Executive Summary

Updating the ASHRAE/ACCA
Residential Heating and Cooling Load

Calculation Procedures and Data
ASHRAE 1199-RP
Final Report
August 20, 2004
Prepared by:
Charles S. Barnaby Jeffrey D. Spitler

Dongyi Xiao
Wrightsoft Corporation Oklahoma State University
Lexington, MA Stillwater, OK
cbarnaby@wrightsoft.com spitler@okstate.edu
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

1791 Tullie Circle, N.E., Atlanta, GA 30329
(404) 636-8400 http://www.ashrae.org
Copyright © 2004 by the American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Inc. (ASHRAE). All rights reserved.
Wrightsoft and Oklahoma State University (“Contractors”) have compiled this report and accompanying
1199-RP Source Code and Data CD (together the “Report”) with care. However, neither the Contractors
nor ASHRAE warrants that the information in the Report is complete or free of errors. ASHRAE does
not necessarily agree with any statement or opinion in the Report. The appearance of any technical data
or editorial material in the Report does not constitute endorsement, warranty, or guaranty by the
Contractors or ASHRAE of any product, service, process, procedure, design, or the like. ASHRAE has
not investigated and ASHRAE expressly disclaims any duty to investigate any product, service, process,
procedure, design, or the like which may be described herein.
THE ENTIRE RISK OF THE USE OF ANY INFORMATION IN THE REPORT IS ASSUMED BY
THE USER.
See also “1199RP_TermsOfUse.txt” on the 1199-RP Source Code and Data CD.
Executive Summary
ASHRAE Research Project 1199-RP, Updating the ASHRAE/ACCA Residential Heating and Cooling
Load Calculation Procedures and Data, developed two new residential heating and cooling loads
calculation procedures:
• Residential Heat Balance (RHB), a detailed heat balance method that requires computer
implementation; and
• Residential Load Factor (RLF), a simplified procedure suitable for hand or spreadsheet use.
RHB is based on first-principles heat balance models. A research-oriented FORTRAN implementation of
RHB, designated ResHB, was developed using ASHRAE Loads Toolkit as a starting point. ResHB
includes algorithms for calculating sensible cooling loads with temperature swing (temperature excursion
above the cooling set point) and to handle master / slave control (room cooling controlled by a thermostat
in another room). ResHB extends Toolkit capabilities to multiple rooms, zones, and systems, thus
supporting load calculations for real buildings. Also added were infiltration, internal gain, and duct loss
models appropriate for residential problems. Because RHB is a first-principles heat balance procedure, it
can be directly validated and refined using empirical data.
ResHB was tested against the general-purpose building model ESP-r. Testing procedures and results are
presented in the final report. Additional analytical, emprical, and inter-model validation of RHB is
underway and will be reported in the literature when complete
The RLF method was developed from RHB results. The form of RLF resembles prior methods.
However, the sensible cooling load procedure was derived using linear regression to find relationships
between design conditions, building characteristics, and peak cooling load predicted by RHB. This
eliminated the need for semi-empirical adjustments, such as averaging, that have been used in the
development of other methods. Results comparing RLF to RHB are presented. The RLF heating load
calculation is also described; it uses the traditional UA∆T formulation with some improvements to
procedures for slab and basement ground losses and infiltration air leakage.
Several conclusions are presented:
• RHB is the method of choice for residential load calculations. Due to its first-principles
approach, it handles virtually any residential configuration, including extreme cases, such as large
or concentrated fenestration area, that are problematic for prior methods. Currently available
computing power is now sufficient to support routine use of heat balance procedures.
• Inter-model validation is useful for testing models, identifying sensitivities, and verifying results.
• Linear regression is an effective and rational approach for development of simplified loads
calculation methods such as RLF.
• Simplified methods, such as prior residential procedures and RLF, are necessarily approximate.
They cannot capture all the effects modeled in RHB without addition of self-defeating complexity
and do not inherently provide indication of when they are inapplicable. For these reasons,
methods of this type should be superceded by rigorous heat balance approaches such as RHB.
Accompanying the report is the 1199-RP Source Code and Data CD containing ResHB source code and
other development material produced during the work.
1199-RP Final Report August 20, 2004 i

Acknowledgements
The authors thank the 1199-RP Project Monitoring Subcommittee – Doug Hittle (chair), Glenn Friedman,
Glenn Hourahan, Curt Pedersen, and Gary Wingfield – for their comments, guidance, and criticism
throughout the project.
In addition, we acknowledge and thank Larry Palmiter for helpful discussions regarding infiltration,
ventilation, and distribution losses; Bruce Wilcox for ideas regarding regression techniques and
fenestration modeling; and Phil Sobolik for development of regression procedures.
Finally, we thank Bill Wright for his overview drawn from long experience with loads calculations and
for Wrightsoft’s support of the project.
This project was supported in part, by the Air Conditioning Contractors of America Educational Institute.
1199-RP Final Report August 20, 2004 ii

Table of Contents
Executive Summary .....................................................................................................................................i
Acknowledgements .....................................................................................................................................ii
1. Introduction............................................................................................................................................ 1
2. The Residential Heat Balance (RHB) Heating and Cooling Load Calculation Method.................. 2
Background and RHB Description ........................................................................................................... 2
Prior methods ....................................................................................................................................... 3
Master / slave control, temperature swing, and cooling load ............................................................... 3
Heat balance for residential applications.............................................................................................. 4
RHB definition ..................................................................................................................................... 5
Calculation Algorithms ............................................................................................................................ 5
Temperature swing ............................................................................................................................... 7
Master/slave control ............................................................................................................................. 7
Models ...................................................................................................................................................... 9
Inside surface convection coefficients.................................................................................................. 9
Elevation effects on convective heat transfer..................................................................................... 10
Buffer spaces ...................................................................................................................................... 11
Infiltration........................................................................................................................................... 12
Distribution losses .............................................................................................................................. 12
Framed constructions ......................................................................................................................... 12
Fenestration and solar gain distribution ............................................................................................. 13
Ground heat transfer........................................................................................................................... 13
Modeling Assumptions........................................................................................................................... 13
Outdoor design conditions.................................................................................................................. 14
Internal gain........................................................................................................................................ 14
Internal mass ...................................................................................................................................... 15
Other assumptions .............................................................................................................................. 17
Toolkit Re-engineering........................................................................................................................... 17
3. Inter-model Validation of the Residential Heat Balance Method ................................................... 19
Introduction ............................................................................................................................................ 19
Background ........................................................................................................................................ 19
Selecting the comparison tool ............................................................................................................ 20
Methodology........................................................................................................................................... 20
Types of comparison .......................................................................................................................... 20
Parametric code .................................................................................................................................. 21
Combined testing process................................................................................................................... 21
RHBGen parametric generator........................................................................................................... 22
ESP-r system ...................................................................................................................................... 22
Design evaluation figure of merit....................................................................................................... 25
1199-RP Final Report August 20, 2004 iii

Model assumptions used in the comparison....................................................................................... 25
Results .................................................................................................................................................... 28
Description of test sets ....................................................................................................................... 28
Ideal load comparison ........................................................................................................................ 29
System design evaluations.................................................................................................................. 33
Conclusions and Recommendations....................................................................................................... 38
4. The Residential Load Factor (RLF) Method..................................................................................... 40
RLF Cooling Load Calculation .............................................................................................................. 40
Total cooling load............................................................................................................................... 41
Opaque surfaces ................................................................................................................................. 41
Fenestration ........................................................................................................................................ 42
Ventilation and infiltration ................................................................................................................. 44
Internal gain........................................................................................................................................ 45
Development Overview.......................................................................................................................... 46
Modeling Assumptions........................................................................................................................... 47
Prototype building .............................................................................................................................. 48
Outdoor design conditions.................................................................................................................. 49
Regression Strategy ................................................................................................................................ 50
Component Models ................................................................................................................................ 51
Ventilation and infiltration ................................................................................................................. 51
Internal gains ...................................................................................................................................... 52
Opaque surfaces ................................................................................................................................. 52
Fenestration ........................................................................................................................................ 53
Heating Loads......................................................................................................................................... 53
Basement wall heat losses .................................................................................................................. 53
Basement floor heat losses ................................................................................................................. 56
Slab-on-grade floor heat losses .......................................................................................................... 57
Verification of Results............................................................................................................................ 58
5. Conclusions........................................................................................................................................... 60
6. Future Research ................................................................................................................................... 62
7. References............................................................................................................................................. 64
Appendix A. The 1199-RP Source Code and Data CD......................................................................... 68
Appendix B. Algorithms and Assumptions............................................................................................ 69
Fenestration ............................................................................................................................................ 69
Internal mass........................................................................................................................................... 70
Framed constructions.............................................................................................................................. 73
Material properties.................................................................................................................................. 75
Appendix C. ResHB Documentation...................................................................................................... 76
Appendix D. RHBGen Documentation ................................................................................................ 108
1199-RP Final Report August 20, 2004 iv

List of Figures
Figure 1. 1199-RP project process .............................................................................................................. 1
Figure 2. Sensible cooling load reduction due to temperature swing.......................................................... 4
Figure 3. Master / slave room temperatures with similar load profiles....................................................... 8
Figure 4. Master / slave room temperatures with mismatched load profiles............................................... 9
Figure 5. Effect of pressure on convection coefficient (see definitions below) .......................................... 11
Figure 6. Roof surface and attic air temperatures predicted by ResHB .................................................... 12
Figure 7. Hourly internal gain (all sources) for a 186 m2 (2000 ft2) house with 4 occupants................... 15
Figure 8. Effect of internal mass, wood floor construction........................................................................ 16
Figure 9. Effect of internal mass, slab floor construction.......................................................................... 16
Figure 10. Overall inter-model testing process ......................................................................................... 22
Figure 11. Predicted Percentage of Dissatisfied (PPD) as a function of Predicted Mean Vote (PMV).... 25
Figure 12. Peak cooling load comparison for the Great room .................................................................. 30
Figure 13. Peak cooling load comparison for the Living room ................................................................. 30
Figure 14. Outside ground surface temperatures from RHB and ESP-r for Alamosa, CO ....................... 31
Figure 15. Comparison of the ambient dry bulb temperature and the outside surface temperatures
calculated from ESP-r (Great room) .................................................................................................. 31
Figure 16. Comparison of cooling load calculated by RHB and ESP-r: with external long wave radiation
............................................................................................................................................................ 32
Figure 17. Comparison of cooling load calculated by RHB and ESP-r: without external long wave
radiation.............................................................................................................................................. 32
Figure 18. Processed PPD showing night time ambient cooling, system designed with zero temperature
swing ................................................................................................................................................... 35
Figure 19. Processed PPD showing system under-sizing and inadequate distribution, system designed
with zero temperature swing ............................................................................................................... 35
Figure 20. Hourly PPD vs. temperature for case 17, system designed with zero temperature swing ....... 36
Figure 21. Hourly PPD vs. temperature for Case 20; system designed with zero temperature swing ...... 36
Figure 22. Processed PPD showing night time ambient cooling, system designed with medium
temperature swing............................................................................................................................... 37
with medium temperature swing ......................................................................................................... 38
Figure 24. Hourly PPD vs. temperature for case 355, system designed with medium temperature swing 38
Figure 25. Predicted infiltration leakage rates, AL = 1000 cm2 (155 in2) and representative range of stack
height, temperature difference, and flue fraction. RLF values from Equation (10), see text. (1 L/s =
2.12 cfm) ............................................................................................................................................. 52
Figure 26. Heat flow from basements (2001 ASHRAE Handbook of Fundamentals)............................... 54
1199-RP Final Report August 20, 2004 v

Figure 27. Definition of basement wall and floor dimensions ................................................................... 55
Figure 28. RLF vs. RHB sensible cooling load comparison. Test building calculated for representative
range of climate and construction conditions (1280 cases)................................................................ 59
Figure 29. Load grouping as function of fenestration classification ......................................................... 69
Figure 30. Effect of internal mass, wood floor house ................................................................................ 71
Figure 31. Effect of internal mass, slab floor house .................................................................................. 71
Figure 32. Effect of internal mass, single-exposure room, wood floor ...................................................... 72
Figure 33. Effect of internal mass, single exposure room, slab floor ........................................................ 72
Figure 34. Wall 1 of interior heat flux calculated with 1145-RP CTF and CTF generated with equivalent
homogeneous layer method................................................................................................................. 74
Figure 35. Wall 18 interior heat flux calculated with 1145-RP CTF and CTF generated with equivalent
homogeneous layer method................................................................................................................. 75
1199-RP Final Report August 20, 2004 vi

List of Tables
Table 1. Inside surface convection coefficient values (W/m2-K) for model alternatives............................ 10
Table 2. Fractional components of internal heat sources.......................................................................... 14
Table 3. Default surface absorptances ...................................................................................................... 17
Table 4. Model assumptions used in ESP-r and RHB for the comparison ................................................ 27
Table 5. Test parameters for the 576 test cases ......................................................................................... 28
Table 6. Algorithm for attributing the PPD Cause .................................................................................... 34
Table 7. Opaque surface coefficients......................................................................................................... 41
Table 8. Fenestration coefficients.............................................................................................................. 42
Table 9. Exterior irradiance (W/m2) .......................................................................................................... 43
Table 10. Exterior attachment transmission .............................................................................................. 44
Table 11. Shade line factors (SLF) ............................................................................................................ 44
Table 12. Duct loss/gain factors ................................................................................................................ 46
Table 13. Prototype building characteristics............................................................................................. 48
Table 14. Prototype surface constructions ................................................................................................ 49
Table 15. Design conditions ...................................................................................................................... 49
Table 16. Site assumptions......................................................................................................................... 50
Table 17. Regression data sets................................................................................................................... 51
Table 18a. Average U-factor for basement walls with uniform insulation (SI units) ................................ 55
Table 19a. Average U-factor for basement floors (SI units) ...................................................................... 56
Table 20a. Heat loss coefficient F2 of slab floor construction (SI units) ................................................... 58
Table 21. Summary of 1145-RP walls, stated U-factors, and U-factors calculated from CTF coefficients
............................................................................................................................................................ 74
1199-RP Final Report August 20, 2004 vii

1. Introduction
This report presents the results of ASHRAE project 1199-RP, “Updating the ASHRAE / ACCA
Residential Heating and Cooling Load Calculation Procedures and Data.”
There were three major aspects to the project:
• Development of the Residential Heat Balance (RHB) heating and cooling load procedure and its
implementation in the ResHB computer program. RHB is based on first-principles heat balance
methods and uses a 24 hour design day simulation. Also developed was the RHBGen application
that generates parametrically varied ResHB test cases.
• Detailed validation of ResHB results by comparison to those produced by ESP-r. This work
included both comparison of overall results and specialized studies to investigate specific models.
• Development of the Residential Load Factor (RLF) method, a hand or spreadsheet tractable
simplified procedure. RLF in some ways resembles prior simplified methods. However, it was
developed from ResHB results using regression techniques, an entirely new approach.
The report sections that follow are devoted to these topics. Later sections present overall conclusions and
highlight topics needing additional research. Several appendices are also included that address
background issues.
Figure 1 represents the process used to conduct 1199-RP.
The overall process was one of “cut and try”. As RHB models were implemented, the updated method
was tested against ESP-r. Successive refinements will be made to RHB until it was deemed to perform
satisfactorily. RHB was then exercised to produce data sets used for the development of RLF.
Start
Planning Develop/refine RHB

