let: Cm.in − Ck .j ) = Cm .out. However. j ) Rearranging to obtain the outlet temperature for the jth surface: Twater. 4/23/05 326 .Twater.

in) to the loop inlet water temperature (Tloop. This amount of heat is then used in the thermal comfort models as shown in Equations (546) for Fanger. If heating would result during cooling mode or cooling during heating mode.out.RADIANT SYSTEM MODELS HIGH TEMPERATURE RADIANT HEATER MODEL An energy balance on the mixing valve-pump group results in the following equation that relates the radiant system inlet temperature (Twater. NOTE: In some “low-energy” applications. EnergyPlus performs a temperature limit check. the overall outlet temperature.out + msystem msystem c p Plugging in the definition of Twater. In EnergyPlus. the user is also allowed the ability to define what fraction of radiation leaving the heater is incident directly on a person within the zone for thermal comfort purposes.out based on the summation of Twater. Rather than specifying an exact location for the radiant heater(s). it is possible that during cooling mode the elevated water temperatures may result in a heat source (or net heating) to the radiant system.in + mrecirc PumpHeat Twater.in). the model will cut-off the flow rate until the inlet water temperature will produce the proper effect. j ) ⎤ ⎦ ⎜ msystem ⎟ j ⎣ ⎝ ⎠ Once the actual water inlet temperature is calculated with this equation. it can be avoided. With the proper establishment of input data. This procedure avoids iteration but is somewhat complex to follow. and the pump heat addition: Twater. (566) for Pierce Two-Node and (595) for 4/23/05 327 . the heater appears much like a lamp or a tube that is suspended from the ceiling of a space.in = mloop msystem Tloop. the high temperature radiant heater model allows the user a reasonable amount of flexibility.out. the radiant system outlet temperature (Twater. This encompasses a wide variety of heaters including both gas-fired and electric.j ⎛ m ⎞ TwiCoeff = ⎜1 − recirc ⎟ ∑ ⎡ FlowFraction j (1 − Cm. the user is allowed to specify the percentage of heat leaving the heater as radiation and then on which surfaces this radiation is incident.out). it is then possible to calculate individual outlet temperatures for each surface. In addition.in + PumpHeat msystem c p Re circTerm = mrecirc msystem ∑ FlowFraction T j j water.in = where: LoopTerm + Re circTerm TwiCoeff LoopTerm = mloop msystem Tloop. In most cases. and finally all of the necessary flow and loop quantities. this second mathematical process is only needed for select cases of the constant flow radiant system when the inlet temperature is not known explicitly. To avoid heating the slab when cooling is desired. and the surface temperatures are high enough that the heaters must be a safe distance away from the occupied portion of the space for safety concerns.j equations results in: Twater. However. High Temperature Radiant Heater Model The high temperature radiant heater model is intended to encapsulate an entire class of heating devices that seek to heat the occupants within a zone by direct radiation.

05. 0. radiant heaters Zone name Availability schedule Zone name (name of zone system is serving) maximum power input (in Watts) type of heater (either gas or electric) combustion efficiency (ignored for elec. this fraction of radiant energy to people does not have a direct impact on any of the surface heat balances.RADIANT SYSTEM MODELS HIGH TEMPERATURE RADIANT HEATER MODEL KSU Two-node. This energy is distributed to people and to the surfaces within the zone. The sum of all of these distribution fractions (the last six lines of input shown in Figure 110) must sum to unity. In addition. Tradiant ⎡ 4 ⎛Q ⎞⎤ = ⎢(TMRT ) + ⎜ heater → person ⎟ ⎥ ⎜ σA ⎟⎥ ⎢ person ⎝ ⎠⎦ ⎣ 0. ! Radiant Operation. The fraction latent is added to the latent energy balance and will affect moisture levels within the zone. ! Zn001:Wall002. and lost with the remainder assumed to be convective energy. ! Program keyword for high temp. The second additive relationship is within the distribution of the radiant fraction.75. 0.! Zone 1 Radiant Heater. 0. the model assumes that any radiation from the high temperature radiant heater that is incident directly on people is accounted for in the thermal comfort model using Equation (315) but is also assumed to be added to the zone air heat balance via convection from people to the surrounding air. These equations are similar in form to the equation promoted by Fanger (1970). This parameter is somewhat sensitive and will have a direct impact on the thermal comfort models.85. Input Description for High Temperature Radiant Heaters The input for the high temperature radiant heater has two additive relationships that are assumed. ! 10000.75. ! 0. the fractions of radiant. ! GAS.05. in order to not “lose” this amount of energy from the perspective of the zone air or the surface heat balances. The input parameters for the high temperature radiant heater model are shown in Table 1. Note that each high temperature radiant heater is allowed to distribute energy to up to 20 surfaces and that radiant energy placed on a surface using these distribution fractions is assumed to be completely absorbed. ! OPERATIVE. ! Heating Setpoints. ! 0. ! Zn001:Wall001. The user is required to enter the fractions radiant. ! Zn001:Wall004. heaters) fraction radiant fraction latent fraction lost temperature control type (MAT. MRT also possible) heating throttling range (in C) schedule of heating setpoint temperatures fraction of radiant energy to people surface/fraction of radiant energy incident on it surface/fraction of radiant energy incident on it surface/fraction of radiant energy incident on it surface/fraction of radiant energy incident on it surface/fraction of radiant energy incident on it Figure 110. however. The fraction lost is assumed to have no impact on the energy balance of the zone and is assumed to be lost or vented to the exterior environment. ! 0. ! Zn001:Flr001.05. one must use caution when determining this fraction since overestimation of this number might lead to predictions of thermal comfort where in fact it does not exist. ! SHOP ZONE.05. Thus.05. latent.25 (315) HIGH TEMP RADIANT SYSTEM. convective. ! Zn001:Wall003. 0.0 . 0. Several things should be noted about the fraction of heat that is radiated directly to people. latent. This is exactly the intent of the high temperature radiant heaters. This 4/23/05 328 .05. the distribution fractions should also take into account any differences in long wavelength absorptivity among the surfaces. ! 0. and lost heat must sum to unity. ! 2. The thermal comfort energy balance is completely separate from and has no bearing on the zone air or the surface heat balances. Thus. ! 0. First.05.

the controlling temperature is allowed to be the mean air. Thesis. M. Fanger. thesis. Myers. Department of Mechanical and Industrial Engineering. Danish Technical Press. F. 115120. Liesen. or the operative temperature. Incropera. the mean radiant. ASHRAE Transactions.C. Bishop. and C. 553-557. Pedersen. Number 11. Strand. Introduction to Heat Transfer. Volume 103. L.O. Draft Publication. and R. Conduction dans un “mur” multicouche avec sources: extension de la notion de quadripole. ASME Journal of Heat Transfer. New York: John Wiley & Sons. Champaign. Degiovanni. and F. Lawrence Berkeley National Laboratories. since the high temperature radiant heater has a direct impact on the surfaces within a zone. 1980. Resolution of Radiant System Response at Varying Time Steps is also used for high temperature radiant heaters. W. pp. D.M. A final “average” heat balance calculation is done after all of the system time steps have been simulated to maintain continuity within the surface heat balances. and A. 2001.P. T. pp. Strand. Buhl. J. Number 1. IL. 1970. R. United States Army Construction Engineering Research Laboratory.P. 1685-1693. Hittle.O. (Operative temperature for radiant system controls is the average of MAT and MRT. Number 3. Liesen.F.O. Radiative Transfer. Maloney. Volume 31. Hittle. D.. and C. Fisher. 1997. “EnergyPlus Input Output Reference”. R. Lee.J. New York.E. Many of the control and integration aspects of the high temperature radiant system model in EnergyPlus are very similar to the low temperature radiant system model. 1997. Volume 103. 4/23/05 329 . 1983. Ph. Long-time solutions to heat conduction transients with time-dependent inputs. International Journal of Heat and Mass Transfer. References Ceylan.S. Copenhagen. H. Department of Mechanical and Industrial Engineering. DeWitt. pp. and R.) Also. TX.. Berkeley. Pedersen. International Journal of Heat and Mass Transfer. “Development of a radiant heater model and the incorporation of thermal comfort considerations into the BLAST energy analysis program”. Winkelmann.C.J. Part 2. “Investigation of Outside Heat Balance Models for Use in a Heat Balance Cooling Load Calculation Procedure”.E. Part 2. McClellan..T.K.D. the surface heat balances are recalculated to determine an approximate response to the radiation from the heater. 1967.F. D. “An Analysis of the Effect of the Building Envelope on Thermal Comfort using the EnergyPlus Program”. 1988. D.C. As with the low temperature radiant system.RADIANT SYSTEM MODELS HIGH TEMPERATURE RADIANT HEATER MODEL guarantees that the people within the space feel the direct radiative effect of the heaters and that this quantity of energy is not “lost” within the heat balance routines.K. Sarofim. 1981. University of Illinois at Urbana-Champaign and Technical Manuscript E-169.K. submitted for publication in the proceedings of the 2001 ACSA (Association of Collegiate Schools of Architecture) Technology Conference. and the setpoint temperature is allowed to vary hourly based on a user defined schedule. Calculating building heating and cooling loads using the frequency response of multilayered slabs. Lawrie. R. 1985. Volume 26. An improved root-finding procedure for use in calculating transient heat flow through multilayered slabs. H. Austin. Variable Flow Low Temperature Radiant System Controls” where the amount of heat generated by the radiant heater varies as a function of the difference between the controlling and the setpoint temperatures. McGraw-Hill. ASHRAE Transactions. and G. University of Illinois at Urbana-Champaign. and D. 1987. The algorithm shown in “Figure 109. “An Evaluation of Inside Surface Heat Balance Models for Cooling Load Calculations”. P. A. Hottel. Volume 102. The controls are the same as shown in “Figure 106. CA.W. Thermal Comfort-Analysis and Applications in Environmental Engineering.

1998. and C. Ph. Conference Proceedings of Building Simulation 1991. France. Volume 103. R. Strand. Pedersen. D. and R. Fisher. Strand. and C. and L.D.J.O. Ph. R. ASHRAE. Pedersen.K. Nice. “Impact of simultaneous simulation of building and mechanical systems in heat balance based energy analysis programs on system response and control”. “Development of a Heat Balance Procedure for Cooling Loads”.O. 2001. Liesen. Part 2. Switzerland..D.J. Liesen. Part 1. “Analytical verification of heat source transfer functions”. Fisher. 1997. Brazil.E.K. R. J. Strand. ASHRAE Transactions.O.O. D.O. 1994.E.O.D. Spitler. C. R. Conference Proceedings of Building Simulation 2001. R. Zürich. International Building Performance Simulation Association. 1987. and C.K.. dissertation.. “Heat source transfer functions and their application to low temperature radiant heating systems”.RADIANT SYSTEM MODELS HIGH TEMPERATURE RADIANT HEATER MODEL Pedersen. Seem. ASHRAE Transactions. Taylor. 1997. Pedersen. Fisher. University of Illinois at Urbana-Champaign. 1991. First Joint Conference of International Simulation Societies. Strand. Thesis. Cooling and Heating Load Calculation Principles. “Implementation of a Radiant Heating and Cooling Model into an Integrated Building Energy Analysis Program”. “Modularization and Simulation Techniques for Heat Balance Based Energy and Load Calculation Programs: the Experience of the ASHRAE LOADS Toolkit and EnergyPlus”.K.D. 4/23/05 330 . Lawrie... and R.E. International Building Performance Simulation Association. C. R. University of Wisconsin-Madison. Department of Mechanical and Industrial Engineering. Modeling of heat transfer in buildings. Rio de Janeiro. Liesen. J. Volume 103. C. 1995. Pedersen. Pedersen. D.

The model is based the following three equations OperatingPartLoadRatio = BoilerLoad BoilerNomCapacity (316) TheoreticalFuelUse = BoilerLoad BoilerEfficiency (317) FuelUsed = TheoreticalFuelUsed C1 + C 2 ∗ OperatingPartLoadRatio + C 3 ∗ OperatingPartLoadRatio 2 (318) A constant efficiency boiler is specified by setting C1=1. and “Plant”. these designations signify where the effects of these models have their first impacts. Table 39. Boilers Boilers: Simple Boiler The EnergyPlus boiler model is “simple” in the sense that it requires the user to supply the theoretical boiler efficiency. The combustion process is not considered in the model. which is input by the user for energy accounting purposes only. average for the simulation time step Pload = parasitic electric load specified by the user (W) The model assumes that this parasitic power does not contribute to heating the water. The SPARK model SBOILER is a steady-state model with no state persistency. The parasitic electric power is calculated based on the user-defined parasitic electric load and the operating part load ratio calculated above. “HVAC”. Boilers: SPARK Simple Boiler The SBOILER model implements the native boiler model BOILER:SIMPLE as a SPARK model.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE BOILERS Simulation Models – Encyclopedic Reference The following descriptions are grouped alphabetically (as much as possible) with some also having additional tags of “Heat Balance”. Consult the engineering documentation of the BOILER:SIMPLE model for more details. The model is independent of the fuel type. and C3=0. Pparasitic = Pload × OperatingPartLoadRatio where: Pparasitic = parasitic electric power (W). The variable mapping between the EnergyPlus variables and the SPARK variables is listed in the next tables for each variable type. Summary of parameter variables in the SBOILER SPARK model SPARK variable 4/23/05 SPARK unit Description 331 . C2=0. With the integrated solution.

Summary of output variables of the SBOILER SPARK model SPARK variable name T_outlet PLR mdot Qdot Qdot_scope. Only used if FlowLockFlag=1 Table 41.001 Table 40.integral FuelUsed 4/23/05 SPARK unit C W/W kg/s W J W Description Boiler outlet temperature Actual operating part load ratio Actual boiler water mass flow rate Boiler operating load Boiler energy integrated over time Boiler fuel used 332 . Summary of input variables in the SBOILER SPARK model SPARK variable name FlowLockFlag Load T_inlet T_setpoint cp mdot_locked SPARK unit 0=false.g.K) kg/s Description Flag indicating whether it is a flow request or the flow is locked Requested load Boiler inlet temperature Boiler outlet setpoint temperature Specific heat capacity of hot water Locked mass flow rate for actual resolved branch flow. forced draft fan) Design nominal capacity of boiler Boiler outlet maximum temperature limit Boiler efficiency at design conditions Boiler water design mass flow rate Minimum significant mass flow rate = 0.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE BOILERS name C1 C2 C3 FlowType PLR_max PLR_min ParasiticElecLoad Qdot_nominal T_limit efficiency mdot_design mdot_tol 1=constant.2=variable W/W W/W W W C W/W kg/s kg/s Coefficient of the fuel consumption/part load ratio curve Coefficients of the fuel consumption/part load ratio curve Coefficients of the fuel consumption/part load ratio curve Flow type Maximum allowed operating part load ratio Minimum allowed operating part load ratio Parasitic electric power (e.1=true W C C J/(kg.

integral ParasiticElecPower ParasiticElecPower_scope. Three sets of coefficients are required. The liquid solution returns to the absorber through a throttling valve whose purpose is to provide a pressure drop to maintain the pressure difference between the generator and absorber. or well water condenser depending on which one is selected by the user based on the physical parameters of the plant. and delivers the liquid to the generator. The evaporator can 4/23/05 333 . The heat from the condenser is rejected to a cooling tower. the coefficient sets establish the ratio of heat power in-to-cooling effect produced as a function of part load. The centrifugal compressor has the incoming fluid entering at the eye of a spinning impeller that throws the fluid by centrifugal force to the periphery of the impeller. The condenser and evaporator are similar to that of a standard chiller. In the generator. After leaving the compressor. Fitting catalog data to third order polynomial equations generates chiller performance curves. A pump receives low-pressure liquid from the absorber. All Absorption Chiller models are based on a polynomial fit of Absorber performance data. The assembly of a generator and absorber provides the compression operation. evaporative condenser. gas. which are both waterto-water heat exchangers. elevates the pressure of the liquid.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS FuelUsed_scope. see the Input/Output Reference Document (Object: Chiller: Absorption). the refrigerant is condensed to liquid in a refrigerant to water condenser. For more information on absorption chillers. heat from a high temperature source drives off the vapor that has been absorbed by the solution. and steam turbines. The heat supplied to the absorber can be waste heat from a diesel jacket or the exhaust heat from diesel. The refrigerant pressure is then dropped through a throttling valve so that fluid can evaporate at a low pressure that provides cooling to the evaporator.integral J W J Boiler Fuel consumed integrated over time Parasitic Electrical Power Parasitic Electrical Consumption Chillers Chillers: Absorption Chiller The absorber is simulated as an empirical model of a standard absorption refrigeration cycle. Low-pressure vapor from the evaporator is absorbed by the liquid solution in the absorber. The BLAST model of an open centrifugal chiller is modeled as standard vapor compression refrigeration cycle with a centrifugal compressor driven by a shaft power from an engine. The ratio of heat-power-in to cooling-effect-produced is the inverse of the coefficient of performance. Chillers: Combustion Turbine Chiller This chiller model (Object name: Chiller:Combustion Turbine) is the empirical model from the Building Loads and System Thermodynamics (BLAST) program. The defining equation is: SteamInputRatio = C1 + C 2 + C 3 ∗ PLR PLR ElectricInputRatio = C1 + C 2 ∗ PLR + C 3 ∗ PLR 2 The Pump Electric Use Part Load Ratio Curve is a quadratic equation that determines the Ratio of the actual absorber pumping power to the nominal pumping power. The gas turbine-driven chiller is an open centrifugal chiller driven directly by a gas turbine. The defining equation is: ElectricInputRatio = C1 + C 2 ∗ PLR + C 3 ∗ PLR 2 Thus. The Steam Use Part Load Ratio Curve is a quadratic equation that determines the Ratio of the steam load on the absorber to the demand on the chiller.

which maintains the capacity at the nominal value. TempEvapOut = Temperature leaving the evaporator. see the Input/Output Reference Document. The defining equation is: FracFullLoadPower = C1 + C2 PartLoadRatio + C3 PartLoadRatio 2 The Fuel Input Curve is a polynomial equation that determines the Ratio of Fuel Input to Energy Output. Shown below are the definitions of the curves that describe this model. The following three fields contain the coefficients for the quadratic equation. TELvrequired = Required leaving evaporator water outlet temperature to maintain rated capacity TELvrated = Rated leaving evaporator water outlet temperature at rated capacity The Capacity Ratio Curve is a quadratic equation that determines the Ratio of Available Capacity to Nominal Capacity. TempRiseCoefficient = User Input from above. The defining equation is: 2 AvailToNominalCapacityRatio = C1 + C2 ∆ temp + C3∆ temp Where the Delta Temperature is defined as: ∆ Temp = TempCondIn . The defining equation is: 2 FuelEnergyInput = PLoad ∗ ( FIC1 + FIC2 RLoad + FIC3 RLoad 2 ) ∗ (TBFIC1 + TBFIC2 ATair + TBFIC3 ATair ) 4/23/05 334 . The equation combines both the Fuel Input Curve Coefficients and the Temperature Based Fuel Input Curve Coefficients. TempCondInDesign = Temp Design Condenser Inlet from User input above. This is calculated as the following ratio: TCEntrequired − TCEntrated TELvrequired − TELvrated Where: TCEntrequiredl = Required entering condenser air or water temperature to maintain rated capacity TCEntrated = Rated entering condenser air or water temperature at rated capacity. For more information.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS chill water that is pumped to chilled water coils in the building. TempEvapOutDesign = Temp Design Evaporator Outlet from User input above.TempCondInDesign − (TempEvapOut . This chiller is modeled like the electric chiller with the same numerical curve fits and then some additional curve fits to model the turbine drive. The defining equation is: FullLoadtoPowerRatio = C1 + C2 AvailToNominalCapRatio + C3 AvailToNominalCapRatio 2 The Full Load Ratio Curve is a quadratic equation that determines the fraction of full load power. The Power Ratio Curve is a quadratic equation that determines the Ratio of Full Load to Power. The chiller’s temperature rise coefficient which is defined as the ratio of the required change in condenser water temperature to a given change in chilled water temperature.TempEvapOutDesign) TempRiseCoefficient TempCondIn = Temperature entering the condenser.

The Exhaust Gas Temperature Curve is a polynomial equation that determines the Exhaust Gas Temperature. 4/23/05 335 . QEvaporator = Load Power = Load / ConstCOPChiller(ChillNum)%COP Then the evaporator temperatures are calculated from the load EvapDeltaTemp = QEvaporator/EvapMassFlowRate/CPwater EvapOutletTemp = Node(EvapInletNode)%Temp . TBC are the Temperature Based Exhaust Gas Temperature Curve Coefficients.EvapDeltaTemp The condenser load and temperatures arecalculated from the evaporator load and the power to the chiller. Rload is the Ratio of Load to Combustion Turbine Engine Capacity. performance does not vary with chilled water temperature or condenser conditions. In this case. and ATair is the difference between the actual ambient and design ambient temperatures. The Exhaust Flow Curve is a quadratic equation that determines the Ratio of Exhaust Gas Flow Rate to Engine Capacity. The defining equation is: 2 ExhaustFlowRate = GTCapacity ∗ (C1 + C2 ATair + C3 ATair ) Where GTCapacity is the Combustion Turbine Engine Capacity.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Where FIC represents the Fuel Input Curve Coefficients. TBFIC represents the Temperature Based Fuel Input Curve Coefficients. The defining equation is: 2 ExhaustTemperature = (C1 + C2 RLoad + C3 RLoad 2 ) ∗ (TBC1 + TBC2 ATair + TBC3 ATair ) − 273. Such a model is useful when the user does not have access to detailed performance data. The nominal capacity of the chiller and the COP are user specified along with the connections to the plant and condenser loop and mass flow rates. The Power from the chiller is calculated from the load divided by the COP. The defining equation is: RecoveryLubeEnergy = PLoad ∗ (C1 + C2 RL + C3 RL2 ) Where Pload is the engine load and RL is the Ratio of Load to Combustion Turbine Engine Capacity The UA is an equation that determines the overall heat transfer coefficient for the exhaust gasses with the stack. and ATair is the difference between the current ambient and design ambient temperatures. constant COP simulation of the chiller. The equation combines both the Exhaust Gas Temperature Curve Coefficients (Based on the Part Load Ratio) and the (Ambient) Temperature Based Exhaust Gas Temperature Curve Coefficients.15 Where C represents the Exhaust Gas Temperature Curve Coefficients. and ATair is the difference between the current ambient and design ambient temperatures. This chiller will meet the load as long as it does not exceed the nominal capacity specified by the user. The defining equation is: UAToCapacityRatio = C1GasTurbineEngineCapacity C2 Chillers: Constant COP Chiller This chiller model (Object: Chiller:CONST COP) is based on a simple. The heat transfer coefficient ultimately helps determine the exhaust stack temperature. RLoad is the Ratio of Load to Combustion Turbine Engine Capacity. The Recovery Lubricant Heat Curve is a quadratic equation that determines the recovery lube energy.

Figure 111. QTot = (mHeat Re c * CpHeat Re c + mCond * CpCond ) *(TAvgOut − TAvgIn ) Then the inlet temperature can be flow ratio averaged to determine its conditions. 4/23/05 336 . The refrigerant pressure is then dropped through a throttling valve so that fluid can evaporate at a low pressure that provides cooling to the evaporator. the refrigerant is condensed to liquid in a refrigerant to water condenser. Chillers: Electric Chiller with Heat Recovery Options (Plant) The Electric Chiller (Object: Chiller:Electric) can have the option of having its condenser hooked up to a heat recovery loop or what is commonly known as a double bundled chiller. After leaving the compressor. Diagram of Chiller:Electric with Heat Recovery Algorithm for the Heat Recovery portion of the chiller needs to be determined from relatively simple inputs to estimate the amount of the heat that is recovered and then send the rest of the heat to the cooling tower. the other path is to the heat recovery. A double bundle condenser involves two separate flow paths through the condenser.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS QCondenser = Power + QEvaporator CondOutletTemp = QCondenser/CondMassFlowRate/CPwater + CondInletTemp See the InputOutput Reference for additional information. QTot = QEvap + QElec = QCond + QHeat Re c QTot = mHeat Re c * CpHeat Re c *(THeat Re cOut − THeat Re cIn ) + mCond * CpCond *(TCondOut − TCondIn ) Approximate the total heat transfer with average temperatures in and out. One of these paths is to a standard cooling tower. The Heat Recovery chiller is simulated as a standard vapor compression refrigeration cycle with a double bundled condenser.

1 chiller model. The EnergyPlus model contains all of the features of the DOE-2. plus additional abilities for modeling evaporatively-cooled condensers and heat recovery for water heating.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS TAvgIn = (mHeat Re c * CpHeat Re c * THeat Re cIn + mCond * CpCond * TCondIn ) (mHeat Re c * CpHeat Re c + mCond * CpCond ) QTot + TAvgIn (mHeat Re c * CpHeat Re c + mCond * CpCond ) TAvgOut = Then the total heat can have an approximate fractional split between heat recovered and rejection through the tower by the following. It also models the power consumption of condenser fans if modeling an air-cooled or evaporatively-cooled condenser. The model is based on the compression chiller model (COMREF) in the DOE-2. the user has to pre-determine the heat recovery split with no relationship to actual simulation conditions. Cooling Tower:Single Speed). Chillers: Electric EIR Chiller Overview This model (object name Chiller:Electric:EIR) simulates the performance of an electric liquid chiller.g. Otherwise. This model does not simulate the thermal performance or the power consumption of associated pumps or cooling towers. This model simulates the thermal performance of the chiller and the power consumption of the compressor(s). The output of this curve is multiplied by the design capacity to give the cooling capacity at specific temperature operating 337 4/23/05 . The three performance curves are: 1) Cooling Capacity Function of Temperature Curve 2) Energy Input to Cooling Output Ratio Function of Temperature Curve 3) Energy Input to Cooling Output Ratio Function of Part Load Ratio Curve • The cooling capacity function of temperature curve is a biquadratic performance curve with two independent variables: the leaving chilled water temperature and the entering condenser fluid temperature. Model Description The chiller model uses user-supplied performance information at design conditions along with three performance curves (curve objects) for cooling capacity and efficiency to determine chiller operation at off-design (part load) conditions. The inputs for the Chiller:Electric to add heat recovery are in the Input/Output Reference.1 building energy simulation program. m * CpHeat Re c *(TAvgOut − THeat Re cIn ) QHeat Re c ≅ FracHeat Re c = Heat Re c QCond mCond * CpCond *(TAvgOut − TCondIn ) Therefore the heat rejected to the water-cooled condenser and heat recovery is: QCond = (1 − FracHeat Re c ) * QTot QHeat Re c = FracHeat Re c * QTot This allows for an approximate split of the heat recovery with some relationship to the mass flow rates and the inlet temperatures of each. This auxiliary equipment must be modeled using other EnergyPlus models (e.

l = leaving chilled water temperature. The output of this curve is multiplied by the design EIR (inverse of the design COP) and the Energy Input to Cooling Output Ratio Function of Temperature Curve to give the EIR at the specific temperatures and part-load ratio at which the chiller is operating. • The energy input to cooling output ratio function of temperature curve is a biquadratic performance curve that parameterizes the variation of the energy input to cooling output ratio (EIR) as a function of the leaving chilled water temperature and the entering condenser fluid temperature. For a water-cooled condenser this will be the water temperature returning from the condenser loop (e. For air. The curve should have a value of 1.0 at the design temperatures and flow rates specified in the input data file by the user.l ) + c (Tcw.e ) w 2 2 here ChillerCapFTemp = cooling capacity factor. It is not imperative that the user utilize all coefficients in the performance curve equations if their performance equation 4/23/05 338 . and the part load ratio is the actual cooling load divided by the chiller’s available cooling capacity.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS conditions (i.0 at the design temperatures and flow rates specified in the input data file by the user. The curve should have a value of 1.l ) + d (Tcond .l ) + d (Tcond . The EIR is the inverse of the COP.l )(Tcond .e ) + e (Tcond . at temperatures different from the design temperatures).e = entering condenser fluid temperature. at temperatures different from the design temperatures). respectively. The output of this curve is multiplied by the design EIR (inverse of the design COP) to give the EIR at specific temperature operating conditions (i. ˚C Tcond. ˚C Tcond. This curve should have a value of 1.e ) where ChillerEIRFTemp = energy input to cooling output factor.e ) + f (Tcw. equal to 1 at design conditions Tcw.or evap-cooled condensers this will be the entering outdoor air dry-bulb or wet-bulb temperature. The EIR is the inverse of the COP.g.e.or evap-cooled condensers this will be the entering outdoor air dry-bulb or wet-bulb temperature. equal to 1 at design conditions PLR = part load ratio = (cooling load) / (chiller’s available cooling capacity) All three of the performance curves are accessed through EnergyPlus’ built-in performance curve equation manager (curve:quadratic and curve:biquadratic)..g. equal to 1 at design conditions Tcw.0 when the part-load ratio equals 1.e.. leaving the cooling tower). ˚C. ˚C. respectively. 2 ChillerEIRFPLR = a + b ( PLR ) + c ( PLR ) where ChillerEIRFPLR = energy input to cooling output factor. For air.l )(Tcond . • The energy input to cooling output ratio function of part load ratio curve is a quadratic performance curve that parameterizes the variation of the energy input ratio (EIR) as a function of the part load ratio..e = entering condenser fluid temperature.l ) + c (Tcw. ChillerCapFTemp = a + b (Tcw.0.e ) + f (Tcw. 2 2 ChillerEIRFTemp = a + b (Tcw. For a water-cooled condenser this will be the water temperature returning from the condenser loop (e..l = leaving chilled water temperature. leaving the cooling tower).e ) + e (Tcond .

the chiller’s available cooling capacity is calculated as follows: Q avail = Q design ( ChillerCapFTemp ) where • • • Q design = chiller capacity at design conditions (design temperatures and flow rates defined by the user).. the exiting chilled water temperature is calculated based on the water temperature entering the evaporator. The part-load ratio is not allowed to be greater than the maximum part-load ratio specified by the user or less than zero as follows: • ⎛ ⎛Q ⎞⎞ . If this calculated heat transfer rate is greater than the heat transfer rate being requested by the PLANT OPERATION SCHEME.e − ⎜ avail • ⎜ mevap C p . For any simulation time step.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS has fewer terms (e.l where ⎛• ⎞ Q ⎟ = Tcw.evap ⎟ ⎝ ⎠ Tcw. kg/s mevap Cp. J/kg-˚C The part-load ratio is then calculated as the ratio of the evaporator heat transfer rate to the available chiller capacity.e • = water temperature leaving the evaporator.e. simply enter the values for a and b.0. W • Q avail = available chiller capacity adjusted for current fluid temperatures. then the evaporator heat transfer rate is reset to the requested cooling rate. The evaporator heat transfer rate is then compared to the available capacity.evap = specific heat of water entering evaporator at Tcw. ˚C = evaporator mass flow rate. the available cooling capacity. if the user’s ChillerEIRFPLR performance curve is linear instead of quadratic. and the evaporator mass flow rate as follows: Tcw. If the available chiller capacity is sufficient to meet the evaporator heat transfer rate. PLRmax ⎟ ⎟ PLR = MAX ⎜ 0. MIN ⎜ evap • ⎜ ⎜ ⎟⎟ Q avail ⎝ ⎠⎠ ⎝ where PLR 4/23/05 = part load ratio 339 .l Tcw.g. the leaving chilled water temperature is set equal to the chilled water set point temperature. ˚C = water temperature entering the evaporator. and set coefficient c equal to zero). For this case. W The model then calculates the evaporator heat transfer rate required to bring the entering chilled water temperature down to the leaving chilled water set point temperature (established using a SET POINT MANAGER object and referenced in the PLANT LOOP object). If the requested evaporator heat transfer rate is larger than the available capacity the chilled water leaving the evaporator is allowed to float upward.