Requirements procedure
Develop evaluation methods

• Prototypes Evaluate procedure
• Simulation models
• Automation Not OK
OK
Develop RLF method
Final deliverables: Reports, Done

papers, CD
Figure 1. 1199-RP project process
1199-RP Final Report August 20, 2004 1

2. The Residential Heat Balance (RHB) Heating and
Cooling Load Calculation Method
The research project Updating the ASHRAE/ACCA Residential Heating and Cooling Load Calculation
Procedures and Data (1199-RP) had two primary products. First, a new fundamental residential heating
and cooling load calculation method was developed and tested. This procedure, called the Residential
Heat Balance (RHB) method, is described in this paper. The second product of 1199-RP is a simpler
procedure, designated the Residential Load Factor (RLF) method. RLF is tractable by hand or can be
straightforwardly implemented using spreadsheet software and is applicable to conventional single-family
detached residences. RLF procedures and data are presented in the Residential Cooling and Heating
Loads Calculation chapter of the 2005 Handbook of Fundamentals (ASHRAE 2005) and RLF
development is documented in Section 4 of this report and Barnaby and Spitler 2005.
RHB is based on heat balance first principles as described by Pedersen et. al. 1997, Pedersen et. al. 1998,
and Chapter 29 of ASHRAE 2001. It uses a computationally intensive 24-hour design-day simulation
that is practical only when implemented in software. Because of its fundamental approach, RHB can be
applied with few restrictions to arbitrarily complex residential buildings, including those with large
fenestration areas, novel construction features, or having non-summer peaks.
The ResHB computer program, developed as part of 1199-RP, is the reference implementation of the
RHB method. ResHB is a PC-based application written in FORTRAN 95. The code is based on the
ASHRAE Loads Toolkit derived from the ASHRAE Loads Toolkit Pedersen et. al. 2001. ResHB is a
research-oriented batch program, taking input from one or more text files and producing various output
reports and data files. ResHB documentation is included in Appendix C. Source code and associated
development procedures are found on the 1199-RP Source Code and Data CD that accompanies this
report (see Appendix A). An additional utility program, RHBGen, was also developed during 1199-RP.
RHBGen generates and runs parametrically varied ResHB cases for testing, research, and RLF
development. Intermodel validation of ResHB is documented in Section 3. Analytical and empirical
validation is underway and will be reported in planned publications.
The sections below describe RHB and its implementation in ResHB. References to RHB and ResHB are
made somewhat interchangeably, because in many ways, ResHB is RHB. Detailed equation-based model
descriptions are not included here; readers are referred to cited sources and the ResHB source code.
Background and RHB Description

Residential heating and cooling load calculations produce information needed for equipment selection and
distribution system design. These results include design values for heating, sensible cooling, and latent
cooling equipment capacity plus room-by-room heating and sensible cooling loads. Experience has
shown that simple procedures are sufficient for heating and latent cooling load calculations. Sensible
cooling load calculations are more problematic. Sensible load results from the combination of several
load components having building- and climate- dependent profiles. Excess sensible capacity increases
first cost and results in performance problems including poor humidity control, excessive power demand,
and noisy operation. Thus using conservative estimates of load components is not acceptable and the
overwhelming focus of the 1199-RP was on calculation of sensible cooling loads.

Prior methods
Prior residential load calculation methods have been published by the Air-Conditioning Contractors of
America (ACCA), including the widely-used Manual J Seventh Edition (ACCA 1986) and Manual J
Eighth Edition (ACCA 2003). The 1989 – 2001 editions of the ASHRAE Handbook of Fundamentals
includes a method based on 342-RP (McQuiston 1984). Canadian Standard CAN/CSA-F280-M90
(HRAI 1996 and CAN/CSA-F280-M90 1990) specifies a cooling method also based on 342-RP and a
heating procedure that includes enhanced ground-loss calculations.
These methods share many features. Their heating load procedures differ only in details: all ignore solar
and internal gains and are based on summing surface UA∆T heat losses, infiltration loss, ventilation loss,
and distribution loss. Sensible cooling loads are similarly derived by summing component contributions
calculated using tabulated or formula-based factors incorporating temperature and solar effects as
appropriate. With the exception of Manual J 8th edition, all perform a single design-condition calculation,
implicitly making assumptions about relative timing of various gains and the zone response that
transforms the gains into load. Recent addenda to Manual J 8th Edition have added an adjustment that
involves evaluation of the full-day room and zone fenestration gain profiles.
The single design-condition calculation of sensible cooling load has long been problematic. Using the
sum of peak component gains as the design load usually produces an excessive result, because the gains
generally occur at different times over the day. To account for gain diversity, factors used in prior
method were derived using semi-empirical adjustments such as multi-hour averaging. However, for
situations with limited exposure (e.g. apartments), the dominant fenestration gains peak simultaneously
and the sum-of-peaks estimate is more appropriate. To handle such configurations, prior methods have
included alternative factors and/or adjustments variously called “multi-family” or “peak” (as opposed to
“single-family” or “average”). User judgment is required to select the applicable condition.
A multi-hour calculation eliminates the average/peak distinction – the design load is simply the peak of
the hourly profile. The only motivation for using a single design-condition is hand tractability.
Implementers of past methods made the decision that an approximate method that would actually be used
was preferable to a more accurate but impractically complex alternative. Given that personal computers
are now ubiquitous, it is reasonable to use a 24-hour calculation for an updated procedure.
Master / slave control, temperature swing, and cooling load

Residential air conditioning applications rely on multi-room constant volume systems controlled by a
single thermostat in one room (master / slave control). Assuming sufficient capacity, good temperature
control occurs in the master (thermostat) room. The slave rooms maintain reasonable temperatures to the
extent they have load profiles similar to that of the master and/or are conditioned by air mixing with
adjacent rooms. In general their temperatures will not be held at the set point even when the system is
operating. The resulting temperature variation, or swing, has the effect of reducing the required capacity.
This has long been recognized as a major consideration in residential cooling load calculations. Its
importance is confirmed by this work.
Temperature swing generally occurs in slave rooms. However, with reduced cooling capacity, a
thermostat room will experience temperature swing as well. As illustration, Figure 2 shows the results of
192 ResHB sensible cooling load calculations for a single room with one exterior exposure in a variety of
climates and in four primary orientations. At a 1.67°C (3°F) swing, the load reduction ranges from 13 to
50%. Some of this reduction is due to the higher average room temperature when swing is allowed.
However, most of the effect results from a portion of peak gains being absorbed in building mass as room
temperature rises. This energy is “carried forward” and is removed at a later time when gains have
moderated and the system has adequate capacity to bring the room back to the set point, re-cooling the
building mass. Permitting a small, short-duration temperature excursion at design conditions usually
results in a significant reduction in required sensible capacity, with associated cost reduction, moisture

removal improvement (as a result of longer run times), and electrical demand reduction. Note that these
are primarily capacity effects – approximately the same amount of energy is removed over the day with or
without temperature swing. Second order considerations, such as higher average part load ratio, may
result in some energy savings when swing is allowed, but their relative magnitude is much smaller than
the capacity savings.
Temperature swing (°F)

0 1.5 3 4.5
1.100
1.000
Relative cooling load
0.900
0.800
0.700
0.600
0.500
0.400
0.00 0.83 1.67 2.50
Temperature swing (°C)
Figure 2. Sensible cooling load reduction due to temperature swing
Heat balance for residential applications

Over the last ten years, ASHRAE non-residential cooling load calculation procedures have moved to the
Heat Balance (HB) method as the fundamental procedure (Pedersen et. al. 1997) and the Radiant Time
Series (RTS) method as a simplified procedure derived from HB (Spitler et. al. 1997). HB and RTS were
evaluated regarding their suitability as the basis for an updated residential procedure. Both are 24-hour
methods. HB was selected because it can readily calculate either load at a known space temperature or
space temperature given a known extraction rate. The latter capability makes HB well suited to the
residential application where room temperature swing is so important.
To handle these floating temperature cases, RTS would have to be modified. An RTS extension,
designated Period Space Air Response Factor (PSARF), was explored during 1199-RP. PSARFs relate
extraction rate to air temperature deviation from a nominal set point and are analogous to the Space Air
Temperature Weighting Factors in the Transfer Function Method (TFM) (McQuiston and Spitler 1992).
The PSARF approach was not ultimately pursued because an extended RTS method would be in essence
a re-invention of TFM, which has been superseded by HB. Given that direct application of HB is now
computationally practical, there is no reason to resort to simplifications.
An important goal for the updated residential procedure was simplicity of required input, preferably
comparable to prior methods. Specifically, a detailed geometric building description was deemed
impractical. Using a simplified geometric model, where surface areas and orientations are known but

their positions are not, implies (1) exact surface-to-surface view factors are not available and (2) room
adjacencies are not known. The heat balance method can model long-wave radiant exchange with good
accuracy without exact view factors using an MRT formulation (Liesen and Pedersen 1997). Room
adjacencies are unnecessary if room loads are calculated independently. Thus, there is a good match
between heat balance and the requirement of simple input.
RHB definition
The Residential Heat Balance Method is a specialized application of the ASHRAE Heat Balance Method.
The following HB changes and extensions define RHB --
• Multi-room, multi-zone, and multi-system. The fundamental RHB modeling unit is the room.
Independent heat balances are performed for each room. Zones and systems are accounting
structures to which loads are accumulated to provide overall results.
• Specialized algorithms. Temperature swing and master / slave control can be modeled to produce
realistic sensible cooling load estimates.
• Residential models and assumptions. Component models and assumptions used for RHB are
appropriate for the residential application.
• Simple heating and latent cooling procedures. As discussed above, simple UA∆T model has
proven satisfactory for heating load calculations. Similarly, latent load can be estimated from
moisture gain from infiltration, ventilation, duct leakage, and occupants. These simple
approaches are retained in RHB.
It should be noted that RHB is not a fully elaborated cooling system design procedure. In particular,
RHB does not specify how temperature swing and master / slave control be considered during the design
process. RHB can model rooms with or without swing, allowing choice on the much-debated question as
to whether systems should be sized to allow swing at the thermostat on the design-day. Slave room
temperature results from a case-specific combination of limited capacity and control profile mismatch, so
its design implications are more complex. It may be that RHB master / slave capabilities should be used
for investigation of zoning options only after primary load calculations are done on an independent room-
by-room basis (with or without temperature swing).
The remaining sections of this paper provide details about the above aspects of RHB in its current form.
One major advantage of a heat balance formulation is that it can be tested and refined via direct
comparison to empirical data. It is expected that RHB will evolve as additional research results become
available.
Calculation Algorithms
The HB method is a design day procedure that requires iteration to find the steady-periodic solution at
which all heat flows correctly balance. RHB adds the additional requirement of finding loads under
floating temperature conditions, in order to handle temperature swing and master/slave control, as
described here.

Calculation sequence and convergence criteria
The fundamental RHB load calculation sequence is:
repeat swing
repeat day
for hour = 1 to 24
for all rooms
repeat
for all surfaces
perform surface heat balance
end for surfaces
perform air heat balance
until room convergence for current hour
end for rooms
end for hours
until day convergence
determine room supply air flow rates for next swing iteration
until swing convergence
The convergence criteria are discussed below. The sequence was modified several times during
development and its logic is worth examining:
• The outer loop handles temperature swing (discussed below). Temperature swing occurs when
cooling capacity is less than required to hold a room at the set point. The swing search algorithm
adjusts each room supply air flow rate and repeats the entire calculation until the specified swing
is achieved.
• The hour loop is outside the room loop. This means that current hour conditions are available for
all rooms (either from the current day iteration or, at worst, from the prior day iteration), allowing
inter-room references.
One of the issues with a design day heat balance procedure is determining when the solution has
converged. A common technique is to continue iteration until calculated or temperatures change a very
small amount between iterations. The difficulty is to determine a “small amount” that truly represents
convergence. Unfortunately, there are cases that change very little iteration to iteration, but will continue
to change, resulting in significant drift in results. Various convergence criteria were attempted for ResHB
and the following are the best found to date:
• Hour. For each room, the current hour calculations are repeated until the sum of the absolute
change in surface temperatures plus air temperature is less than .0005 °K (.0009 °F), indicating
that a fully simultaneous solution has been closely achieved.
• Day. The day calculations are repeated until all rooms meet a) the fractional difference between
daily total inside and outside surface flux is less than .005 and b) the area-weighted total absolute
temperature change for all surfaces plus air is less than .0002 °K (.00036 °F). Note that the all-
room requirement means that some rooms will be iterated beyond this point.
• Swing. The swing search is continued until swings for all rooms are within .01 °K (.018 °F) of
specified. Each room has a specified swing. This allows different swings in master and slave
rooms, for example. Again, the all-room requirement means extra iteration for some rooms.
In addition to these basic criteria, there are various safety checks that detect oscillation or excessive
iteration and attempt to find an adequate result.

The criteria have been tuned to balance reasonable performance against successful convergence. In spite
of this, there remain cases that converge slowly or not at all. For example, heavy buildings that are well
insulated at the outside surface take hundreds of days to converge. These cases are the subject of on-
going attempts to find better initialization and/or iteration strategies.
One discovery is that hour and day convergence strongly interact with the swing search algorithm
(discussed below). An initial implementation used relaxed criteria during early swing search steps, the
idea being that approximate results were adequate during the early going. However, this resulted in
unstable estimates for next step supply air flow rate. The current stringent day and hour criteria were
found to be necessary for correct swing convergence.
The criterion of equal total daily heat flux on the inside and outside faces of all surfaces is useful because
it is absolute. Given the periodic nature of the calculation and the fact that there are no heat sources
within surfaces, all heat going into one face of a surface must come out the other. There was some
concern that although this criterion is true in theory, the numeric methods used in the Toolkit CTF
formulation would not be sufficiently precise to allow this test to be effective. These worries are
groundless – inside and outside face fluxes readily balance within 1 part in 107 in a well-converged room.
Temperature swing
ResHB uses a secant method search algorithm to search for the load when temperature swing is permitted.
Note that the calculations are based on varying system air volume flow rate with an assumed supply
temperature. This means the maximum extraction rate varies as the room temperature changes and sub-
cooling is self-limiting. If the modeling were done in terms of heat extraction, room air temperature
could be driven below the supply air temperature, which is impossible except under naturally floating
conditions.
The following calculation sequence is used:
• The required cooling air flow rate is found first for the 0 swing situation (that is, maximum
available air supply volume is unlimited and room temperature held at the set-point or floating
below it with no supply air flow).
• This maximum supply air flow rate is then reduced by 20% per °K (11% per °F) of target swing
and the room is calculated again. Generally, this reduction in supply air flow rate will produce a
significant temperature swing.
• The supply air flow rate is iteratively adjusted in proportion to the error in temperature swing, as
indicated by the secant method.
This algorithm is extremely efficient, because local linearity allows each subsequent estimate of supply
air flow rate to be much better than the prior one. Convergence to within .01 °K (.018 °F) of the target
swing usually occurs in less than 10 cycles. However, specific room characteristics can cause the search
to fail. As noted above, if room convergence is not essentially perfect, the secant method can produce
wildly unstable supply flow rate estimates. Several limits have been implemented to reduce the number
of cases that fail. Also attempted was an alternative algorithm based on a least-squares fit to the last
several points (as opposed to only the last two in the secant method); this approach did not work as well
as the secant method.
Master/slave control
The modeling of master/slave control is handled by the RHB temperature swing algorithm. By definition,
a slave room has the same air supply flow rate profile as its master, so the problem reduces to finding the
peak slave room supply air volume flow rate such that the maximum room temperature is the set point
(plus allowed swing if any). At each swing iteration, the peak flow rate is adjusted up or down using the

temperature swing search described above (the search is used even if the specified swing is 0). Then the
flow rates for all hours are set by applying the master room profile and the next day iteration proceeds
without any further adjustment of air flow.
In situations where the master and slave rooms have significantly different load profiles, sub-cooling can
occur in the slave room. This is illustrated in Figure 3 and Figure 4. These plots show room temperatures
for a two room building. The cooling set point is 24°C (75.2°F) and a 1°C (1.8°F) temperature swing is
allowed. In the first case, both rooms have a 10 m2 (107.6 ft2) window facing south; the master and slave
rooms have essentially the same temperature profile. In the second case, the master room has a west
facing window and the slave room has an east facing window. When the master room load peaks in the
late afternoon, the slave room is uncomfortably cold.
Because RHB models rooms independently (neglecting air mixing between rooms), the sub-cooling effect
shown in Figure 4 is exaggerated. However, it is clear that the ResHB master/slave capability can be used
to identify the need for zoning and, with further development, allow optimization of zoned designs.
26 78.8
25 77
24 75.2
Room air temperature (°C)
Room air temperature (°F)

23 73.4
22 71.6
21 69.8
20 68
19 66.2
18 64.4
17 62.6
0 4 8 12 16 20 24
Hour Master (south)
Slave (south)
Figure 3. Master / slave room temperatures with similar load profiles

26 78.8
25 77
24 75.2
Room air temperature (°F)

Room air temperature (°C)
23 73.4
22 71.6
21 69.8
20 68
19 66.2
18 64.4
17 62.6
0 4 8 12 16 20 24
Hour
Master (west)
Slave (east)
Figure 4. Master / slave room temperatures with mismatched load profiles
Models
RHB development involved review, refinement, and extension of Loads Toolkit models, as described in
the following sections.
Inside surface convection coefficients

In the heat balance method, the load is the total convective transfer from all space surfaces plus
convective gain from other sources. Thus, the choice of surface convective coefficient values is
particularly crucial when implementing the procedure. The Toolkit offers a number of convective models
and additional alternatives are found in the literature. Several models are summarized in Table 1. The
Toolkit ASHRAE, TARP detailed, and Fisher models are documented in the Loads Toolkit (Pedersen et.
al. 2001). The TARP simple model description and additional information about TARP detailed are
found in Section III of Walton 1983.

Table 1. Inside surface convection coefficient values (W/m2-K) for model alternatives
ta = room air temperature (°C), ts = surface temperature (°C), ach = air change rate (h-1); 1 W/m2-K = .176 Btu/h-ft2-F
Ceiling Floor
Model Wall
Heat flow up Heat flow down Heat flow up Heat flow down
Toolkit
1.25 4.68 4.37
ASHRAE
Sys
6.14 6.14 6.14
TARP on
simple Sys
4.043 .920 3.078 4.043 .920
off
TARP
detailed
1.52 ⋅ 3 ta − ts .76 ⋅ 3 ta − ts 1.31 ⋅ 3 ta − ts 1.52 ⋅ 3 ta − ts .76⋅ 3 ta − ts
Fisher .49 ⋅ ach 0.8 .19 ⋅ ach 0.8 .13 ⋅ ach 0.8
Sys
RHB 5 5 5
on
(see
Sys
text) 4.043 .920 3.078 4.043 .920
off
Comparison studies were done for a range of conditions, with the expected result that loads depend
strongly on coefficient model. The Toolkit ASHRAE model was eliminated because it lacks sensitivity to
heat flow direction. The Fisher correlations are appropriate for ceiling diffuser configurations that are
typical in only a fraction of residential buildings.
The model selected for RHB is a variant of TARP simple. The “sys off” values are ASHRAE-based
natural convection values. To improve convergence stability, the transition between heat flow up and
down values is made linearly over 2°C (3.6°F), rather than abruptly. The “sys on” value is enhanced due
to air motion in the room. 5 W/m2-K (.88 Btu/h-ft2-F) was chosen as the “sys on” value based on analysis
of experimental data from ASHRAE research projects 529-RP and 664-RP for air change rates of
approximately 8 ACH (typical for residential systems). For each hour, the coefficient used in the heat
balance is the system-run-fraction-weighted combination of the “sys on” and “sys off” values.
Elevation effects on convective heat transfer

In ResHB, care has been taken to adjust air volumetric heat capacity as a function of site elevation. This
adjustment results in significant changes in predicted infiltration, ventilation, and HVAC heat transfers.
Thus the question arises as to the effect of elevation on convection coefficients. The convection
correlation given by Clear et. al. 2001 for horizontal roofs was implemented in a spreadsheet, with air
pressure and density as variables. A 10 m by 15 m (32.8 ft by 49.2 ft) roof was modeled with a surface
temperature of 60°C (140°F) and ambient air at 30°C (86°F). Two cases were analyzed: (1) forced
convection, 3.35 m/s (7.5 mph) wind parallel to the short roof dimension of the roof (Gr/Re2 ≈ 0.02) and
(2) natural convection, 0.045 m/s (0.1 mph). Elevation was varied from 0 m to 2250 m (7382 ft) and
standard atmosphere relationships were used to derive pressure and density from elevation. The
convective coefficients predicted by the Clear et. al. 2001 correlation were normalized to their 0 elevation
values, yielding the results shown in Figure 5 (where P is the site atmospheric pressure and P0 is sea level
atmospheric pressure).

1.050
1.000
0.950
h/h0
0.900
0.850
0.800
0.750
0.000 0.050 0.100 0.150 0.200 0.250 0.300
1 - P/P0 Forced
Natural
.24+.76 * P/P0
Figure 5. Effect of pressure on convection coefficient (see definitions below)

Given the magnitude of this effect (about 13% for Denver), applying an elevation correction to
convection coefficients has a significant effect on predicted loads for high elevation locations. A simple
linear approximation was developed and used in RHB (and also shown in Figure 5):
⎛ P ⎞
h = h0 ⋅ ⎜ .24 + .76 ⎟ (1)
⎝ P0 ⎠
where
h = convective coefficient at pressure P (units consistent with h0)
h0 = convective coefficient at sea level pressure
P = atmosphere pressure at site elevation (units consistent with P0)
P0 = sea level atmospheric pressure
Buffer spaces
One of the many advantages of the heat balance approach is that buffer space temperatures can be
predicted by simply modeling an unconditioned room. These temperatures can be used as outside
boundary conditions for surfaces of adjacent conditioned spaces. Figure 6 shows typical ResHB results
for an attic with a dark asphalt roof. Changes in roof solar absorptance and inside surface long wave
emissivity (to represent radiant barriers) have the expected effects on predicted temperatures.

70 158
60 140
Temperature (°F)
Temperature (°C)
50 122
40 104
30 86
20 68
10 50
0 32
0 4 8 12 16 20 24
Hour Roof surface
Attic air
Figure 6. Roof surface and attic air temperatures predicted by ResHB
Infiltration
After review of available models, the AIM-2 model was selected for RHB (Walker and Wilson 1990,
Walker and Wilson 1998, and “enhanced model” in Chapter 26, ASHRAE 2001). As with other
simplified residential models, AIM-2 requires several input values that are difficult to determine,
including effective leakage area, leakage area distribution, and wind shelter parameters. ResHB provides
typical default values for these inputs. Leakage area can be specified based on pressurization test or
defaulted based on leakage classes defined by ANSI/ASHRAE Standard 119 (ASHRAE 1994).
AIM-2 is a single zone model. Infiltration leakage is determined for the entire building. In RHB, this
overall rate is allocated to rooms in proportion to volume – that is, the same air change rate is assumed to
apply to all rooms. Prior methods have variously allocated infiltration in proportion to exposed surface or
window area. Actual room leakage can be inward or outward and depends on room position relative to
the building neutral level and wind-induced pressure field. Thus there is no simple method for allocating
overall leakage other than using the average for all rooms.
Modeling of the interaction between mechanical ventilation and infiltration follows Palmiter and Bond
(1991) and Sherman (1992).
Distribution losses
ResHB duct losses are calculated using models specified in ANSI/ASHRAE 152-2004, Method of Test for
Determining the Design and Seasonal Efficiencies of Residential Thermal Distribution Systems and
Palmiter and Francisco 1997. These models are fully implemented in the RHB method, allowing room-
specific losses and gains to be included in estimates of air requirements and equipment capacity.
Framed constructions
The Loads Toolkit CTF-based conduction model assumes one-dimensional heat flow and thus requires
layer-by-layer construction descriptions as input. Framed constructions are common in residential

buildings. ResHB includes an algorithm that derives fictitious material properties for a homogeneous
layer that corresponds to a framed layer. The resistance of the layer is chosen to preserve the overall U-
factor of the construction. Density and specific heat are the volumetric averages of the framed layer
components. Heat fluxes calculated with equivalent-construction CTFs were compared to those
calculated with CTFs found with detailed 3-D methods (Carpenter et. al. 2003). Differences of 5% or less
were found.
Fenestration and solar gain distribution

The Loads Toolkit fenestration model requires input of angular SHGC and absorptance values. This is
not practical for a routine use. ResHB implements fenestration class, which can be thought of as a family
of fenestration types that exhibit common behavior. The fenestration class embodies the ratio of
transmission to absorption and the angular characteristics of the fenestration system. An actual
fenestration is specified by its U-factor, SHGC, and its fenestration class. The required angular
characteristics are taken from the fenestration class and are scaled by the ratio of rated SHGC to nominal
(fenestration class) SHGC. ResHB includes built-in fenestration class definitions for common residential
glazing types.
ResHB uses the ASHRAE Interior Attenuation Coefficient (IAC) and Exterior Attenuation Coefficient
(EAC) models to represent interior and exterior shading treatments (Chapter 30, ASHRAE 2001).
Overhang and fin shading is modeled with Loads Toolkit methods. ResHB additionally allows shading
scheduled by hour of the day.
The conversion of radiant solar gain into cooling load occurs when surfaces are heated by incident radiant
gain and energy is transferred convectively to room air. The distribution of solar gains to the various
room surfaces is thus an important determinant of cooling load. The actual distribution within a room
involves multiple surface inter-reflections and is impractical to compute. RHB uses a modified version of
the Loads Toolkit BLAST model, which distributes radiative gains in proportion to surface area-
absorptance product. Beam gain is assumed to hit floor surfaces. The RHB enhancement to this model is
that internal mass surfaces are assumed to be “half floor” with respect to beam radiation, based on the
idea that furnishings typically intercept some of the incoming beam. As with the BLAST model,
fenestration surfaces are back-transmit some incident radiation, so a room has an overall cavity
absorptance less than 1 (although typically very close to 1 except when glazing fraction is large).
Ground heat transfer

There is speculation that net heat flow into the ground in slab-on-grade houses can have a significant
impact on sensible cooling loads. However, no truly simple ground loss models are available and
determination of soil properties is problematic. Several approaches were investigated, notably
Beausoleil-Morrison and Mitalas 1997. Pending further research, in RHB, slabs are modeled with 300
mm (1 ft) of earth and adiabatic boundary conditions. This construction captures some of the diurnal heat
storage effects of slab construction, but not net conduction to the ground. For heating load calculations,
some improvements were achieved for calculation of ground losses, as discussed in Section 4.
Modeling Assumptions
Practical application of the HB method requires that fixed or default values be established for as many
inputs as possible, both as a practical aid to the user and to achieve consistent results. The following
sections document the assumptions developed for RHB.

Outdoor design conditions
RHB requires hourly outdoor conditions for the design day. The Loads Toolkit requires user input of 24
hour profiles for these values, which is impractical. While ResHB retains the ability to accept full
profiles for testing purposes, it can also automatically generates profiles from design dry-bulb
temperature, daily range of dry-bulb temperature, coincident wet-bulb temperature, site coordinates, and
site elevation, as follows:
• Dry-bulb temperature. The design dry bulb and daily range are expanded to a 24 hours using the
generic profile from Table 17, Chapter 29 of ASHRAE 2001. The generic profile is shifted 1
hour later when daylight savings is specified.
• Wet-bulb temperature and other moisture-related values. The design dry bulb and coincident wet
bulb are used to determine the design dew point temperature. The hourly dew point is the
minimum of the design dew point and the hourly dry bulb (that is, constant absolute humidity is
assumed, limited by saturation). Other hourly psychrometric values (wet bulb temperature,
humidity ratio, and enthalpy) are derived from the hourly dry bulb and dew point temperatures.
• Solar radiation. Hourly incident solar is derived using the ASHRAE clear sky model (Chapters
29 and 30, ASHRAE 2001) with updated coefficients per Machler and Iqbal 1985.
• Sky temperature. Sky temperature is required for calculation of exterior surface long wave
radiant exchange. The model of Berdahl and Martin 1984 is used to calculate hourly sky
temperature from hourly dry-bulb and dew point temperatures (cloud cover assumed to be 0).
All psychrometric calculations are done using Loads Toolkit procedures (originally from Brandemuehl et.
al. 1993) assuming a constant barometric pressure determined from site elevation according to a standard
atmosphere relationship (Eqn (3), Chapter 6, ASHRAE 2001).
Internal gain
RHB internal gain assumptions are based on Building America 2003, which provides gain intensities and
schedules for significant residential end uses as a function of building floor area and number of
occupants. When estimating residential internal gains, care must be taken to distinguish between energy
consumption and space gain. For example, a clothes dryer uses significant energy, but most is exhausted
outside the space. In addition, RHB requires the radiant/convective/latent split for each gain source,
which Building America 2003 does not fully define. Estimates were developed from ASHRAE 2001 and
other sources as needed. These values are shown in Table 2 and have been incorporated into ResHB.
Note that both sensible and latent heat gain due to hot water use are neglected because they are not
covered in Building America 2003; these gains are probably small due to intermittent use and shower
exhaust fans but deserve further investigation.
Table 2. Fractional components of internal heat sources
Source Internal gain (to space)
Exhausted
Radiant Convective Latent
Refrigerator 0 1 0 0
Range .24 .16 .30 .30
Dishwasher .51 .34 .15 0
Clothes washer .40 .60 0 0
Clothes dryer .09 .06 .05 .80
Lighting .79 .21 0 0
Other appliances and plug loads .54 .36 .1 0
People (living) .33 .22 .45 0
People (sleeping) .30 .30 .40 0

A typical 24 hour internal gain profile is shown in Figure 7. Note that the sensible gain values during late
afternoon peak cooling hours are quite moderate. This is consistent with the traditional values used in
prior methods. For example, for the Figure 7 case Manual J 8th Edition specifies internal gain of 621 W
(2120 Btu/h, 1200 Btu/h for appliances plus 230 Btu/h per occupant). Because of the relative timing of
typical peak solar and internal gains, the contribution of the latter to the total sensible load is generally
surprisingly small.
1800 6143
1600 5461
1400 4778
1200 4095
Gain (Btu/h)
Gain (W)
1000 3413
800 2730
600 2048
400 1365
200 683
0 0
0 3 6 9 12 15 18 21 24
Hour Sensible
Latent
Figure 7. Hourly internal gain (all sources) for a 186 m2 (2000 ft2) house with 4 occupants
The gains specified in Building America 2003 are average values derived from energy consumption
measurements. It could be argued that higher values should be used for loads calculations, given that 600
W is less than the output of a single range burner. However, gains from normal residential activity are
intermittent and are absorbed by small space temperature swings. Higher assumptions are appropriate
only in cases where a significant gain is routinely expected. Occasional high gain situations, such as
social functions, should not be considered or handled with a parallel system designed for that condition.
Internal mass
The presence of internal mass (such as furniture) in a room has two effects on cooling load: the load is
increased due to enhanced surface area available for convective exchange and decreased due to the
storage (depending on its construction, space temperature swing, and presence of other mass such as a
slab floor). The ultimate impact of internal mass depends on which of these effects dominates in a
particular case. Note also that residential buildings typically have relatively small rooms and thus the
partition-to-floor area ratio can easily exceed 1 and often exceed 2. Partitions function as additional
internal mass.
ResHB parametric studies showed the impact of internal mass on sensible cooling load to be –2% to
+23%. Typical results are shown in Figure 8 and Figure 9 for a 162.6 m2 (1750 ft2) house in Washington,
DC with various amounts of 12 mm (.5 in) wood internal mass.