The result is available as the output variable Chiller Part Load Ratio. PLR = MIN ( PLR.g. ChillerCyclingRatio = MIN ⎛ PLR . W = maximum part load ratio (specified by the user in the input data file) PLRmax The model assumes that the cooling load is met through chiller unloading down to the minimum unloading ratio. The amount of false loading on the chiller is calculated using this revised PLR and is reported as an output variable as follows: Q falseloading = Q avail • ( ) • ( PLR )( ChillerCyclingRatio ) − Q evap • The electrical power consumption for the chiller compressor(s) for any simulation time step is then calculated using the following equation: ⎞ 1 ⎛• ⎞⎛ Pchiller = ⎜ Q avail ⎟ ⎜ ⎟ ( ChillerEIRFTemp )( ChillerEIRFPLR )( ChillerCyclingRatio ) ⎝ ⎠ ⎜ COPdesign ⎟ ⎝ ⎠ where Pchiller = chiller compressor power. hot-gas bypass) is assumed to occur between the minimum unloading ratio and the minimum part load ratio yielding constant electrical power consumption under these conditions. Below the minimum part load ratio. For electric chillers with hermetic compressors. False loading (e. the chiller cycles on and off to meet very small loads and the power consumption during the on cycle is the same as when the chiller is operating at the minimum part load ratio..0). W COPdesign = design coefficient of performance.0 ⎞ ⎜ ⎟ PLRmin ⎝ ⎠ To properly account for chiller electric power consumption when PLR is less than the minimum unloading ratio. only a portion of the compressor energy use is rejected by the condenser. the PLR is reset to the greater of the PLR calculated above and the PLR at the minimum unloading ratio. W/W Heat rejected by the chiller condenser includes the heat transferred in the evaporator plus a portion or all of the compressor electrical energy consumption. all compressor energy consumption is rejected by the condenser (compressor motor efficiency = effmotor = 1. The heat transfer rate for the chiller condenser is calculated as follows: Q cond = ( Pchiller ∗ eff motor ) + Q evap + Q falseloading 4/23/05 340 • • • .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • CHILLERS Q evap = load to be met by the chiller. When the chiller part load ratio is less than the minimum part load ratio. 1. For chillers with semi-hermetic or open compressors. PLRMinUnloadRatio ) This revised PLR accounts for the “false loading” (e. the on-off cycling ratio of the chiller is calculated as follows and is available as an output variable.g. hot gas bypass) that is assumed to occur whenever the PLR (based on cooling load divided by available capacity) is less than the minimum unloading ratio specified.

J/kg-˚C mcond C p .cond For air. the water temperature leaving the condenser is then calculated as shown below.l Tcond. as well as the total electric energy consumed by the chiller compressor motor(s) and condenser fan(s). ˚C = water temperature entering the condenser.cond • ) where: Tcond. The amount of condenser fan energy is assumed to be proportional to the chiller cycling ratio and is calculated as follows: Pcond = Q design ( Pcondfanratio ) ( ChillerCyclingRatio ) where • Pcond = chiller condenser fan electric power. W Pcondfanratio = condenser fan power ratio. kg/s = specific heat of water entering the condenser at Tcond. the exiting air temperature is not calculated and is set equal to the entering air or wet-bulb temperature. Qcond = Q cond * TimeStepSys *3600 Qevap = Q evap * TimeStepSys * 3600 • • Echiller = Pchiller * TimeStepSys * 3600 Econd = Pcond ∗ TimeStepSys * 3600 where 4/23/05 341 . The model then calculates the condenser fan energy for air. The results are available as output variables. ˚C = mass flow rate through the condenser.e. respectively. W = compressor motor efficiency = fraction of compressor electrical energy consumption rejected as condenser heat For water-cooled chillers.e • = water temperature leaving the condenser. • Tcond .and evaporatively-cooled condensers.l = Tcond .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS where • Q cond eff motor = condenser heat transfer rate.and evaporatively-cooled condensers. W/W The final calculations determine the total heat transfer energy for the condenser and evaporator.e + Q cond ( mcond ∗ C p .

Refer to the section entitled Chiller:Electric with Heat Recovery Options (Plant) for details. and the evaporator and condenser heat transfer coefficients. It calculates the cooling capacity and the compressor consumption for the specified operating conditions (i. a compressor and a condenser with the refrigerant loop. Chillers: SPARK Electric Chiller The ECHILLER SPARK model implements the PISSIM1 model of the ASHRAE Primary HVAC Toolkit with the addition of a simple control law to regulate the volumetric flow rate through the compressor in order to meet the requested load. The control law essentially regulates the volumetric flow rate through the compressor so that the cooling capacity satisfies the requested load by enforcing a residual relation of the form: 4/23/05 342 .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Qcond = chiller condenser heat transfer. J TimeStepSys = HVAC system simulation time step. Figure 112.e. the cooled water loads and the heated water loads). Schematic of the ECHILLER model with feedback loop. The algorithms are identical to those used for Chiller:Electric. J Qevap = chiller evaporator heat transfer. the compressor parameters. sec/hr Electric EIR Chiller with Heat Recovery Option Heat from the electric EIR chiller condenser may be recovered when a water-cooled condenser is selected for simulation. The PISSIM1 model calculates the cooling capacity as a function of the volumetric flow rate through the compressor: Qev = Qev (Vcomp ) . The following inputs need to be specified to characterize the chiller: the refrigerant fluid. The PISSIM1 model is a steady-state.. physical model describing an electric chiller consisting of an evaporator. hr 3600 = conversion factor. J Econd = chiller condenser fan electric consumption. The heat recovery water flow rate is specified by the user along with the input and output nodes connected to the heat recovery loop. J Echiller = chiller (compressor) electric consumption.

Summary of parameter variables in the ECHILLER SPARK model SPARK variable name chiller. The volumetric flow rate through the compressor can be viewed as a time-varying extension of the geometric displacement of the compressor in the original ASHRAE model. Table 42.Cf compressor.Acl refrigerant_properties.K) J/(kg.To refrigerant_properties. Although a simplification of the real-world operation of a chiller. Parameters are set only once at the onset of the simulation and remain constant for the duration of the simulation.b_vap refrigerant_properties.AU refrigerant_properties.Tb refrigerant_properties.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Qev − Load = 0 The control strategy assumes that the volumetric flow rate can be adapted continuously over the whole operating range.MinCap compressor.Tc refrigerant_properties. Model description The ECHILLER model is a steady-state SPARK model with no internal state.hfo 4/23/05 standard boiling temperature critical temperature Reference temperature of refrigerant constant coefficient used vaporization enthalpy in calculation of J/(kg.gamma refrigerant_properties.cpliq refrigerant_properties.AU evaporator.Wlo compressor. this approach has the benefits of requiring fewer inputs than a full description of the compressor configuration such as the number of cylinders and the various modes of operation.hfgb refrigerant_properties. The variable mapping between the EnergyPlus variables and the SPARK variables is listed in the next tables for each variable type.cpvap refrigerant_properties. The input variables are set at each EnergyPlus time step to reflect the changing boundary conditions.MaxCap chiller.Bcl refrigerant_properties.K) mean specific heat in saturated liquid state mean specific heat at constant pressure in superheated vapor state mean isentropic coefficient of refrigerant J/kg J/kg specific vaporization enthalpy at standard boiling point specific enthalpy of the saturated liquid at the f t t 343 .alpha W SPARK unit W W Description maximum operating capacity of chiller minimum operating capacity of chiller clearance factor of compressor constant part of the electromechanical losses in compressor loss factor allowing to define another electromechnical loss which is assumed to be proportional to isentropic power W/K W/K evaporator heat transfer coefficient condenser heat transfer coefficient first coefficient in Clausius-Clapeyron equation K K K K second coefficient equation in Clausius-Clapeyron condenser.

zeta J/(kg. Table 44. energy from heat transfer rate) must be computed at the time scale used in the SPARK model. Note that the outlet temperatures are converted from their internal representation in degree Kelvin back into degree Celsius as expected by the EnergyPlus simulation program. Table 43..cp condenser.r refrigerant_properties.mdot EvapInletTemp evaporator. the energy quantity calculated in the SPARK model corresponds to the energy accumulated over the current EnergyPlus time step. water side loop demand component will meet The output variables of the ECHILLER model are the variables computed by the PISSIM1 model along with the volumetric flow rate through the compressor returned by the control law.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS reference temperature refrigerant_properties. Finally.cp evaporator. the ECHILLER model expects the inlet water temperatures to be specified in degree Celsius to conform to the unit requirements in EnergyPlus.K) gas constant compressibility factor Although the model temperatures are expressed internally in degree Kelvin. Summary of output variables of the ECHILLER SPARK model SPARK variable name evaporator. Thus.Qdot W SPARK unit Description evaporator effectiveness rate of heat transfer to the evaporator coil 11 The integration calculation must be performed inside the SPARK model as opposed to in the EnergyPlus component module because the SPARK models can be simulated using a time step that is smaller than the EnergyPlus time step.K) kg/s C J/(kg. Summary of input variables in the ECHILLER SPARK model SPARK variable name CondInletTemp condenser. water side evaporator inlet temperature specific heat capacity of cold water on evaporator side evaporator mass flow rate.eff evaporator. the heat transfer rates to the evaporator and condenser coils as well as the rate of energy use in the compressor are integrated over each EnergyPlus step to produce the resulting energies for the EnergyPlus reports.K) kg/s W Description condenser inlet temperature specific heat capacity of cold water on condenser side condenser mass flow rate.g.mdot Load SPARK unit C J/(kg.11 Energy = Tn+1 Tn ∫ Q ( t ) dt where Ti indicates time values in EnergyPlus and ti indicates time values in the SPARK model. To ensure a consistent resolution of the differing time scales the integrated quantities (e. 344 4/23/05 .

water side volumetric efficiency of compressor W J m3/s kg/s rate of chiller energy use chiller energy use geometric displacement of compressor refrigerant mass flow rate coefficient of performance PISSIM1 model The simulation of the PISSIM1 model without the controller produces the following cooling capacity (Qdot_ev) and electric energy consumption (W_comp) when varying the volumetric flow rate through the compressor. Figure 113.integral compressor. Volumetric Flow for PISSIM1 model The resulting temperature profiles for the evaporator and condenser water nodes are shown in the following graph.COP J C heat transfer to the evaporator coil evaporator outlet temperature. water side condenser effectiveness W J C rate of heat transfer to the condenser coil heat transfer to the condenser coil condenser outlet temperature. The boundary conditions (i.Vdots refrigerant_cycle.eff condenser. 4/23/05 345 .mdot chiller.e..integral CondOutletTemp compressor.effvol compressor.W CompW_scope.Qdot CondQ_scope. water mass flow rates and water inlet temperatures for the evaporator and condenser) are kept constant for the sensitivity analysis.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS EvapQ_scope.integral EvapOutletTemp condenser.

Evaporator and Condenser Temperature Profiles Finally. the last graph shows the actual refrigerant mass flow rate for the varying volumetric flow rate through the compressor. 4/23/05 346 .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Figure 114.

The following equation describes the control strategy implemented in the ECHILLER model. Thus. The minimum capacity value specified in the IDF is used to that effect. which in turn will stop the EnergyPlus simulation. This will make the resulting nonlinear problem singular and therefore not solvable numerically. Then. as long as the maximum capacity reflects the actual capacity of the chiller model the ECHILLER model will never attempt to solve for an non-attainable operating condition. The non-zero load ensures that the Jacobian remains non-singular.. The residual formulation of the control law achieves just that by finding the intersection of the requested load with the cooling characteristic achievable with the current operating conditions (i. Another pathologic problem inherent to the system of equations under consideration arises when trying to solve for the no load condition (Load = 0 W). albeit probably poorly conditioned. Refrigerant Mass Flow Control strategy The task of the controller is to find the volumetric flow rate through the compressor that produces the cooling capacity requested by the plant manager. Another built-in safeguard consists in enforcing a lower bound on the requested load used in the residual control equation.e. inlet temperatures and mass flow rates on the evaporator and condenser sides). To avoid such a situation the control law enforces an upper bound on the requested load that corresponds to the maximum capacity values as specified in the IDF. the nonlinear solver produces a very small refrigerant flow.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Figure 115. If it is zero. This situation is dealt with by preventing from simulating the ECHILLER model altogether when the chiller is off. consistent with the no-load assumption. This results in a nonlinear system of algebraic equations that needs to be solved iteratively until convergence is achieved. 4/23/05 347 . then a hard-coded lower bound of 1 W is used. If the requested load is not attainable with the current operating conditions it will not be possible to make the residual control equation small enough to achieve convergence and the ECHILLER model will return with an error.

Vdots compressor. Break variable refrigerant_cycle. However. max ( Load . ( ) refrigerant_cycle. MinCap ) ) = 0 Solution methodology The SPARK model identifies three independent subsets of equations that are solved sequentially. Thus the theoretical speedup achieved with SPARK for this problem is of the order of (44/6) = 394 ! 4/23/05 348 12 3 . Initialized to 0 m3/s at the onset of the simulation.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CHILLERS Qev − min MaxCap.Qdot Cooling capacity (W).05 m3/kg at the beginning of each step. Then uses previous solution as predicted value. The first and last components in the solution sequence are weakly connected set of equations that are re-arranged so that they can be solved in a non-iterative manner by forward propagation.h3 compressor. the second component is a strongly connected set of nonlinear equations that must be solved in a iterative manner until convergence. The solution sequence derived by SPARK produces an efficient decomposition of the problem into smaller sub-problems.Wis evaporator. Then uses previous solution as predicted value. Evaporating temperature (K) of refrigerant fluid. To achieve this and to ensure numerical stability the following break variables have been selected among all possible candidates. compression Predicted value Initialized to 0. Then uses previous solution as predicted value. Volumetric flow rate (m3/s) through compressor. Initialized to the requested load (Load) at each step.12 The break variables need to be initialized with meaningful predicted values to ensure that the nonlinear solved finds a solution. Initialized to 280 K at the onset of the simulation. corresponds to a subgraph that depends only input variables and the unknowns of a previously solved component. Initialized to 100000 J/kg at the onset of the simulation. SPARK graph-analysis identifies a subset of so called break variables that breaks all cycles in the sub-graph for the second component. Each such subset. or component in the SPARK jargon. The relative tolerance is set to 10-4. List of break variables in ECHILLER model. To facilitate the prediction process it helps choosing variables for which we can easily specify predicted values.T1 refrigerant_cycle. Isentropic power (W).v1p Description Specific volume (m3/kg) of refrigerant fluid after heating-up. The resulting linear system is The computational time is approximately proportional to N whereby N is the number of simultaneous equations being 3 solved. Then uses previous solution as predicted value. This drastically reduces the computation time compared to a conventional solver. Initialized to 10000 W at the onset of the simulation. max (1. thus ensuring that the solution will be accurate to at least 4 significant digits. Specific enthalpy (J/kg) of refrigerant fluid at condenser exhaust. The original system of 44 simultaneous equations is reduced to an equivalent core system of 6 equations only. A finite-differences scheme is used to estimate the Jacobian matrix. Table 45. The strongly connected component is solved using the Newton-Raphson iteration method along with the line search backtracking step control strategy.

should be derivable from specific manufacturer's data. a coil model that predicts changes in air and water flow variables across the coil based on the coil geometry is required. equations are written for both the dry and wet regions of the coil. ma Ta1 Ha1 Ta2 Ha2 Ta3 Ha3 Ts1 Tw3 Tw2 Tw1 mw Figure 116. or it may have wet and dry sections. The UA parameter can be calculated from the parameters in the following table. a modular program also designed for energy analysis of building systems. Elmahdy and Mitalas state that crossflow counterflow coils with at four rows or more are approximated well by this model. For each region the heat transfer rate from air to water may be defined by the rate of enthalpy change in the air and in the water. The actual condition of the coil surface depends on the humidity and temperature of the air passing over the coil and the coil surface temperature. A greatly simplified schematic of enthalpy and temperature conditions in a counterflow cooling/dehumidifying coil is shown in the following schematic figure. This does not constitute a major limitation since cooling and dehumidifying coils typically have more than four rows. The partly wetpartly dry case represents the most general scenario for the coil surface conditions. in most cases. the variables required to model a cooling/dehumidifying coils and their definitions are extensively listed in “Table 46. The model solves the equations for the dry and wet sections of the coil using log mean temperature and log mean enthalpy differences between the liquid and the air streams. 4/23/05 349 . Coil Geometry and flow variables for coils”. 1985). The rates must balance between each medium for energy to be conserved. Simplified Schematic of Cooling/Dehumidifying Coil Heat Transfer and Energy Balance The cooling coil may be completely dry. Finally. Equations (319) through (322) express the energy balance between the water and the air for the case of dry and wet coils respectively. Equations (323) and (324) represent the heat transfer rate between water and air based on the actual performance of the coil. The all dry and all wet cases can be considered as limiting solutions of the wet or dry areas respectively going to zero. In the general case.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS solved using Gaussian elimination with affine invariant scaling. completely wet with condensation.Detailed Cooling Coil (HVAC) In order to provide this simulation capability. Coils Coil Model -. In addition. due to the high nonlinearity of the equations the solver also checks for bad numerics at each iteration and if such a situation is detected the relaxation coefficient is halved successively until the system of equations produces valid numerical values for the current iteration. The input required to model the coil includes a complete geometric description that. The coil simulation model is essentially the one presented by Elmahdy and Mitalas (1977) and implemented in HVACSIM+ (Clark.

effective diameter hydraulic diameter on air side dry region thickness spacing heat transfer film coefficient variable in fin eff. bb C1. Qd = ma Cpa (Ta . water side.2 − Tw. ord 0 mod Bessel fn. Ta.1 − Ta . 1st kind. ma . Tw. 3 Equations (319) through (324) represent two sets of three equations with 7 unknowns: Qd . Tw. ord 1 mod Bessel fn. 2nd kind. * wet area variables in fin eff. air side coeff.3 ) Qw = mwCpw (Tw. ord 1 inside (water side) variable in sol'n form of eq. outside of metal Stanton number temperature water tube dry heat xfer coeff. 2.2 − H a . so that the system of equations is effectively closed. water flow rate. ue V w wa Z 1. thermal conductivity length LMHD LMTD log mean enthalpy difference log mean temperature difference mass flow rate metal and fouling viscosity outside (air side) Prandtl number heat transfer rate overall thermal resistance Reynolds number ratio of diameters surface. or wet region humidity ratio variables in sol'n form of eq. in enthalpy approximation coeff. for example: inlet water temperature.3. 2nd kind. C2 Cp D Dhdr D δ ∆ F Fai fin.1.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Table 46. normally at least four of these variables are specified.2. mw .3 − Tw. However.2 ) Qd = (UAdry ) ( LMTD ) Qw = ma ( H a . calculation air side fin geometry enthalpy efficiency mod Bessel fn. * dry area wet heat xfer coeff. in air side film coeff. 1st kind. Ta. fins H η I0() I1() K0() K1() I K1 K L area air. calculation average velocity water. air flow rate. positions (see diagram) m mf µ o Pr Q R Re ρ s St T tube UAdry UcAw ub. specific heat diameter.2. Coil Geometry and flow variables for coils A A aa. ord 0 mod Bessel fn.1 ) 4/23/05 (319) (320) (321) (322) (323) 350 . outlet air temperature.2 ) Qd = mwCpw (Tw.

2 ln a .2 = ⎝ mwCpw − (1 − Z ) K1 Cpa bb − Z ma ⎛ mwCpw ⎞ ⎟ ma ⎠ (328) An alternative solution method is to define the coil leaving air temperature as an input with a variable water flow rate. In this case Equations (329) and (330) are more convenient.1 ) + Z Tw.3 − H w.1 − aa − K1 Cpa Ta . (1 − Z ) ( H a .3 Ta .2 − H w. a linear approximation of the enthalpy of saturated air over the range of surface temperature is made using Equation (327). Finally.1 − Tw.2 ) T −T ln a .2 − Tw.1 w. the log mean temperature and enthalpy differences are expanded as shown in Equations (325) and (326). (1 − Z ) ( H a .3 ) − (Ta . the coil does not have a controlled outlet air temperature.2 = ⎛ mwCpw ⎞ − bb Z ⎟ ⎝ ma ⎠ mwCpw − bb ma (329) ( Z d − 1) Ta1Cpa + Tw.1 ) H − H w.1 ⎜ Tw. Equations (331) through (333) define terms that are used to simplify Equations (328).3 ⎜ Cpa − Z d Tw.2 (326) H w = aa + bb Tw (327) Equation (328) is derived from the above equations and is used to solve for the coil conditions when all of the inlet conditions are given as input.2 H a .3 − aa ) + Tw.2 ) − ( H a .1 ⎜ bb − Tw.2 − Tw. (329) and (330).2 = ⎝ ⎛ m Cp ⎞ Z d ⎜ Cpa − w w ⎟ ma ⎠ ⎝ ⎛ mwCpw ⎞ ⎟ ma ⎠ (330) 4/23/05 351 .3 − H w.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Qw = (U c Aw )( LMHD ) (324) In order to manipulate these equations.1 (325) LMHD = (H a . Note that in Equation (326) Hw refers to the enthalpy of saturated air at the water temperature. LMTD = (T a . Operating in this manner.

The definition of overall inside thermal resistance follows directly as shown in Equation(335). The coefficients in Equation (336) and (337) are calculated by Equations (339) and (340) that are functions of the coil geometry.2 (334) This is valid for Reynolds numbers greater than 3100 based on water flow velocity and pipe inside diameter and is given in Elmahdy and Mitalas (1977) as recommended in the standard issued by the Air-Conditioning and Refrigeration Institute (1972) for air-cooling coils.63 ×10−7 Rea ) −0. the air side film coefficient is modified according to Equation (338). is valid for Reynolds numbers between 400 and 1500. and Assumptions (333) Overall heat transfer coefficients are calculated from the specified coil geometry and by using empirical correlations from fluid mechanics and heat transfer. Ri = 1 fi Ai (335) Equation (336) gives the film coefficient for the air side. Another form of the same equation is Equation (337).425 − 5. w = f o (1. For the water side. which is familiar from the data presented in Kays and London (1984).1× 10−4 Rea + 2.141 (339) 352 .065 (337) (338) ⎛ δ fin ⎞ C1 = 0.429 (1 + 0. Equation (334) gives the film heat transfer coefficient in SI units: fi = 1.159 ⎜ ⎟ ⎝ Dhdr ⎠ 4/23/05 ⎛ δ fin ⎜ ⎜L ⎝ fin ⎞ ⎟ ⎟ ⎠ 0.0146 Tw ) Vw0. a function of air Reynolds number. Equations (341) through (344) show definitions and values of common parameters and properties. The correction term. For coil sections that have a wet surface due to condensation. Properties.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS ⎛ ⎛ 1 bb ⎞ ⎞ Z = exp ⎜ U c Aw ⎜ − ⎟⎟ ⎜ ⎟ ⎝ ma mwCpw ⎠ ⎠ ⎝ K1 = Zd −1 m Cp Zd − a a mwCpw (331) (332) ⎛ ⎛ 1 1 ⎞⎞ Z d = exp ⎜ U c Adry ⎜ − ⎟⎟ ⎜ ⎟ ⎝ ma Cpa mwCpw ⎠ ⎠ ⎝ Underlying Correlations. C f o = C1 Rea 2 ma Aa_min_flow 2 Cpa Pra 3 2 (336) C C1 Rea 2 = Sta Pra 3 2 f o .8 Di−0. Elmahdy (1977) explains the modifier for the wet surface and coefficients for the film coefficient.

the fin efficiency must also be considered in calculating the overall thermal resistance on the outside. η fin = −2 ρ ⎡ I1 (ub ) K1 (ue ) − K1 (ub ) I1 (ue ) ⎤ ⎢ ⎥ fai (1 + ρ ) ⎣ I 0 (ub ) K1 (ue ) + K 0 (ub ) I1 (ue ) ⎦ (345) fai = ( D fin − Dtube ) 2 k finδ fin 2 fo (346) ρ= Dtube D fin (347) ue = fai 1− ρ (348) ub = ue ρ (349) ηo = 1 − (1 − η fin ) Afins As _ total (350) The definition of overall outside thermal resistance is given in Equation (351) as a function of fin efficiency and film coefficient.549 ⎛ δ fin ⎜ ⎜∆ ⎝ fins ⎞ ⎟ ⎟ ⎠ −0.323 ⎜ ⎜ L ⎝ fin ⎞ ⎟ ⎟ ⎠ 0.028 (340) Dhdr = 4 Aa_min_flowδ coil As _ total 4δ coil (1 + wa ) ma As _ total µa (341) Rea = (342) Pra = 0.that of circular fins with the same fin area -.846 × 10−5 The film coefficients above act on the extended surface of the air side. Note that the efficiency is found by the same equations for the wet surface using the wet surface film coefficient. The overall efficiency of the surface is shown by Equation (350). To model a coil with flat fins. This equation is based on circular fins of constant thickness.is used. For a wet coil surface the resistance must be defined 4/23/05 353 . that is the area of the fins and the tubes.049 ⎛ D fin ⎜ ⎜∆ ⎝ tube _ rows ⎞ ⎟ ⎟ ⎠ 0. Therefore. Gardner (1945) gives the derivation of Equation (345).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS ⎛ ∆ fins C2 = −0.733 (343) (344) µa = 1. Equations (346) through (349) define variables used in Equation (345). used as a curve fit to find the fin efficiency as a function of film coefficient. an effective diameter -.

total ⎢ Ri + Rmf + Ro ⎥ ⎣ ⎦ Aw As . wηo .total Cpa (352) Equation (353) gives the last two overall components of thermal resistance. For the controlled coil.total ⎡ ⎤ 1 bb ⎢ ⎥ ⎢ Ri + Rmf + Ro . The fouling factor. was the motivation for this approach. The coil is solved as if it were all wet before the general case is attempted. and the water flow rate and boundary position are adjusted until the two equations agree. the TYPE12 subroutine. no further solution is required. The MODSIM implementation of a cooling coil. w ⎥ ⎣ ⎦ (354) U c Aw = (355) Solution Method of Model The complicated equations derived above were implemented in a successive substitution solution procedure to calculate the coil performance based on the input parameters. Several simple equations in the loop adjust the boundary point until the dry surface temperature at the boundary is equal to the dew point of the inlet air. Special cases occur when the coil is all wet or all dry. In the solution of the controlled coil the outlet air enthalpy.total (351) Ro . the cooling coil is only partially wet. All components of thermal resistance are added in series to produce the overall heat transfer coefficients shown in Equations (354) and (355). due to deposits of dirt and corrosion of the tube inside surfaces. Equation (328) is used to find the water temperature at the boundary. 4/23/05 354 . the general solution is used to calculate the unknowns. to ensure a continuous solution as flow variables are changed. However. Rmf = δ tube ktube Ai + Fl Ai (353) UAdry = ⎤ Adry ⎡ 1 ⎢ ⎥ As . must correspond to the enthalpy at the specified outlet air temperature. Clark (1985) contains notes about the MODSIM routine. If the wet surface temperatures at the coil inlet and outlet are both below the dew point. Ro = 1 f oηo As . They represent the metal tube wall and internal fouling. as shown in Equation (352). is assumed to be 5x10-5 m2·K/W.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS differently because the heat transfer equations are based on enthalpy rather than temperature differences. For an uncontrolled coil. Equations (329) and (330) give two calculations of the boundary temperature. given some resulting dehumidification. w As . In the general case. w = bb f o . when the surface is all dry or when it is wet with only the dry surface equations yielding a surface temperature below the dew point at the water outlet. the method used there has been retained with modifications for the uncontrolled coil model.