1.300
1.250
1.200
1.150
1.100
1.050
1.000
0.950
0.900
0 0.5 1 1.5 2 2.5
IM area / floor area Swing = 0
Swing = 0.83°C (1.5°F)
Swing = 1.67°C (3°F)
Swing = 2.5°C (4.5°F)
Figure 8. Effect of internal mass, wood floor construction
1.300
1.250
1.200
1.150
1.100
1.050
1.000
0.950
0.900
0 0.5 1 1.5 2 2.5
IM area / floor area Swing = 0
Swing = 0.83°C (1.5°F)
Swing = 1.67°C (3°F)
Swing = 2.5°C (4.5°F)
Figure 9. Effect of internal mass, slab floor construction

No source was found for documented information about the amount or composition of internal mass
found in typical residences. Using judgment, a recommended assumption was selected for RHB: each
room should be modeled including internal mass having surface area equal to room floor area and
consisting of 12 mm (.5 in) wood exposed on one side (adiabatic outside surface conditions). This
surface should be radiantly coupled to all room surfaces. ResHB implements a special surface type “IM”
having this behavior.
Other assumptions
Surface absorptance. Table 3 summarizes absorptance values recommended for typical load calculations.
Other values should be used as appropriate, particularly for roof outside absorptance if a reflective color
is specified.
Table 3. Default surface absorptances
Surface Short wave (solar) Long wave (thermal)
Outside Inside Outside Inside
Roof / ceiling (tilt < 60 °) .8 .45 .9 .9
Wall (180° < tilt ≤ 60°) .6 .45 .9 .9
Fenestration per FenClass per FenClass .84 .84
Floor (tilt = 180°) 0 .6 .9 .9
Material properties. ResHB includes default material properties gathered and reconciled from multiple
sources, as documented in Appendix A.
Toolkit Re-engineering
As stated above, the starting point for ResHB was the ASHRAE Loads Toolkit (Pedersen et. al. 2001).
Specifically, ResHB was created via incremental re-engineering of the Toolkit successive substitution
sample zone example. That module is a fully functional single zone model that uses the conduction
transfer function (CTF) calculations for building surfaces and the MRTBal radiant exchange model
(Liesen and Pedersen 1997; Walton 1980). Changes were introduced one at a time. Comparison of
output from each incremental change minimized the possibility of unintended changes.
The Toolkit is intended to allow exploration, understanding, and comparison of alternative loads models.
Its structure emphasizes clarity and modularity over efficiency. These objectives are secondary for
ResHB and ultimately an almost entirely new application resulted as the original code was adapted to the
residential loads application. Some of the significant changes include --
• Application framework. ResHB includes various utility capabilities such as error handling,
command line processing, and common report formatting functions.
• Streamlined input processing. The data dictionary file (IDD) is now simplified and embedded in
the application, making ResHB.exe a complete, standalone package. Unused input processing
features have been removed. An INCLUDE capability is now available, allowing common input
to be maintained in a single file. In addition, a rudimentary PARAMETER scheme has been
added.
• Unused models removed. As ResHB algorithms are determined, the alternative methods provided
in the Toolkit have been dropped. This makes the code smaller and eases development.
• Residential models added. Models not included in the Toolkit have been added for infiltration
and duct losses. Other models have been enhanced.
• RHB calculation algorithms. Temperature swing and master / slave control algorithms were
added.

• Constant air properties. The Toolkit performs detailed psychrometric calculations to obtain an
exact energy balance on the room air. ResHB uses constant indoor air properties derived during
initialization from site elevation and nominal room conditions. This change causes minor
changes in results.
• Data structure generalization. In order to support multiple systems, zones, and rooms, global
variables have been removed and replaced by arrays allocated as needed. The result is that
ResHB problem size is limited only by available memory.
• Efficient initialization. All temperature-independent calculations are done only once in ResHB,
including solar geometry, solar gains, and internal gains.
• Refactoring constructions. In ResHB, several changes have been made in the logical structure of
surface construction. This revised structure is deemed more convenient since it does not require
multiple definitions for the same material.
o Surfaces can be described with a simple U-factor or with a layered construction.
Absorptivities are surface properties, not construction properties. Also, surfaces may
reference a “reversed” construction, eliminating the need for duplicate definition of a
shared construction (a room ceiling and an attic floor, for example).
o Constructions are a series of layers specified by thickness and material.
o Materials are characterized by conductivity, density, and specific heat (or pure
resistance). This contrasts with the Toolkit MaterialLayer formulation, which includes
thickness. A default thickness material thickness is available that is convenient for
defined-thickness layers, such as carpet.
• Temperatures in °C. The Toolkit uses absolute temperature (°K) for all calculations. ResHB was
converted to operate in °C with appropriate conversion to absolute temperature as needed (e.g. for
radiant exchange calculations).
ResHB loads calculations for multi-room buildings generally take less than 1 second on a 2.4 GHz PC.

3. Inter-model Validation of the Residential Heat
Balance Method
Introduction
Background
An important aspect of 1199RP is demonstrating the validity of results, diagnosing deficiencies in the
calculation method or its computer implementation, and evaluating the system designs resulting from the
procedure.
A number of systematic validation and assessment methods have been developed for building
simulation/energy calculation programs in the last two decades, e.g. Judkoff, et al. (1983); Judkoff and
Neymark (1995), BRE (1988) and Bloomfield et al. (1988). Much less emphasis has been placed on
design load calculation procedures, perhaps since design load calculation methods have historically relied
less heavily on computer implementation than annual energy calculation (Rees and Spitler 1999)
Three types of test methods applied to whole building energy simulation programs can be identified
(Judkoff et al. 1983). These are experimental validation, inter-model comparison and analytical testing. In
any type of test, where results of a given code are compared with data from another source, three things
are implicitly tested, each of which may contribute to the overall ‘error’ in the results:
• The interpretation of the input data
• The model(s) or algorithm(s)
• The computer implementation of the algorithm(s)
Being most abstracted from the full complexities of real building simulation problems, analytical testing
has the advantage of offering the most certain form of reference or ‘truth’ model with which comparisons
can be made. The nature of analytical testing also makes it only applicable to limited cases for which
analytical solutions can be derived. Errors arising from the integrated performance of all the sub-models
and algorithms in a program are beyond the scope of an analytical test.
Experimental validation offers an approximate truth standard within the accuracy of the data acquisition
system and can be applied with any level of complexity. However, detailed, high quality, measurements
are usually expensive and time-consuming, even for one or a few cases.
Inter-model comparison cannot directly address the issue of a truth standard, but can be a very powerful
way of identifying errors by doing many comparisons quickly and inexpensively. Inter-model
comparisons of results from annual energy analysis codes have been attempted by a number of groups.
One notable attempt has also been made at devising a systematic diagnosis tool, BESTEST, based on
inter-model comparisons (Judkoff and Neymark 1995).
A further type of inter-model comparison can be identified that consists of a large number of test cases (of
the order one thousand) where certain parameters are systematically varied. Such a study using ‘design
day’ cooling load calculation methods was performed as part of ASHRAE 942-RP (Rees et al. 1998;
Spitler and Rees 1998). In order to check the residential heat balance (RHB) procedure, it is desirable
that this type of inter-model comparison be made in a highly-automated fashion, so that the RHB
procedure can be tested and evaluated throughout the development cycle.

Selecting the comparison tool
In choosing the building simulation program to be used for comparison in the inter-model validation, a
number of features have been considered. These include:
• Technical basis: heat balance
• Accuracy, as demonstrated by analytical tests and experimental validation. In addition to the
analytical tests developed for ASHRAE 1052-RP (Spitler, et al. 2001), tests developed for testing
inter-zone heat transfer, inter-zone airflow, and master-slave zone control have been developed.
(Xiao, et al. 2002)
• Capabilities for master-slave zone control, interzone conduction heat transfer, and interzone
airflow.
• Feasibility to extend the code, if necessary
• Automated parametric run feasibility.
• Cost for licensing multiple copies to run on several machines simultaneously.
A systematic review of candidate computer programs has been made. Candidates in the review included:
SPARK (LBL and Ayres Sowell Associates 1997), IDA Indoor Climate and Energy (IDA ICE 2004),
HLITE/VLITE (Walton. 1993), HTB2 (Alexander 1997), ESP-r (ESRU 2000), HVACSIM+ (Park, et al.
1986), BLAST (BLAST Support Office 1995), and EnergyPlus (EnergyPlus 2004). A detailed review
was given by Xiao (2001). After a broad review of the candidate programs, preliminary analytical tests,
based on ASHRAE 1052-RP, were performed with the two most promising candidates: ESP-r and
EnergyPlus. After weighing the advantages and disadvantages of each candidate, ESP-r was selected as
the comparison program.
The ESP-r System for Building Energy Simulation has been under development since the 1970s (Clarke
1977). With more than two decades of development and use, ESP-r has been extensively validated. ESP-r
is a finite volume and thermal network method based transient energy simulation program. (The thermal
network method discretizes the building into a network of nodes and applies heat balance to each of the
nodes in the simulation. It may be considered a refinement of the heat balance method in many respects.)
ESP-r employs a partitioned solution approach, which applies customized solvers to each model domain
(thermal, inter-zone air flow, electric power flow, et.). Interdependencies are handled by passing
information between solution domains on a time-step basis (Beausoleil-Morrison 2000). This manner
enables the optimal treatment of each model domain and the global solution to evolve in a coupled
manner.
Methodology
Types of comparison
Two types of inter-model comparison between the residential heat balance (RHB) load calculation
procedure and ESP-r are performed:
• Load Comparisons. In these comparisons, loads predicted by RHB are compared to those
calculated by ESP-r. These comparisons serve to verify RHB calculations and to reconcile RHB
and ESP-r input used for system design evaluation.
• System Design Evaluation. In these comparisons, the RHB procedure is used to generate design
system capacities. In ESP-r, the buildings are simulated with a residential cooling system, with
system capacity and room airflows taken from the RHB procedure. Room temperatures and

comfort levels are then evaluated. These comparisons provide guidance for refinement of the
RHB procedure.
The RHB method is implemented in a computer program, ResHB.
Parametric code
Both types of inter-model comparisons depend on a parametric code that identifies the specific case to be
run by both programs. A typical code is:
R2HsLCASDO000Q00m00AmmmmCwmhtmWwmmG1000mmmmPwwMwmFwm0m
Each sub-field of the code controls one aspect of input file generation. In this example, “Hs” indicates
that building “s” (shoebox) should be used, “LCASD” means that the location is San Diego, CA, and
O000 specifies building orientation of 0 (building front facing north). “Q00m00” means zero internal
gain, zero infiltration, medium cooling set point, zero room temperature swing, and zero set point setoff in
the controls. “Ammmm” means all interior surfaces have medium short wave and long wave
absorptivities. Other fields determine surface construction and configration, fenestration type and amount,
shadings, internal mass, etc. (e.g. “Cwmhtm” means the exterior ceiling is of wood frame construction
with medium insulation, and the roof above the hip attic has tile construction with medium short wave
absorptivity.) A complete description of the parametric code is found in the RHBGen documentation.
See Appendix D.
There are literally millions of combinations that can be generated using this scheme. In the results
presented below, cases have been chosen to capture typical residential building features and cover a
variety of applications.
Combined testing process
Figure 10 illustrates the combined test process, which includes both types of comparisons. It starts with
the parametric run generator RHBGen (described below), which takes as input the basic parametric case
list, a description of the parameter variations and building prototypes. It then generates an expanded case
list and RHB input files, and runs RHB. Then, the ESP-r input generator (described below) takes in the
expanded case list and the system capacities and design airflow rates determined by RHB, and creates the
ESP-r input files. Perl scripts (described below) will then run the ESP-r simulations, both with an ideal
system (to determine the loads for a load comparison) and with a residential cooling system with RHB-
determined system capacities and design airflow rates (for the system design evaluation.) ESP-r results
are then processed by the ESP-r output processor and post processor (described below).

Parametric
Case list
Variants RHBGen RHB RHB

Input files
Prototypes
Expanded System
designs Loads
case list
ESP-r input generator

ESP-r
results
ESP-r
Input files ESP-r output processor &
post processor
ESP-r
Design evaluation Loads comparison
Figure 10. Overall inter-model testing process
RHBGen parametric generator

RHBGen is a FORTRAN application built with the same general framework as RHB. Starting with a
parametric case list, a description of the parameter variations and building prototypes, it generates and
runs large numbers of RHB cases. In the inter-model comparison, RHBGen is not only used to generate
and run RHB cases, but also to generate a list of codes of the cases to be compared, loads for comparison
to ESP-r and system capacities and airflow rates as inputs to ESP-r.
An important RHBGen feature is its ability to apply variants to a given parametric code. For example, the
variant “O=000|090|180|270” causes each base case to be expanded into 4 cases – one for each
orientation. Variants are applied in a nested fashion, so “L=CASD|TNME|FLKW, O=000|090|180|270”
generates 12 runs. (In this case, it would generate all combinations of locations San Diego, Memphis and
Key West, and orientations of north, east, south, and west.) This makes it straightforward to generate
large case sets in an automated fashion.
ESP-r system
ESP-r input generator

The ESP-r input generator starts with a set of base case input files for each house prototype that is going
to be tested. In the base case input files, there are no windows, shading, internal mass, internal gains or
infiltrations defined. Given the parametric code for the specific case, the ESP-r input generator will
modify the corresponding input parameters.
The ESP-r input generator is created taking advantage of the text mode interface of ESP-r. It exists in the
form of shell scripts managed by a set of Perl scripts, rather than a FORTRAN program as in RHBGen.