Some simplifications can be made.78 ⎪ ⎜ ⎪exp ⎢ − ⎜ ⎟ { NTU } ⎥ − 1⎬ ⎟ ⎨ ⎜ max {C p .out = Ta . however. the product of heat transfer coefficient and coil area. it was concluded that. w mw } ⎟ ⎜ ⎟ ⎝ ⎠ ⎝ ⎠ (357) The parameter NTU is the number of transfer units and is defined as a function of the UA value of the coil as follows: NTU = UA min ( Cp. In order to allow heating and cooling coils to be specified using the same geometric parameters. a coil model that predicts changes in air and water flow variables across the coil based on the coil geometry is required. There is no condensation in a heating coil. Fin spacing etc. C p . w mw } ⎞ ⎤ ⎫⎞ 0. The coil simulation model is essentially the modification of one presented by Elmahdy and Mitalas 4/23/05 355 . a ma .out ⎛ C p .22 ⎜ ⎟ ⎜ ⎟ { NTU } ⎜ max {C p .in + (Tw. since the enthalpy change of the air flowing over a heating coil is entirely sensible. a heating coil simulation was developed from the cooling coil model described above by eliminating the wet surface analysis. In addition. an option was provided in IBLAST for a heating coil design using only the UA value of the coil.in − Tain ) ε ⎜ ⎜ ⎟ C p . a ma . The same kind of loops exist to provide hot water to the heating coils from the boilers as exist to supply the cooling coils with chilled water from the chillers. C p . C p . w mw ) ⎞ ⎟ Ta . A good simulation program should ask the user bare minimum number of inputs keeping into account accuracy of the model.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Application of Cooling Coil Model to Heating Coils The implementation of detailed heating coil models in IBLAST was another important aspect of the system/plant integration.w w ⎠ (356) where the coil effectiveness is given by: ⎛⎧ ⎡ ⎛ min {C p . This model was largely based on the TYPE10 subroutine implemented in MODSIM. w mw } ⎞ −.out = Tw.a ma ⎝ ⎠ Tw.w m w ) (358) Coil Model – Water Cooling Coil (HVAC) In order to provide this simulation capability. but for the user to input all the details of the coil geometry: Area. The input required to model the coil includes only a set of Thermodynamic design inputs. since much simpler and less computationally expensive heating coil simulations are possible. The equations used to model the performance of the TYPE10 heating coil are as follows: ⎛ min ( C p . Number of tubes.in − (Ta . w mw } ⎟ ⎜⎪ ⎢ ⎝ ⎥ ⎪⎟ ⎠ ⎣ ⎦ ⎭⎟ ⎜⎩ ε = 1 − exp ⎜ ⎟ ⎛ min {C p .Cp.a m a .a ma . A greatly simplified schematic of enthalpy and temperature conditions in a counter flow cooling/dehumidifying coil is shown in the following schematic figure 1. C p . hence a model is required which is able to convert the large number of coil geometric details into a small set of auto sizable thermodynamic inputs. is a very cumbersome task especially in early stages of simulation. C p .a ma .a ma .a ma ⎞ − Tain ) ⎜ ⎜C m ⎟ ⎟ ⎝ p. which require no specific manufacturer's data.

The rates must balance between each medium for energy to be conserved. the conditions at which the coil operates. 1970. The part-wet part-dry case represents the most general scenario for the coil surface conditions. completely wet with condensation. Design Inputs (User Inputs) Input Field DesWaterVolFlowRate: DesAirVolFlowRate: DesInletWaterTemp: DesInletAirTemp: DesOutletAirTemp: DesInletAirHumRat: DesOutletAirHumRat: 4/23/05 Description Maximum Water Volume Flow Rate Maximum Air Volume Flow Rate Inlet Water Temperature at Design Condition Inlet Air Temperature at Design Condition Outlet Air Temperature at Design Condition Inlet Air Humidity Ratio at Design Conditions Outlet Air Humidity Ratio at Design Conditions. and a subroutine that iterates between the dry and wet subroutines to calculate the fraction surface wet of the coil. which is widely applicable in HVAC systems and the second being Counter flow mode. however the execution time for detailed mode is noticeably higher. There are subroutines present in the model for both the dry and wet regions of the coil. The operating block is the one containing the operating conditions. J. Following is the list of Design and Operating inputs and subsequently the Design and Operating variables used in the model. There are two modes of flow operation for this model: Cross Flow. TRNSYS. For each region the heat transfer rate from air to water may be defined by the rate of enthalpy change in the air and in the water. Table 47. Simplified Schematic of Cooling/Dehumidifying Coil Heat Transfer and Energy Balance The cooling coil may be completely dry. This block calculates the Design UA values required by the model. 356 . The default value in program is set up for Cross Flow. Model Description The Model has two blocks: 1st = Design Block with the Design Inputs. Wet and Part Wet & Part Dry Coil and iterates between the Dry and Wet Coil to output the fraction wet. or it may have wet and dry sections. ma Ta1 Ha1 Ta2 Ha2 Ta3 Ha3 Ts1 Tw3 Tw2 Tw1 mw Figure 117. The actual condition of the coil surface depends on the humidity and temperature of the air passing over the coil and the coil surface temperature. The default input for the Analysis is SimpleAnalysis. The model calculates the UA values required for a Dry. Using these UA values the model simulates the operating conditions. 1990 and Threlkeld. In addition the coil has two modes of analysis: Simple Analysis and Detailed Analysis.L.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS (1977). The Simple analysis mode operates the coil as either wet or dry while the detailed mode simulates the coil as part wet part-dry.

Calculating Coil UA Enthalpy Based: Heat Transfer in Wet Coil is based on Enthalpy rather than temperature to take into account latent effects.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Table 48. The aim of the Design Block is to calculate the Coil UA for use in the operating Block. Design Block Calculations: The design block has the code for calculating the six Coil UA values required by the operating block. The model is basically divided into Two Blocks: the Design Block and the Operating Block. Q = DesUACoilTotalEnth * (H1-H2). Where DesUACoilTotalEnth = CoilUA / cp and CoilUA is the conventional heat transfer coefficient and cp = specific heat across enthalpy difference. Water )Capacity *(DesInletAirEnth-DesSatEnthAtWaterInTemp) • (359) 4/23/05 357 . Following are few important calculations to understand the working of the model. The various UA are: Table 49. which are explained later in the document.not user inputs) Condition Variable InletWaterMassFlowRate: InletWaterTemp: InletAirMassFlowRate: InletAirTemp: InletAirHumRat: Description Entering Water Mass Flow Rate at operating condition Inlet Water Temperature at operating condition Entering Air Mass Flow Rate at operating condition Inlet Air Temperature at operating condition Entering air humidity ratio at operating conditions Intermediate calculated UA’s: The Crux of the Model The various UA values required by this model are calculated from the above inputs. and a temperature difference. While heat transfer rates are commonly expressed as the product of an overall heat transfer coefficient. UA Descriptions of Model UA Variable Name Description UACoilTotal: Overall heat transfer coefficient (W/C) UACoilInternal: Overall internal UA (W/C) UACoilExternal: Overall external UA (W/C) UACoilInternalPerUnitArea: Internal overall heat transfer coefficient (W/m2 C) UAWetExtPerUnitArea: External overall heat transfer coefficient (W/m2 C) UADryExtPerUnitArea: External overall heat transfer coefficient (W/m2 C) The UA values are calculated assuming Wet Coil at Design Condition. Q Maximum = Min ( Air . We use the following two heat transfer equations for calculating the DesUACoilTotalEnth. The Design Block is a One Time Calculation and in an Hourly simulation the code is skipped subsequently. UA. Reasonable assumptions have been made in the calculations to maintain the simplicity of the model. Operating Conditions (From Nodes -. The value of Q is calculated using product of air mass flow rate and difference in inlet and outlet air enthalpies at design conditions. the use of enthalpy-based heat transfer calculations requires an enthalpy-based heat transfer coefficient which is defined as = DesUACoilTotalEnth and hence the equation.

The above equation (0. DesSatEnthAtWaterInTemp = PsyHFnTdbW(DesInletWaterTemp.equation (362) and the initial guess value of DesAirTempApparatusDewPt till convergence is achieved. We now calculate the value of Coil ByPass Factor based on enthalpies of the stream at various corresponding temperatures. which is the airside coefficient. 4/23/05 358 . as follows. The Enthalpy Based UA is calculated as shown in equation (359) and (360) and airside coefficient which is the Coil UA external is calculated in equation (364) and (365). Calculating Coil UA External: The procedure for calculating the variable Coil UA External is best described by the familiar Fortran programming -. The initial guess value of the above variable is DesAirTempApparatusDewPt = PsyTdpFnWPb(DesOutletAirHumRat. IF (DesAirApparatusDewPtEnth ≤ DesSatEnthAtWaterInTemp) THEN CoilUA External = DesCoilUA EnthalpyBased * Cp Air ELSE (364) CoilUA External = − Log ( ByPassFactor ) *(Cap Air ) (365) END IF Where Cap_Air is the product of Design Air Mass Flow rate and Specific Heat Capacity of air and Cp_air is the specific heat capacity of air at particular condition. External UA. and W=Humidity Ratio (363) Iterate between -. DesByPassFactor = DesOutletAirEnth-DesAirApparatusDewPtEnth DesInletAirEnth-DesAirApparatusDewPtEnth (361) The variable DesAirApparatusDewPtEnth is calculated from DesAirTempApparatusDewPt which is intern calculated by the following procedure as shown below.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • COILS Q Maximum = CoilUAEnthalpyBased *(DesInletAirEnth-DesSatEnthAtWaterInTemp) (360) Iterate on equation (359) and (360). Calculating Coil UA Internal: The liquid-side coefficient. DesHumRatioAirInlet) Psy**Fn** are the Psychrometric functions from Energy Plus for calculating Psychrometric properties of air.If Then Else block. on the two value of maximum Heat Transfer till convergence achieved to calculate the value Coil UA Enthalpy Based.e. The variables used in above equations are calculated as below. i. is calculated assuming that surface temperature is at apparatus dew point. UA Coil Internal is calculated from enthalpy-based overall coefficient and airside coefficient.1) and (0. Temp ApparatusDewPtEstimate = DesInletAirTemp − Slope × (W AirIn − W ApparatusDewPt ) Where Slope = ∗ ∗ (362) DesInlet AirTemp − DesOutlet AirTemp DesInlet AirHumRatio − DesOutlet AirHumRatio .. Patm) Where the variables used have usual meaning.2) give the value for Coil UA Enthalpy Based.

Counter − Flow (369) All the variables in the above equation are calculated from design conditions. similar to thermal resistances. Hence the equation for UA Coil Total is as follows. In the case of our coil model. The variable CpSat_Intermediate is an intermediate fictitious specific heat and is estimated using entering air dew point and water temperature. Calculating Overall Heat Transfer Coefficient: The overall heat transfer coefficient is calculated from a combination of series or parallel resistances. Calculating Coil UA Internal/Unit area of Coil: Internal overall heat transfer coefficient (W/m2 C) is approximated as overall internal UA divided by Total surface area of coil. Coil UA External. CoilUA Total = 1 1 1 − CoilUAInternal CoilUAExternal (368) In order to calculate the value of Heat Transfer coefficients per unit area we need to get a good approximate for the coil area required. The default set up in the program is for cross flow since most air-conditioning applications are in that mode. total UA is a parallel combination of External and Internal UA values which are calculated in equations (364) to equation (366). Coil UA Enthalpy Based have already been calculated and are know from previous equations. the variable LMTD can be calculated for Cross and Counter flow mode of coil depending on choice. The calculation is shown below. CoilUAInternalPerUnitArea = CoilUAInternal Area (370) Calculating Dry UA External/Unit area of Coil: The value of Dry UA external is calculated from rearranging the equation (371) 4/23/05 359 . CpSat Intermediate = EnthAirDesDewPo int − EnthSat DesWaterInTemp DewPt DesTempAir − TempWaterDesInlet (367) The variables used in the above equation are calculated from known and earlier calculated temperatures.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS CoilUAInternal = CpSat Intermediate Cp Air 1 − DesCoilUAEnthalpyBased CoilUAExternal (366) In the above equation the variables Cp_air. This is achieved by using the simple heat transfer correlation QDes = CoilUATotal × CoilArea × LMTDCross . using the psychrometric functions in Energy Plus.

These inputs are automatically generated from the node connections in Energy Plus. WetUAExternalPerUnit Area = CoilUA Internal Area (373) The Design Block calculations end at this part. They are the Simple analysis mode and the detailed analysis mode. Part Wet Part Dry Mode: This is the usual/frequent mode of operation of coil. The operating block requires 5 inputs. The operating block for Detailed Mode Analysis of this coil model is divided into three modes of coil performance. We get the following equation (372) for the Dry UA external per unit area for the coil: DryUAExtPerUnit Area = CoilUATotal × CoilUAInternal Area × CoilUAInternal − CoilUATotal (372) Calculating Wet UA External/Unit Area: Since the Coil UA’s are calculated assuming it is Wet Coil at design condition.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS DryUA = Area 1 DryUAExtPerUnitArea 1 − CoilUAInternal (371) Rearranging equation (371) and making a reasonable assumption that DryUA of Coil is equivalent of the Coil UA Total for this model. we make a reasonable assumption as in equation (373) to calculate the value of Wet UA External per unit area. on the other hand the detailed mode checks for part wet part dry mode of operation and reports surface area wet fraction of coil. The simple analysis mode assumes the coil to be either all wet or either all dry and execute the model . Operating Block Calculations: There are two modes of coil analysis in the operating block. This subroutine iterates between the Dry Coil and the Wet Coil to give outputs. as shown in figure 1. Coil is completely wet: The entire coil is wet due to complete condensation on the surface of the coil. which calculates all UA based on precise geometric inputs. The Part Wet Part Dry Mode of operation is essentially a function the Coil Completely Dry and Coil Completely Wet mode. the outlet states of water and air given the inlet conditions and the Heat Transfer: sensible and total. a detailed explanation is given later in the document. This is an extreme condition when the entering air has very low humidity ratio or is dry air. 4/23/05 360 . The modes being Coil is completely dry: There is no moisture condensation on the coil surface and the coil is a dry coil. The user does not have to input any information to run this coil model. we now have all the essential UA values to proceed with the 5 operating conditions to calculate the coils performance. however the program execution time in detailed mode is noticeably higher. where part of the coil at entry of air is dry and as air cools condensation occurs and part of the coil becomes wet. This assumption gives remarkably close values to what is expected if detailed geometric variables are used to calculate the same wet coil UA per unit area. This assumption yields reasonably close results to the detailed coil. which are mentioned earlier in the document. Hence the design block in the coil model is essentially being utilized to calculate the coil UA so that the operating block could use those values of UA.

In this case outlet temperature of air would be higher than the air dew point and hence there would be no condensation. IF (Temperature Dew Point Air < Water Inlet Temperature) THEN the coil is Dry and we call the Subroutine Coil Completely Dry. the two modes being Coil is completely dry: There is no moisture condensation on the coil surface and the coil is a dry coil. it is assumed that moisture condensation occurs over completely surface of the coil. IF (Temperature Dew Point Air > Water Inlet Temperature) THEN the coil is completely wet.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS The option to identify which mode of operation the coil should perform ie. IF (AirDewPointTemp < AirInletCoilSurfTemp) THEN. IF (Temperature Dew Point Air < Water Inlet Temperature) THEN the coil is Dry and we call the Subroutine Coil Completely Dry. the coil is Partially Wet because there is possibility that air temperature will go below its dew point and moisture will condense on latter part of the cooling coil. for a given set of inputs would the coil be Dry or Wet is decided by set of conditions described below. However we go ahead and check for the coil being partially wet with the following condition. Effectiveness Equations: There are two modes of flow for the coil. default set up is as cross flow since most air condition applications have cross flow heat exchangers. Counter Flow mode or the Cross Flow mode. it is assumed that moisture condensation occurs over completely surface of the coil. Counter Flow Heat Exchanger: Effectiveness Equation: ηCounterFlow = (1 − Exp(− NTU × (1 − RatioStreamCapacity ))) 1 − RatioStreamCapacity × Exp(− NTU × (1 − RatioStreamCapacity )) (374) In Equation (374) the variable Ratio_StreamCapacity is defined as below 4/23/05 361 . In this case outlet temperature of air would be higher than the air dew point and hence there would be no condensation. The above is a simple mode of analysis and the results are very slightly different from the detailed mode of analysis. According to the mode of flow the following NTU . Coil is completely wet: The entire coil is wet due to complete condensation on the surface of the coil.Effectiveness relationships are used to calculate coil effectiveness. Part Wet) for calculating air outlet conditions and heat transfer. Following are the relations used for calculating effectiveness equation for the Heat exchangers. Wet. The program defaults to simple mode of analysis for enabling higher execution speed. which is used later by all the three modes (Dry.0 for wet or dry coil respectively. call subroutine Coil Completely Wet. Wet or Part Wet Part Dry. However we go ahead and check for the coil being partially wet with the following condition. The option to identify which mode of operation the Simple mode analysis should perform ie. for a given set of inputs would the coil be Dry. call subroutine Coil Completely Wet. The surface area wet fraction in the coil is reported as 1. is decided by set of conditions described below. IF (Temperature Dew Point Air > Water Inlet Temperature) THEN the coil is completely wet. The Operating Block for Simple Mode Analysis is divided into two modes of coil performance.0 or 0. The algorithms used in Simple mode and the Detailed mode are identically similar.

Wet.22 RatioStreamCapacity × NTU ⎪ ⎪ ⎩ ⎭ (378) The variables in the above equation have already been defined earlier. as in equation (382) and (383). Temperature of air and water at outlet to the coil is given as in following equations TempAirOut = TempAirinlet − ηcross . Depending on the mode of operation of the coil model the cross or the counter flow equations are used to calculate the effectiveness. which in our case is defined as CoilUA_total. is defined as the Number of Transfer Units. i. Max) CapacityStream = ( MassFlowRate × Cp ) air . Part Wet) is calling upon equation (374). Equation (377) gives definition for NTU. The Coil UA is a variable in this equation and depends on which mode of the coil operation (Dry. the sensible load is equal to latent load and the same with the humidity ratios at inlet and outlet. QSensibleDryCoil = QTotalDryCoil 4/23/05 (382) 362 .counter × (380) In the above equations (379) and (380) the maximum heat transfer is calculated as shown in the following equation MaxHeatTransfer = MinStreamCapacity × (TempAirInlet − TempWaterInlet ) Coil Completely Dry Calculations: (operating block) (381) Since the coil is dry.22 ⎧ ⎫ ) − 1⎪ ⎪ Exp(− NTU × RatioStreamCapacity × NTU ⎬ −0. The energy difference between the outlet and inlet stream conditions gives the amount of heat transfer that has actually take place..e. if it is Coil Completely Dry calling upon the effectiveness equation with the value of Dry UA total. Coil Outlet Conditions: Calculating the Outlet Stream Conditions using the effectiveness value from equation (374) or (378) depending on the mode of flow.water (376) NTU: equation (374).counter × MaxHeatTransfer StreamCapacity Air MaxHeatTransfer StreamCapacityWater (379) TempWaterOut = TempWaterInlet + ηCross . it is a function of Coil UA and the Minimum Capacity of Stream.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS RatioStreamCapacity = MinCapacityStream MaxCapacityStream (375) In equation (375) capacity of stream is defined as below in equation (376) ( Min. NTU = CoilUA MinStreamCapacity (377) Cross Flow Heat Exchanger: Effectiveness Equation: ηCrossFlow = 1 − EXP ⎨ −0.

Similar to equations (379) and (380) we calculate the air outlet enthalpy and water outlet enthalpy ie by replacing temperature with enthalpy of the respective streams. hence calculations are done using enthalpy of air and water instead of stream temperatures Hence we need to define coil UA for the wet coil based on enthalpy of the operating streams and not design streams. η = 1 − Exp ⎨− 4/23/05 ⎧ CoilUAExternal ⎫ ⎬ ⎩ Capaci tan ceAir ⎭ (388) 363 . The input variable for Coil UA in equation (377) for calculating NTU. and outlet temperature is a function of outlet air enthalpy as below OutletAirTemp = PsyTdbFnHW (EnthalpyAirOutlet. OutletAirHumRat) and OutletAirHumRat = InletAirHumRat ENDIF Effectiveness eta used in equation (387) is defined in equation (388) and Condensation Temperature is calculated using Psychrometric function as in equation (389) and (390).EnthAirOutlet) ELSE There is no condensation and hence the inlet and outlet Hum Ratios are equal .Patm)) THEN AirTempOut = AirTempinlet − ( AirTempinlet − CondensationTemp ) ×η (387) and OutletAirHumdityRatio = PsyWFnTdbH(OutletAirTemp. IF (TempCondensation < PsyTdpFnWPb(InletAirHumRat . in this case it would be enthalpy based and is given as shown in equation (385) CoilUA EnthalpyBased = 1 CpSatint ermediate Cp Air ( ) + UACoilInternal UACoilExternal (385) Total Coil Load in case of Wet Coil is the product of mass flow rate of air and enthalpy difference between the inlet and outlet streams as given in the following equation QTotal = M air × ( EnthAirInlet − EnthAirOutlet ) ∗ (386) Once the enthalpy is known the outlet temperatures and outlet humidity ratios of the wet coil are calculated as in equations below.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS HumRatioInlet = HumRatioOutlet Total Heat Transfer in dry coil is as follows: (383) QTotalDryCoil = CapacityAir × ( AirTempIn − AirTempOutlet ) (384) The variables in the above equation are calculated in the earlier in equations (379) and (380) to give the total cooling load on the coil. Coil Completely Wet Calculations: (operating block) In wet coil we need to account for latent heat transfer.

Patm) EnthAirCondensateTemp = EnthAirInlet − ( EnthAirInlet − EnthAirOutlet ) (389) η (390) Once the air outlet temperature are known. Iterate between the Dry Coil and the Wet Coil. First calculate Coil Completely Dry performance by estimating the wet dry interface water temperature using equation (392) and inputting this variable as the water inlet temperature to dry Coil.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS CondensationTemp = PsyTsatFnHPb(Enth AirCondensateTemp . which are calculated as below. where the Wet part area is the product of Surface fraction Wet and Total Coil Area. WetDryInterfaceWaterTemp = WaterTemp Inlet +Area WetFraction *(WaterTempOutlet -WaterTemp Inlet ) (392) The value of Surface Area Wet fraction is estimated initially as follows Area WetFractionEstimate = AirDewPt Temp − InletWaterTemp OutletWater Temp − InletWaterTemp (393) For the above mentioned iteration the value of Coil UA for Wet and Dry part need to be varied according to the new respective area of the wet and dry parts. DryUA external = CoilUADry SurfArea 1 ExternalPerUnitArea Dry 1 + CoilUAInternal PerUnitArea (394) Where Surface Area Dry =(Total Coil Area – Wet Part Area). This estimate of Wet and Dry area is a product of the estimated Surface Area Fraction and total coil external area. based on the area of the dry and wet part respectively. WetPartUA External = CoilUAWet ExternalPerUnitArea × SurfaceAreaWet (395) 4/23/05 364 . UA value for Dry part of the Coil is estimated as below. This problem is solved utilizing the fact that the Exit conditions from the Dry Part of the Coil would become the inlet conditions to the wet part of the coil (Figure-1) and the coil model determines by iteration what fraction of the coil is wet and based on that it calculates the areas and subsequently the UA values of that dry and wet part. Explained below are the steps followed to the estimating the wet dry behavior of the coil. as in equation (391) QSensible = Capaci tan ceAir × ( AirTempInlet − AirTempOutlet ) Coil Part Wet Part Dry Calculations: (operating block) (391) The Coil would perform under part wet part dry conditions when Air Dew Point Temperature is less than Coil surface temperature at inlet to air. UA value for the Wet part of the Coil requires Wet UA external and Wet UA Internal. then sensible load is calculated as a product of capacitance of air and temperature difference at inlet and outlet. which keeps varying as will be explained further in the document. In this case part of the coil used value of Dry UA for heat transfer and part the coil used Wet UA value for heat transfer.

WetDryInterfcAirTemp. and in the Wet Coil the Outlet Air Temperature from dry Coil is the inlet air temperature to Wet Coil. The key is to have the difference between the variables (WetDryInterfcWaterTemp – OutletWaterTemp) in Dry Coil equal to (InletWaterTemp-EstimatedWetDryInterfcWaterTemp) in Wet Coil. & EstimateSurfAreaWetFraction. All values of variable used have been normalized. which is the input to Dry Coil. & WetCoilTotalHeatTransfer. 4/23/05 365 . WetDryInterfcSurfTemp) Iterate Between the above two Wet and Dry Coil calls until the two variables in blue ie WetDryInterfcWaterTemp = EstimateWetDryInterfcWaterTemp. After the above convergence check for the coil being dry otherwise iterate to calculate surface fraction area wet. WetDryInterfcHumRat WetPartUAInternal. OutletAirHumRat. Graphs Showing the Performance of the coil model at optimum operating conditions are shown below.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS WetPartUA Internal = CoilUAWet InternalPerUnitArea × SurfaceAreaWet (396) It is essential to remember that the mode of calculation for the coils remains the same as in completely wet and completely dry mode. & DryCoilHeatTranfer). WetDryInterfcAirTemp. Start with the initially guess value of surface area fraction (equation (393)) wet and iterate on the entire loop starting from equation (393) until the Wet Dry Interface Temperature equals the Air Dew Point Temperature. The value of Surface Area fraction wet at which the interface air temperature equals is dew point is the transition point from wet to dry and gives the % of coil that is dry and % that is wet.0) and (WetDryInterfaceSurfTemp > AirDewPt )} (397) THEN CoilCompletelyDry If Equation (397) is satisfied then Coil is Dry and simply output the value for Dry Coil calculated else the coil is partially wet and then iterate to find the surface fraction area wet. Now Iterate between the Dry Coil and wet Coil with the above respective UA. WetDryInterfcHumRat. DryCoilUA. IF {( AreaFraction Wet ≤ 0. and usual operating inputs except the variable water inlet temperature for dry Coil is replaced with Wet Dry Interface Water temperature. InletAirTemp. The iteration proceeds till the Outlet Water Temperature from Wet Coil equals the Wet Dry Interface Water Temp. air outlet and inlet values change. & EstimateWetDryInterfcWaterTemp. only the UA values and water.WetPartUAExternal. WetCoilSensibleHeatTransfer. CALLCoilCompletelyWet (InletWaterTemp. OutletAirTemp. Input the calculated values calculated by Dry Coil above into Wet Coil below. This equality quantized the relative part of coil that is dry and part that is wet on the basis of heat transfer that has occurred. Dry Part Inputs: (changed operating inputs) :Iteration Case 1: Explained In Programming Fashion: CALL CoilCompletelyDry (WetDryInterfcWaterTemp. & OutletWaterTemp. The variables have been highlighted in color red and blue.

2 0.7 0.1 0.8 Figure 119.3 0.4 0.3 0.6 0.6 0.4 Air Mass Flow Rate 0.7 0.3 0 0.5 0.6 Figure 118. Air Outlet Temperature Vs Air Mass Flow Rate 1 QSensible 0.6 0.1 Air Outlet Temp Air Outlet Temperature TA irLvgNewM o del Air Mass Flow Rate 0.5 0. Sensible Load variations Vs Air mass Flow Rate 4/23/05 366 .4 0.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS 0.2 0.45 0.55 0.5 0.5 0.65 0.75 0.3 0.35 0.9 0.4 0.8 0.7 Sensible Heat Transfer Rate QSensibleNewM o del 0.8 0.

8 1 Air Mass Flow Rate Figure 121. M& IE Dept UIUC.8 FWetNewM o del 0.2 0 0 0. Richard J Liesen M&IE . Rahul J Chillar.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS 1 Total & Sensible Heat Transfer Rate: Modified Model 0. W. Colarado Boulder: An Improvement of Ashrae Secondary HVAC toolkit Simple Cooling Coil Model for Building Simulation.6 QTo tSimNewM o del QSensibleNewM o del 0.6 0. 4/23/05 367 .2 Fraction Wet Dry: 1 Fraction Wet Dry 0.UIUC.4 0.6 0.8 Heat Transfer Rate 0.4 0.2 Air Mass Flow Rate 0 0 0.4 0.4 0. Surface Area Fraction Wet Vs Air mass Flow Rate References IBPSA BuildSim-2004. Total and Sensible Load variations Vs Air mass Flow Rate 1.6 0.8 1 Figure 120.Stoecker: ME 423 Class Notes . Design of Thermal Systems.2 0.F.2 0.

Sensible/latent capacity splits are determined by the rated sensible heat ratio (SHR) and the apparatus dew point (ADP)/bypass factor (BF) approach. 1975. 4.R. 1977. 2nd Edition. London. 1977 . D. Elmahdy.g. The total amount of heat rejected by the condenser is also calculated and stored for use by other waste heat recovery models (e. pp. 83. ASHRAE 1993). energy input ratio and part-load fraction to determine the performance of the unit at part-load conditions (Henderson et al.H. Ottawa.P.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS HVAC2 Toolkit: Algorithms and Subroutines for Secondary HVAC Systems Energy Calculations. coefficient of performance (COP) and the volumetric air flow rate across the cooling coil at rated conditions.83 Part 2. Brandemeuhl.P. "A Simple Model for Cooling and Dehumidifying Coils for Use in Calculating Energy Requirements for Buildings. Analytical and Experimental Multi-Row Finned-Tube Heat Exchanger Performance During Cooling and Dehumidifying Processes. Kays. Solar Energy Laboratory. January. and these models can be easily linked with the DX system model described here to simulate the entire DX air conditioner being considered (e. 1992. 1993. pp. 1970.g. this DX system model does not account for the thermal effects or electric power consumption of the indoor supply air fan.. Threlkeld. A. Canada. EnergyPlus contains separate models for simulating the performance of various indoor fan configurations.. AUTO fan. excluding any thermal or energy impacts due to the indoor supply air fan. The model uses performance information at rated conditions along with curve fits for variations in total capacity. No.DX Cooling Coil Model (HVAC) Overview This model (object names Coil: DX: CoolingBypassFactorEmpirical and COIL: DX: MultiMode: CoolingEmpirical with CoilPerformance: DX: CoolingBypassFactorEmpirical) simulates the performance of an air-cooled or evaporative-cooled direct expansion (DX) air conditioner. M. New York: McGraw-Hill. G. see Furnace:HeatCool. A. Ph. 103-117. Pub. 1985 Elmahdy. Coil:Desuperheater:Heating).L. Wisconsin-Madison. and A. crankcase heater and evap condenser water pump). SHR and COP inputs should be “gross” values. extended to a cooling and dehumidifying coil. Clark..D. Elmahdy.S.).H. 103-117. and Mitalas..Compact Heat Exchangers. The rated conditions are 4/23/05 368 ..8-1 .H. This model simulates the thermal performance of the DX cooling coil and the power consumption of the outdoor condensing unit (compressor. pp. W. 1990. Englewood Cliffs: Prentice-Hall.8-12. J. 2nd Edition.6. pp. A Transient System Simulation Program: Reference Manual. Carleton University. variable air volume.L. Air Conditioner:Window:Cycling or UnitarySystem:HeatPump). Thesis. ASHRAE Secondary HVAC Toolkit TRNSYS. This approach is analogous to the NTU-effectiveness calculations used for sensible-only heat exchanger calculations.M. sensible heat ratio (SHR). and Mitalas. 1964. Thermal Environmental Engineering. G. HVACSIM+ Building Systems and Equipment Simulation Program Reference Manual.. J. constant fan vs. National Bureau of Standards. fan." ASHRAE Transactions . The performance of the indoor supply air fan varies widely from system to system depending on control strategy (e. Vol. constant air volume vs.. NBSIR 84-2996. etc.. 254-270.4. The capacity.Inc. ASHRAE. Univ. fan motor efficiency and pressure losses through the air distribution system. Department of Commerce. A. U. Therefore. Model Description The user must input the total cooling capacity. December. Unitary System:HeatCool."A Simple Model for Cooling and Dehumidifying Coils for Use In Calculating Energy Requirements for Buildings ASHRAE Transactions.g.6. Vol. fan type. Part 2. Coil Model -.

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

considered to be air entering the cooling coil at 26.7°C drybulb/19.4°C wetbulb and air entering the outdoor condenser coil at 35°C drybulb/23.9°C wetbulb. The rated volumetric air flow should be between 0.00004699 m3/s and 0.00006041 m3/s per watt of rated total cooling capacity (350 – 450 cfm/ton). The user must also input five performance curves that describe the change in total cooling capacity and efficiency at part-load conditions:

1) Total cooling capacity modifier curve (function of temperature) 2) Total cooling capacity modifier curve (function of flow fraction) 3) Energy input ratio (EIR) modifier curve (function of temperature) 4) Energy input ratio (EIR) modifier curve (function of flow fraction) 5) Part load fraction correlation (function of part load ratio)

•

The total cooling capacity modifier curve (function of temperature) is a biquadratic curve with two independent variables: wet-bulb temperature of the air entering the cooling coil, and dry-bulb temperature of the air entering the air-cooled condenser coil (wet-bulb temperature if modeling an evaporative-cooled condenser). The output of this curve is multiplied by the rated total cooling capacity to give the total cooling capacity at the specific entering air temperatures at which the DX coil unit is operating (i.e., at temperatures different from the rating point temperatures).

2 2

**TotCapTempModFac = a + b (Twb ,i ) + c (Twb ,i ) + d (Tc ,i ) + e (Tc ,i ) + f (Twb ,i )(Tc ,i )
**

where

(398)

Twb ,i = wet-bulb temperature of the air entering the cooling coil, °C Tc ,i = dry-bulb temperature of the air entering an air-cooled condenser or wet-bulb

temperature of the air entering an evaporative-cooled condenser, °C • The total cooling capacity modifier curve (function of flow fraction) is a quadratic (or cubic) curve with the independent variable being the ratio of the actual air flow rate across the cooling coil to the rated air flow rate (i.e., fraction of full load flow). The output of this curve is multiplied by the rated total cooling capacity and the total cooling capacity modifier curve (function of temperature) to give the total cooling capacity at the specific temperature and air flow conditions at which the DX unit is operating.

2

TotCapFlowModFac = a + b ( ff ) + c ( ff )

or

(399)

TotCapFlowModFac = a + b ( ff ) + c ( ff ) + d ( ff )

2

3

where

**⎛ Actual air mass flow rate ⎞ ff = flow fraction = ⎜ ⎟ ⎝ Rated air mass flow rate ⎠
**

Note: The actual volumetric air flow rate through the cooling coil for any simulation time step where the DX unit is operating must be between 0.00003356 m3/s and .00006713 m3/s per watt of rated total cooling capacity (250 - 500 cfm/ton). The simulation will issue a warning message if this air flow range is exceeded.

4/23/05 369

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

•

The energy input ratio (EIR) modifier curve (function of temperature) is a biquadratic curve with two independent variables: wet-bulb temperature of the air entering the cooling coil, and dry-bulb temperature of the air entering the air-cooled condenser coil (wet-bulb temperature if modeling an evaporative-cooled condenser). The output of this curve is multiplied by the rated EIR (inverse of the rated COP) to give the EIR at the specific entering air temperatures at which the DX coil unit is operating (i.e., at temperatures different from the rating point temperatures).

2 2

**EIRTempModFac = a + b (Twb ,i ) + c (Twb ,i ) + d (Tc ,i ) + e (Tc ,i ) + f (Twb ,i )(Tc ,i )
**

where

(400)

Twb ,i = wet-bulb temperature of the air entering the cooling coil, °C Tc ,i = dry-bulb temperature of the air entering an air-cooled condenser or wet-bulb

temperature of the air entering an evaporative-cooled condenser, °C • The energy input ratio (EIR) modifier curve (function of flow fraction) is a quadratic (or cubic) curve with the independent variable being the ratio of the actual air flow rate across the cooling coil to the rated air flow rate (i.e., fraction of full load flow). The output of this curve is multiplied by the rated EIR (inverse of the rated COP) and the EIR modifier curve (function of temperature) to give the EIR at the specific temperature and air flow conditions at which the DX unit is operating.

EIRFlowModFac = a + b ( ff ) + c ( ff )

or

2

(401)

EIRFlowModFac = a + b ( ff ) + c ( ff ) + d ( ff )

2

3

where

**⎛ Actual air mass flow rate ⎞ ff = flow fraction = ⎜ ⎟ ⎝ Rated air mass flow rate ⎠
**

• The part load fraction correlation (function of part load ratio) is a quadratic or a cubic curve with the independent variable being part load ratio (sensible cooling load / steady-state sensible cooling capacity). The output of this curve is used in combination with the rated EIR and EIR modifier curves to give the “effective” EIR for a given simulation time step. The part load fraction (PLF) correlation accounts for efficiency losses due to compressor cycling.

2

**PartLoadFrac = PLF = a + b ( PLR ) + c ( PLR )
**

or

(402)

**PartLoadFrac = a + b ( PLR ) + c ( PLR ) + d ( PLR )
**

2

3

(403)

where

4/23/05

370

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

**⎛ ⎞ sensible cooling load PLR = part − load ratio = ⎜ ⎟ ⎝ steady − state sensible cooling capacity ⎠
**

The part-load fraction correlation should be normalized to a value of 1.0 when the part load ratio equals 1.0 (i.e., no efficiency losses when the compressor(s) run continuously for the simulation time step). For PLR values between 0 and 1 (0 <= PLR < 1), the following rules apply: PLF >= 0.7 and PLF >= PLR If PLF < 0.7 a warning message is issued, the program resets the PLF value to 0.7, and the simulation proceeds. The runtime fraction of the coil is defined as PLR/PLF. If PLF < PLR, then a warning message is issued and the runtime fraction of the coil is limited to 1.0. A typical part load fraction correlation for a conventional, single-speed DX cooling coil (e.g., residential or small commercial unit) would be: PLF = 0.75 + 0.25(PLR) All five part-load curves are accessed through EnergyPlus’ built-in performance curve equation manager (curve: quadratic, curve:cubic and curve:biquadratic). It is not imperative that the user utilize all coefficients shown in equations (398) through (402) if their performance equation has fewer terms (e.g., if the user’s PartLoadFrac performance curve is linear instead of quadratic, simply enter the values for a and b, and set coefficient c equal to zero). For any simulation time step, the total (gross) cooling capacity of the DX unit is calculated as follows:

Q total = Q total , rated (TotCapTempModFac )(TotCapFlowModFac )

•

•

(404)

In a similar fashion, the electrical power consumed by the DX unit (compressors plus outdoor condenser fans) for any simulation time step is calculated using the following equation:

**⎛• ⎞ Power = ⎜ Q total ⎟ ( EIR )( RTF ) ⎝ ⎠
**

where

•

(405)

**Q total = Total cooling capacity, W -- ref. equation (404)
**

⎛ ⎞ 1 EIR = Energy input ratio = ⎜ ⎟ ( EIRTempModFac )( EIRFlowModFac ) ⎝ COPrated ⎠

**COPrated = Coefficient of performance at rated conditions (user input)
**

RTF =

**( PLR PartLoadFrac ) = runtime fraction of the cooling coil
**

•

The total amount of heat rejected by the condenser is then calculated and stored for use by other waste heat recovery models (e.g., Coil:Desuperheater:Heating).

Q cond = Q total (1 + EIR )

•

4/23/05

371

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

where

•

**Q cond = total amount of heat rejected by the condenser (W)
**

The crankcase heater is assumed to operate when the cooling coil’s compressor is OFF and the outdoor dry-bulb temperature is below the maximum outdoor temperature for crankcase heater operation. The average crankcase heater power for the simulation time step is calculated as follows:

**Pcrankcase = Q cap ,crankcase (1 − RTF )
**

where

•

(406)

**Pcrankcase = DX cooling coil crankcase heater power, W
**

•

**Q cap ,crankcase = crankcase heater capacity, W
**

If this cooling coil is used as part of an air-to-air heat pump (Ref. UnitarySystem:HeatPump:AirToAir), the crankcase heater defined for this DX cooling coil is disregarded and the associated output variable is omitted. Instead, the crankcase heater defined for the DX heating coil (Coil:DX:HeatingEmpirical) is enabled during the time that the compressor is not running for either heating or cooling. In this instance, RTF in the above equations would be the runtime fraction of the heat pump’s heating coil or cooling coil, whichever is greater.

In addition to calculating the total cooling capacity provided by the DX air conditioner, it is important to properly determine the break down of total cooling capacity into its sensible (temperature) and latent (dehumidification) components. The model computes the sensible/ latent split using the rated SHR and the ADP/BF approach (Carrier et al. 1959). When the DX coil model is initially called during an EnergyPlus simulation, the rated total capacity and rated SHR are used to calculate the coil bypass factor (BF) at rated conditions. The rated total capacity and rated SHR are first used to determine the ratio of change in air humidity ratio to air dry-bulb temperature:

**⎛ ω −ω SlopeRated = ⎜ in out ⎜T −T ⎝ db ,in db ,out
**

where

⎞ ⎟ rated ⎟ ⎠

(407)

ωin = humidity ratio of the air entering the cooling coil at rated conditions, kg/kg ωout = humidity ratio of the air leaving the cooling coil at rated conditions, kg/kg Tdb,in = dry-bulb temperature of the air entering the cooling coil at rated conditions, °C Tdb,out = dry-bulb temperature of the air leaving the cooling coil at rated conditions, °C

Along with the rated entering air conditions, the algorithm then searches along the saturation curve of the psychrometric chart until the slope of the line between the point on the saturation curve and the inlet air conditions matches SlopeRated. Once this point, the apparatus dew point, is found on the saturation curve the coil bypass factor at rated conditions is calculated as follows:

4/23/05

372

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

BFrated =

where

hout ,rated − hADP hin , rated − hADP

(408)

hout,rated = enthalpy of the air leaving the cooling coil at rated conditions, J/kg hin,rated = enthalpy of the air entering the cooling coil at rated conditions, J/kg hADP = enthalpy of saturated air at the coil apparatus dew point, J/kg

The coil bypass factor is analogous to the “ineffectiveness” (1-ε) of a heat exchanger, and can be described in terms of the number of transfer of unit (NTU).

BF = e − NTU = e

⎛ UA ⎞ −⎜ ⎟ m ⎝ cp ⎠

= e − Ao m

(409)

For a given coil geometry, the bypass factor is only a function of air mass flow rate. The model calculates the parameter Ao in equation (409) based on BFrated and the rated air mass flow rate. With Ao known, the coil BF can be determined for non-rated air flow rates. For each simulation time step when the DX air conditioner operates to meet a cooling load, the total cooling capacity at the actual operating conditions is calculated using equation (404) and the coil bypass factor is calculated based on equation (409). The coil bypass factor is used to calculate the operating sensible heat ratio (SHR) of the cooling coil using equations (410) and (411).

•

(Q / m) hADP = hin − total 1 − BF

(410)

**⎛⎛ h −h ⎞ ⎞ SHR = Minimum ⎜ ⎜ Tin , wADP ADP ⎟ , 1⎟ ⎜ ⎟ ⎝ ⎝ hin − hADP ⎠ ⎠
**

where

(411)

hin hADP

J/kg

= enthalpy of the air entering the cooling coil, J/kg = enthalpy of air at the apparatus dew point condition, J/kg

hTin , wADP = enthalpy of air at the entering coil dry-bulb temperature and humidity ratio at ADP,

m

= air mass flow rate, kg/s

With the SHR for the coil at the current operating conditions, the properties of the air leaving the cooling coil are calculated using the following equations:

•

hout = hin −

Qtotal

m

**(412) (413) (414)
**

373

hTin ,ωout = hin − (1 − SHR )( hin − hout )

**ωout = PsyWFnTdbH (Tin , hTin,ωout )
**

4/23/05

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

**Tdb ,out = PsyTdbFnHW (hout , ωout )
**

where

(415)

hout

ratio, J/kg

= enthalpy of the air leaving the cooling coil, J/kg

hTin ,ωout = enthalpy of air at the entering coil dry-bulb temperature and leaving air humidity

ωout

Tdb ,out

= leaving air humidity ratio, kg/kg = leaving air dry-bulb temperature, °C

PsyWFnTdbH = EnergyPlus psychrometric function, returns humidity ratio given dry-bulb temperature and enthalpy PsyTdbFnHW = EnergyPlus psychrometric function, returns dry-bulb temperature given enthalpy and humidity ratio

Dry Coil Conditions If the model determines that the cooling coil is dry (ωin < ωADP), then equations (404) and (405) are invalid since they are functions of entering wet-bulb temperature. Under dry-coil conditions, coil performance is a function of dry-bulb temperature rather than wet-bulb temperature. In this case, the model recalculates the performance of the DX cooling unit using the calculation procedure described above but with ωin = ωdry, where ωdry is the inlet air humidity ratio at the coil dry-out point (SHR = 1.0). Condenser Options: Air Cooled vs. Evaporative Cooled As described previously, this model can simulate the performance of air-cooled or evaporative-cooled DX air conditioners. The following paragraphs describe three modeling options. If the user wants to model an air-cooled condenser, they should simply specify AIR COOLED in the field Condenser Type. In this case, the Total Cooling Capacity Modifier Curve (function of temperature) and the Energy Input Ratio Modifier Curve (function of temperature) (equations (398) and (400) above) will utilize the outdoor dry-bulb temperature. If the user wishes to model an evaporative-cooled condenser AND they have performance curves that are a function of the wet-bulb temperature of air entering the condenser coil, then the user should specify Condenser Type = Evap Cooled and the evaporative condenser effectiveness value should be entered as 1.0. In this case, the Total Cooling Capacity Modifier Curve (function of temperature) and the Energy Input Ratio Modifier Curve (function of temperature) (equations (398) and (400) above) will utilize the outdoor wet-bulb temperature. If the user wishes to model an air-cooled condenser that has evaporative media placed in front of it to cool the air entering the condenser coil, then the user should specify Condenser Type = Evap Cooled. The user must also enter the appropriate evaporative effectiveness for the media. In this case, the Total Cooling Capacity Modifier Curve (function of temperature) and the Energy Input Ratio Modifier Curve (function of temperature) will utilize the condenser inlet air temperature as calculated below:

**Tc ,i = (Twb ,o ) + (1 − EvapCondEffectiveness ) (Tdb ,o − Twb ,o )
**

where

Tc,i = the temperature of the air entering the condenser coil, °C Twb,o = the wet-bulb temperature of the outdoor air, °C

4/23/05 374

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

**Tdb,o = the dry-bulb temperature of the outdoor air, °C
**

In this case, the Total Cooling Capacity Modifier Curve (function of temperature) and the Energy Input Ratio Modifier Curve (function of temperature) input fields for this object should reference performance curves that are a function of outdoor dry-bulb temperature. Be aware that the evaporative media will significantly reduce the dry-bulb temperature of the air entering the condenser coil, so the Total Cooling Capacity and EIR Modifier Curves must be valid for the expected range of dry-bulb temperatures that will be entering the condenser coil. If an evaporative-cooled condenser is modeled, the power requirements for the water pump are calculated as follows:

**Pevapcondpump = Q cap ,evapcondpump ( RTF )
**

where

•

**Pevapcondpump = DX cooling coil evap condenser pump electric power, W
**

•

**Q cap ,evapcondpump = evaporative condenser pump rated power consumption, W
**

Water consumption for the evaporative-cooled condenser is calculated using the difference in air humidity level across the evaporative media and the condenser air mass flow rate:

Vwater =

where

m air (ωevapcond ,out − ωevapcond ,in )

•

ρ water

( RTF )(TimeStepSys )( 3600.)

Vwater

•

= DX cooling coil evap condenser water consumption, m3 = evaporative condenser air mass flow rate, kg/s = humidity ratio of outdoor air entering the evap condenser, kg/kg = humidity ratio of air leaving the evap condenser, kg/kg = density of water at the outdoor dry-bulb temperature, kg/m3

m air

ωevapcond ,in ωevapcond ,out ρ water

**TimeStepSys = HVAC system simulation time step, hr
**

Supply Air Fan Control: Cycling vs. Continuous One of the inputs to the DX cooling coil model is the supply air fan operation mode: cycling fan, cycling compressor (CycFanCycComp) or continuous fan, cycling compressor (ContFanCycComp). The first operation mode is frequently referred to as “AUTO fan”, where the compressor(s) and supply air fan operate in unison to meet the zone cooling load, and cycle off together when the cooling load has been met. The second operation mode is often referred to as “fan ON”, where the compressor(s) cycle on and off to meet the zone cooling load but the supply air fan operates continuously regardless of compressor operation. The EnergyPlus methodology for determining the impact that HVAC equipment has on an air stream is to calculate the mass flow rate and air properties (e.g., enthalpy, dry-bulb temperature, humidity ratio) exiting the equipment. These exiting conditions are passed along

4/23/05 375

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

as inlet conditions to the next component model in the air stream. Eventually the flow rate and properties of the air being supplied to the conditioned zone are used in the zone energy balance to determine the resulting zone air temperature and humidity ratio. With this methodology, the determination of the air mass flow rate and air properties for the two different supply air fan operation modes is slightly different. For the case of cycling fan/cycling compressor, the conditions of the air leaving the cooling coil are the steady-state values calculated using equations (412), (414) and (415) above. However the air mass flow rate passed along to the next component (and eventually to the conditioned zone) is the average air mass flow rate for the system simulation time step (determined by the cooling system; see Air Conditioner:Window, Furnace:HeatCool, Unitary System:HeatCool or UnitarySystem:HeatPump). For the case of continuous fan/cycling compressor, the air mass flow rate is constant. However, the air properties leaving the cooling coil are calculated as the average conditions during the system simulation time step. The model assumes that the exiting air conditions are the steady-state values calculated using equations (412), (414) and (415) above when the compressor(s) operate. For the remainder of the system simulation time step, it is assumed that the air exiting the DX coil has the same properties as the air entering the coil. For this supply air fan operating strategy, the leaving air properties are calculated as follows:

hout ,ContFanCycComp = hout ( PLR ) + hin (1 − PLR )

(416) (417) (418)

**ωout ,ContFanCycComp = ωout ( PLR ) + ωin (1 − PLR )
**

Tdb ,out ,ContFanCycComp = PsyTdbFnHW (hout , ContFanCycComp , ωout ,ContFanCycComp )

Latent Capacity Degradation with Continuous Supply Air Fan Operation

The latent (dehumidification) capacity of a direct-expansion (DX) cooling coil is strongly affected by part-load, or cyclic, operation. This is especially true in applications where the supply air fan operates continuously while the cooling coil cycles on and off to meet the cooling load. During constant fan operation, moisture condenses on the cooling coil when the compressor operates, but part or all of the moisture that is held by the coil evaporates back into the airstream when the cooling coil is deactivated (Figure 122). The net effect is that the amount of moisture removed from the air is degraded at part-load conditions as compared to steady-state conditions when the compressor operates continuously (Figure 123). EnergyPlus is able to model latent capacity degradation based on algorithms developed by Henderson and Rengarajan (1996). The model is applicable to single-stage cooling units, like residential and small commercial air conditioners or heat pumps with less than 19 kW of nominal cooling capacity. The model inputs are described in the EnergyPlus Input/Output Reference for the object Coil:DX:CoolingBypassFactorEmpirical. The model is enabled only if the four numerical inputs are defined (values greater than zero, see IO Reference) and the field “Supply Air Fan Operation Mode” must be “ContFanCycComp”.

4/23/05

376

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

Figure 122. Transient Sensible and Latent Capacity of a Cooling Coil Over an Operating Cycle

Figure 123. Field Data Showing the Net Impact of Part-Load Operation on Sensible Heat Ratio

4/23/05

377

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

Figure 124. Concepts of Moisture Buildup and Evaporation Figure 124 graphically depicts the latent degradation concepts and defines several key model parameters. After the cooling coil starts to operate, the coil temperature is eventually reduced below the dewpoint temperature of the entering air. Moisture from the air then builds on the surface of the coil until time to has elapsed and the total moisture mass on the coil is Mo. After this time (to), moisture begins to fall from the coil and all of the latent capacity provided by the coil is “useful” since this condensate is collected and removed from the unit. When the coil cycles off and the supply air fan continues to operate, the initial moisture mass buildup on the coil (Mo) evaporates back into the supply air stream. If the cooling coil cycles back on before all of the moisture has evaporated, then the time until the first condensate removal (to) is shorter for this cooling cycle since the coil is already partially wetted. Figure 124 also shows several parameters that are used in the latent degradation model. The

•

ratio of the coil’s moisture holding capacity (Mo) and the steady-state latent capacity ( Q L ) is defined as twet : the nominal time for moisture to fall from the coil (ignoring transient effects at

•

startup and starting with a dry coil). The ratio of the initial moisture evaporation rate ( Q e ) and

•

the steady-state latent capacity ( Q L ) is defined as

γ

. Both twet and

γ

at the rated air volume

flow rate and temperature conditions are required model inputs. Two other model inputs are the Maximum ON/OFF Cycling Rate (cycles per hour, Nmax) and the time constant ( τ , in seconds) for the cooling coil’s latent capacity to reach steady state after startup. The

4/23/05 378

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE

COILS

development of the latent degradation model is fully described by Henderson and Rengarajan (1996). The model implemented in EnergyPlus is for their “linear decay” evaporation model. During the simulation, all of the steady-state calculations described previously in equations (398) through (411) are completed. The latent degradation model then modifies the steadystate sensible heat ratio for the coil as shown below. The value of twet at the current air volume flow rate and entering air conditions is first calculated based on the rated value of twet entered by the user:

• ⎛ ⎛Q = Minimum ⎜ twet ,rated ⎜ latent ,rated • ⎜ ⎜ Q latent ⎝ ⎝

twet

⎞ ⎞ ⎟ ⎟, t ⎟ wet ,max ⎟ ⎠ ⎠

where

**twet = nominal time for condensate removal to begin at the current airflow and entering air
**

conditions, starting with a dry coil (sec)

twet,rated = nominal time for condensate removal to begin at the coil’s rated airflow and entering

air conditions, starting with a dry coil (sec)

•

**Q latent ,rated = cooling coil latent capacity at the rated airflow and temperature conditions, W
**

•

**Q latent = cooling coil latent capacity at the current airflow and temperature conditions, W
**

twet,max = maximum allowed value for twet (9999.0 sec)

Likewise, the value of γ at the current air volume flow rate and entering air conditions is calculated based on the rated value of γ entered by the user:

• ⎛Q γ = γ rated ⎜ latent ,rated • ⎜ Q latent ⎝

⎞⎛ T −T db ,i wb ,i ⎟⎜ ⎜T ⎟ ⎝ db ,rated − Twb , rated ⎠

⎞ ⎟ ⎟ ⎠

where: = ratio of the initial moisture evaporation rate from the cooling coil (when the compressor first turns off, in Watts) and the coil’s steady-state latent capacity (Watts) at the current air volume flow rate and entering air conditions

γ

γ rated

=

γ

at rated air flow and entering air conditions

Tdb,i = dry-bulb temperature of the air entering the cooling coil, °C Twb,i = wet-bulb temperature of the air entering the cooling coil, °C Tdb,rated = dry-bulb temperature of air entering the cooling coil at rated conditions (26.7°C) Twb,rated = wet-bulb temperature of air entering the cooing coil at rated conditions (19.4°C)

The cooling coil on and off times are then calculated based on the maximum number of cycles per hour and the calculated run-time fraction for the coil.

ton =

4/23/05

3600 4 N max (1 − X )

379

0) SHRss = steady-state sensible heat ratio (from eqn.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS toff = where 3600 4 N max X ton Nmax X toff = duration of cooling coil on-cycle (sec) = maximum on/off cycles per hour (cph) = cooling coil runtime fraction (-) = duration of cooling coil off-cycle (sec) The equation for calculating the time to when moisture first begins to fall from the cooling coil is shown below. ⎛⎛ ⎞ ⎞ ⎜⎜ ⎟ ⎟ ⎜⎜ ⎟ ⎟ ton − to LHR = Maximum ⎜ ⎜ ⎟ . 0.0 ⎟ LHRss ⎛ ⎛ − ton ⎞ ⎞⎟ ⎜ ⎟ ⎜⎜ ⎟ ton + τ ⎜ e⎝ τ ⎠ − 1⎟ ⎟ ⎜⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎝ ⎠⎠ ⎝⎝ ⎠ ⎛ LHR ⎞ SHReff = 1 − (1 − SHRss ) ⎜ ⎟ ⎝ LHRss ⎠ where LHR = part-load latent heat ratio LHRss = latent heat ratio at steady-state conditions ( 1 − SHRss with SHRss from eqn. which is in turn used to calculate the “effective” sensible heat ratio of the cooling including part-load latent degradation effects. and is solved iteratively by EnergyPlus: t j +1 o = γ toff ⎛ 2 γ −⎜ ⎜ 4t ⎝ wet j ⎞ ⎛ to ⎞ ⎟ 2 − τ ⎜ τ − 1⎟ . ton and τ . (411)) 4/23/05 380 . (411)) SHReff = part-load sensible heat ratio ( SHRss ≤ SHReff ≤ 1. ⎜e ⎟ ⎟ t off ⎝ ⎠ ⎠ t off ⎛ 2t ⎞ ≤ ⎜ wet ⎟ ⎝ γ ⎠ where to = time where condensate removal begins (sec) τ j = latent capacity time constant at start-up (sec) = iteration number The part-load latent heat ratio of the cooling coil is then calculated with to .