There are two types of shell scripts: one type modifies the prototype dependent input parameters; the
other type modifies the prototype independent input parameters. There is a set of the first type of shell
scripts for each house prototype and a set of the second type of shell scripts for all prototypes.
Each shell script serves to modify a group of input parameters and is manipulated by the Perl scripts,
which manage the entire ESP-r test procedure. The parameters are grouped to be modified by individual
shell scripts as follows:
• Location: The location information (weather file name and latitude, longitude difference) is
modified by the shell script Climate. (The ESP-r weather file for each location is created
beforehand by a separate weather file generator.)
• Orientation: The orientation of the building is modified by the shell script Rotate, which rotates
the building to the required orientation. When external shading obstructions are present, they are
rotated together.
• Fenestration: The fenestration area, type, and internal shading information are modified by the
shell script WinAreaTyIsh. External shading, if present, is added by the shell script ExtSh. All
fenestration types are predefined in the ESP-r optics database and are selected by the database
index.
• Internal gain and infiltration: The internal gains and infiltration are modified by the shell scripts
IntGain and Infiltration.
• Construction and surface properties: The construction and surface properties of the external walls,
external ceilings and roofs, external floors, internal partitions and internal floors/ceilings are
modified by the shell script Modcon. The surface properties (short wave absorptivity and long
wave emissivity) are stored in the material property database in ESP-r and then incorporated into
the construction type. So for each combination of construction and surface property, a
construction type needs to be defined in the ESP-r construction database. The construction index
in the database is assigned to a specific fabric component using the “construction anchor” of ESP-
r, which can be used to attribute a property to a group of components.
• Internal thermal mass: If present, the internal thermal mass and its construction type are added by
the shell script ThermalMass.
• Floor configuration: the external floor boundary condition is modified by the shell script
FloorConfig.
• Set point: In the system design evaluation mode, the internal air set point in the master zone is
modified by the shell script Setpoint. In an ideal load comparison mode, the set points for all
zones are modified by the shell script Setpoint_Ideal.
• After the above parameters are modified, the view factors for internal long wave radiation, the
solar shading and insolation files, the convection coefficient files (if fixed convection coefficients
is being used) will be updated by the shell scripts ViewFactor, Shade and ShInsolation (or
Insolation), and HcConv.
The input generator modifies the parameters in an order that will not cause any conflicts between various
input files (since the ESP-r input is a multi-file system) and ensures that all information is updated. The
order that has been found to work well is as follows:
• House Model (select base case)
• Location (Climate)
• Floor boundary condition configuration

• Constructions except for windows and internal thermal mass
• Thermal mass (construction, surface/floor area ratio)
• Window area, type, internal shading
• Window external shading (overhangs, fins)
• Internal gain level
• Infiltration level
• Orientation (rotate building)
• Update internal surface view factors
• Update shading/insolation files
• Update convection coefficient files if necessary
• Set point modification
ESP-r test procedure

As shown in the overall flow diagram, the ESP-r test procedure reads in the parametric codes and creates
corresponding input files; takes the system capacities designed by RHB; runs the simulations; and
processes the results. PERL scripts are written to manage the whole procedure. Specifically, the PERL
scripts:
• Read in the parametric code for each case. (The case code list is generated by RHBGen.)
• Make a copy of the base case input files.
• Call the input generator (shell scripts) to modify input parameters.
• Read in the system inputs from the RHB result file (generated by RHBGen).
• Call shell scripts to modify the system airflow rates if in the design evaluation mode.
• Call shell scripts to run the simulation.
• Call shell scripts to analyze the result.
ESP-r output processor and post processor

The ESP-r output processor extracts and summarizes the results from each test case and writes the results
of interest to files for further analysis and plotting. The ESP-r output processor is also written in Perl. All
output files are in comma-separated format for convenience of post processing in a spreadsheet.
For load comparisons, both the peak loads and hourly loads of ESP-r and RHB are extracted and output
into four separate files. (ESP-r peak, ESP-r hourly, RHB peak, and RHB hourly.)
For system design evaluations, the output processor writes a file containing the maximum, minimum, and
mean values of room temperature and Predicted Percentage Dissatisfied (PPD). Times of occurrence for
the minimum and maximum values are also output. In addition, hourly room temperature and PPD values
are output in two separate files.
Some output from ESP-r (such as PPD values) needs to be further processed in order to present the
evaluation results clearly. This is dealt with by the ESP-r post processor (in the form of PERL scripts).
Having a separate post-processor allows experimentation with the post-processing algorithm without
having to rerun the time-consuming ESP-r simulations.

Design evaluation figure of merit
For purposes of the design evaluation testing procedure, it is desirable to have a figure of merit for each
system design (designed for a specific building case with a given version of the load calculation
procedure). The figure of merit should have the following characteristics:
• Feasible to calculate with ESP-r, or to calculate with results given by ESP-r.
• Somehow capture the overall suitability of the design, as experienced by the occupants.
• Useful in identifying bad systems – ones that have too much temperature swing or too much
temperature variation between rooms.
• Simple – preferably one number.
Some thermal comfort related parameter would seem to be ideal, as it should be useful in representing the
suitability of the design as experienced by the occupants. ESP-r can give the predicted mean vote (PMV)
or the predicted percentage of dissatisfied” (PPD) at each time step in the simulation or averaged over
each hour. (Time steps in ESP-r may be as short as a minute.) While PMV has the advantage of giving
information as to whether the space is too cold or too hot, if a simple figure of merit is to be reached, it is
preferable to further simplify the thermal comfort information by using the PPD. As shown in Figure 11
below (Figure 13 from Chapter 8 of the 2001 ASHRAE HOF), there is a direct relationship between PPD
and PMV.
Figure 11. Predicted Percentage of Dissatisfied (PPD) as a function of Predicted Mean Vote (PMV).
Source: Ch. 8, Figure 13, 2001 ASHRAE Handbook of Fundamentals
This relationship is given by Equation 64 in Chapter 8 of the 2001 HOF:
[(
PPD = 100 − 95 exp − 0.03353PMV 4 + 0.2179 PMV 2 )]
Being able to determine PPD for any given room at any given time step is useful, but does not yet allow
us to develop a single figure of merit. Unfortunately, no single figure of merit was identified that
distinguished between a “good” system and a “bad” system. As described in the section “System design
evaluations”, a procedure has been developed which gives integrated PPD values that distinguish between
undersizing, inadequate distribution / control problems, and nighttime ambient overcooling.
Model assumptions used in the comparison

When comparing the results and determining the source of possible discrepancies between ESP-r and
ResHB, it is necessary to look into the building sub-models used in each program. Both ESP-r and ResHB

may have one or several options for each sub -model. In the load comparison mode, we verified the
ResHB calculations and reconciled ResHB and ESP-r input used for design evaluation. In order to
reconcile the differences, it was necessary to use the same sub-models in both programs whenever
possible.
In the design evaluation mode, we evaluated the system design produced by ResHB through the relatively
detailed simulation of ESP-r so that guidance could be provided for the refinement of ResHB. Therefore,
in ESP-r, the “best available” sub-models were chosen. Table 4 shows the sub-models used for the
comparison results presented below. (Models that have not caused big discrepancies in our detailed
studies were ignored in the table.)

Table 4. Model assumptions used in ESP-r and RHB for the comparison
Models ESP-r RHB
ASHRAE clear sky model ASHRAE clear sky model with

Clear sky solar with Machler/Iqbal revised A, Machler/Iqbal revised A, B,
B, and C coefficients (1985) and C coefficients (1985)
Solar
irradiance
Sloped surface incident diffuse
solar (i.e. short wave sky Perez (1990, 1987) Perez (1990, 1987)
model)
Internal solar distribution Beam on the floor Beam on the floor

Clear sky temperature model Martin and Berdahl (1984) Martin and Berdahl (1984)
Outside ground surface 1 Equals outside air
ESP-r algorithm
temperature temperature
External Surrounding buildings’ surface 2 Equals outside air
ESP-r algorithm
long wave temperature temperature
radiation To the sky 0.45 0.35
View factor
for vertical To the ground 0.45 0.5
surfaces To surrounding
0.1 0.15
buildings
Fixed- Fixed-
Ceiling: 1.25 W/m2-K Ceiling: 1.25 W/m2-K
2 2
(0.22 Btu/hr-ft -F) (0.22 Btu/hr-ft -F)
Load
Inside Walls: 4.7 W/m2-K Walls: 4.7 W/m2-K
comparison 2 2
(0.82 Btu/hr-ft -F) (0.82 Btu/hr-ft -F)
mode 2 2
Floor: 4.4 W/m -K Floor: 4.4 W/m -K
(0.77 Btu/hr-ft2-F) (0.77 Btu/hr-ft2-F)
Fixed-12.5 W/m2-K Fixed-12.5 W/m -K
2
Outside
(2.2 Btu/hr-ft2-F) 2
(2.2 Btu/hr-ft -F)
Runtime weighted. System
Convection 2 2
on: 5 W/m -K (0.88 Btu/hr-ft -
coefficients System on: correlation based
F) on all surfaces
on data from ASHRAE 559-
System off: ASHRAE still air
RP and 664-RP
Design Inside values with flow-direction
System off: ESP-r default
evaluation enhancement; linear
natural convection (Alamdari
mode transition of value from ∆T = -
and Hammond, 1983)
1°C - +1°C
(-1.8°F - +1.8°F)
MoWiTT (Yazdanian and MoWiTT (Yazdanian and
Outside
Klems, 1994) Klems, 1994)
Elevation adjustment No No
1
Ground temperature at 1.2 m (4 ft) depth calculated from the procedure of Kusuda and Achenbach (1965) was
input to ESP-r. ESP-r calculates the ground surface temperature with a 6-node finite volume scheme, assuming the
1.2 m (4 ft) depth ground temperature as a constant.
2
The average temperature of the surrounding external buildings for a surface is estimated by using the temperature
of the zone surface, which has an azimuth difference of 180 Deg. This method is assumed to be applicable for
surfaces with elevations > -30 and < 30 deg. For all other surfaces (ie elevations < -30 and > 30 deg.) use a mean
weighted average of all external surfaces in all zones under simulation. (From ESP-r code)

Results
Description of test sets
The results presented below are for the two-zone “shoebox” prototype. This prototype has two adjacent
rectangular-shaped zones of dimension 8 m x 6 m x 2.4 m (26.2 ft x 19.7 ft x 7.9 ft). With orientation
equal to zero, the east-side zone is called the “living room”, the west-side zone is the “great room.” (The
two zones are connected on the 8 m x 2.4m 26.2 ft x 7.9 ft)facade. All other facades are exposed to the
outside.) If present, the windows will be at the front and left façades of the living room, and at the back
and right facades of the great room. The thermostat is located in the great room (the master zone). The
two rooms are connected by a doorway and the return grille is in the master zone, so that all return air
from the living room passes through the doorway into the great room.
A set of 576 test cases was created for the two-zone shoebox prototype. The test parameters, summarized
in Table 5, include 9 locations, 4 orientations, 2 fenestration types, 2 fenestration areas, 2 constructions
(light weight and heavy weight) for the roof, external wall and floor, and 2 floor configurations.
Three test sets of 576 cases were created. One set was used for load comparisons -- this set used a fixed
setpoint and fixed convection coefficients. The other two test sets, for system design evaluation, used the
convection correlations summarized in Table 4. Two test sets were created by varying the allowable
temperature setpoint swing – in one set, the RHB procedure sized the system and airflows using a fixed
setpoint; in the other set the system capacity and airflows were determined which allowed a maximum
temperature swing of 1.67º C (3.0ºF).
Table 5. Test parameters for the 576 test cases
Number of
Parameters Variants
Variants
CASD, TNME, GAAT, AZPH, CASA, COAL,
FLKW, MNMI, TXDA (San Diego; Memphis;
Location 9
Atlanta; Phoenix; Sacramento; Alamosa, Colorado;
Key West; Minneapolis; Dallas)
Changing
parameters Orientation 4 0, 90, 180, 270
1: single glazing clear;
Fenestration type 2
x: double glazing low-E
Fenestration area multiplier 2 l: 0.5; m: 1.0
Roof, wall, floor construction 2 w: wood frame; c: solid concrete
Floor configuration 2 0: adiabatic; a: exterior
House model 1 Shoebox
Roof, wall, floor insulation 1 m
Interior mass floor area ratio 1 m: 100%
Internal gains, Infiltrations 1 0
Set point 1 m:24ºC (75.2ºF)
0: 0ºC (0ºF) or m: 1.67ºC ºC (3.0ºF), depending on
Temperature swing 1
the test mode
Other Temperature setpoint offset 1 0: none
parameters
Exterior SW absorptivity 1 m: 0.6
Interior SW absorptivity 1 m: 0.6
Interior LW absorptivity 1 m: 0.9 (fenestration: 0.84)
Fenestration interior shading 1 0
Fenestration exterior shading 1 0
Interior partition construction 1 w: wood frame
Interior mass construction 1 w: wood frame

Ideal load comparison
A comparison of the ideal peak loads for the 576 test cases is shown in Figure 12 (for the Great room) and
Figure 13 (for the Living room). In both plots, the RHB-calculated peak load is plotted against the ESP-r
calculated peak load. With the exception of a few cases, all of the load differences fall within a +/- 15%
band. Almost all of the error is caused by differences in the long wave radiation models used by the two
programs.
As summarized in Table 4, each building surface is assumed to exchange longwave radiation with three
different source/sinks –the sky, ground, and surrounding buildings. Both the values of the source/sink
temperatures and the view factors from the surface to each source/sink are important in determining the
longwave radiation flux. Comparing the model assumptions in each program, the view factors are
slightly different; the sky temperatures are the same; but the temperatures assumed for the surrounding
buildings and ground are somewhat different.
In the RHB program, both the outside ground surface temperature and the surrounding buildings’
temperature are assumed to be equal to the ambient dry bulb temperature. In ESP-r, the user may specify
a ground temperature at 1.2 m (4ft) depth; then a 1-d finite volume procedure is used to calculate the
ground surface temperature. We used the Kusuda and Achenbach (1965) procedure to determine the
ground temperature at 1.2 m (4 ft) depth. It should be noted that the 1-d finite volume procedure only
accounts for conduction in the soil and convection, solar radiation, and longwave radiation at the surface.
Evapo-transpiration effects are not included – hence, this is roughly equivalent to the ground being
covered with pavement rather than vegetation. Vegetation should lower the ground surface temperature
substantially. As shown in Figure 14 for Alamosa, Colorado, the ESP-r ground surface temperature is
quite a bit higher than the RHB ground surface temperature, which is the same as the air temperature.
This causes ESP-r to predict higher radiative heat fluxes from the surrounding ground to the building
surface; higher surface temperatures; increased heat gains; and higher cooling loads.
The surface temperature of the surrounding buildings in RHB is also assumed to be the ambient dry bulb
temperature. In ESP-r, the average temperature of the surrounding buildings is estimated by using the
temperature of a zone surface that has an azimuth difference of 180 degrees (for surfaces with elevations
between –30 ~ 30 degree). For all other surfaces (with elevations less than -30 or greater than 30 degree),
a mean weighted average of the temperatures of all external surfaces in all zones is used. Figure 15
shows the comparison of the ambient dry bulb temperature and the outside surface temperatures
calculated from ESP-r for one of the rooms in Alamosa, Colorado. The room has light weight (wood
frame) construction with medium insulation, and 25% glazing (low-E double pane)area on the south and
west facades. As can be seen, during the peak load hours, the surrounding buildings’ surface
temperatures are higher than the ambient dry bulb temperature, which leads to higher external long wave
radiation from the surrounding buildings. Again, this leads to higher cooling loads in ESP-r.
For the case shown in Figure 15, Figure 16 shows a comparison of the hourly cooling loads. The peak
cooling load from the RHB program is about 10% lower than the peak cooling load calculated by ESP-r.
For comparison purposes, the load calculation was redone with both programs after setting the longwave
emissivity to 0.01. This effectively eliminates the longwave radiation. The results are shown in Figure
17. Note the peak cooling loads are now within 2%. The conclusion is that the longwave radiation model
is the significant difference between the two calculation procedures. Astute readers may also note that the
peak cooling load went up after the longwave emissivity was set to 0.01. In Alamosa, Colorado, the sky
temperature varies between –11.7ºC (10.9ºF) and 3.0ºC(37.4ºF). Therefore, radiation from the roof to the
sky and, to a lesser extent, from the walls to the sky, significantly reduces the cooling load. When the
longwave radiation is effectively eliminated, the cooling load shows a corresponding increase.

10000
9000
8000
15%
7000
RHB Loads (W)
6000
5000 -15%
4000
3000
2000
1000
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
ESP-r loads (W)
Figure 12. Peak cooling load comparison for the Great room
10000
9000
8000
15%
7000
6000
RHB Loads (W)
-15%
5000
4000
3000
2000
1000
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
ESP-r loads (W)
Figure 13. Peak cooling load comparison for the Living room

45
40
35
Temperature (C)
30
25
20
15
10
3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00
Hour
RHB ESP-r
Figure 14. Outside ground surface temperatures from RHB and ESP-r for Alamosa, CO
60
50
40
Temperature (C)
30
20
10
0
0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00
Hour
Ambient dry bulb Great south surface Great north surface Great west surface
Figure 15. Comparison of the ambient dry bulb temperature and the outside surface temperatures
calculated from ESP-r (Great room)

2500
2000
Cooling Load (W)
1500
1000
500
0
0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00
Hour
RHB ESP-r
Figure 16. Comparison of cooling load calculated by RHB and ESP-r: with external long wave radiation
2500
2000
Cooling Load (W)
1500
1000
500
0
0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00
Hour
RHB ESP-r
Figure 17. Comparison of cooling load calculated by RHB and ESP-r: without external long wave
radiation
In summary, ESP-r predicts higher cooling loads than RHB because of the different longwave radiation
source/sink temperature models. RHB uses very simple models, assuming that both other building
temperatures and the ground surface temperature are the same as the air temperature. ESP-r has

comparatively complex models, and the model for the other building temperatures seems likely to give
more accurate answers. With regards to the ground surface temperature model, it should be noted that the
ESP-r model is roughly equivalent to being surrounded by an empty parking lot. The ground surface
temperature model does not account for evapo-transpiration from the ground, nor does it account for the
effects of vegetation. Presumably, the effective surface temperature for a building surrounded by lawn
may be better approximated by the air temperature than a parking-lot surface temperature. While this
tends to have a significant effect only on some lightly-insulated buildings, it should be a topic of further
research.
System design evaluations

As mentioned above, one challenge in evaluating system designs is to come up with either a single figure
of merit or a small number of figures of merit. Because thermal discomfort can occur for several reasons,
which vary in their cause and acceptability, it was thought desirable to develop figures of merit that
differentiate between the causes:
• Night time ambient overcooling. In some locations, e.g., Alamosa, Colorado, the house will cool
below the set point at night. Although this results in increased PPD, it is thought to be
acceptable. In any case, it is not the “fault” of a poor air-conditioning system.
• Undersizing. In some cases, the system designed by the RHB procedure will be undersized,
allowing the temperature to rise above the setpoint in the master zone (i.e. the zone with the
thermostat.) While this is acceptable at some level, the PPD that results from this is tracked
separately.
• Inadequate distribution / control limitations. In other cases, the master zone will maintain the
setpoint, but the other zones (i.e. slave zones) may drift above the setpoint or drift below the
setpoint. Again, the PPD resulting from this phenomenon is tracked separately.
• Other. Finally, a small amount of PPD may result even when the system setpoint is being held
within a reasonable tolerance, δT. δT=+/- 0.15ºC (0.27ºF) was chosen as reasonable. This was
based on the ESP-r simulation of the on-off control, which will typically hold the thermostatically
controlled zone within 0.15ºC (0.27ºF).
The algorithm may be described as follows. At each hour, the master zone temperature and slave zone
temperature are checked. PPD in the master zone is attributed to one of the above causes, depending on
the master zone temperature, as shown in Table 6. δT is taken to be 0.15 C (0.27ºF). For each of the
slave zones, the PPD is attributed to one of the above causes, depending on both the master zone
temperature and the slave zone temperature. Further consideration of this algorithm will reveal that it is
necessarily approximate – there are some conditions that are ambiguous. For example, when the
temperature in the master zone exceeds the setpoint by more than δT and the temperature in the slave
zone exceeds the setpoint by more than δT, the slave zone PPD is attributed to undersizing. It may also
be due to inadequate distribution, but further refinement of the algorithm is probably not warranted.
At this point, the PPD has been attributed to specific causes for each zone for each hour. In order to
reduce the results to just a few figures of merit, two additional steps are taken:
1. For each zone, the PPD attributed to each cause is summed for all 24 hours and the total is
divided by 24. This gives a mean contribution of each cause to the discomfort.
2. These values are area-weighted to find a mean value for the building.

Table 6. Algorithm for attributing the PPD Cause
Slave zone status
Master zone
Master zone temperature
status Tslave < (Tsetpoint – δT) ≤ Tslave ≤ Tslave >
(Tsetpoint – δT) (Tsetpoint + δT) (Tsetpoint + δT)
Night time
Night time ambient Inadequate
Tmaster < (Tsetpoint – δT) ambient Other
cooling distribution
cooling
(Tsetpoint – ∆T) ≤ Tmaster ≤ Other
Inadequate
Other
Inadequate
(Tsetpoint + δT) distribution distribution
Inadequate
Tmaster > (Tsetpoint + δT) Under-sizing Other Undersizing
distribution
Figure 18 through Figure 21 show the result of the system design evaluations for 576 cases where the
system capacities and airflow rates were designed with the RHB program, using zero temperature swing.
Figure 18 shows the processed PPD values for the category of nighttime ambient cooling. For
convenience in identifying specific cases, they are plotted against case number. For most cases, the
processed PPD values fall in the range of 0-10%. The group of high values in this figure is for the light-
weight construction cases from Alamosa, Colorado, which has the cooling design condition of 27.8ºC
(82.0ºF) maximum dry bulb temperature and a 17.3ºC (31.1ºF) daily dry bulb range. As the outdoor air
temperature drops to 10.5ºC (50.9ºF) at night, and the sky temperature drops to –11.7ºC (-11.0ºF) at night,
it is not surprising that a lightweight house falls well below the setpoint of 24ºC (75.2ºF) at night.
Figure 19 shows the processed PPD values for the categories of system under-sizing and inadequate
distribution. For most cases, the processed PPD values of the category of system under-sizing fall in the
range of 0-2%; the processed PPD values for the category of inadequate distribution fall in the range of 1-
6%.
To understand the meaning of these values, Figure 20 and Figure 21 plot out the hourly PPD values and
hourly room air temperatures for two particular cases (case 17, case 20). In Case 17, the PPD attributed
to undersizing is zero; the PPD attributed to inadequate distribution is 6.3%, the worst case of all 576.
Looking at Figure 20, we can see that the actual hourly distribution of temperatures and PPD shows
problems in the slave zone (the living room).
On the other hand, Case 20, which has PPD attributed to undersizing equal to zero; and PPD attributed to
inadequate distribution of 2.7%, a more typical value. In Figure 21, we see some overcooling of the
living room in the morning, and a slight drift above the setpoint in the early evening.
Case 17 and 20 are both wood frame, medium insulation cases located in San Diego. For both cases, the
great room (master zone) has a north and an east window; the living room (slave zone) has a south and a
west window. However, Case 17 has single pane glazing with twice the area of the low-e double pane
glazing used in Case 20. This results in a significantly higher required system capacity – the Case 17
system capacity is over three times as high as the Case 20 system capacity. The combination of higher
transient heat gains, with different profiles in the two rooms, and larger system capacity causes Case 17 to
have significantly higher thermal discomfort.

35
30
25
20
PPD (%)
15
10
0
0 100 200 300 400 500 600
Case number
Floor area weighted
Figure 18. Processed PPD showing night time ambient cooling, system designed with zero temperature
swing
case 17
6
case 449
4
PPD (%)
2
case 20
0
0 100 200 300 400 500 600
Case number
Floor area weighted Undersizing Floor area weighted Inadequate distribution
with zero temperature swing

50 30
40 24
Temperature (C)
30 18
PPD (%)
20 12
10 6
0 0
1 3 5 7 9 11 13 15 17 19 21 23
Hour
Great PPD Living PPD Great DB Living DB
Figure 20. Hourly PPD vs. temperature for case 17, system designed with zero temperature swing
50 30
40 24
Temperature (C)
30 18
PPD (%)
20 12
10 6
0 0
1 3 5 7 9 11 13 15 17 19 21 23
Hour
Figure 21. Hourly PPD vs. temperature for Case 20; system designed with zero temperature swing
Figure 22 to Figure 24 show the results of the test set with RHB sizing the system capacities and room
airflow rates allowing a 1.67ºC (3.0ºF) temperature swing.
Figure 22 shows the processed PPD values for the category of nighttime ambient cooling. The processed
PPD values are slightly lower than those with zero design temperature swing. This is due to the fact that
the smaller system capacities cause the zone temperatures to swing upward during the day. Presumably,
the stored energy slightly reduces the downward temperature swing at night, resulting in slightly lower
values of PPD due to nighttime ambient cooling.

Figure 23 shows the processed PPD values for the categories of system under-sizing and inadequate
distribution. Compared to the corresponding values in the zero design temperature swing cases, the
processed PPD values for the category of system under-sizing are significantly higher, because of the
decrease in system design capacities resulting from permitting temperature swing in the design
calculation. The processed PPD values for the category of inadequate distribution are similar or slightly
lower than the zero swing cases.
Figure 24 plots the hourly PPD values and hourly room air temperatures for Case 355, which has a
processed PPD due to undersizing of 7.2%, the worst case. The processed PPD due to inadequate
distribution is 0.4%. Looking at Figure 24, we can see that the undersizing, while evident for a number of
hours, never becomes severe. This is due to the fact that the location is Key West, FL, which has only a
4.5ºC (8.1ºF) daily dry bulb range. Furthermore, this case has low fenestration area and concrete
construction, both of which tend to limit the diurnal changes in cooling load.
35
30
25
20
PPD (%)
15
10
0
0 100 200 300 400 500 600
Case number
Floor area weighted
Figure 22. Processed PPD showing night time ambient cooling, system designed with medium
temperature swing

8
case 355
7
case 149
6
5
PPD (%)
0
0 100 200 300 400 500 600
Case number
Floor area weighted Undersizing Floor area weighted Inadequate distribution
with medium temperature swing
50 30
40 24
Temperature (C)
30 18
PPD (%)
20 12
10 6
0 0
1 3 5 7 9 11 13 15 17 19 21 23
Hour
Figure 24. Hourly PPD vs. temperature for case 355, system designed with medium temperature swing
Conclusions and Recommendations

Inter-model validation of the ASHRAE residential heat balance load calculation method (RHB) was
performed using ESP-r. The testing process was automated through parametric generation and simulation
of large sets of test cases for both RHB and ESP-r.

The load comparison mode of the testing process was found effective in confirming the RHB load
calculation and reconciling RHB and ESP-r input used for design evaluation. Excluding input
discrepancies, load comparison results are largely dependent on the sub-models used in both programs.
After review of the results, the load differences are considered reasonable, with the main source of
difference being the different external long wave radiation models used in each program. However, the
sensitivity of the peak load calculations to longwave radiation suggests that this is a likely topic for
further research and refinement of the methods.
The system design evaluation mode was useful for showing the quality of designs made with the RHB
procedure. Two processed PPD values are used to identify system under-sizing and inadequate
distribution/control limitation problems. The amount of inadequate distribution depends on the level of
the load profile difference between the master and slave zones. Increased difference in the load profiles
will result in more inadequate distribution. Also, increased system under-sizing is predicted if
temperature swing is permitted in the RHB system design.

4. The Residential Load Factor (RLF) Method
The research project Updating the ASHRAE/ACCA Residential Heating and Cooling Load Calculation
Procedures and Data (1199-RP) had two primary products. First, a new fundamental residential heating
and cooling load calculation method was developed and tested. This procedure, called the Residential
Heat Balance (RHB) method, is based on heat balance first principles as described by Pedersen et. al.
1997, Pedersen et. al. 1998, and ASHRAE 2001. RHB is documented in Section 2 of this report. It uses
a computationally intensive 24-hour design-day simulation that is practical only when implemented in
software. Because of its fundamental approach, RHB can be applied with few restrictions to arbitrarily
complex residential buildings, including those with large fenestration areas, novel construction features,
or having non-summer peaks.
The ResHB computer program, developed as part of 1199-RP, implements the RHB method. ResHB is a
batch-driven FORTRAN-90 application derived from the ASHRAE Loads Toolkit (Pedersen et. al. 2001)
that operates on Windows-based PCs. Several key ResHB features are noted here. First, ResHB models
room temperature swing: in addition to the standard fixed-set point capability, ResHB can find the
sensible cooling extraction rate that results in a specified temperature swing above the thermostat set
point. Second, ResHB incorporates the updated models identified in 1199-RP as appropriate for
residential loads calculation. Third, ResHB is multi-room and multi-zone, allowing application to real
buildings as well as simple test cases. Finally, ResHB can model typical residential master-slave control,
where a thermostat in one room controls the cooling delivery in another, with resulting imperfect
temperature control in the slave room.
The second product of 1199-RP is a simpler procedure, designated Residential Load Factor (RLF)
method. RLF is tractable by hand or can be straightforwardly implemented using spreadsheet software.
This simplification is achieved at the expense of generality – RLF is applicable only to conventionally
constructed residences with typical space-conditioning requirements. The procedures and data required to
use RLF are presented in the Residential Heating and Cooling Loads Calculation chapter of the 2005
Handbook of Fundamentals (ASHRAE 2005).
This paper discusses the design of RLF and documents the methodology used in its development. Some
testing results are also presented. While RLF includes both cooling and heating load procedures, the
heating calculations rely on the traditional UA∆T model that has proven satisfactory for decades.
Improvements have been introduced in relation infiltration leakage rate and to ground heat loss.
The RLF cooling procedure resembles and builds upon prior methods, but was developed using a linear
regression approach that avoids some semi-empirical derivations used in the past. Prior methods have
been published by the Air-Conditioning Contractors of America (ACCA), including the widely-used
Manual J Seventh Edition (ACCA 1986) and Manual J Eighth Edition (ACCA 2003). The 1989 – 2001
editions of the ASHRAE Handbook of Fundamentals included a method based on 342-RP (McQuiston
1984). Canadian Standard CAN/CSA-F280-M90 (HRAI 1996 and CAN/CSA-F280-M90 1990) specifies
a cooling method also based on 342-RP and a heating procedure that includes enhanced ground-loss
calculations.
RLF Cooling Load Calculation