If stage 1 operates for less than the full time step. Refrigerating and Air-Conditioning Engineers. W. September. A dehumidification mode flag is passed to the coil model which is used to select a different set of performance specifications from a CoilPerformance:DX:CoolingBypassFactorEmpirical input object.H.E. Rengarajan. Understanding the Dehumidification Performance of AirConditioning Equipment at Part-Load Conditions. Jr. Any mode may specify that a portion of the total airflow is bypassed. 1998. only one of the stages will be degraded. Then the simulation of the DX coil proceeds as described above including any bypass or multi-stage operation. Presented at ASHRAE’s IAQ & Energy ‘98 conference. Paper presented at Joint CIBSE/ASHRAE Conference. equations (416) through (418) are used to calculate the average leaving air conditions (average when the coil is on and off) for the simulation time step. The impact of comfort control on air conditioner energy use in humid climates. H. 2003. Modern air conditioning. New York: Pitman Publishing Corporation. 3d ed. 1992. then stage 1 is running for the full time step and latent degrdataion is applied only to the portion of the latent load which is attributed to stage 2 operation. Henderson. Shirey. Henderson. 1993. A Model to Predict the Latent Capacity of Air Conditioners and Heat Pumps at Part-Load Conditions with Constant Fan Operation. LA. New Orleans. 1996. H. and D. 1959. and D. Rengarajan. including the impacts of latent degradation. Multi-Stage Operation Multi-stage control is modeled with the DX coil.. ASHRAE Transactions 98(2): 104-113. References ASHRAE. Scotland. H. Building Sustainability. ASHRAE Transactions 102(2): 266-274. Edinburgh. Shirey III. and K. Jr.Electric Heating (HVAC) The Electric heating coil (object name: Coil:Electric:Heating) is a simple capacity model with a user-inputted efficiency. then latent degradation is applied.A. If stage 2 operates. In many cases. Henderson. Atlanta: American Society of Heating. equations (412) through (415) are then used to calculate the properties of the air leaving the cooling coil when it operates. and W. Cherne. heating and ventilating. W. Special Calculations for COIL:DX:MultiMode:CoolingEmpirical with CoilPerformance:DX:CoolingBypassFactorEmpirical The multimode DX coil uses 1 to 4 set of performance specifications which all feed into the main DX coil model as described above.. The coil can be used in the air loop simulation or in the zone equipment as a reheat coil. Carrier.H. Henderson. Finally. The impact of part load air conditioner operation on dehumidification performance: Validating a latent capacity degradation model. HVAC2 Toolkit: A Toolkit for Secondary HVAC System Energy Calculation.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS With the “effective” SHR for the coil at the current operating conditions. Then the bypassed air stream is mixed with the conditioned air stream and these conditions are place on the coil outlet node. stage 1 is assumed to operate first and continue to operate if stage 2 is required.B. The multimode DX coil may have 1 or 2 capacity stages and may have 1 enhanced dehumidification mode in addition to its "normal" or base mode performance. Grant. Value and Profit. Bypass Operation When a mode has coil bypass. Depending on where it is used 4/23/05 381 . Enhanced Dehumidification Mode If enhanced dehumidification mode is available. This coil only has air nodes to connect it in the system. October. the non-bypassed air flow fraction is used to model the DX coil performance as described above.I. Coil Model -. Inc. this efficiency for the electric coil will be 100%. H.I. Roberts. R. K. If latent degradation is active. this is controlled by the parent object of the DX coil. If 2-stage operation has been specified. such as DXSystem:AirLoop.

If used in the air loop simulation it will be controlled to a specified temperature scheduled from the Setpoint Manager.GT. The gas coil has additional features that can add a part load correction and a parasitic gas or electric load.gt.and. set ! output to zero. IF(QCoilReq > HeatingCoil(CoilNum)%NominalCapacity) Then QCoilCap = HeatingCoil(CoilNum)%NominalCapacity Else QCoilCap = QCoilReq End IF TempAirOut=TempAirIn + QCoilCap/CapacitanceAir HeatingCoilLoad = QCoilCap !The HeatingCoilLoad is the change in the enthalpy of the Heating HeatingCoil(CoilNum)%ElecUseLoad = HeatingCoilLoad/Effic For a temperature setpoint coil the delta temperature from the coil inlet temperature to the setpoint is determined and the capacity of the coil is calculated and is met if less than the user specified capacity. & (QCoilReq .0 .and.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS determines if this coil is temperature or capacity controlled. & (GetCurrentScheduleValue(HeatingCoil(CoilNum)%SchedPtr) .LE. ! Control coil output to meet a setpoint temperature. 4/23/05 382 .TempAirIn) ! check to see if set point above enetering temperature.gt. Else IF(QCoilCap > HeatingCoil(CoilNum)%NominalCapacity) Then QCoilCap = HeatingCoil(CoilNum)%NominalCapacity TempAirOut=TempAirIn + QCoilCap/CapacitanceAir Else TempAirOut = TempSetPoint End IF HeatingCoilLoad = QCoilCap !The HeatingCoilLoad is the change in the enthalpy of the Heating HeatingCoil(CoilNum)%ElecUseLoad = HeatingCoilLoad/Effic Coil Model – Gas Heating (HVAC) The Gas heating coil (object name: Coil:Gas:Heating) is a simple capacity model with user inputted gas burner efficiency. The default for the gas burner efficiency is 80%.0)) THEN !check to see if the Required heating capacity is greater than the user! specified capacity.GT. 0. TempControlTol) ) THEN QCoilCap = CapacitanceAir*(TempSetPoint . If it is used in zone equipment. 0.0 TempAirOut = TempAirIn !check to see if the Required heating capacity is greater than the user ! specified capacity. For a coil that is controlled to meet the zone demand and will meet the capacity necessary for that zone unless it exceeds the capacity of the coil specified by the user. IF(QCoilCap . it will be controlled from the zone thermostat by meeting the zone demand. If used in the air loop simulation it will be controlled to a specified temperature on the setpoint node by the Setpoint Manager. HeatingCoil(CoilNum)%NominalCapacity > 0. 0. it will be controlled from the zone thermostat by meeting the zone demand.0 . & (GetCurrentScheduleValue(HeatingCoil(CoilNum)%SchedPtr) . (TempSetPoint == 0.0) THEN QCoilCap = 0. This coil only has air nodes to connect it in the system.AND. 0.and.and. The coil can be used in the air loop simulation or in the zone equipment as a reheat coil. ! Control output to meet load QCoilReq IF((AirMassFlow .0) . HeatingCoil(CoilNum)%NominalCapacity > 0.0) .0) .0) .gt.0) .AND. & (ABS(TempSetPoint-TempAirIn) . 0. If it is used in zone equipment.and. Else IF((AirMassFlow . 0.and. If not.gt. Depending on where it is used determines if this coil is temperature or capacity controlled. & (QCoilReq == 0.0) . See Input Output Reference for additional information.

This model simulates the thermal performance of the indoor DX heating coil. The capacity and COP 4/23/05 383 . and the power consumption of the outdoor unit (compressors. The part-load fraction correlation should be normalized to a value of 1.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS The parasitic electric load associated with the gas coil operation.2i PLR For a better understanding of how the coil meets the temperature setpoint in the air loop or the zone demand as zone equipment. and these models can be easily linked with the DX system model described here to simulate the entire DX system being considered (e. The performance of the indoor supply air fan varies widely from system to system depending on control strategy (e. crankcase heaters and defrost heaters).. Coil Model -. The parasitic gas load associated with the gas coil’s operation (Watts).e. fan motor efficiency and pressure losses through the air distribution system. The part-load curve accounts for efficiency losses due to transient coil operation. The runtime fraction of the heating coil is defined a PLR/PLF. then a warning message is issues and the runtime fraction of the coil is limited to 1. For any simulation time step. variable air volume. Model Inputs The user must input the total heating capacity. This will be modified by the PLR (or coil runtime fraction if a part-load fraction correlation is provided in the next input field) to reflect the time of operation in a simulation time step. the program resets the PLF value to 0. such as an inducer fan. Also see Input Output Reference for additional input information. the nominal gas consumption rate (heating load/burner efficiency) is divided by the part-load fraction (PLF) if a part-load curve has been defined. constant fan vs. A typical part load fraction correlation for a conventional gas heating coil (e. The model assumes that this parasitic load is consumed only for the portion of the simulation time step where the gas heating coil is not operating. If PLF < PLR. this DX system model does not account for the thermal effects or electric power consumption of the indoor supply air fan. energy input ratio and part-load fraction to determine the performance of the unit at part-load conditions (DOE 1982). coefficient of performance (COP) and the volumetric airflow rate across the heating coil at rated conditions. The model uses performance information at rated conditions along with curve fits for variations in total capacity. For PLR values between 0 and 1 ( 0 <= PLR < 1). no efficiency losses when the heating coil runs continuously for the simulation time step). Adjustment factors are applied to total capacity and input power to account for frost formation on the outdoor coil.0. see Unitary System:HeatPump:AirToAir). EnergyPlus contains separate models for simulating the performance of various indoor fan configurations.. Therefore. see Coil:Electric:Heating for additional information.. AUTO fan.7 a warning message is issued. Field: Part Load Fraction Correlation (function of part load ratio) The part load correction defines the name of a quadratic or cubic performance curve (Ref: Performance Curves) that parameterizes the variation of gas consumption rate by the heating coil as a function of the part load ratio (PLR.g. fan type.g.. and the simulation proceeds.DX Heating Coil Model (HVAC) Overview This model (object name Coil:DX:HeatingEmpirical) simulates the performance of an air-to-air direct expansion (DX) heating system.0 when the part load ratio equals 1. such as a standing pilot light. constant air volume vs. fans.7 and PLF >= PLR If PLF < 0.g. etc.0 (i.7. etc.8 + 0.). residential furnace) would be: PLF = 0. sensible heating load/nominal capacity of the heating coil). the following rules apply: PLF >= 0.

i Tdb . excluding any thermal or energy impacts due to the indoor supply air fan.o = dry-bulb temperature of the air entering the indoor coil. The biquadratic curve is recommended if sufficient manufacturer data is available as it provides sensitivity to the indoor air dry-bulb temperature and a more realistic output. Depending on the defrost strategy that is selected.o ) (421) where Tdb . User has the choice of a bi-quadratic curve with two independent variables or a quadratic curve as well as a cubic curve with a single independent variable.i ) + d (Tdb ..o ) or 2 (419) TotCapTempModFac = a + b (Tdb .00006041 m3/s per watt of rated total heating capacity (350 – 450 cfm/ton). at an outdoor or indoor air temperature different from the rating point temperature).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS inputs should be “gross” values.55C wet bulb temperatures. The output of this curve is multiplied by the rated total heating capacity to give the total heating capacity at specific temperature operating conditions (i.o ) + e (Tdb . TotCapTempModFac = a + b (Tdb .o ) + d (Tdb . The rating condition is considered to be outdoor air at 8. The rated volumetric air flow across the DX heating coil should be between 0.o ) + f (Tdb .11C wet bulb temperatures (i.e.i )(Tdb . the user must also input up to six performance curves that describe the change in total heating capacity and efficiency at partload conditions.11C dry bulb and 15.o ) 2 2 2 3 (420) TotCapTempModFac = a + b (Tdb .o ) + c (Tdb .i ) + c (Tdb . fraction of full load flow). °C 2) The total heating capacity modifier curve (function of flow fraction) is a quadratic or cubic curve with the independent variable being the ratio of the actual air flow rate across the heating coil to the rated air flow rate (i. with air entering the indoor DX heating coil at 21.e.e.o ) + c (Tdb .33C dry bulb and 6. TotCapFlowModFac = a + b ( ff ) + c ( ff ) or 2 (422) TotCapFlowModFac = a + b ( ff ) + c ( ff ) + d ( ff ) 2 3 (423) where 4/23/05 384 .. and efficiency during reverse-cycle defrosting: 1) The total heating capacity modifier curve (function of temperature) can be a function of both the outdoor and indoor air dry-bulb temperature or only the outdoor air dry-bulb temperature. air entering the outdoor coil). The output of this curve is multiplied by the rated total heating capacity and the total heating capacity modifier curve (function of temperature) to give the total heating capacity at the specific temperature and air flow conditions at which the coil is operating.00004699 m3/s and 0.. °C = dry-bulb temperature of the air entering the outdoor coil.

.o ) or 2 (424) EIRTempModFac = a + b (Tdb .e.o ) + c (Tdb .o ) + e (Tdb .o ) 2 2 (426) 4) The energy input ratio (EIR) modifier curve (function of flow fraction) is a quadratic or cubic curve with the independent variable being the ratio of the actual air flow rate across the heating coil to the rated air flow rate (i. PartLoadFrac = PLF = a + b ( PLR ) + c ( PLR ) 4/23/05 2 (429) 385 . at an outdoor or indoor air temperature different from the rating point temperature).i ) + c (Tdb .i ) + d (Tdb .. The output of this curve is used in combination with the rated EIR and EIR modifier curves to give the “effective” EIR for a given simulation time step.00003356 m3/s and . 3) The energy input ratio (EIR) modifier curve (function of temperature) can be a function of both the outdoor and indoor air dry-bulb temperature or only the outdoor air dry-bulb temperature.o ) + c (Tdb .o ) 2 3 (425) or EIRTempModFac = a + b (Tdb . The biquadratic curve is recommended if sufficient manufacturer data is available as it provides sensitivity to the indoor air dry-bulb temperature and a more realistic output. fraction of full load flow).00006713 m3/s per watt of rated total heating capacity (250 .500 cfm/ton).o ) + d (Tdb .o ) + f (Tdb .i )(Tdb .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS ⎛ Actual air mass flow rate ⎞ ff = flow fraction = ⎜ ⎟ ⎝ Rated air mass flow rate ⎠ Note: The actual volumetric airflow rate through the heating coil for any simulation time step where the DX unit is operating should be between 0. The simulation will issue a warning message if this airflow range is exceeded. The output of this curve is multiplied by the rated EIR (inverse of the rated COP) and the EIR modifier curve (function of temperature) to give the EIR at the specific temperature and air flow conditions at which the coil is operating. The partload fraction (PLF) correlation accounts for efficiency losses due to compressor cycling. The output of this curve is multiplied by the rated EIR (inverse of the rated COP) to give the EIR at specific temperature operating conditions (i. User has the choice of a bi-quadratic curve with two independent variables or a quadratic curve as well as a cubic curve with a single independent variable.e. EIRFlowModFac = a + b ( ff ) + c ( ff ) or 2 (427) EIRFlowModFac = a + b ( ff ) + c ( ff ) + d ( ff ) 2 3 (428) 5) The part-load fraction correlation (function of part-load ratio) is a quadratic or cubic curve with the independent variable being part-load ratio (sensible heating load / steady-state heating capacity). EIRTempModFac = a + b (Tdb .

A typical part load fraction correlation for a conventional.e. the fractional defrost time period and the runtime fraction of the heating coil to give the defrost power at the specific temperatures at which the coil is operating.0.0 (i. It is not imperative that the user utilize all coefficients shown in the preceding equations {(419) through (431)} in items (1) through (6) if their performance equation has fewer terms (e.75 + 0.. then a warning message is issued and the runtime fraction of the coil is limited to 1. The next two inputs define the minimum outdoor dry-bulb temperature that the heat pump compressor will operate and the maximum outdoor dry-bulb temperature for defrost operation. Crankcase heater capacity and crankcase heater cutout temperature are entered in the following two inputs.i ) + c (Twb . The output of this curve is multiplied by the heating coil capacity. The next input item for the HeatingEmpirical DX coil is the supply air fan operation mode.i ) + d (Tdb .0 when the part load ratio equals 1. DefrostEIRTempModFac = a + b (Twb .i )(Tdb . the program resets the PLF value to 0. defrost control (timed or on-demand).o = wet-bulb temperature of the air entering the indoor heating coil. and the simulation proceeds. no efficiency losses when the compressor(s) run continuously for the simulation time step). the fractional defrost time period (timed defrost control only).7 a warning message is issued. single-speed DX heating coil (e. The runtime fraction of the coil is defined a PLR/PLF. the following rules apply: PLF >= 0. For PLR values between 0 and 1 ( 0 <= PLR < 1). °C = dry-bulb temperature of the air entering the outdoor coil.7 and PLF >= PLR If PLF < 0. curve:cubic and curve:biquadratic)..g. This curve is only required when a reverse-cycle defrost strategy is specified. 4/23/05 386 .7. residential heat pump) would be: PLF = 0. °C All six curves are accessed through EnergyPlus’ built-in performance curve equation manager (curve:quadratic. The final four inputs cover the type of defrost strategy (reverse-cycle or resistive).o ) + e (Tdb . and set the remaining coefficients to zero).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS or PartLoadFrac = a + b ( PLR ) + c ( PLR ) + d ( PLR ) 2 3 (430) where ⎛ ⎞ sensible heating load PLR = part − load ratio = ⎜ ⎟ ⎝ steady − state sensible heating capacity ⎠ The part-load fraction correlation should be normalized to a value of 1. and the resistive defrost heater capacity if a resistive defrost strategy is selected.o ) + f (Twb . Either the supply air fan runs continuously while the DX coil cycles on/off.25(PLR) 6) The defrost energy input ratio (EIR) modifier curve (function of temperature) is a bi- quadratic curve with two independent variables: outdoor air dry-bulb temperature and the heating coil entering air wet-bulb temperature. simply enter the appropriate values for the coefficients a and b..o ) 2 2 (431) where Twb . if the user’s PartLoadFrac performance curve is linear instead of quadratic or cubic.g. or the fan and coil cycle on/off together.i Tdb . If PLF < PLR.

82Tdb. If the outdoor air dry-bulb temperature is below the specified maximum temperature for defrost operation. ωoutdoor − ωsat (Tcoil .589 (432) The difference between the outdoor air humidity ratio (from the weather file) and the saturated air humidity ratio at the estimated outdoor coil temperature is then calculated. depending on the defrost strategy and defrost control type specified for the heating coil. has a significant impact on heating capacity and energy use by the DX heating system. and crankcase heater power.out = MAX ⎡1.1E (ESTSC 2001. scenario # 2 above).out . Using the rated heating capacity and COP. input power multiplier and fractional defrost time period depending on the defrost strategy and defrost control type specified for the heating coil. then: a. the part-load curves specified for the DX heating coil.out or fractional defrost time period as the independent variable. b. total DX coil heating rate. The fraction of compressor runtime when the coil is being defrosted is either entered by the user (for timed defrost) or is calculated by the model (for on-demand defrost) using an empirical equation and ∆ω coil . 4/23/05 387 . then the model calculates adjustment factors for heating capacity and input power due to frost formation. and this value is used as an indication of frost formation on the outdoor coil. ∆ω coil .e. determine the following: heating coil exiting air conditions (dry-bulb temperature. and the part-load ratio that is being requested of the heating coil. 2) If the outdoor air dry-bulb temperature is above the specified minimum temperature for compressor operation and the DX heating coil is scheduled to operate.o − 8.out = 0. and the fractional defrost time period. If the outdoor dry-bulb temperature is below the specified maximum outdoor drybulb temperature for defrost operation. and set crankcase heater power equal to the crankcase heater capacity value specified by the input file. the defrost multipliers calculated above (if applicable).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Model Description The general flow of the model is as follows: 1) If the outdoor air dry-bulb temperature is below the specified minimum temperature for compressor operation or the DX heating coil is not scheduled to operate. Tcoil . Frost Adjustment Factors Frost formation on the outdoor coil. electric power during heating (compressors and outdoor fans). The model first estimates the outdoor coil temperature according to a linear empirical relationship with outdoor air dry-bulb temperature as the independent variable.. humidity ratio and enthalpy). and the need to periodically defrost this coil. The defrost time period fraction and adjustment factors due to frost formation on the outdoor coil vary depending on the defrost control type as shown below. OutBaroPress ) ⎤ ⎣ ⎦ (433) Frost formation on the outdoor coil must be periodically removed. calculate a heating capacity multiplier. set power and heating rates equal to zero. This method of accounting for the impacts of frosting/defrost was taken from the model used in DOE-2. simply pass through the heating coil inlet air conditions as the coil outlet conditions.0 E − 6. The following paragraphs give a detailed description of the model calculations that are performed when the DX heating coil is operating (i. electric power during defrost. Adjustment factors to total heating coil capacity and input power due to frost formation on the outdoor coil are also calculated by empirical models with ∆ω coil .out . Miller and Jaster 1985).

45 ( ∆ω coil .01446 ⎞ 1+ ⎜ ⎜ ∆ω coil . the additional heating load due to defrost (indoor cooling during defrost) is also calculated so that it may be added to the existing heating load when calculating input power for the compressor(s) and outdoor coil fan(s).222 − Tdb.out ⎟ ⎟ ⎝ ⎠ = t frac . In the case of reverse-cycle defrost.875 (1 − t frac .33 ( ∆ω coil .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Timed Defrost: Fractional Defrost Time = time period specified by user = t frac .9 − 36.01667 ⎠ ⎛Q ⎞ Pdefrost = DefrostEIRTempModFac ⎜ total .01667 ⎠ Resistive: (440) (441) Qdefrost = 0.rated ⎟ ( t frac .o ) ⎜ total .out ) Input Power Multiplier = 0.defrost ) (438) (439) If the outdoor air dry-bulb temperature is above the specified maximum temperature for defrost operation. then the model calculates the electrical power used during defrost.0 Pdefrost = ( Qcap .e. defrost ) ( RTF ) ⎝ 1.defrost ) Input Power Multiplier = 0.. the fractional defrost time period is set to zero and the heating capacity/input power multipliers are set to unity. Reverse-Cycle: ⎛Q ⎞ Qdefrost = 0.defrost (437) Heating Capacity Multiplier = 0.defrost )( t frac .01( t frac .909 − 107.defrost ) ( RTF ) where: 4/23/05 (442) (443) 388 .954 (1 − t frac . Defrost Operation If the fractional defrost time period is greater than zero for the simulation time step. reverse-cycle or resistive). defrost Heating Capacity Multiplier = 0. defrost ) ( 7.out ) On-Demand Defrost: (434) (435) (436) Fractional Defrost Time = 1 ⎛ 0.rated ⎟ ⎝ 1. The method for calculating defrost power varies based on the defrost strategy specified (i.

Equation (429) or (430)) is used in the calculation of electrical power draw to account for efficiency losses due to compressor cycling. the total heating capacity of the DX unit is calculated as follows: Qtotal = Qtotal .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Qdefrost Qtotal . If a resistive defrost strategy is selected. rated (TotCapTempModFac )(TotCapFlowModFac ) (444) If the outdoor air dry-bulb temperature is below the maximum temperature for defrost operation. Qdefrost = 0. the additional heating load (Qdefrost) generated during defrost operation is added to the heating load being requested by adjusting the part-load ratio. PLR + ⎜ defrost Qtotal ⎟ ⎟ ⎝ ⎠⎠ ⎝ PartLoadFrac = a + b ( PLR ) + c ( PLR ) + d ( PLR ) 2 3 (446) (430) Pheating = where ( Qtotal )( EIR )( PLR ) × InputPowerMultiplier PartLoadFrac (447) Pheating = average compressor and outdoor fan power for the simulation time step(W) Qtotal = total heating capacity W. ⎛ ⎛Q ⎞⎞ PLR = MIN ⎜1. Eqn. (445) ⎛ ⎞ 1 EIR = Energy input ratio = ⎜ ⎟ ( EIRTempModFac )( EIRFlowModFac ) ⎝ COPrated ⎠ COPrated = coefficient of performance at rated conditions (user input) 4/23/05 389 . the electrical power draw by the DX unit (compressors plus outdoor coil fans) for any simulation time step is calculated.0. defrost = additional indoor heating load due to reverse-cycle defrost (W) = total full-load heating capacity of the coil at rated conditions (W) = average defrost power for the simulation time step (W) = capacity of the resistive defrost heating element (W) DefrostEIRTempModFac = energy input ratio modifier curve applicable during defrost RTF = PLR ( PartLoadFrac ) = runtime fraction of the heating coil Heating Operation For any simulation time step. rated Pdefrost Qcap . Qtotal = Qtotal ( HeatingCapacityMultiplier ) (445) In a similar fashion. The part-load fraction correlation for the heating coil (user input. For a reverse-cycle defrost strategy. then the total heating capacity is further adjusted due to outdoor coil frost formation based on the results of Equation (444) and Equation (435) or (438).

whichever is greater. returns dry-bulb temp given enthalpy and humidity ratio Supply Air Fan Control: Cycling vs. RTF in the above equations would be the runtime fraction of the heat pump’s heating coil or cooling coil. In this instance. The properties of the air leaving the heating coil at full-load operation are calculated using the following equations: houtlet = hinlet + Qtotal • (450) m (451) (452) ωoutlet = ωinlet Tdb . The first operation mode is frequently referred to as “AUTO fan”.outlet = PsyTdbFnHW (houtlet . the crankcase heater defined for this DX heating coil is enabled during the time that the compressor is not running for either heating or cooling (and the crankcase heater power defined in the DX cooling coil object is disregarded in this case).crankcase (1 − RTF ) (448) RTF = PLR where ( PartLoadFrac ) = runtime fraction of the heating coil (449) Pcrankcase = average crankcase heater power for the simulation time step (W) Qcap . Continuous One of the inputs to the DX coil model is the supply air fan operation mode: cycling fan. UnitarySystem:HeatPump:AirToAir).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS InputPowerMultiplier = power adjustment due to frost if applicable -Eqn. where the compressor(s) cycle on and off to meet the zone heating load but the supply air fan operates continuously regardless of compressor operation. (436) or (439) The crankcase heater is assumed to operate when the heating coil’s compressor is OFF. cycling compressor (ContFanCycComp). 4/23/05 390 .crankcase = crankcase heater capacity (W) If this heating coil is used as part of an air-to-air heat pump (Ref. and the average crankcase heater power for the simulation time step is calculated as follows: Pcrankcase = Qcap .outlet PsyTdbFnHW = EnergyPlus psychrometric function. The second operation mode is often referred to as “fan ON”. ωoutlet ) where houtlet = enthalpy of the air leaving the heating coil (J/kg) = leaving air humidity ratio (kg/kg) = leaving air dry-bulb temperature (°C) ωoutlet Tdb . cycling compressor (CycFanCycComp) or continuous fan. where the compressor(s) and supply air fan operate in unison to meet the zone heating load. and cycle off together when the heating load has been met.

version 2. ContFanCycComp . For the case of cycling fan/cycling compressor. humidity ratio) exiting the equipment. (451) and (452) above. Approximately 25-30% of the energy rejected by typical 4/23/05 391 . and Jaster.g. DOE-2.Desuperheater Heating Coil Overview The Coil:Desuperheater:Heating model simulates the thermal performance of a refrigerant-toair heating coil and the parasitic electric consumption of its control valves or other auxiliary devices. ESTSC. H.ContFanCycComp ) (455) References DOE. Performance of Air-Source Heat Pumps. These exiting conditions are passed along as inlet conditions to the next component model in the air stream. EM-4226. enthalpy.ContFanCycComp = PsyTdbFnHW (houtlet . For this fan control type. DOE-2 engineers manual. the air mass flow rate is constant.1A.1E Version 110 (source code). For this supply air fan operating strategy. see Unitary System:HeatPump:AirToAir). For the case of continuous fan/cycling compressor. 1982. Berkeley. With this methodology. it is assumed that the air exiting the DX coil has the same properties as the air entering the coil.. The three objects from which this coil can obtain its heating energy are: COIL:DX:CoolingBypassFactorEmpirical COIL:DX:MultiSpeed:CoolingEmpirical COIL:DX:MultiMode:CoolingEmpirical Compressor Rack:Refrigerated Case The heat reclaim recovery efficiency (specified by the user) defines the amount of heat available for use by this coil. However. the heating coil part-load fraction (Equation (429) or (430)) is also passed to Fan:Simple:OnOff (if used) to properly calculate the supply air fan power and associated fan heat. R. For the remainder of the system simulation time step. Oak Ridge. the determination of the air mass flow rate and air properties for the two different supply air fan operation modes is slightly different. The model assumes that the heating energy provided by this coil is reclaimed from the superheated refrigerant gas leaving a compressor and does not impact the performance of the compressor. 1985.outlet . CA: Electric Power Research Institute. the air properties leaving the heating coil are calculated as the average conditions during the system simulation time step.ContFanCycComp = houtlet ( PLR ) + hinlet (1 − PLR ) (453) (454) ωoutlet .ContFanCycComp = ωoutlet ( PLR ) + ωinlet (1 − PLR ) Tdb . Coil Model -. the leaving air properties are calculated as follows: houtlet . However the air mass flow rate passed along to the next component (and eventually to the conditioned zone) is the average air mass flow rate for the system simulation time step (determined by the heating system. the conditions of the air leaving the heating coil are the steady-state values calculated using equations (450). Palo Alto. ωoutlet . Miller. 2001. LBL-11353. Eventually the flow rate and properties of the air being supplied to the conditioned zone are used in the zone energy balance to determine the resulting zone air temperature and humidity ratio. dry-bulb temperature.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS The EnergyPlus methodology for determining the impact that HVAC equipment has on an air stream is to calculate the mass flow rate and air properties (e. TN: Energy Science and Technology Software Center. CA: Lawrence Berkeley National Laboratory. (451) and (452) above when the compressor(s) operate.L. The model assumes that the exiting air conditions are the steady-state values calculated using equations (450).