The RLF cooling load calculation is based on the idea of independent load components, as are prior
simplified methods. The load contributions from various sources are separately evaluated and then

summed. The following sections summarize the method, showing both sensible and latent components as
applicable. Later sections document the derivation of the component models and coefficients.
In RLF, surfaces have associated load factors (LFs) or load contribution per unit area. These are
designated CFs for cooling and HFs for heating. For the most part, HF values are simply UA∆T. CF
values depend on surface construction, climate, and, in some cases, surface orientation, solar absorptance,
or other characteristics. Each unique LF needs to be evaluated once for a given set of site and
construction conditions, and then is applied repeatedly to building elements of the same type. This two-
step process is convenient for hand or spreadsheet application. Note that LFs are the functional
equivalent of Manual J’s Heat Transfer Multipliers (HTMs) but are derived differently and in general do
not have the same values.
Total cooling load

qs = ∑ Ai ⋅ CFi + qvi , s + qig , s (2)
ql = qvi ,l + qig ,l (3)
where
qs = sensible cooling load, W (Btu/h)
ql = latent cooling load, W (Btu/h)
Ai = area of ith surface, m2 (ft2)
CFi = cooling factor of ith surface, W/m2 (Btu/h-ft2)
Opaque surfaces
The cooling load per unit area of opaque walls, ceilings, and non-slab floors is calculated as follows:
CFopq = U ⋅ (OFt ⋅ ∆T + OFb + OFr ⋅ DR ) (4)
where
CFopq = opaque surface cooling factor, W/m2 (Btu/h-ft2-F)
U = construction U-factor, W/m2-K (Btu/h-ft2-F)
∆T = design dry-bulb temperature difference (outdoor – indoor), °K (°F)
DR = daily range of outdoor dry-bulb temperature, °K (°F)
OFt, OFb, OFr = coefficients from Table 7
Table 7. Opaque surface coefficients
Surface type Construction OFt OFb OFr
Ceiling or knee wall Wood frame .62 14.1 ⋅ ∝roof – -0.23
adjacent to vented attic 4.3
Ceiling/roof assembly Wood frame 1 39 ⋅ ∝roof – 6.8 -.42
Wall Wood frame 1 7.9 -.34
Floor over ambient Wood frame 1 0 -.10

Floor over crawlspace Wood frame .32 0 -.46
αroof = roof solar absorptance
Slab floors produce a slight reduction in cooling load, as follows:

CFslab = 1.6 − 1.4 ⋅ hsrf (5)
where
CFslab = slab cooling factor, W/m2 (Btu/h-ft2-F)

hsrf = effective surface conductance, including resistance of slab covering material such as
carpet, W/m2-K (Btu/h-ft2-F). 1/(Rcvr+.12) W/m2-K or 1/(Rcvr+.68) Btu/h-ft2-F
Fenestration
Fenestration cooling factors are calculated as follows:
CFfen = U ⋅ ( ∆T − .49 ⋅ DR ) + FFs ⋅ PXI ⋅ SHGC ⋅ IAC (6)
where
CFfen = fenestration cooling factor, W/m2 (Btu/h-ft2)
U = fenestration NFRC heating U-factor, W/m2-°K (Btu/h-ft2-F)
DR = daily range of outdoor dry bulb temperature, °K (°F)
∆T = cooling design temperature difference, °K (°F)
FFs = load factor (see Table 8)
PXI = peak exterior irradiance, including shading modifications (see below), W/m2 (Btu/h-ft2-F)
SHGC = fenestration rated or estimated NFRC Solar Heat Gain Coefficient
IAC = interior shading attenuation coefficient
Table 8. Fenestration coefficients
Exposure FFs
N .17
NE .09
E .17
SE .25
S .45
SW .54
W .48
NW .34
Horiz .66
Peak exterior irradiance (PXI) is the hourly maximum solar gain incident on the surface.
PXI = Tx Et (unshaded) (7)
PXI = Tx ( E d +(1 − Fshd ) ED ) (shaded) (8)
where
PXI = peak exterior irradiance for exposure, W/m2 (Btu/h-ft2-F)
Et, Ed, ED = peak total, diffuse, and direct irradiance for exposure, W/m2 (Btu/h-ft2-F)
Tx = Transmission of exterior attachment, see Table 10
Fshd = fraction of fenestration shaded by permanent overhangs, fins, or environmental obstacles
For horizontal or vertical surfaces, irradiance values can be obtained from Table 9 for primary exposures
or Algorithm 1 for any exposure. Skylights with slope less than 30° from horizontal should be treated as
horizontal. Steeper slopes, other than vertical, are not supported by the RLF method.

Table 9. Exterior irradiance (W/m2)
Latitude ( N or °S)
Exp 20 25 30 35 40 45 50 55 60
N ED 132 117 106 101 103 110 124 145 172
Ed 136 122 109 98 88 79 70 63 55
Et 269 238 215 199 190 189 194 207 227
NE/NW ED 541 532 522 511 501 490 480 470 461
Ed 163 154 147 140 135 130 126 123 120
Et 704 686 668 652 636 621 606 593 580
E/W ED 627 640 650 657 662 663 662 659 653
Ed 173 169 166 163 162 161 161 161 162
Et 800 809 816 821 824 825 823 820 815
SE/SW ED 334 380 422 460 494 525 553 577 598
Ed 174 173 174 175 177 180 183 187 191
Et 508 553 595 635 672 705 736 764 788
S ED 0 65 146 223 297 368 436 501 563
Ed 149 171 175 180 186 192 198 205 212
Et 149 236 321 403 482 559 634 705 774
Hor ED 906 901 888 867 838 801 756 703 642
Ed 124 124 124 124 124 124 124 124 124
Et 1030 1025 1012 991 962 925 880 827 766
Note: multiply value by .317 to convert to Btu/h-ft2
Algorithm 1. Exterior irradiance
Horizontal surfaces
Et = 970 + 6.2 L - 0.16 L Ed = MIN(Et ,124 )
2
E D = Et − E d
Vertical surfaces
ψ
Ψ = (normalized exposure)
180
Et = 462.2 + 1625Ψ + 6183Ψ + 3869Ψ + 32.38ΨL
3 4
2
1.040L
+.3237ΨL − 12.56L − .8959L +
2 2
Ψ +1
121 4 L
Ed = MIN( Et , 392.1 - 138.6Ψ + 2.107Ψ L − )
Ψ +1
E D = Et − E d
where
Et, Ed, ED = peak hourly total, diffuse, and direct irradiance, W/m2 (multiply by
.317 to convert to Btu/h-ft2)
L = site latitude, °N or °S
ψ = exposure (surface azimuth), ° from south (-180 - +180)

Table 10. Exterior attachment transmission
Attachment Tx
Exterior insect screen .6
Shade screen Manufacturer SC value,
typically .4 to .6
Note: see Brunger et. al. 1999 re insect screens
The shaded fraction, Fshd, can be taken as 0 for fenestration in full sun and 1 for any fenestration that is
shaded by adjacent structures or other obstacles during peak hours. Fshd for simple overhang
configurations can be calculated as follows (more complex configurations should be analyzed with the
RHB method):
⎛ SLF ⋅ Doh − X oh ⎞
Fshd = MIN ⎜1, MAX (0, )⎟ (9)
⎝ H ⎠
where
SLF = Shade line factor from Table 11
Doh = Depth of overhang (from plane of fenestration), m (ft)
Xoh = Vertical distance from top of fenestration to overhang, m (ft)
H = Height of fenestration, m (ft)
The shade line factor (SLF) is the ratio of the distance a shadow falls beneath the edge of an overhang to
the width of the overhang (Table 11). Therefore the shade line equals the SLF times the overhang depth.
The tabulated values are the average of the shade line values for 5 h of maximum solar intensity on
August 1 on each wall exposure shown. North-, northeast-, and northwest-facing windows are not
effectively protected by roof overhangs; in most cases, they should not be considered shaded.
Table 11. Shade line factors (SLF)
Window Latitude, degrees N
exposure 24 32 36 40 44 48 52
East 0.8 0.8 0.8 0.8 0.8 0.8 0.8
SE 1.8 1.6 1.4 1.3 1.1 1.0 0.9
South 9.2 5.0 3.4 2.6 2.1 1.8 1.5
SW 1.8 1.6 1.4 1.3 1.1 1.0 0.9
West 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Note: Shadow length below the overhang equals the shade line factor times the overhang depth.
Ventilation and infiltration

Infiltration air flow is calculated as follows --
AL ⎡ AL , flue ⎤
Qinf = ⋅ ⎢ I 0 + H ⋅ ∆T ⋅ ( I1 + I 2 ⋅ )⎥ (10)
1000 ⎣ AL ⎦
Qinf = infiltration air flow rate, L/s (cfm)
AL = building effective leakage area (including flue) at 4 Pa assuming CD = 1, cm2 (in2)
I0 – I2 = coefficients, as follows:
Cooling Heating
windspeed = 3.4 windspeed = 6.7
m/s (7.5 mph) m/s (15 mph)
I0 25 (343) 51 (698)
I1 .38 (.88) .35 (.81)
I2 .12 (.28) .23 (.53)
H = building average stack height, m (ft) (approximately 2.5 m (8 ft) per story)
∆T = indoor – outdoor temperature difference, °K (°F)
AL,flue = flue effective leakage area at 4 Pa assuming CD = 1, cm2 (in2)

The ventilation air flow rate is determined according to the installed or planned ventilation equipment that
is expected to be operating at design conditions. Generally intermittently operated exhaust fans are not
included. Overall supply and exhaust flow rates are determined and divided into “balanced” and
“unbalanced” components and combined with infiltration.
Qbal = MIN (Qsup , Qexh ) (11)
Qunbal = MAX (Qsup , Qexh ) − Qbal (12)
Qvi = MIN ( Qunbal , Qinf + 0.5 ⋅ Qunbal ) (13)
where
Qbal = balanced ventilation air flow rate, L/s (cfm)
Qsup = total ventilation supply air flow rate, L/s (cfm)
Qexh = total ventilation exhaust air flow rate (including any combustion air requirements), L/s
(cfm)
Qunbal = unbalanced air flow rate, L/s (cfm)
Qvi = combined infiltration/ventilation flow rate (not including balanced component), L/s (cfm)
Note that unbalanced duct leakage can produce additional pressurization or depressurization. This effect
is included in distribution losses, discussed below.
The cooling (or heating) load due to ventilation and infiltration is calculated as follows, taking into
account the effects of heat / energy recovery ventilation (HRV / ERV) equipment:
qvi ,s = Cs ⋅ (Qvi + (1 − ε s ) ⋅ Qbal ,hr + Qbal ,oth ) ⋅ ∆T (14)
qvi ,l = Cl ⋅ ( Qvi + Qbal ,oth ) ⋅ ∆W (no HRV/ERV) (15)
qvi ,t = Ct ⋅ ( Qvi + (1 − ε t ) ⋅ Qbal ,hr + Qbal ,oth ) ⋅ ∆h (16)

qvi ,l = qvi ,t − qvi ,s (17)
where
qvi,s = sensible ventilation/infiltration load, W (Btu/h)
Cs = air sensible heat factor, 1.23 W/(L/s)-K (1.1 Btu/h-cfm-F) at sea level
εs = HRV/ERV sensible effectiveness
Qbal,hr = balanced ventilation flow rate supplied via HRV/ERV equipment, L/s (cfm)
Qbal,oth = other balanced ventilation supply air flow rate, L/s (cfm)
∆T = indoor-outdoor temperature difference, °K (°F)
qvi,l = latent ventilation/infiltration load, W (Btu/h)
Cl = air latent heat factor, 3010 W/(L/s) (4840 Btu/h-cfm) at sea level
∆W = indoor-outdoor humidity ratio difference
qvi,t = total ventilation/infiltration load, W (Btu/h)
Ct = air total heat factor, 1.2 W/(L/s)-(kJ/kg) (4.5 Btu/h-cfm-(Btu/lb)) at sea level
εt = HRV/ERV total effectiveness
∆h = indoor-outdoor enthalpy difference, kJ/kg (Btu/lb)
Internal gain
The contributions of internal gains to peak sensible and latent loads are:
qig , s = G0, s + Gcf , s ⋅ Acf + Goc , s ⋅ N oc (18)
qig ,l = G0,l + Gcf ,l ⋅ Acf + Goc ,l ⋅ N oc (19)
where

qig,s = sensible cooling load due to internal gains, W (Btu/h)
qig,l = latent cooling load due to internal gains, W (Btu/h)
Gx = coefficients, as follows:
Sensible Latent
G0 136 (464) 20 (68)

Gcf 2.2 (.7) .22 (.07)
Goc 22 (75) 12 (41)
Acf = conditioned floor area of building, m2 (ft2)
Noc = number of occupants, if not known, estimate as (number of bedrooms + 1)
Distribution losses
The allowance for distribution losses is calculated as follows:
qdl = Fdl ⋅ qs (20)
where
qdl = distribution loss, W (Btu/h)
Fdl = distribution loss factor, from Table 12
qs = building sensible load, W (Btu/h)
Table 12. Duct loss/gain factors