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS refrigeration system condensers is to reduce the superheated refrigerant vapor temperature to the condensing temperature. Figure 125 shows a schematic diagram of the desuperheater heating coil used in conjunction with the DXSystem:AirLoop object. For this reason. The desuperheater heating coil can be used in air loop simulations for various air heating applications. the DX cooling coil could be used to the maintain its discharge air temperature at 11°C for zone dehumidification. the desuperheater coil is controlled based on the calculated heating load to maintain the zone temperature and humidity setpoints (load-based control). Desuperheating heating coil placement is unrestricted when reclaiming heat from a refrigerated case compressor rack. Figure 125. the supply air fan control should be either variable volume or constant fan since the desuperheater heating coil will typically be available the entire simulation time step. Furnace:Blowthru. Recovery efficiencies higher than 30% may cause the refrigerant gas to condense which in turn impacts the performance of the refrigeration system. The source of reclaimed heat could be the direct expansion (DX) cooling coil itself (Coil:DX:CoolingBypassFactorEmpirical) or a refrigerated case compressor rack (Compressor Rack:Refrigerated Case). The desuperheater heating coil could then raise the air dry-bulb temperature to 16°C providing a dry supply air stream at a temperature which does not require much additional heating by terminal units to meet the zone temperature set point. The desuperheater heating coil must be placed downstream of the DX cooling coil when reclaiming heat from that cooling coil’s compressor(s). In this example. For example. When the coil’s heating source is a direct expansion cooling coil (Coil:DX: BypassFactorEmpirical. For this application. For example. the selection of fan control for the air loop equipment is essential for proper modeling. The configuration in Figure 125 shows the heating coil being controlled via a temperaturebased strategy. When the heating source is a compressor rack for refrigerated cases. the air loop’s fan control mode may be auto fan (cycling fan cycling coil). High Humidity Control With HeatCool Configuration). The desuperheater heating coil can also be used with cooling/heating systems that maintain specific air loop (deck) temperatures. or variable volume since the desuperheater heating coil cycles on/off with the DX cooling coil. constant fan. Desuperheater heating coil used as a reheat coil with DXSystem:AirLoop Since the heating provided by the desuperheater coil is based on available waste heat from another system. or Coil:DX:MultiMode:CooingEmpirical). 4/23/05 392 . it can be used as an air reheat coil for high humidity control in the compound objects Furnace:Blowthru:HeatCool and UnitarySystem:Blowthru:HeatCool (Ref. Coil:DX:MultiSpeed:CooingEmpirical. the maximum heat reclaim recovery efficiency for this coil is 30%.

then the amount of reclaimed heat (average heating rate over the simulation time step) is set equal to the requested load and the desuperheater heating coil will cycle off when the requested load is satisfied. The latter of these optional fields is the parasitic electric load. The first of these is the coil temperature setpoint node name.g. with a range of 0% to 30%). and the heat reclaim recovery efficiency (default of 25%. Furnace:Blowthru:HeatCool).g. If the heating capacity is greater than the heating load. If the temperature setpoint node name is entered. This method is used when the heating coil is controlled to maintain a dry-bulb temperature setpoint in an air loop. The user must also enter the desuperheater heat source type and name which are validated when the model inputs are read into the program. 4/23/05 393 • • . Model Inputs A minimum of seven inputs must be defined for the desuperheater heating coil. The next two inputs are the heating coil’s inlet and outlet air node names. The model first calculates the heating coil capacity based on the total condenser waste heat rejected by the heating source. the control type is temperature-based operation. an availability schedule name. Load-Based Operation When load-based operation is selected. control valves). The user must input the coil’s name. Setpoint Manager) is used to place a temperature set point on this air node. When a temperature setpoint node name is not entered. This field is used when the desuperheater heating coil is controlled based on an air loop temperature and a Setpoint Manager (Ref. The following sections describe the calculations used for both the load-based and temperature-based strategies that can control the desuperheater heating coil.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS NOTE: Use of the desuperheater heating coil in variable air volume systems should be done with caution since the model assumption of a fixed heat reclaim recovery efficiency may not be valid if the air flow rate over the coil varies significantly.. the control type is assumed to be load-based operation. Q coil = Q cond × ε reclaim • • where: • Q coil • = heating capacity of the desuperheater coil (W) = total amount of condenser waste heat rejected by the heating source (W) = heat reclaim recovery efficiency ε Q cond reclaim The model then compares the coil’s heating capacity to the heating load being requested. a load to be met by the heating coil is requested by the parent object (e. This field allows the user to define the electric energy use of parasitic equipment associated with the desuperheater heating coil (e. Model Description The model calculates the thermal performance of the heating coil based on the control strategy used. Q reclaim = Q load Otherwise the amount of reclaimed heat is set equal to the coil’s heating capacity and the desuperheater heating coil will operate the entire time that the waste heat source operates. The final two fields are optional. This method should be used when this coil is specified as the air reheat coil for high humidity control with the compound object Furnace:Blowthru: HeatCool or Unitary System:Blowthru:HeatCool.

4/23/05 394 . returns enthalpy given dry-bulb temperature and humidity ratio Temperature-Based Operation When temperature-based operation is selected. the coil’s leaving air temperature is then calculated based on the amount of heat recovered and the air mass flow rate through the coil. RTFcoil where: • ⎛Q ⎞ = RTFsource ⎜ •reclaim ⎟ ⎜ Q ⎟ ⎝ coil ⎠ RTFcoil RTFsource = runtime fraction of the desuperheater heating coil = runtime fraction of the desuperheater heating source The heating coil’s outlet air humidity ratio and air mass flow rate are simply set equal to the coil inlet air values. Q coil = Q cond × ε reclaim • • The model then calculates the heating load required to reach the desired set point temperature. J/kg = humidity ratio of the air leaving the heating coil. • Tout = Tin + where: Q reclaim • m Cp Tout = coil outlet air temperature (°C) Tin = coil inlet air temperature (°C) • m = air mass flow rate through the heating coil (kg/s) C p = specific heat of air entering the heating coil (J/kg-C) A final calculation is made to determine the runtime fraction of the desuperheater heating coil. Since the maximum amount of reclaim heat available is dependent on the runtime fraction of the waste heat source. ωout ) where hout = enthalpy of the air leaving the heating coil. hout = PsyHFnTdbW (Tout .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • • COILS Q reclaim = Q coil In either case. the model first calculates the heating coil capacity based on the total condenser waste heat rejected by the heating source (same as for load-based operation). the runtime fraction of the desuperheater heating coil is the product of the waste heat source’s runtime fraction and the ratio of the amount of heat reclaimed to the desuperheater coil’s heating capacity. kg/kg ωout PsyHFnTdbW = EnergyPlus psychrometric function. The outlet air enthalpy is calculated based on the outlet air dry-bulb temperature and the outlet air humidity ratio.

• • RTFcoil • ⎛Q ⎞ = RTFsource ⎜ •reclaim ⎟ ⎜ Q ⎟ ⎝ coil ⎠ The heating coil’s outlet air humidity ratio and air mass flow rate are simply set equal to the coil inlet air values. If the calculated load is less than or equal to 0. Q reclaim = Q load Tout = Tsetpoint A final calculation is made to determine the runtime fraction of the desuperheater heating coil. the amount of reclaimed heat is set equal to the calculated heating load and the outlet air dry-bulb temperature is set equal to the set point temperature.Tin ) where: • Q load Tsetpoint = heating load to meet the desired dry-bulb temperature set point (W) = desired dry-bulb temperature set point (°C) A comparison is then made to determine if the desuperheater heating coil should be energized. hout = PsyHFnTdbW (Tout . In this case the desuperheater heating coil will cycle off when the requested load is satisfied. The outlet air enthalpy is calculated based on the outlet air dry-bulb temperature and the outlet air humidity ratio. the runtime fraction of the desuperheater heating coil is the product of the waste heat source’s runtime fraction and the ratio of the amount of heat reclaimed to the desuperheater coil’s heating capacity. Q reclaim = 0 Tout = Tin If the amount of heat required to meet the set point is larger than the desuperheater coil’s heating capacity. 4/23/05 395 . ωout ) Parasitic Electric Load The parasitic electric load attributed to the desuperheater heating coil is calculated using the user-supplied input value ( Pparasitic ) and the desuperheater heating coil runtime fraction calculated above. Since the maximum amount of reclaim heat available is dependent on the runtime fraction of the waste heat source.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • • COILS Q load = m C p (Tsetpoint . then the amount of reclaimed heat is set equal to the coil heating capacity and the outlet air dry-bulb temperature is calculated. the coil’s inlet air dry-bulb temperature is at or above the set point. • Q reclaim = Q coil • • • Tout = Tin + Q reclaim • m Cp If the amount of heat required to meet the set point is less than the desuperheater coil’s heating capacity. In this case the desuperheater heating coil is not active and the outlet air dry-bulb temperature is set equal to the inlet air dry-bulb temperature.

and reuses this energy to reheat the supply air leaving the cooling coil. The heat exchanger and cooling coil are then successively modeled using the calculation routines specific to the type of heat exchanger and cooling coil selected. Convergence is reached when the change in this air temperature for successive iterations is within a 4/23/05 396 . The air temperature exiting the cooling coil is compared with the air temperature exiting the cooling coil on the previous modeling iteration for this simulation time step. the air-to-air heat exchanger precools the air entering the cooling coil. The input syntax for these compound objects can be found in the EnergyPlus Input/Output Reference. As shown in Figure 126. When modeling the air-to-air heat exchanger during the first modeling iteration. Figure 126. EnergyPlus has two compound objects to model this scenario: Coil:DX:CoolingHeatExchangerAssisted and Coil:Water:CoolingHeatExchangerAssisted. This heat exchange process improves the latent removal performance of the cooling coil by allowing it to dedicate more of its cooling capacity toward dehumidification (lower sensible heat ratio). Modeling of the heat exchanger assisted cooling coil begins by initializing the air mass flow rate (based on the air mass flow rate placed on the compound object’s inlet air node) and passing this value to the exhaust air inlet node of the air-to-air heat exchanger.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COILS Pelec = Pparasitic × RTFcoil The model assumes that this electric load is small and does not contribute to heating the air stream. Coil Model -. The detailed modeling calculations for the individual components (air-to-air heat exchangers and cooling coils) are described elsewhere in this document.Heat Exchanger Assisted Cooling Coils (HVAC) An air-to-air heat exchanger can be used to enhance the dehumidification performance of a conventional cooling coil. Schematic of a heat exchanger assisted cooling coil Modeling of the heat exchanger assisted cooling coil is performed by consecutively modeling the air-to-air heat exchanger and the cooling coil until convergence on a solution is achieved. the cooling coil exiting air temperature from the previous simulation time step is used as the heat exchanger’s exhaust air inlet condition.

the controller monitors two set point values.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CONTROLLERS (HVAC) specified tolerence (0. the chilled water flow rate is increased. The simple controller may also be used to control both high temperature and high humidity levels by controlling the water flow rate through a chilled water coil. HeatPump: Water To Air (HVAC) See discussion at Unitary System: See discussion at Unitary System: Controllers (HVAC) Controller:Simple The simple controller is really a solution inverter. In this case. The action parameter input is set to REVERSE for chilled water cooling coils and NORMAL for hot water heating coils. this “controller” will numerically step through all of the water flow possibilities by an interval-halving technique until the mass flow rate is determined to meet the specified outlet air temperature within a specified user tolerance. HeatPump: Water To Air (HVAC) Coil Model – Water to Air HeatPump. Simple Controller used with Central Chilled Water Coil In this case. Coil Model – Water to Air HeatPump. The figure below illustrates the use of a simple controller used with a central chilled water coil (control variable TEMP). one for temperature control and the other for high humidity 4/23/05 397 . Setting the controller’s control variable to TEMPandHUMRAT accomplishes this task. The controller reads the desired temperature set point from the control node (established by a Set Point Manager) and modulates the chilled water flow rate at the actuator node in order to meet the desired supply (coil outlet) air temperature. Cooling This is object name: Coil:WaterToAirHP:Cooling.001°C). Consecutive modeling of the heat exchanger and cooling coil is terminated and a warning message is issued if the number of modeling iterations exceeds 15. If the sensed temperature is above the desired set point temperature. Figure 127. the controller simply senses the temperature at the control node and compares this value with the desired temperature set point. Heating This is object name: Coil:WaterToAirHP:Heating. Thus. For a water coil the simulation cannot be inverted where the mass flow rate of the water through the coil can be solved directly given an air temperature.

Figure 128. Pa Tdp = supply (outlet) air dew point temperature. If high humidity control is the limiting case then colder supply air will be delivered by the cooling coil to achieve proper dehumidification. and some form of air reheat may be required to avoid overcooling of the zones being served by this air loop.. P ) Tapproach = TSA − Tdp where: Tapproach = approach temperature. The limiting case for either temperature or high humidity control (i.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CONTROLLERS (HVAC) control. ˚C ωSA = supply (outlet) air humidity ratio. The model first calculates the approach temperature using the dry-bulb temperature and dew point temperature of the air leaving the water coil: Tdp = PsyTdpFnWPb (ωSA . Note that two set point managers must be used to establish these set points as shown in the figure below. the minimum supply air temperature required to meet both set points) is used for controlling the water flow rate through the chilled water coil. Two Set Point managers used in Controller:Simple Model Description TEMPandHUMRAT When the control variable TEMPandHUMRAT is used. the controller modulates water flow through a chilled water coil to meet both a temperature and a humidity ratio set point. These two set points are placed on the control node by set point managers. ˚C 4/23/05 398 . kg/kg P = outdoor barometric pressure.e.

SP = dew point temperature corresponding to ωSP. ˚C TTemp . TSP = MIN (TTemp . SP = Tdp . returns dew point temperature given humidity ratio and barometric pressure The supply air dew point temperature required to meet the humidity ratio set point (placed on the control node by Set Point Manager:Single Zone Max Hum) is then calculated as a function of the humidity ratio set point and barometric pressure as follows: Tdp . P ) where: Tdp . SP + Tapproach where: THR . SP = supply air temperature set point required for zone temperature control. SP = PsyTdpFnWPb (ωSP . the controller varies the chilled water flow rate through the coil using an interval-halving technique until the actual supply air temperature reaches Tsp within the specified tolerance: TActual − TSP ≤ Controller ConvergenceTolerance where: TActual = actual air temperature leaving the cooling coil. ˚C As described previously. and the minimum of these two set point values is used as the set point temperature for controlling the chilled water coil. SP + Tapproach : THR . ˚C ωSP = humidity ratio set point. SP = supply air dry-bulb temperature set point required to achieve the specified humidity ratio set point. SP the drybulb temperature of air leaving the cooling coil must be at Tdp . THR . ˚C PsyTdpFnWPb = EnergyPlus psychrometric function. ˚C The supply air temperature set point required to achieve the specified humidity ratio set point is then compared to the set point temperature required for zone temperature control. SP ) where: TSP = chilled water coil supply air temperature set point. ˚C 4/23/05 399 . SP . kg/kg In order for the dew point temperature of the coil’s supply (outlet) air to reach Tdp .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CONTROLLERS (HVAC) TSA = supply (outlet) air dry-bulb temperature.

without a traditional air distribution system) while the Controller:Outside Air is used with systems that utilize an air loop to provide conditioned air to one or more zones. except the addition of a few calculations that are unique for this stand alone ERV controller. In this way. the logic for this controller follows that established for the object Controller:Outside Air.e. The air flow rate through the stand alone ERV remains the same regardless of whether the controller is signaling for economizer (free cooling) operation or not. In some instances local variables used in the 4/23/05 400 . Controller Logic In many ways. Nearly the same computations (source code) are used for this controller as for Controller:Outside Air. this controller is only used with the stand alone energy recovery ventilator object (dedicated to serving a single zone.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE CONTROLLERS (HVAC) Controller:Outside Air Controller:Stand Alone ERV The stand alone energy recovery ventilator (ERV) controller is used solely in conjunction with a stand alone energy recovery ventilator (see figure below). However. this controller only signals the object Heat Exchanger:Air to Air:Generic that favorable conditions are available for free cooling and heat exchange should be suspended (i. air flow is fully bypassed around a fixed-plate heat exchanger or the rotation of a rotary heat exchanger is stopped). this controller is very similar to Controller:Outside Air with the field EconomizerChoice set to “BYPASS”. Figure 129. However.. Schematic of the Energy Recovery Ventilator:Stand Alone compound object This controller object mimics virtually all of the control logic available for a conventional airside economizer as embodied in the object Controller:Outside Air.

differential temperature control. when the tower fan is off but the water pump remains on. Model Description Based on Merkel’s theory. The model will also account for tower performance in the “free convection” regime. The inputs for this controller specify temperature and/or enthalpy conditions that are used to trigger economizer operation. The user can enter a high enthalpy limit.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) Controller:Outside air computations are set to specific values for Controller:Stand Alone ERV to allow the same computations and logic to be used for both controllers. If the user wishes to model differential temperature control. Bourdouxhe et al. Outdoor Air Temp > Exhaust Air Temp) THEN EconomizerOperationFlag = FALSE IF (EXHAUST AIR ENTHALPY LIMIT . single-point enthalpy and differential point enthalpy). Similar logic can be used with air enthalpy. Outdoor Air Enthalpy > Exhaust Air Enthalpy) THEN EconomizerOperationFlag = FALSE Cooling Towers (Plant) Overview The EnergyPlus cooling tower model is based on Merkel’s theory (Merkel 1925). As explained above the controller senses when air conditions are such that heat exchange by the air-to-air heat exchanger should be suspended to provide free cooling to the zone. The user can enter a high and low temperature limit for economizer operation. the steady-state total heat transfer between the air and water entering the tower can be defined by the following equation: 4/23/05 401 . EXHAUST AIR ENTHALPY LIMIT should be specified in the proper input field. thereby reducing the amount of mechanical cooling that needs to be provided by other equipment.AND. the model assumes a simple linear interpolation between two steady-state regimes without accounting for any cycling losses. The user still needs to set the low temperature limit to restart the heat exchange process when the outdoor temperature falls too low. and then checks each one of the limits that the user has set (single-point temperature control. The logic that is being applied for Controller:Stand Alone ERV is presented below. This logic represents a conventional single-point temperature economizer control. Regardless of modeling single-point enthalpy or differential enthalpy control.AND. If the user wishes to model differential enthalpy control. EXHAUST AIR TEMP LIMIT should be specified in the proper input field. In this case. then economizer operation is terminated and heat exchange between the supply and exhaust air streams is modeled: EconomizerOperationFlag = TRUE IF (Outdoor Air Temp < Temperature Low Limit) EconomizerOperationFlag = FALSE IF (Outdoor Air Temp > Temperature High Limit) EconomizerOperationFlag = FALSE IF (Outdoor Air Enthalpy > Enthalpy High Limit) ) EconomizerOperationFlag = FALSE IF (EXHAUST AIR TEMP LIMIT . heat exchange is suspended while air flow rates remain unchanged. and checks all limits established by the user in the object’s input data. 1994). The model is flexible. The model can be used to simulate the performance of both single speed and two speed mechanical-draft cooling towers. which is also the basis for the tower model included in ASHRAE’s HVAC1 Toolkit for primary HVAC system energy calculations (ASHRAE 1999. When the supply inlet (outdoor) air is between these two values. the user still needs to set the low temperature limit to restart the heat exchange process when the outdoor temperature falls too low. Cooling tower performance is modeled using effectiveness-NTU relationships for counterflow heat exchangers. For part-load operation. and heat exchange between the supply and exhaust air streams will be suspended when the outdoor air enthalpy falls below this value. heat exchange is suspended whenever the temperature of the exhaust air is greater than the temperature of the outdoor air. The model initially assumes that heat exchange should be suspended. This logic represents single-point enthalpy economizer control. If any of the limits entered by the user is exceeded.

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) dQtotal = where UdA ( hs − ha ) cp (456) hs = enthalpy of saturated air at the wetted-surface temperature. J/kg-oC U = cooling tower overall heat transfer coefficient. the wettedsurface temperature is assumed to be equal to the water temperature. it is also assumed that the moist air enthalpy is solely a function of the wet-bulb temperature and that the moist air can be treated as an equivalent ideal gas with its mean specific heat defined by the following equation: cpe = ∆h ∆Twb (457) where ∆h = enthalpy difference between the air entering and leaving the tower. m2 Equation (456) is based on several assumptions: • • • • air and water vapor behave as ideal gases the effect of water evaporation is neglected fan heat is neglected the interfacial air film is assumed to be saturated • the Lewis number is equal to 1 In this model. W/m2. oC Since the liquid side conductance is much greater than the gas side conductance. the expression for total heat transfer becomes: dQtotal = UedA (Tw − Twb ) where (458) Ue = U cpe cp Twb = wet-bulb temperature of the air. kg/s 4/23/05 402 .oC A = heat transfer surface area. oC An energy balance on the water and air sides of the air/water interface yields the following equations: dQtotal = mwcpw dTw dQtotal = macpe dTwb where (459) (460) mw = mass flow rate of water. J/kg ha = enthalpy of air in the free stream. Based on this assumption and equations (456) and (457). J/kg ∆Twb = wet-bulb temperature difference between the air entering and leaving the tower. oC Tw = temperature of the water. J/kg cp = specific heat of moist air.

and combining the result with equation (461) provides the following expression for cooling tower effectiveness: i ⎧ ⎡ ⎛i ⎞⎤ ⎫ 1 − exp ⎨− NTU ⎢1 − ⎜ C w / C a ⎟ ⎥ ⎬ ⎠⎦ ⎭ ⎣ ⎝ ⎩ ε = i i i i ⎧ ⎡ ⎛ ⎛ ⎞ ⎞⎤ ⎫ 1 − ⎜ C w / C a ⎟ exp ⎨− NTU ⎢1 − ⎜ C w / C a ⎟ ⎥ ⎬ ⎝ ⎠ ⎠⎦ ⎭ ⎣ ⎝ ⎩ (462) where C w = mwcpw and C a = macpe i i NTU = Number of TransferUnits = UAe i Cw This equation is identical to the expression for effectiveness of an indirect contact (i. by an equivalent counterflow heat exchanger as shown in the following figure. the cooling tower can be modeled. kg/s Assuming that the heat capacity rate ( mcp ) for the cooling tower water is less than that for the air. Therefore. oC Twout = outlet water temperature. oC Twbin = wet-bulb temperature of the inlet air.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) ma = mass flow rate of air. (459). and (460) and integrating over the entire heat transfer surface area. in the steady-state regime. 4/23/05 403 . the effectiveness of the cooling tower can be defined by analogy to the effectiveness of a simple heat exchanger: ε= Twin − Twout Twin − Twbin (461) where ε = heat exchanger effectiveness Twin = inlet water temperature. oC Combining equations (458).e.. fluids separated by a solid wall) counterflow heat exchanger (Incropera and DeWitt 1981).

Based on these inlet and outlet air conditions. Therefore. repeated here: cpe = ∆h ∆Twb With the overall heat transfer coefficient-area product for the cooling tower entered by the user. The exiting water temperature is then calculated based on an energy balance that assumes that the energy absorbed by the air is equivalent to the energy removed from the water. Since an iterative solution is required.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) Tw out water Tw in UAe equivalent fluid Twb in Twb out Figure 130. the effective heat transfer coefficient-area product is calculated by rearranging equation (463): 4/23/05 404 . it is assumed that the moist air enthalpy can be defined by the wetbulb temperature alone. its overall heat transfer coefficient-area product UAe. As explained previously. Method for Calculating Steady-State Exiting Water Temperature The objective of the cooling tower model is to predict the exiting water temperature and the fan power required to meet the exiting water set point temperature. the first step in the procedure is to calculate the enthalpy of moist air entering the cooling tower based on the ambient wet-bulb temperature from the weather file. Cooling Tower Schematic The first fluid is water and the second fluid is an equivalent fluid entering the heat exchanger at temperature Twbin and specific heat cpe . The procedure for calculating the steadystate. Since only the inlet air and inlet water temperatures are known at any simulation time step. the mean specific heat of the air is calculated based on equation (457). The actual cooling tower heat transfer coefficient-area product is related to UAe by the following expression: UA = UAe cp cpe (463) This heat transfer coefficient-area product is assumed to be a function of the air mass flow rate only and can be estimated from laboratory test results or manufacturers’ catalog data. an iterative procedure is required to determine the exiting fluid temperatures using the equations defined in the previous section. The heat exchanger is characterized by a single parameter. exiting air wet-bulb temperature is outlined below. the iterations are performed to determine the exiting wet-bulb temperature of the air. In the case of the EnergyPlus model. a first guess of the outlet air wet-bulb temperature is then made and the enthalpy of this estimated outlet air wet-bulb temperature is calculated.

the outlet water temperature is calculated as follows: Twout = Twin + Qtotal • Cw Calculating the Actual Exiting Water Temperature and Fan Power The previous section describes the methodology used for calculating the steady-state temperature of the water leaving the cooling tower. The cooling tower model seeks to maintain the temperature of the water exiting the cooling tower at (or below) a set point.C a ) and C max = Maximum (C w. The model first checks to determine the impact of “free convection”. The exiting water temperature calculations use the fluid flow rates (water and air) and the UAvalues entered by the user for each regime. Finally. the effectiveness of the heat exchanger is then calculated: i ⎧ ⎡ ⎛i ⎞⎤ ⎫ 1 − exp ⎨− NTU ⎢1 − ⎜ C min/ C max ⎟ ⎥ ⎬ ⎠⎦ ⎭ ⎣ ⎝ ⎩ ε = i i ⎧ ⎡ ⎛i ⎛i ⎞ ⎞⎤ ⎫ 1 − ⎜ C min/ C max ⎟ exp ⎨− NTU ⎢1 − ⎜ C min/ C max ⎟ ⎥ ⎬ ⎝ ⎠ ⎠⎦ ⎭ ⎣ ⎝ ⎩ where C min = Minimum (C w.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) UAe = UA cpe cp With cpe and UAe known. tower fan off) and with the tower fan operating (including low and high fan speed for the two-speed tower). If free convection is not specified by the user. then the exiting water temperature is initially set equal to the entering tower water temperature. If the user 4/23/05 405 . on the tower exiting water temperature. if specified by the user. This methodology is used to calculate the exiting water temperature in the free convection regime (water pump on.C a ) C w = mwcpw and C a = macpe i i i i i i i i NTU = Number of TransferUnits = UAe i C min The heat transfer rate is then calculated as follows: i Qtotal = ε C min (Twin − Twbin ) The outlet air wet-bulb temperature is then recalculated: Twbout = Twbin + Qtotal i Ca The iterative process of calculating Twbout continues until convergence is reached. The set point schedule is defined by the field “Loop Temperature Setpoint Schedule Name” for the CONDENSER LOOP object.

then the tower fan is increased to its high speed and the steady-state exiting water temperature is calculated. The model first checks to see if “free convection” is specified and if the resulting exiting water temperature is below the set point temperature. high) + (1 − ω )( Powerfan . off (464) Tset = exiting water set point temperature. oC The average fan power for the simulation time step is calculated by multiplying ω by the steady-state fan power specified by the user. HVAC1 Toolkit: A Toolkit for Primary HVAC System Energy Calculations. on = exiting water temperature with tower fan on.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE COOLING TOWERS (PLANT) specifies “free convection” and the steady-state exiting water temperature based on “free convection” is at or below the set point. Cyclic losses are not taken into account. If low-speed fan operation is unable to reduce the exiting water temperature below the set point. oC Twout . The calculation method for the two-speed tower is similar to that for the single-speed tower example described above. tower fan on for the entire simulation time step and tower fan off for the entire simulation time step). Refrigerating and Air-Conditioning Engineers. If the exiting water temperature at low fan speed is below the set point temperature. oC Twout . low ) References (466) ASHRAE 1999. oC The average fan power for the simulation time step is calculated for the two-speed cooling tower as follows: Powerfan .g. The model assumes that part-load operation is represented by a simple linear interpolation between two steady-state regimes (e. Atlanta: American Society of Heating. low speed fan power specified by the user. then a modified version of equation (464) is used to calculate runtime at high fan speed: ω= where Tset − Twout . low = exiting water temperature with tower fan at low speed. then the tower fan is not turned on. then the average fan power is calculated based on the result of equation (464) and the steady-state. avg = ω ( Powerfan. The fraction of time that the tower fan must operate is calculated based on the following equation: ω= where Tset − Twout . high − Twout . low Twout . then the model calculates the steady-state exiting water temperature with the tower fan at low speed. If the exiting water temperature remains above the set point after “free convection” is modeled. low (465) Tset = exiting water set point temperature. off Twout . high = exiting water temperature with tower fan at high speed. on − Twout . If not. then the tower fan is turned on to reduce the exiting water temperature to the set point. If this temperature is below the set point.. oC Twout . oC Twout . Inc. 4/23/05 406 . off = exiting water temperature with tower fan off.

field EconomizerChoice). 1925.g. CO2-based DCV.P. Cooling tower model developed in a toolkit for primary HVAC system energy calculation: part 1.. The VENTILATION:MECHANICAL object simplifies the procedure for calculating these outside air ventilation requirements and reseting them based on varying occupancy levels. Liege (Belgium).1 (ASHRAE 2003) currently requires outdoor air ventilation rates to be determined based on the floor area of each occupied zone plus the number of people in each zone. CO2-based demand controlled ventilation (DCV) is being increasingly used to modulate outside ventilation air based on real-time occupancy (Emmerich and Persily 1997. Schell and Int-Hout 2001). and any components and controllers associated with the system relief air and outside air streams.. December 5-7. Fundamentals of Heat Transfer. Incropera. Grodent. VDI Forschungarbeiten. Economizer operation can also be specified to increase the outside air flow above the minimum flow rate to provide free cooling when conditions permit (Controller:Outside Air.P. using the Ventilation:Mechanical object in conjunction with the Controller:Outside Air object. CONTROLLER:OUTSIDE AIR controls the amount of outside ventilation air introduced via the mixed air box based on several user inputs. Proceedings of the fourth international conference on system simulation in buildings. New York: John Wiley & Sons. J. Demand Controlled Ventilation ASHRAE Standard 62. Determining the outside air ventilation rate proportional to occupancy and introducing this ventilation via the mixed air box are accomplished by the mechanical ventilation and outside air controller objects. 1981. for a variable-air volume system) or a fixed minimum outside air flow rate (not as a percentage but a fixed value) (field MinimumLimit). variations in occupancy).. This is particularly useful for large air distribution systems that serve a number of different zone types with varying occupancy levels. The user must include the following three objects in their input data file in order to model CO2-based DCV: OUTSIDE AIR SYSTEM to simulate the mixed air box of the air loop VENTILATION:MECHANICAL to determine the minimum outside air flow rate to be provided by the mixed air box CONTROLLER:OUTSIDE AIR to control the outside air flow rate introduced via the mixed air box The Outside Air System is a subsystem of an Air Primary Loop which handles the mixed air portion of the primary air system: the system relief air. Verduftungskuhlung. 1994. contains provisions that allow building ventilation systems to vary the amount of outside ventilation air delivered to occupied zones based on feedback from sensors that monitor various indoor air contaminants (ASHRAE 2001). Modulating the outside ventilation air while maintaining proper indoor air quality has the potential for large energy savings compared to constant rate ventilation systems that are typically designed to provide outside ventilation air based on maximum anticipated occupancy. Silva. carbon dioxide (CO2) levels can be monitored as an indicator of building occupancy and the associated human bioeffluent concentration. Schell et al.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE DEMAND CONTROLLED VENTILATION Bourdouxhe. EnergyPlus currently models CO2-based DCV by the simplified method of providing outside ventilation air in proportion to the number of people located in the zones being served by an air distribution system (air loop). F. Lebrun and C. The Ventilation Rate Procedure of ASHRAE Standard 62.P. Merkel. DeWitt. The user can define the minimum outside air flow rate as a percentage of the system’s supply air flow rate (e. Berlin. The inputs for this object are fully described in the EnergyPlus Input Output Reference. the outside air inlet.g. and D. 4/23/05 407 . 1998. J. 1994. F. Ventilation for Acceptable Indoor Air Quality. No 275. allows a third option for setting the minimum outside air flow. Although not a contaminant of concern in most buildings. The outside air ventilation rate can be reset dynamically as operating conditions change (e. M.

6. • ⎛m ⎞ ⎟ OutAirMinFrac = ⎜ min OA • ⎜ m max branch ⎟ ⎝ ⎠ 4/23/05 408 .204 kg/m3) adjusted for the local barometric pressure (standard barometric pressure corrected for altitude.1).Air Flow Paths The minimum outside air flow rate is first calculated based on the minimum outside air flow rate and the minimum outside air schedule value as defined by the user inputs for the object CONTROLLER:OUTSIDE AIR: m min OA = V min OA ∗ MinOAScheduleValue ∗ RhoStd where: • • • m min OA • = minimum outside air flow rate for this time step. The method used to calculate the outside ventilation air flow rate for each system simulation time step is described in more detail below. Demand Control Ventilation -. The figure below schematically illustrates air flow paths used in the calculation of outside air flow rate.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE DEMAND CONTROLLED VENTILATION EnergyPlus uses the largest outside air flow rate calculated by the various methods described above when modeling system performance (as long this rate doesn’t exceed the maximum flow rate specified for the main air loop branch or for the outside air controller itself). Exhaust Air Supply Air Air Loop Equipment Mixed Air Zone Return Air OA Mixer Outside Air Relief Air Figure 131. The outside air minimum fraction is then calculated as the ratio of the minimum outside air flow rate calculated above to the maximum air flow rate defined in the BRANCH statement for the main air loop (converted to mass flow rate). m3/s identified in CONTROLLER:OUTSIDE AIR V min OA MinOAScheduleValue = minimum outside air schedule value defined by the schedule RhoStd = standard air density (1. ASHRAE 1997 HOF pg. kg/s = minimum outside air flow rate defined in CONTROLLER:OUTSIDE AIR.

m2 = ventilation rate per occupant for zone (or zone list) “i”.i Pi ⎞ ⎥ ∗ RhoStd ⎜ ⎟ ⎠⎦ ⎣ i =1 ⎝ • m max branch where: MechVentOutsideAirMinFrac = outside air minimum fraction based on all zones specified in the Ventilation:Mechanical object • V A . determined automatically by the model based on the zone name or zone list name provided.i Pi i The algorithms then uses the larger of these outside air minimum fractions in subsequent calculations. kg/s The program then calculates the minimum outside air flow fraction based on the information provided in the object VENTILATION:MECHANICAL and the maximum branch air flow rate as follows: • ⎡ All Zones • ⎤ MechVentOutsideAirMinFrac = ⎢ ∑ ⎛V A. OAFrac = MAX (OutAirMinFrac. 0. people = index for the zone name Ventilation:Mechanical object or zone list names specified in the V P .i Ai • = ventilation rate per unit floor area for zone (or zone list) “i”. the outside air flow fraction is increased beyond the minimum fraction to meet the mixed air setpoint temperature (set point temperature assigned to the node defined in field “Control_Node” of CONTROLLER:OUTSIDE AIR). EconomizerOAFrac) The mass flow rate of outside air is then calculated based on the outside air fraction determined above and the mixed (supply) air mass flow rate: mOutsideAir = OAFrac * m MixedAir where: • • 4/23/05 409 .SIMULATION MODELS – ENCYCLOPEDIC REFERENCE DEMAND CONTROLLED VENTILATION where: • m max branch = Max branch air volume flow rate times RhoStd.1. determined automatically by the model based on the zone name or zone list name provided. If conditions permit economizer operation.0 ) The algorithm goes on to determine if economizer operation is possible based on the user inputs and the current conditions of the outdoor air and return air. m3/s-person = number of occupants for zone (or zone list) “i”.i Ai +V P . OutAirMinFrac = MAX ( OutAirMinFrac. m3/s-m2 = floor area for zone (or zone list) “i”. MechVentOutsideAirMinFrac ) OutAirMinFrac = MIN ( MAX ( OutAirMinFrac.0 ) . and also makes sure that the resulting fraction is between 0 and 1.

and A. ASHRAE.ashrae. Atlanta: American Society of Heating.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • DEMAND CONTROLLED VENTILATION mOutsideAir OAFrac • = mass flow rate of outside air. the relief air flow rate is calculated as the difference between the outside and exhaust air mass flow rates. Refrigerating and Air-Conditioning Engineers.J. kg/s = fraction of outside air in the mixed (supply) air stream = mass flow rate of the mixture of return air and outside ventilation air. ASHRAE Transactions 103(2):229-243. m MixedAir • ( • • The outside air mass flow rate should also be less than or equal to the maximum outside air flow rate specified by the user. m Re liefAir = MAX m OutsideAir − m ExhaustAir . m min OA • ( • • ) ) The outside air mass flow rate should be less than or equal to the mixed (supply) air flow rate. 2001. Addendum n to Standard 62-2001. m ExhaustAir • ( • • ) If a fixed minimum outside air flow rate is specified (field MinimumLimit in Controller:Outside Air) for a continuous air flow system. Literature review on CO2-based demand-controlled ventilation.0 References • ( • • ) ASHRAE. and the outside air flow rate is reset if necessary.Ventilation for acceptable indoor air quality.K. m OutsideAir = MAX m OutsideAir . and the outside air flow rate is reset if necessary. Persily. Emmerich. m OutsideAir = MIN m OutsideAir . ANSI/ASHRAE Standard 62-2001. S. m OutsideAir = MAX m OutsideAir . m max OA where: • ( • • ) = maximum outside air mass flow rate. kg/s m MixedAir The algorithm checks to make sure the calculated outside air mass flow rate is greater than or equal to the air flow rate being exhausted. 1997. kg/s = maximum outside air volume flow rate from CONTROLLER:OUTSIDE AIR times RhoStd Finally. 0. Inc. http://www. the program makes sure that the outside air mass flow rate is greater than or equal to the minimum outside air flow rate specified by the user. 4/23/05 410 . m OutsideAir = MIN m OutsideAir .org . 2003.

The electric load center takes all of the building. The purchased electrical demand limit is the user input for the demand limit above which the generators will try and meet the entire electrical load on the building minus the photovoltaic array if available. and “TRACK ELECTRICAL”.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE ELECTRIC LOAD CENTER DISTRIBUTION MANAGER Schell. Electric Load Center Generators The electric load center generators (object name: Electric Load Center:Generators) provide a set of scheduled electric generators for electric power generation. “DEMAND LIMIT”. ASHRAE Journal. February. R. system loads and sums them and provides operating schemes for the Generators. Application of CO2-based demand controlled ventilation using ASHRAE Standard 62-1989. Schell. M. 4/23/05 411 . The total electric purchased are both in Power and Energy units. Electric Load Center Distribution Manager The electric load center distribution manager (object name: Electric Load Center:Distribution) operates generators specified in the simulation and reports the amount of generated and purchased electrical for the input file. The DEMAND LIMIT scheme limits the amount of purchased electrical from the utility to the amount specified in the input. An optional controller (see object Controller:Stand Alone ERV) may be used to simulate economizer (free cooling) operation. Therefore.O.C. This value is positive when the amount of energy is purchased from the utility. This compound object consists of 3 required components: a generic air-toair heat exchanger (see object Heat Exchanger:Air to Air:Generic). Here is where the user lists what generators and PV’s are available at any given time. 2001. and an exhaust air fan (see object Fan:Simple:OnOff). The Demand Limit scheme tries to have the generators meet all of the demand above the purchased electric limit set by the user in the next field. The excess will either be available for storage or sell back to the power company. If a generator is needed in the simulation for a small load and it is less than the minimum part load ratio the generator will operate at the minimum part load ratio and the excess will either reduce demand or the excess energy will be available for storage or sell back to the power company. The BASELOAD scheme operates the generators at their rated capacity when the generator is scheduled ON. 1998. M. The internal meters used by EnergyPlus for reporting do all of the demand tracking. Therefore only “1” Electric Load Center:Distribution object is allowed per input file and all electric equipment is considered to be on this one electric loop. The available operating schemes are “BASELOAD”. Int-Hout.Stand Alone (HVAC) The stand alone energy recovery ventilator (ERV) (object name: Energy Recovery Ventilator:Stand Alone) is a single-zone HVAC component used for exhaust air heat recovery (see figure below). Demand control ventilation using CO2. The TRACK ELECTRICAL scheme tries to have the generators meet all of the electrical demand for the building.. then purchased electrical will offset the difference. Shim. a supply air fan. The electric produced from photovoltaic arrays will be reported in the electric produced report variable and will reduce the demand that the generators will try to meet for that time step. For more details on the individual inputs required see the InputOutput Reference Document. Turner.B. and Photo Voltaic cells to supply. Energy Recovery Ventilator. This value can be negative when the total electric produced is greater than the facility electrical needs. For more details on the individual inputs required see the InputOutput Reference Document. and D. S. if DEMAND LIMIT or TRACK ELECTRICAL is utilized and the available generators are not enough to meet demand. What is not provided by the on-site generation equipment is met by purchasing off-site electric. All demand not met by available generator capacity will be met by purchased electrical. Both the DEMAND LIMIT and TRACK ELECTRICAL schemes will sequentially load the available generators. ASHRAE Transactions 104(2):12131225.

exhaust air fan and heat exchanger parasitics). Ref.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE ENERGY RECOVERY VENTILATOR. The stand alone ERV object can be used in conjunction with an economizer feature whereby heat exchange is suspended whenever free cooling is available (i. Heat Exchanger:Air to Air:Generic for additional guidance) CONTROLLED ZONE EQUIP CONFIGURATION ZONE EQUIPMENT LIST OUTSIDE AIR INLET NODE LIST Model Description The purpose of this compound component is to simply call the individual component models and optional controller for each energy recovery ventilator. sensible and latent energy supplied to the zone. 4/23/05 412 . the compound object sets the supply and exhaust air mass flow rates through the ventilator.e. the input data file should include the following objects: ENERGY RECOVERY VENTILATOR:STAND ALONE HEAT EXCHANGER:AIR TO AIR:GENERIC FAN:SIMPLE:ONOFF (supply air) FAN:SIMPLE:ONOFF (exhaust air) CONTROLLER:STAND ALONE ERV (if economizer [free cooling] operation is desired) SET POINT MANAGER:SCHEDULED (if supply air outlet temperature control is used. as well as the total electrical energy consumed by all of the individual components (supply air fan. Schematic of the Energy Recovery Ventilator:Stand Alone compound object This compound object models the basic operation of supply and exhaust air fans and an airto-air heat exchanger. To model a stand alone ERV connected to a single zone. This compound object is also used to report the total. The stand alone ERV operates whenever the unit is scheduled to be available (Availability schedule). air flow is fully bypassed around a fixed-plate heat exchanger or the rotation of a rotary heat exchanger is stopped).STAND ALONE (HVAC) Energy Recovery Ventilator:Stand Alone Heat Exchanger:Air to Air:Generic Zone Supply Air Node Supply Air Fan Supply (Outside) Air Inlet Node Zone Zone Exhaust Air Node Exhaust Air Inlet Node Exhaust Air Fan Controller:Stand Alone ERV (Optional) Figure 132.. Since this equipment is not associated with an air loop.

the compound object then calls the generic air-to-air heat exchanger model to determine its supply air and exhaust air exiting conditions based on the inputs specified in the heat exchanger object. The sensible heat transfer rate to the zone by the stand alone ventilator is then calculated as follows: Q Sensible = m Supply ( hSupplyOutlet − hExhaustInlet ) HR min where: • • • Q Sensible • = sensible energy transfer rate to the zone. J/kg = enthalpy of the air being exhausted from the zone through the ventilator. the minimum humidity ratio of the supply air outlet or the exhaust air inlet 413 m Supply hSupplyOutlet hExhaustInlet HRmin 4/23/05 .STAND ALONE (HVAC) During each similation time step. m3/s = volumetric flow rate of the exhaust air stream. The heat exchanger and fan models are described in detail elsewhere in this document (reference: Heat Exchanger:Air to Air:Generic and Fan:Simple:OnOff).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE ENERGY RECOVERY VENTILATOR. kg/s = density of dry air at local barometric pressure (adjusted for altitude) and 20 ºC. kg/s = enthalpy of the air being supplied to the zone. m3/s m Exhaust ρ • V Supply • V Exhaust With the supply and exhaust inlet air mass flow rates set.0 • m Supply • = mass flow rate of the supply air stream. W = mass flow rate of the supply air stream. The supply air and exhaust air fans are then modeled to determine the final conditions of the air streams exiting the stand alone energy recovery ventilator. the air mass flow rate at the supply air and exhaust air inlets is set based on the stand alone ERV’s availablility schedule and the specified volumetric air flow rates as follows: IF (availability schedule value > 0) THEN m Supply = ρV Supply m Exhaust = ρV Exhaust ELSE • • • • • m Supply where: • = m Exhaust = 0. kg/m3 = volumetric flow rate of the supply air stream. J/kg = enthalpies evaluated at a constant humidity ratio. kg/s = mass flow rate of the exhaust air stream.

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE ENERGY RECOVERY VENTILATOR. W = sensible energy transfer rate to the zone. a controller is coupled to the stand alone ERV by providing the name of the controller object in the ERV Controller input field.. as well as the electric power and consumption by the ventilator. W = mass flow rate of the supply air stream. When economizer (free cooling) operation is desired. the sensible. Controller:Stand Alone ERV).STAND ALONE (HVAC) The resulting sensible energy transfer rate is passed to the zone equipment manager and added to the zone load to be met by other heating or cooling equipment. the stand alone heat exchanger must be modeled first in the list of zone equipment. At the end of each HVAC simulation time step. this compound object reports the heating or cooling rate and energy delivered to the zone. J/kg = enthalpy of the air being exhausted from the zone through the ventilator. Since the stand alone ERV is intended to reduce the outdoor air load through heat exchange and not meet that load completely. latent and total energy transfer rate to the zone is calculated as follows: QTotal = m Supply ( hSupplyOutlet − hExhaustInlet ) Q Sensible = m Supply ( hSupplyOutlet − hExhaustInlet ) HR min • • • • • Q Latent where: • =Q • Total −Q • Sensible QTotal • = total energy transfer rate to the zone. This is accomplished by setting the stand alone ERV priority for cooling and heating higher than that of other zone cooling or heating equipment (reference: ZONE EQUIPMENT LIST).e. This controller determines when the air-side economizer is active (i. The following calculations are representative of what is done for each of the energy transfer rates: • IF ( Q Total < 0.0 ) THEN 4/23/05 414 . J/kg = enthalpies evaluated at a constant humidity ratio. air flow is fully bypassed around a fixed-plate heat exchanger or the rotation of a rotary heat exchanger is stopped) based on the controller inputs (Ref. kg/s = enthalpy of the air being supplied to the zone. the minimum humidity ratio of the supply air outlet or the exhaust air inlet Q Sensible • Q Latent • m Supply hSupplyOutlet hExhaustInlet HRmin Since each of these energy transfer rates can be calculated as positive or negative values. W = latent energy transfer rate to the zone. In terms of thermal energy delivered to the zone. individual reporting variables are established for cooling and heating and only positive values are reported.

J’ TimeStepSys = HVAC system simulation time step. for water feed by gravity from above the media. shown in the figure below. W’ QTotalHeating In addition to heating and cooling rates.0 Total ) QTotalHeating ELSE • QTotalCooling • = 0. The water is pumped from the reservoir to a water distribution header. The following example for total cooling energy is representative of what is done for the sensible and latent energy as well as the heating counterparts.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE • • EVAPORATIVE COOLERS QTotalCooling • = ABS ( Q = 0. consists of a rigid media evaporative pad. with water recirculated from a reservoir. the heating and cooling energy supplied to the zone is also calculated for the time step being reported.0 • QTotalHeating where: • = QTotal QTotalCooling • = output variable ‘Stand Alone ERV Zone Total Cooling Rate. The direct stage is used for comfort cooling in a building where adding humidity to the air can be tolerated. the moisture content of the leaving air is higher than the entering condition. The evaporative pad provides the area for the adiabatic saturation of the air. While the process provides a lower dry bulb temperature. EvapCooler:Direct:CelDekPad The direct stage. W’ = output variable ‘Stand Alone ERV Zone Total Heating Rate. 4/23/05 415 . QTotalCooling = QTotalCooling ∗ TimeStepSys ∗ 3600. hr This group of objects describes the properties and configuration for the evaporative coolers models for the HVAC section. where: • QTotalCooling Evaporative Coolers = output variable ‘Stand Alone ERV Zone Total Cooling Energy.

or adiabatic cooling. it is assumed that the wet bulb temperature is constant across the direct evaporative stage. Psychrometric Chart -.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE Evaporative Media EVAPORATIVE COOLERS Supply Air Flow A B Sump Figure 133. Since the deviation of the constant wet-bulb line and the constant enthalpy line is small. Saturation Efficiency Since the evaporative process is not 100% efficient the saturation efficiency is defined by. the leaving dry bulb temperature would equal the entering wet bulb temperature. Direct Stage Evaporative Cooler The thermodynamic process is a simultaneous heat and mass transfer. the leaving dry bulb temperature can be expressed by the following equation. The efficiency of the direct evaporative process is less than 100% and by defining saturation efficiency (εse) for the direct stage or evaporative pad.Constant Enthalpy If the direct evaporative process were 100% efficient. and follows a constant enthalpy line on the psychrometric chart. it is shown in the figure below as a process from A to B. 4/23/05 416 . 180 Dry Air] Humidity Ratio [Grains/Lb 80 160 140 120 70 Wet-Bulb Temp [F] B 100 60 50 40 30 A 80 60 40 20 0 30 40 50 60 70 80 Dry-Bulb Temp [F] 90 100 110 Figure 134.

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE EVAPORATIVE COOLERS The saturation efficiency is determined from manufacturer's data. The pad geometry is constant throughout the simulation. adiabatic saturation in evaporative cooling is only an approximation. the leaving dry bulb temperature can be determined directly. 4/23/05 417 . since the water recirculates rapidly and approximates the wet bulb temperature at steady state operation. This gives the dry bulb temperature leaving the evaporative cooler. the evaporative process follows the constant wet bulb line or the constant enthalpy line on the psychrometric chart. therefore the wet bulb temperature is constant from inlet to outlet. friction from water re-circulation. The saturation efficiency of the direct evaporative cooler is a function of the pad geometry and airflow rate. with the leaving dry bulb temperature and assuming adiabatic heat transfer across the direct stage. Therefore. Using the saturation efficiency (εse) for the direct stage evaporative pad. heat transfer from surroundings. The Figure below shows a typical graph of manufacturer's data for the saturation efficiency. A multi-variate least squares curve fit of the data was used to generate saturation efficiency functions for the evaporative models that use the CelDek rigid media pad. The evaporative process approximately follows a constant wet bulb line. Curve Fitting Evaporative Media The saturation efficiency is usually reported as a function of airflow. Some things that can cause departure from the ideal adiabatic saturation process in the direct evaporative cooler are: • • • makeup water entering the sump. but the airflow rate can change from hour to hour when the evaporative cooler is used with an air economizer. and pad thickness. Assuming adiabatic heat transfer across the direct stage. however the adiabatic saturation assumption in the rigid-media cooler is good. and the least squares curve fit is discussed in Curve Fitting Evaporative Media section. Thus. pad face velocity. The saturation efficiency would then be determined from the flow for that hour with the geometry of the direct evaporative cooler. the outlet conditions for the direct stage are known. • solar radiation (sun upon a cooler).

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE EVAPORATIVE COOLERS Saturation Efficiency vs. has a rigid media pad. with water recirculated from a reservoir.0 Saturation Efficiency 0.5 2.13137 (Depth3 * Airvel) + 0. The water is pumped from the reservoir to a water distribution header.1 1. The model uses the air velocity (Airvel) through the pad and the depth of the media (Depth). where the adiabatic cooling takes place.914869 (Depth * Airvel) . The least squares routine produced the following model that is used for the evaporative cooling rigid media pad.5 4.018992 (Airvel3 * Depth) + 1.6 0.0.03215 (Depth2) + 0.0. The moist secondary air is then exhausted to the environment. εse = 0.1.1542 m Pad 0.5 1.0 0.5 0. The least squares routine produced an eleven-term multi-variate fit using a third order quadratic. Graph of Saturation Efficiency The curve fit for saturation efficiency was obtained using the functions listed below.0327622 (Airvel3 * Depth2) .0 3. The secondary air leaves the rigid media pad and enters an air-to-air heat exchanger where it cools the supply air flowing through the heat exchanger tubes. shown in the figure below.4572 m Pad 0.145384 (Depth3 * Airvel2) Where Airvel is in meters per second and Depth is in meters.8 0.1.5 5.48241 (Airvel * Depth2) . The secondary air stream has its own fan.0.0 4.0 1.0 2. for water feed by gravity from above the media.7 0.consists of a rigid media evaporative pad. 4/23/05 418 . EvapCooler:InDirect:CelDekPad The dry coil indirect evaporative cooler.0 Air Velocity [m/sec] Figure 135.5 3.958569 (Depth) .6096 m Pad 0.9 0. similar to the direct evaporative stage. This curve fit is used for the rigid media in the EvapCooler:Direct:CelDekPad and EvapCooler:InDirect:CelDekPad. Pad Thickness [m] 1.3048 m Pad 0. The evaporative pad provides the area for the adiabatic saturation of the air.0262659 (Airvel2) + 0.25193 (Airvel) .792714 + 0.

is shown by the dashed lines in the following figure.Indirect Dry Coil The process that the secondary air goes through. Secondary Air Process -.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE EVAPORATIVE COOLERS Secondary Fan D Supply Air Flow A Air-To-Air Heat Exchanger B Ambient Conditions Tdb. 4/23/05 419 . All the evaporative cooling takes place in the rigid media pad. thus no mineral deposits are left on the heat exchange surface to reduce the efficiency of the heat exchanger. The following equations are used to determine the dry-bulb temperature leaving the evaporative media. Dry Coil Indirect Evaporative Cooler The dry coil indirect evaporative cooler has a rigid media pad similar to the direct evaporative stage. 180 Dry Air] Humidity Ratio [Grains/Lb 80 160 140 120 70 Wet-Bulb Temp [F] C D 100 B 60 50 40 30 A 80 60 40 20 0 30 40 50 60 70 80 Dry-Bulb Temp [F] 90 100 110 Figure 137. Twb C Evaporative Media A Secondary Air Flow Sump Figure 136. The heat transfer in the heat exchanger can be determined with the effectiveness of the heat exchanger according. given pad geometry and secondary airflow information. The rigid media pads are designed to flush the mineral deposits to the sump. where it cools the supply air flowing through the heat exchanger tubes. A to C to D.Indirect Dry Coil Evap Cooler The advantage of the dry coil heat exchanger is that the heat exchanger does not have the evaporation taking place on the outside of the tubes. Then the air enters the shell side of the heat exchanger and is sensibly heated from C to D by the warm supply air passing through the tube side. The secondary air leaves the rigid media pad and enters the air-to-air heat exchanger. Process A to C is adiabatic cooling in the rigid media pad. Evaporative Cooler -. so the saturation efficiency of the pad stays relatively constant.

C Secondary Fan T sec out T sup in A Ambient Conditions Tdb. A to C on the following figure. The effectiveness of the heat exchanger is determined from a parameter estimation using manufacturer's performance data.εse*(Todb . has water sprayed directly on the tubes of the heat exchanger where latent cooling takes place. but it does not follow a line of constant enthalpy as in the direct stage. and barometric pressure. The secondary air stream has its own fan. since humidity ratio is constant for the supply air across the indirect stage.Indirect Wet Coil The process that the secondary air goes through. Tdb sup out = Tdb sup in - QHx rair * c pair * CFM sup ply The wet-bulb temperature is determined from psychrometric routines using the leaving drybulb temperature. 4/23/05 420 . EvapCooler:InDirect:WetCoil The wetted coil evaporative cooler shown in the figure below.67 represented reasonable default effectiveness. For the indirect evaporative cooler it was found that a value of 0. The process is not adiabatic due to the heat gain from the supply air flowing through the tubes of the heat exchanger. Evaporative Cooler -. The vaporization of the water on the outside of the heat exchanger tubes allows the simultaneous heat and mass transfer which removes heat from the supply air on the tube side.Towb ) QHx = εHx * Min( CFMsec . Twb Wetted Coil Heat Exchange r T sup out B T sec in A Sump Figure 138. CFMsupply) * ρair * cp air * ( Todb .Tdb sec out ) After the heat transfer for the heat exchanger has been determined. an energy balance is done on the supply airside to determine the dry-bulb temperature leaving the indirect evaporative cooler. humidity ratio. Then the moist secondary air is exhausted. is a path of simultaneous heat and mass transfer.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE EVAPORATIVE COOLERS Tdb sup out = Tdb sup in .

SIMULATION MODELS – ENCYCLOPEDIC REFERENCE 180 Dry Air] Humidity Ratio [Grains/Lb EVAPORATIVE COOLERS 80 160 140 120 C 100 B 70 Wet-Bulb Temp [F] 60 50 40 30 A 80 60 40 20 0 30 40 50 60 70 80 Dry-Bulb Temp [F] 90 100 110 Figure 139. εHx. the secondary or wet side air stream acts as a heat sink for the supply air. εse. and the saturation efficiency on the wet streamside. So the efficiency of the indirect evaporative cooler is defined by: εind = (Tdbsupin − Tdbsupout ) (Todb − Towb ) . Over the operating lifetime of the heat exchanger. Todb − Towb where Tdb sup in = Tdb sec in for the indirect cooler. The maximum heat transfer possible would be obtained if the supply stream was cooled all the way to the wet bulb temperature. Efficiencies of the Indirect Stage In an indirect stage of an evaporative cooler. These are expressed as: εHx = q qmax = Csup (Tsupin − Tsupout ) Cmin (Tsecin − Tsecout ) . Secondary Air Process – Indirect Wet Coil Evap Cooler The wet coil heat exchanger can have a higher stage efficiency than the dry coil due to a higher heat transfer rate on the outside of the heat exchanger tubes. The efficiency of the indirect stage is given as the effectiveness of the sensible heat exchange. εse = Tdbsecin − Tdbsecout . the vaporization taking place on the heat exchange surface can leave mineral deposits that will decrease the effectiveness of the heat exchanger. Using the combination of the effectiveness and saturation efficiency. In many cases Csup = Cmin and the efficiency of the indirect stage reduces to: εind = εHx εse. Wet Coil Indirect Evaporative Cooler The wetted coil indirect stage is different from the dry coil indirect stage in that water recirculated from a reservoir is sprayed directly on the air-to-air heat exchanger shown 4/23/05 421 . the total efficiency of the indirect stage can be expressed by: εind = εHx εse Csup Cmin .

the mineral deposits that are left can decrease the effectiveness of the heat exchanger unless appropriate maintenance has taken place. but is not adiabatic due to the heat gain from the supply air through the tubes of the heat exchanger. This configuration is mainly used for total comfort cooling for a building and would not normally be used as a pre-cooler for a refrigeration coil.67 * 0. This would make the efficiency tend to go to the maximum efficiency. The figure below shows a dry coil indirect evaporative cooler with a direct evaporative cooler. 4/23/05 422 . This model can be used for all indirect models by curve fitting data from the evaporative cooler of interest.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE EVAPORATIVE COOLERS above.8 compared to the dry coil indirect value of approximately 0.C1 * ( CFM sup CFM sec ) C1 is the "Flow Efficiency Ratio" and is determined from performance data. The vaporization of the water on the outside of the heat exchanger tubes allows the simultaneous heat and mass transfer which removes heat from the supply air on the tube side. Two Stage: Indirect Staged with a Direct The two stage can be either a wet coil or the dry coil indirect evaporative cooler staged with a direct evaporative cooler. The constant C1 tells how quickly the efficiency of the stage would decrease with a mismatch of the supply and secondary flows. and the overall efficiency of the stage would approach the maximum. since the direct stage would increase the latent load on a refrigeration coil. Physically this would be true since the convective terms for heat and mass transfer would increase on the outside of the tube with the additional mass flow rate. This value can be higher than the dry coil overall efficiency since the convective coefficients on the outside of the tube are larger. The maximum efficiency for the dry coil indirect was determined at the condition where flow through the evaporative pad in the secondary air stream approached zero. As the magnitude of the secondary flow increases. The maximum efficiency of the stage is a combination of the efficiency due to the simultaneous heat and mass transfer on the outside of the tube and the efficiency of the heat exchanger. The model development starts with the total efficiency of the indirect evaporative cooler: εind = (Tdbsupin − Tdbsupout ) (Todb − Towb ) Solving for T db sup out gives the leaving conditions of the supply air at a constant humidity ratio: T db sup out = Tdb sup in . An intuitive model determining the performance of the wet coil indirect model was developed.εind * (Todb . for a 12-inch thick pad.Towb) A form for the efficiency of the indirect stage was devised using a maximum efficiency with a coefficient to reduce the efficiency depending on the ratio of the airflows. A check of limits will verify that it makes physical sense. εind = εmax .65 (0. the second term of the equation above becomes small. The process that the secondary air goes through involves simultaneous heat and mass transfer. It should be noted again that over the operating life of the wet coil heat exchanger. For example. Now if the supply air flow goes to zero with a constant secondary air flow. the second term of equation above becomes smaller. a least squares fit for the maximum efficiency showed this value was approximately 0.97).

going from A to B to C. The process from B to C is simultaneous heat and mass transfer following a constant enthalpy line. The air leaving the final stage has a lower dry bulb and wet bulb temperature. EvapCooler:InDirect:WetCoil in series in any combination the user desires in the supply air loop. A fan model must calculate not only the fan energy usage – often a significant portion of the building energy consumption – but also the temperature rise in the 4/23/05 423 . Fans Overview The fan component is the prime mover in most of the air loop and zonal air conditioning systems in EnergyPlus. and an increase in moisture from the direct stage. 180 Dry Air] Humidity Ratio [Grains/Lb 80 160 140 120 70 Wet-Bulb Temp [F] C B 100 60 50 40 30 A 80 60 40 20 0 30 40 50 60 70 80 Dry-Bulb Temp [F] 90 100 110 Figure 141. Two Stage Evaporative Cooler The thermodynamic process for the supply air is shown below.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FANS Secondary Air Supply Air Supply Air A Secondary Air B C Figure 140. Thermodynamic process for Supply Air Two stage evaporative coolers can be accomplished by putting the EvapCooler:Direct:CelDekPad. EvapCooler:InDirect:CelDekPad. The process from A to B is sensible cooling in the indirect stage.

Simulation The following equations define the model for the simple (single speed) fan. Model The fan model is a forward model: the model inputs are the conditions of the air at the fan inlet. the independent variable is the fan power part load ratio. If the flag TurnFansOn is true. Inputs and Data The user describes the fan by inputting design values for the design pressure increase across the fan. However the fan maximum and minimum air flow rates act as absolute limits on the air stream flow rate. The dependent variable is the mass flow fraction. The fan schedule can be overruled by flags set by system availability managers. waste heat from the motor will also act to raise the air stream temperature. a zero fan schedule value will be overridden and the fan will be turned on. This is an on/off schedule associated with each fan: a value of 1 indicates the fan is on. The model algorithms and data are contained in the Fans module in EnergyPlus. If the motor is in the air stream. If the flag TurnFansOff is true the fan will be forced off. For variable speed fans the user must also enter the coefficients of a 4th order polynomial that relates the fan mass flow rate to the fan power consumption. Control The model must decide whether the fan is on or off. a value of 0 indicates the fan is off. uses it in its calculations of energy consumption and temperature rise. Qtot = mi∆P /(etot i ρ air ) Qshaft = emotor iQtot Qtoair = Qshaft + (Qtot − Qshaft )i f motortoair hout = hin + Qtoair / m wout = win Tout = PsyTdbFnHW (hout . In central air systems this type of model is only appropriate for systems with a fixed static pressure setpoint.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FANS air stream caused by the static pressure increase as the air goes through the fan. The primary on/off trigger is the fan schedule. Generally the fan is a passive component: it accepts the mass flow rate on its inlet node. 4/23/05 424 . The inlet air mass flow rate must be greater than zero for the fan to be on. the design volumetric flow rate. the outputs are the conditions at the fan outlet and the fan electrical power consumption. and passes it to the outlet node. wout ) The model for the variable speed fan is similar except that a part load factor that multiplies the fan power consumption. the fan total efficiency and the fan motor efficiency. The user also needs to specify the fraction of the fan waste heat that will enter the air stream (usually 0 or 1). EnergyPlus uses a simple curve-fit model to describe the relation between the air mass flow rate and the fan power.

HVAC 2 Toolkit: A Toolkit for Secondary HVAC System Energy Calculations. m is the air mass flow in kg/s. f flow is the flow fraction. The DXSystem:AirLoop component controls the associated DX cooling coil which is specified in a separate object. References ASHRAE 1993. Atlanta: American Society of Heating. etot is the fan total efficiency. Heat Pumps and Unitary Systems DXSystem:AirLoop Overview The DX System is a “virtual” component that consists of a DX cooling coil component as shown in the Figure below. HEAT PUMPS AND UNITARY SYSTEMS f flow = m / mdesign 2 3 4 f pl = c1 + c2 ⋅ f flow + c3 ⋅ f flow + c4 ⋅ f flow + c5 ⋅ f flow Qtot = f pl imi∆P /(etot i ρ air ) The rest of the calculation is the same as for the simple fan. hin . This coil can be specified anywhere in the air loop simulation and is controlled by the setpoint(s) on the control node. Tout is the outlet air temperature in degrees C. win . f pl is the part load factor. Furnaces. hout are the inlet and outlet air stream specific enthalpies in J/kg. emotor is the motor efficiency. Refrigerating and Air-Conditioning Engineers.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. Qshaft is the fan shaft power in watts. PsyTdbFnHW is the EnergyPlus psychrometric routine relating enthalpy and humidity ratio to temperature. Here Qtot is the fan power in watts. wout are the inlet and outlet air stream humidity ratios. ρ air is the air density at standard conditions in kg/m3. This system does not need any plant specification and will provide the electric consumption necessary to operate the compressor and the condenser fan. Inc. 4/23/05 425 . Qtoair is the power entering the air in watts. Optional dehumidification controls may also be specified. ∆P is the fan design pressure increase in Pascals.

Schematic of EnergyPlus DX System in Air Loop for a Blow-Thru Application Controls The system calculates the current sensible load using the temperature of the inlet node and the System Node Setpoint Temp on the control node.0. Step 1 – Meet Sensible Load Requirement The controls first attempt to meet the sensible requirement. If the humidity requirement is already met.. This is a semi-passive approach to dehumidification which may fall short or may exceed the dehumidification requirement. This is compared with the desired outlet node temperature and a sensible PLR is calculated. Step 2a – Humidity Control = MULTIMODE If the humidity control type is MULTIMODE. the current latent load is calculated using the humidity ratio of the inlet node and the System Node Humidity Ratio Max on the control node. then the coil is re-simulated depending on the type of humidity control. a Regula-Falsi iteration routine is called to determine the coil run-time fraction which results in the desired outlet node temperature.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. then the coil's enhanced dehumidification mode is activated and Step 1 above is repeated to meet the sensible load using different coil performance.. If the PLR is <1. then no further control action is taken. If the humidity requirement has not been met. HEAT PUMPS AND UNITARY SYSTEMS DX System in Air Loop Air Loop Zone Equipment Fan Dx Cooling Coil Zone 1 Terminal Reheat Unit Zone 2 etc. If the control node is not the outlet node. the leaving humidity ratio resulting from operation to meet the sensible load (Step 1 above) is compared with the desired outlet node humidity ratio. Step 2 – Meet Latent Load Requirement (if activated) If dehumidification controls are active. The controls determine the required coil run-time fraction and dehumidification mode (if applicable) using the steps outlined below.0 to determine the full-load output of the coil. The specified DX coil model is called with a part-load ratio (PLR) of 1. 4/23/05 426 . Likewise. the desired outlet node temperature is adjusted for the current temperature difference between the outlet node and the control node. Terminal Reheat Unit Figure 142.

returns dew point temperature given humidity ratio and barometric pressure The supply air dew point temperature required to meet the desired outlet node humidity ratio is then calculated as a function of the desired outlet humidity ratio and barometric pressure as follows: Tdp . HEAT PUMPS AND UNITARY SYSTEMS Step 2b – Humidity Control = COOLREHEAT If the humidity control type is COOLREHEAT. 4/23/05 427 . and the minimum of these two values is used. Pa Tdp = outlet node dew point temperature from Step 1. SP = PsyTdpFnWPb (ωSP . ˚C PsyTdpFnWPb = EnergyPlus psychrometric function. SP = Tdp . SP = outlet node dry-bulb temperature required to achieve the desired outlet humidity ratio. SP + Tapproach where: THR . The model first calculates the coil approach temperature using the dry-bulb temperature and dew point temperature of the air leaving the coil: Tdp = PsyTdpFnWPb (ωSA . kg/kg In order for the dew point temperature of the coil outlet air to reach Tdp . P ) where: Tdp . the desired outlet node temperature is adjusted to achieve the desired outlet node humidity ratio. kg/kg P = outdoor barometric pressure. SP the dry-bulb temperature of air leaving the cooling coil must be at Tdp . P ) Tapproach = TSA − Tdp where: Tapproach = approach temperature.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. SP + Tapproach : THR . ˚C TSA = outlet node dry-bulb temperature from Step 1. SP = dew point temperature corresponding to ωSP. ˚C ωSA = outlet node humidity ratio from Step 1. ˚C The outlet temperature required to achieve the desired outlet humidity ratio is then compared to the outlet temperature required to meet the sensible load. ˚C ωSP = desired outlet humidity ratio.

The module then 4/23/05 428 . system operation is controlled by a thermostat located in a single “control” zone. The furnace module scales the calculated load for the control zone upward based on this fraction to determine the total load to be met by the furnace. THR . ˚C Step 1 above is repeated to meet the new desiared outlet temperature. SP ) where: TSP = desired outlet temperature. HeatCool Only Furnace Air Loop Fan Heating Coil Cooling Coil Zone Equipment Control Zone Direct Air Unit Zone Direct Air Unit Figure 143. SP = outlet node dry-bulb temperature required to meet sensible load. an optional reheat coil may also be modeled for controlling high zone humidity levels and the furnace’s configuration when specifying this option is shown in Figure 144 below. For the HeatCool version. This is an active approach to dehumidification control and will be able to meet the latent load up to the limits of the coil full-load capacity.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. HEAT PUMPS AND UNITARY SYSTEMS TSP = MIN (TTemp . One of the key input parameters for the furnace component is the fraction of the total system volumetric air flow that goes through the controlling zone. then a DX cooling coil is also modeled as part of the system as shown in Figure 143 below. ˚C TTemp . SP . If the HeatCool version is selected. Furnace : BlowThru : HeatOnly or HeatCool (HVAC) Overview The EnergyPlus blowthru furnace is a “virtual” component that consists of an on/off or constant volume fan component and a GAS or ELECTRIC heating coil component. Schematic of the EnergyPlus Blowthru Furnace While the furnace may be configured to serve multiple zones.

) is done by the individual system components (fan. Calculating the sensible heating energy rate involves modeling the supply air fan (and associated fan heat) and the heating coil. for the HeatCool version. The furnace component is able to model supply air fan operation in two modes: cycling fan – cycling coil (i. The heating coil does not require fan operation mode as an input for either the HeatOnly or HeatCool configurations. a cooling coil with an optional reheat coil.. AUTO fan) and continuous fan – cycling coil (i. The fan part-load ratio is defined as the actual air mass flow rate through the system for the time step divided by the design mass flow rate specified for the furnace ( mactual mdesign ). For each simulation time step when the performance of a heat-only furnace is being modeled. the sensible heating energy rate delivered by the furnace is calculated as follows: 4/23/05 429 . the furnace is a “virtual” component consisting of a fan. See the section “Summary of Predictor-Corrector Procedure” elsewhere in this document for more details regarding load calculations. the first step is to retrieve the heating load required to meet the thermostat setpoint for the “control” zone (see Figure 143. then the sensible heating energy rate is zero when the heating coil is OFF. The only output variables reported by the furnace object are the fan part-load ratio and the compressor part-load ratio (HeatCool only). energy consumption.e. the total heating load to be met by the furnace is determined from the following equation: Furnace Heating Load = Control Zone Heating Load Control Zone Air Flow Fraction (467) The model then calculates the furnace’s sensible heating energy rate delivered to the zones being served when the system runs at full-load conditions and when the heating coil is OFF. the compressor partload ratio is reported as the ratio of the actual cooling load to the full-load sensible capacity (see Eqn. If the supply air fan cycles with the heater. For each of these cases (full load and heating coil OFF). The sole purpose of the furnace model is to properly coordinate the operation of the various system components. as well as the differences between cycling and continuous supply air fan operation. (475)). The fan operation mode specified for the furnace must be similarly specified for the DX cooling coil if the HeatCool system is being modeled (see the IO Reference Manual for details).. heating coil and. For the HeatCool version. heating coil and DX cooling coil). Reporting of other variables of interest for the furnace (heating rate. The following sections describe the flow of information within the model for both the HeatOnly and HeatCool configurations. etc. The last section describes the optional control of high zone humidity with a reheat coil for the HeatCool configuration. then the sensible heating energy rate will not be zero when the heating coil is OFF. HeatOnly Configuration The HeatOnly configuration consists of an on/off or constant volume fan and an electric or gas heating coil. while Fan:Simple:OnOff or Fan:Simple:ConstVolume can be used to model fan ON. fan ON). HEAT PUMPS AND UNITARY SYSTEMS proceeds to calculate the required part-load ratio for the system coil and the supply air fan to meet this total load. all of the input data specified for each furnace in the input data file are read into data structures for use throughout the remainder of the simulation. Schematic of the EnergyPlus Blowthru Furnace). Fan:Simple:OnOff must be used to model AUTO fan. Since the furnace may be specified to serve several zones but controlled based on the load calculated for the “control” zone. cooling rate. The heating or cooling capacity delivered by the furnace is distributed to all of the zones served by this system via the direct air units that supply air to each zone.e. When the model is first called during an EnergyPlus simulation.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. Model Description As described previously. However if the fan is configured to run continuously regardless of coil operation.

the operating mass flow rate is specified as m design .0. kg/s hout. Qfurnace = ⎜ m furnace ⎟ ( hout . HEAT PUMPS AND UNITARY SYSTEMS Full Heat Output = ( Mass Flow Rate full load )( hout . coil off − hcontrol zone ) where HR min (468) HR min (469) Mass Flow Ratefull load = air mass flow rate through furnace at full-load conditions. If the fan cycles on and off with the heating coil (AUTO fan). the part-load ratio for the furnace is estimated. actual = enthalpy of air exiting the furnace (J/kg) hout.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. full load − hcontrol zone ) No Heat Output = ( Mass Flow Ratecoil off )( hout . J/kg With the calculated sensible heating energy rates and the total sensible heating load to be met by the system. ⎛ ABS ( Furnace Heating Load − NoHeatOutput ) ⎞ PartLoadRatio = MAX ⎜ 0. 4/23/05 430 . J/kg HRmin = enthalpies evaluated at a constant humidity ratio. and the supply air fan heat varies based on heating coil operation for the case of cycling fan/cycling coil (AUTO fan). coil off = enthalpy of air exiting the furnace with the heating coil OFF. The furnace exiting air conditions and energy consumption are calculated and reported by the individual component models (fan and heating coil). ⎟ ⎜ ⎟ ABS ( FullHeatOutput − NoHeatOutput ) ⎝ ⎠ (470) The part-load ratio calculated above is used to determine the required heating coil capacity as Qheating coil = Qdesign* PartLoadRatio . full load = enthalpy of air exiting the furnace at full-load conditions. The furnace’s fan and heating coil are then re-simulated to determine the furnace’s delivered sensible heating capacity at the above calculated part-load ratio. the final part-load ratio for the heating coil and fan are determined through iterative calculations (successive modeling of the heating coil and fan) until the furnace’s heating output matches the heating load to be met within the heating convergence tolerance that is specified. fan • • ON).e. then this part-load ratio is also used to determine the operating mass flow rate of • the furnace as m furnace = m design * PartLoadRatio . kg/s hout. control zone = enthalpy of air in the control zone (J/kg) HRmin = enthalpies evaluated at a constant humidity ratio. actual − hcontrol zone ) ⎜ ⎟ ⎛ ⎝ • ⎞ ⎠ HR min (471) where: Qfurnace • = sensible heating capacity delivered by the furnace (W) = air mass flow rate through the furnace (kg/s) m furnace hout. If the fan operates continuously (i. the minimum humidity ratio of the furnace exiting air or the air in the control zone Since the part-load performance of the heating coil can be non-linear. J/kg hcontrol zone = enthalpy of air in the control zone (where thermostat is located). the minimum humidity ratio of the furnace exiting air or the air in the control zone Mass Flow Ratecoil off = air mass flow rate through the furnace with the heating coil OFF.

If a cooling load is calculated by EnergyPlus for the control zone. but for the case with no heating load the heating coil model simply passes the inlet air conditions and mass flow rate from its inlet node to its outlet node. then the sensible cooling energy rate is zero when the cooling coil is OFF. where the zone heat balance is performed to determine the resulting zone air conditions. then the sensible cooling energy rate will not be zero when the cooling coil is OFF. full load − hcontrol zone ) No Cool Output = ( Mass Flow Ratecoil off where HR min (473) )( h out . J/kg 4/23/05 431 . the supply air fan model is still called to determine the fan exiting air conditions. the heating coil (simply to pass the air properties and mass flow rate from its inlet node to its outlet node) and the DX cooling coil. full load = enthalpy of air exiting the furnace at full-load conditions. then the furnace’s design air mass flow rate is multiplied by PartLoadRatio to determine the average air mass flow rate for the system simulation time step. the sensible cooling energy rate delivered by the furnace is calculated as follows: Full Cool Output = ( Mass Flow Rate full load )( hout . an electric or gas heating coil. The air exiting the heating coil is then sent to the direct air units for distribution to each zone served by the furnace. then the air mass flow rate remains at the furnace’s design air mass flow rate. However if the fan is configured to run continuously regardless of coil operation.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. Calculating the sensible cooling energy rate involves modeling the supply air fan (and associated fan heat). and a DX cooling coil. the model follows the exact same computational steps as described in the HeatOnly Configuration section above. For each of these cases (full load and DX cooling coil OFF). If EnergyPlus determines that the furnace must supply cooling to the control zone to meet the zone air temperature set point. coil off − hcontrol zone ) HR min (474) Mass Flow Ratefull load = air mass flow rate through furnace at full-load conditions.e. The air conditions at nodes downstream of the heating coil represent the full-load (steady-state) values when the coil is operating. HEAT PUMPS AND UNITARY SYSTEMS If the furnace has been specified with cycling fan/cycling coil (AUTO fan). the solution methodology is virtually identical and is described here for completeness. For the cases where a heating load is calculated for the control zone or no heating/cooling load is calculated for the control zone.. Furnace Cooling Load = Control Zone Cooling Load Control Zone Air Flow Fraction (472) The model then calculates the furnace’s sensible cooling energy rate delivered to the zones being served when the system runs at full-load conditions and when the DX cooling coil is OFF. For the case where the furnace is scheduled to operate with continuous supply air fan operation. In this case. kg/s hout. then the model computes the total sensible cooling load to be met by the furnace based on the control zone sensible cooling load and the fraction of the furnace air flow that goes through the control zone. If the supply air fan cycles with the compressor. The heating coil model is also called. the weighted average of full-load conditions when the coil is operating and inlet air conditions when the coil is OFF). HeatCool Configuration The HeatCool configuration consists of an on/off or constant volume fan. If the supply air fan is specified to run continuously (fan ON). but no heating load is required to meet the set point temperature in the control zone. the air conditions at nodes downstream of the heating coil are calculated as the average conditions over the simulation time step (i.

kg/s hout.. the weighted average of full-load conditions when the coil is operating and inlet air conditions when the coil is OFF). ⎛ ABS ( Furnace Cooling Load − NoCoolOutput ) ⎞ PartLoadRatio = MAX ⎜ 0. J/kg With the calculated sensible cooling energy rates and the total sensible cooling load to be met by the system. the air conditions at nodes downstream of the cooling coil are calculated as the average conditions over the simulation time step (i. High Humidity Control with HeatCool Configuration An optional reheat coil can be specified with the HeatCool configuration to allow the furnace to control high zone humidity levels. the minimum humidity ratio of the furnace exiting air or the air in the control zone Mass Flow Ratecoil off = air mass flow rate through the furnace with the cooling coil OFF. The system is controlled to keep the relative humidity in the control zone from exceeding the set point specified in the object ZONE CONTROL:HUMIDISTAT. and the supply air fan heat varies based on cooling coil operation for the case of cycling fan/cycling coil (AUTO fan). HEAT PUMPS AND UNITARY SYSTEMS hcontrol zone = enthalpy of air in the control zone (where thermostat is located). The model then calculates the LatentPartLoadRatio required to meet the humidistat set point. If a moisture (latent) load exists because the control zone humidity has exceeded the set point.e. If the furnace has been specified with cycling fan/cycling coil (AUTO fan). SystemMoistureLoad = Control Zone Moisture Load Control Zone Air Flow Fraction (476) ⎛ ABS ( SystemMoistureLoad − NoLatentOutput ) ⎞ LatentPartLoadRatio = MAX ⎜ MinPLR. then the air mass flow rate remains at the furnace’s design air mass flow rate. then the furnace’s design air mass flow rate is multiplied by PartLoadRatio to determine the average air mass flow rate for the system simulation time step. ⎟ ⎜ ⎟ ABS ( FullCoolOutput − NoCoolOutput ) ⎝ ⎠ (475) Since the part-load performance of the DX cooling coil is frequently non-linear (Ref: DX Cooling Coil Model). If the supply air fan is specified to run continuously (fan ON). In this case. the final part-load ratio for the cooling coil compressor and fan are determined through iterative calculations (successive modeling of the cooling coil and fan) until the furnace’s cooling output matches the cooling load to be met within the cooling convergence tolerance that is specified. J/kg HRmin = enthalpies evaluated at a constant humidity ratio. coil off = enthalpy of air exiting the furnace with the cooling coil OFF. The air conditions at nodes downstream of the cooling coil represent the full-load (steady-state) values when the coil is operating.0. (475)) to maintain the dry-bulb temperature set point in the control zone. This option is only available with a fan operating mode of ContFanCycCoil (supply air fan ON continuously). The model first calculates the PartLoadRatio required to meet the sensible cooling load as described above (see Eqn.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. ⎟ ⎜ ABS ( FullLatentOutput − NoLatentOutput ) ⎟ ⎝ ⎠ 4/23/05 (477) 432 . The specific configuration of the blowthru HeatCool Furnace with high humidity control option is shown in Figure 144. the total moisture load to be met by the HeatCool furnace (SystemMoistureLoad) is calculated based on the control zone moisture load and the control zone air flow fraction. the part-load ratio for the furnace is estimated. The furnace exiting air conditions and energy consumption are calculated and reported by the individual component models (fan and DX cooling coil).

. heating coil rate. the model calculates the difference between the actual sensible cooling energy rate delivered by the unit (with LatentPartLoadRatio) and the sensible cooling load to be met (i. In this case. Furnace Air Loop Fan Cooling Coil Heating Coil Reheat Coil Zone Equipment Control Zone Direct Air Unit Zone Direct Air Unit Figure 144. and heating coil. HEAT PUMPS AND UNITARY SYSTEMS where FullLatentOutput = the furnace’s latent cooling energy rate at full-load conditions.g. heating coil energy.e. this value is the minimum part-load ratio at which the cooling coil will dehumidify the air. As previously described. DX cooling coil. W NoLatentOutput = the furnace’s latent cooling energy rate with the cooling coil OFF. which is usually 0. The model first checks to see if a sensible cooling load or sensible heating load exists for the current simulation time step. Schematic for Blow Thru Furnace with High Humidity Control If the model determines that the LatentPartLoadRatio is to be used as the operating part-load ratio of the furnace’s cooling coil. the reheat coil is used to offset the excess sensible capacity provided by the unit. the reheat coil is used to offset this excess sensible cooling energy provided by the DX cooling coil. If a sensible cooling load exists.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. the sensible cooling required to reach or maintain the dry-bulb temperature set point). iterations are performed to converge on the solution within the specified cooling convergence tolerance. As with the fan. For the case when the latent capacity degradation model is used (Ref: DX Cooling Coil Model).0. W MinPLR = the minimum part-load ratio. report variables associated with reheat coil performance (e. the reheat coil is used to offset the entire sensible cooling energy rate of the DX cooling coil (to meet the humidistat set point) and the heating coil is used to meet the entire heating load as described in the HeatOnly configuration section above. If a heating load exists. The model uses the greater of the two part-load ratios. heating coil gas or electric 4/23/05 433 . to determine the operating part-load ratio of the furnace’s DX cooling coil. PartLoadRatio or LatentPartLoadRatio..

The heating or cooling capacity delivered by the heat pump is distributed to all of the zones served by this system via the direct air units that supply air to each zone. DX DX Cooling Heating Heating Coil Coil Coil Zone Equipment Control Zone Direct Air Unit Zone Direct Air Unit Figure 145. One of the key input parameters for the heat pump component is the fraction of the total system volumetric airflow that goes through the controlling zone.e. a DX heating coil. UnitarySystem : HeatPump : AirToAir (HVAC) Overview The EnergyPlus air-to-air heat pump is a “virtual” component that consists of an on/off or constant volume fan component.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. the fan is located between the DX heating coil and the supplemental heating coil. HEAT PUMPS AND UNITARY SYSTEMS consumption. fan ON). The specific configuration of the blowthru heat pump is shown in the following figure.) are managed in the reheat (heating) coil object. The module then proceeds to calculate the required part-load ratio for the system coil and the supply air fan to meet this total load.. Heat Pump Air Loop Fan Supp. heating coil runtime fraction. Schematic of a Blowthru Air-to-Air Heat Pump While the heat pump may be configured to serve multiple zones. and a gas or electric supplemental heating coil.. The fan operation mode specified 4/23/05 434 . system operation is controlled by a thermostat located in a single “control” zone. while Fan:Simple:OnOff or Fan:Simple:ConstVolume can be used to model fan ON. UnitarySystem : BlowThru : HeatOnly or HeatCool (HVAC) The Unitary System models are identical to the equivalently named Furnace models. etc. Fan:Simple:OnOff must be used to model AUTO fan.e. a DX cooling coil. AUTO fan) and continuous fan – cycling coil (i. Please reference the previous section for details. The heat pump component is able to model supply air fan operation in two modes: cycling fan – cycling coil (i. For a drawthru heat pump. The heat pump module scales the calculated load for the control zone upward based on this fraction to determine the total load to be met by the heat pump.

However if the fan is configured to run continuously regardless of coil operation. If the supply air fan cycles with the compressor. Model Description As described previously. as well as the differences between cycling and continuous supply air fan operation. full load − h control zone ) No Cool Output = ( Mass Flow Ratecoil off where HR min (479) )( h out . DX heating coil and a supplemental heating coil. J/kg 4/23/05 435 . cooling rate. full load = enthalpy of air exiting the heat pump at full-load conditions. The following sections describe the flow of information within the model. DX cooling coil.) is done by the individual system components (fan. kg/s hout. then the heat pump model computes the total sensible cooling load to be delivered to the zones being served based on the control zone sensible cooling load and the fraction of the heat pump air flow that goes through the control zone. Compressor part-load ratio is the actual load for the time step divided by the full-load sensible capacity (see Eqn.SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. DX cooling coil. Fan part-load ratio is defined as the actual air mass flow rate through the system for the time step divided by the design mass flow rate specified for the heat pump • • ( m actual m design ). coil off − hcontrol zone ) HR min (480) Mass Flow Ratefull load = air mass flow rate through heat pump at full-load conditions. then the sensible cooling energy rate will not be zero when the cooling coil is OFF. energy consumption. For each of these cases (full load and DX cooling coil OFF). J/kg hcontrol zone = enthalpy of air leaving the control zone (where thermostat is located). Calculating the sensible cooling energy rate involves modeling the supply air fan (and associated fan heat) and the DX cooling coil. but only to pass the air properties and mass flow rate from their inlet nodes to their outlet nodes. etc. HEAT PUMPS AND UNITARY SYSTEMS for the heat pump must be similarly specified for the DX cooling and DX heating coils being modeled (see the IO Reference Manual for details). and supplemental heating coil). then the sensible cooling energy rate is zero when the cooling coil is OFF. Reporting of other variables of interest for the heat pump (heating rate. the sensible cooling energy rate delivered by the heat pump is calculated as follows: Full Cool Output = ( Mass Flow Rate full load )( h out . The DX heating coil and the supplemental heating coil are also modeled.(485)). Cooling Operation If EnergyPlus determines that the heat pump must supply cooling to the control zone to meet the zone air temperature set point. (481) or Eqn. Heat Pump Cooling Load = Control Zone Cooling Load Control Zone Air Flow Fraction (478) The model then calculates the heat pump’s sensible cooling energy rate delivered to the zones being served when the system runs at full-load conditions and when the DX cooling coil is OFF. The output variables reported by the heat pump object are fan part-load ratio and compressor part-load ratio. DX heating coil. The sole purpose of the heat pump model is to properly coordinate the operation of the various system components. the heat pump is a “virtual” component consisting of a fan.

If the supply air fan cycles with the compressor. the DX cooling coil (simply to pass the air properties and mass flow rate from its inlet node to its outlet node). and the supplemental heating coil (simply to pass the air properties and mass flow rate from its inlet node to its outlet node).. Heat Pump Heating Load = Control Zone Heating Load Control Zone Air Flow Fraction (482) The model then calculates the heat pump’s sensible heating energy rate delivered to the zones being served when the system runs at full-load conditions and when the DX heating coil is OFF (without supplemental heater operation in either case).SIMULATION MODELS – ENCYCLOPEDIC REFERENCE FURNACES. the final part-load ratio for the cooling coil compressor and fan are determined through iterative calculations (successive modeling of the cooling coil and fan) until the heat pump’s cooling output matches the cooling load to be met within the cooling convergence tolerance that is specified. then the air mass flow rate remains at the heat pump’s design air mass flow rate. additional calculations are necessary to properly meet the total heating load for the zones being served. the part-load ratio for the heat pump is estimated. The heat pump exiting air conditions and energy consumption are calculated and reported by the individual component models (fan and DX cooling coil). off HRmin = enthalpies evaluated at a constant humidity ratio. the DX heating coil. the weighted average of full-load conditions when the coil is opera