Heating
Cooling
Furnace Heat pump
Duct tightness Unsealed Sealed Unsealed Sealed Unsealed Sealed
Duct insulation R 0 .7 (4) 1.4 (8) 0 .7 (4) 1.4 (8) 0 .7 (4) 1.4 (8) 0 .7 (4) 1.4 (8) 0 .7 (4) 1.4 (8) 0 .7 (4) 1.4 (8)
(m2-K/W (h-ft2-
F/Btu))
Duct location
Conditioned
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
space
Attic 2.01 1.35 1.23 0.83 0.42 0.35 0.68 0.46 0.41 0.41 0.20 0.16 0.78 0.57 0.53 0.40 0.22 0.18
Basement 0.20 0.17 0.16 0.09 0.06 0.05 0.43 0.31 0.28 0.25 0.13 0.11 0.35 0.27 0.26 0.18 0.11 0.09
Enclosed
0.25 0.20 0.20 0.12 0.08 0.07 0.68 0.46 0.41 0.41 0.20 0.16 0.78 0.57 0.53 0.40 0.22 0.18
crawlspace
Development Overview
The RLF formulation is conceptually transparent and hand-tractable: the “loads” from each wall, window,
and other gain sources are calculated and summed to get the total load. Unfortunately, however, it is not
possible to find invariant models for each load component because of interactions among them. For
example, a major interaction occurs between opaque surfaces and fenestration – the load resulting from
solar gain is lagged and moderated by differing amounts depending on surface construction. Even simple
convective gains, such as infiltration and ventilation, present difficulty because they should be evaluated
at the building-dependent peak hour. Note that RHB completely avoids these difficulties: 24 hour
calculations allow gains to combine according to their case-specific profiles and the heat balance
procedure accurately represents component interactions.
Development of a load-component method such as RLF requires that the significant interactions be
identified and addressed (via configuration-specific load-component calculations), eliminated (by
restricting the configurations to which the method is applicable), or neglected if the effects are deemed

small. Addressing interactive effects introduces more complexity in the method, which defeats its
purpose. Given the availability of RHB to handle essentially any configuration, RLF applicability is
restricted to typical residential construction.
Prior methods assumed the independence of load components and developed models for each. The
component models were in many cases the obvious choice (e.g. infiltration load derived directly from an
air leakage rate). However, excessive loads are predicted by simply using maximum fenestration and
opaque surface heat gain rates. For these components, semi-empirical factors or adjustments were
invoked to make the results consistent with experience. In particular, a common strategy was to use
factors equal to multi-hour averages of calculated instantaneous gains. While the averaging approach has
some intuitive appeal, it has no rigorous basis, as is acknowledged in older editions of the Handbook of
Fundamentals (ASHRAE 1972) –
Averaging in this manner was found to give results compatible with measured residential loads. Hence, these are
averages only in the sense that combining numbers in this manner results in accurate factors for calculating window
loads of residential structures.
Room temperature swing is one reason adjustment is required. Assuming a fixed indoor temperature, as
is typically done in non-residential procedures, results in excessive loads for the residential case. Better
overall system performance and cost effectiveness results when equipment is sized to allow some
temperature variation at design conditions. Averaging of gains derived assuming fixed room temperature
mitigates their excessive peak.
The RLF development procedure avoided adjustments by relying on RHB cooling loads calculated with
temperature swing and deriving required factors using linear regression. Equation (2) was treated as a
model for which sub-models and coefficients were needed. Later sections of the paper present the
approaches used for each load component. The regression approach has two advantages. First,
significant independent variables and efficient model forms are naturally identified by the regression
process. If a model does not accurately predict load, that is revealed by poor statistical figures of merit.
Second, no averaging or other semi-empirical adjustments are required.
From a processing point of view, RLF was developed using three PC applications: ResHB as described
above (loads calculations), RHBGen (parametric case generator), and the R statistics package.
RHBGen was developed to support ResHB testing and RLF development (see Appendix D and the
accompanying 1199-RP Source Code and Data CD). RHBGen generates and runs ResHB input files
under the control of multi-character parametric codes. Fields within the code control various aspects of
the case to be run, such as base prototype (fundamental building geometry), location, orientation,
constructions, fenestration type and area, and so forth. RHBGen combinatorically varies code fields,
allowing sets of hundreds or thousands of ResHB runs to be constructed and executed. On typical
Pentium-based PCs, the RHBGen / ResHB system can complete several hundred cases per minute.
ResHB writes suitable results files for regression analysis and other post-processing.
The R package (R 2004) is an open-source system with extensive statistical and data visualization
capabilities. For RLF development, the linear regression and data plotting procedures were used. R is
particularly suitable for RLF development because it includes a script language for automation of
complex analysis sequences. The R scripts used in this work can be found on the 1199-RP Source Code
and Data CD that accompanies this report.
Modeling Assumptions
As discussed above, RLF cooling load procedures were developed by using linear regression analysis of
sets of ResHB results. To generate the input data for regression, specific combinations of inputs were
varied while others were maintained at typical values. For most variables, three levels were identified: L
= minimum, M = typical, and H = maximum. Even with only three levels and relatively few variables,

the total number of possible combinations is impractically large. To limit the number of cases, the range
of RLF applicability was restricted to conventionally constructed and occupied single family detached
wood-frame buildings at latitude 20° - 60° and at modest elevation. In addition, load components were
analyzed separately, as is discussed below under Regression Strategy.
Prototype building
Cooling loads for many variations of a single prototype building were the basis for the regression
analysis; Table 13 summarizes the prototype characteristics and Table 14 shows construction details. The
floor plan of the prototype building was square with 4 rooms, one in each corner. Note that ResHB
surfaces need not be geometrically consistent, allowing exterior wall area to be based on an assumed
typical-width rectangular plan and that area to be distributed equally on all facades. The 4 room plan was
chosen so there was a reasonable ratio of interior partition to floor area and to limit radiant transfer among
exterior walls. The sensible cooling load used as the regression independent variable was the maximum
value of the combined 24 hour profile derived by summing the room loads for each hour.
Table 13. Prototype building characteristics
Item Value Notes
Conditioned floor area 168 m2 (1808 ft2) Typical size
Height 2.5 m (8.2 ft) Single story
Exterior wall area 142.4 m2 (1533 ft2) Average width assumed to be 8.5 m (28 ft),
yielding perimeter = 57 m (187 ft)
Interior partition area 140 m2 (459 ft2) 83% of conditioned floor area
Nominal fenestration 27.2 m2 (89.2 ft2) windows window area = 16% of floor area
2 2
1.68 m (18 ft ) skylight skylight area = 1% of floor area
clear double glazed (U = 2.73
2 2
W/m K (.48 Btu/h-ft -F), SHGC
= 0.76)
Fenestration variation All cases run with 200%
nominal area. IAC values
varied L = 0, M = 0.5, H = 1.
Internal mass 168 m2 (1808 ft2) of 12 mm 1199-RP default
(0.5 in) wood
Indoor design 24 °C (75.2 °F)
temperature
Indoor temperature 1.67 °K (3 °F)
swing
Infiltration Leakage class E (normalized Reasonably tight contemporary construction
leakage = .34) (ASHRAE Standard 119)
Internal gain Default Based on Building America 2003, see below.
Surface exterior solar walls: 0.6
absorptance roof: 0 – 1 (varied)
Surface interior Beam solar gain: floor: 0.6,
absortance internal mass: 0.3, other: 0
Diffuse solar gain: all surfaces:
0.6
Orientation 0 ° and 45 ° All 8 primary orientations considered.

Table 14. Prototype surface constructions
Surface Construction Framing Insulation U-factor Insulation description
2
fraction level W/m -K
(Btu/h-ft2-
F)
Ceiling Roof / ceiling, 10% L 1.48005 none
asphalt shingles, (0.26066)
plywood deck, 2 M 0.21835 190 mm (7.5 in) fiberglass between
x 8 framing, (0.03846) framing
gypsum board H 0.11879 190 mm (7.5 in) fiberglass between
(0.02092) framing plus 140 mm (5.5 in) overlay
Attic / ceiling, 2 10% L 2.24413 none
x 8 framing, (0.39523)
gypsum board M 0.23148 190 mm (7.5 in) fiberglass between
(0.04077) framing
H 0.12204 190 mm (7.5 in) fiberglass between
(0.02149) framing plus 140 mm (5.5 in) overlay
Wall Wood frame, 25% L 1.57621 none
plywood, 2 x 4 (0.27760)
framing, gypsum M 0.50209 90 mm (3.5 in) fiberglass between
board (0.08843) framing
(0.04570) framing plus 25 mm (1 in) foam at
outside of framing
Floor Wood frame, 10% L 0.80772 none
oak floor, (0.31837)
plywood deck, 2 M 0.22512 190 mm (7.5 in) fiberglass between
x 8 framing (0.03965) framing
(0.02772) framing plus 50 mm (2 in) foam at
outside surface
Slab: 100 mm (4 n/a L n/a Bare slab
in) concrete, 300 M n/a Additional surface resistance = .185
2 2
mm (12 in) soil, (m -K)/W (1.05 (ft -F-h)/Btu), light carpet
adiabatic H n/a Additional surface resistance = .370
2 2
exterior (m -K)/W (2.10 (ft -F-h)/Btu), heavy
boundary carpet
conditions
Outdoor design conditions

Eight combinations of outdoor design dry-bulb temperature and daily range were selected to span a broad
range of design conditions. Coincident wet-bulb temperatures were chosen by inspection of actual sites
having design conditions similar to those of each combination. Table 15 summarizes the temperature
assumptions. Other site-related assumptions are shown in Table 16 most of which were held constant.
Table 15. Design conditions
Case Design dry-bulb Daily range of dry- Design wet-bulb
temperature °C (°F) bulb temperature, temperature, °C (°F)
°C (°F)
LL 24 (75.2) 4 (7.2) 19 (66.2)
LM 24 (75.2) 11 (19.8) 16 (60.8)
ML 33 (91.4) 4 (7.2) 29 (84.2)
MM 33 (91.4) 11 (19.8) 23 (73.4)
MH 33 (91.4) 21 (37.8) 15 (59)
HL 43 (109.4) 4 (7.2) 24 (75.2)
HM 43 (109.4) 11 (19.8) 22 (71.6)
HH 43 (109.4) 21 (37.8) 21 (69.8)

Table 16. Site assumptions
Item Value
Latitude 20°N, 40°N, 60°N
Longitude 75 °W
Time zone -5 hr
Elevation 50 m (164 ft)
Date July 21
Time Daylight savings
Clearness 1
The ResHB application uses these inputs to generate 24-hour design sequences that drive the heat balance
simulation. Hourly incident solar radiation was calculated using the ASHRAE clear sky model
(ASHRAE 2001) with updated coefficients (Machler and Iqbal 1985).
The combination of 8 design conditions and 3 latitudes resulted in 24 runs for each prototype variant.
Regression Strategy
It was not practical to perform one regression analysis to identify all RLF coefficients because the
overwhelming number of case combinations that would have been required. Instead, an iterative series of
linked regressions was performed. Equation (2) was applied to ResHB loads results and rearranged to
isolate the envelope load component:
qenv = qs ,rhb − qig ,rhb − qvi ,rhb (21)
where
qenv = Envelope cooling load component = ∑ Ai ⋅ CFi in Equation (2), W (Btu/h)
qs,rhb = ResHB sensible cooling load, W (Btu/h)
qig,rhb = ResHB sensible internal gain at peak hour (simultaneous with qs,rhb), W (Btu/h)
qvi,rhb = ResHB sensible ventilation/infiltration at peak hour (simultaneous with qs,rhb), W (Btu/h)
The envelope cooling load is the sum of the load components from the various envelope elements:
qenv = q fen + qceil + qwall + q floor (22)
Each component term of Equation (22) was estimated using a separate data set described in Table 17.
Each data set contains ResHB loads based on varying inputs relating to the term under consideration
while fixing other inputs at M or nominal values. The component regressions were performed in the
sequence shown in Equations (23) - (26) and the results of each were applied to the next step. (The
details of each component model are discussed below.) Initial (iteration 0) estimates were set by hand
using suitable prior results. It was determined by trial and error that 5 iterations achieved essentially
complete convergence.
+1
qˆ ifen = qenv − qˆceil
i
− qˆwall
i
− qˆ ifloor (23)
i+1
qˆ
ceil = qenv − qˆ i+1
fen − qˆ i
wall − qˆ i
floor (24)
i +1 i +1 i +1
qˆ
wall = qenv − qˆ fen − qˆ ceil − qˆ i
floor (25)
i +1 i +1 i +1 i +1
qˆ floor = qenv − qˆ fen − qˆ ceil − qˆ wall (26)
where
qˆ xi = ith iteration estimated load component for fenestration, ceiling, wall, or floor, W (Btu/h)

Table 17. Regression data sets
Component Fenestration Ceiling Wall Floor Total cases
(24 design
conditions, 2
orientations)
qfen 41 combinations of LMH Attic, M Wood frame, Crawlspace, M 5904
on 4 facades plus skylight ∝roof = .85 M Exposed, M
(Box-Behnkin 5 factor Slab, M
design)
qceil Nominal Roof/ceiling Wood frame, Crawlspace, M 2592
LMH, Attic M Exposed, M
LMH, each Slab, M
with ∝roof = 0,
.6, 1
qwall Nominal Attic, M Wood frame, Crawlspace, M 432
∝roof = .85 LMH Exposed, M
Slab, M
qfloor Nominal Attic, M Wood frame, Crawlspace, 432
∝roof = .85 M LMH
Exposed, LMH
Slab, LMH
Component Models
Ventilation and infiltration
As discussed above, typical infiltration was included in the ResHB runs used to generate regression data,
but the cooling load induced by this air leakage was subtracted from the load used in the envelope
regressions. Thus the loads predicted by the regression models implicitly assume 0 air leakage.
Equation (10) was developed to provide a simple method for estimating infiltration leakage for RLF.
ResHB calculates infiltration using the AIM-2 model (Walker and Wilson 1990, Walker and Wilson
1998), which is too complex for practical hand application. The AIM-2 model was exercised over a range
of temperature differences and building heights. Other assumptions included shelter class 4, flue shelter
class 2, and wind speed multiplier values from Table 10, Chapter 26, ASHRAE 2001. Leakage
distribution was assumed to be walls=.5, ceiling=.25, floor=.25 (R=1, X=0), all proportionately reduced if
flue is present. The maximum flue leakage fraction considered was 0.5. Regression was used to find the
form of Equation (10) and the Ix coefficients. The underlying functional form of the AIM-2 model is not
linear but the simple form of Equation (10) was maintained for ease of application. The regression model
yielded an adjusted R2 of .94. Figure 25 compares results from the regression to those from AIM-2.
Because of minimal air density dependence, Equation (10) is valid at any elevation.
The procedure for combining mechanical ventilation with infiltration air flows, shown in Equations (11) -
(13), follows Palmiter and Bond 1991 and Sherman 1992.

200
150
RLF infiltration (L/s)
100
50
0
0 50 100 150 200
AIM-2 infiltration (L/s)
Figure 25. Predicted infiltration leakage rates, AL = 1000 cm2 (155 in2) and representative range of
stack height, temperature difference, and flue fraction. RLF values from Equation (10), see text. (1 L/s =
2.12 cfm)
Internal gains
RHB internal gains are based on Building America 2003, which specifies gain intensities and schedules
for residential appliances, lighting and occupants. Experiments with these gains and schedules in ResHB
revealed that the sensible cooling load attributable to internal gains is generally approximated by the total
sensible internal gain during the peak cooling hour. This is not necessarily expected, since a significant
fraction of the gain is radiant and has a delayed load impact. The removal of load due to internal gain in
Equation (21) is based on this approximation.
For RLF, Equations (18) and (19) are the aggregated Building America gains using 4 PM schedule
values, that time being a common peak cooling hour for typical residential construction. Consideration
was given to developing a model that predicts the peak cooling hour so a more accurate internal gains
formulation could be included. However, such an addition to RLF was deemed excessively complex.
Opaque surfaces
The model forms for opaque surface CFs were found by experimentation. Prior methods, both residential
and non-residential, have used an Equivalent Temperature Difference (ETD) or Cooling Load
Temperature Difference (CLTD) form, where ETD or CLTD = A + ∆T - DR/2 (where A is a constant, ∆T
is the outdoor-indoor temperature difference, and DR is the daily range). This was taken as a starting
point for RLF. A coefficient was added for ∆T and multipliers other than 0.5 were allowed for DR. In
some cases, the ∆T coefficient was found to be a value very close to 1, in which case it was dropped from
the regression and forced to be 1. In other cases, coefficients were found to be not significant and

dropped. The DR coefficient takes many values, indicating that the traditional 0.5 is perhaps not ideal.
The final coefficient values are shown in Table 7. Adjusted R2 values for all regressions were above 0.96.
A major design consideration was how many surface types to include. It was decided to limit RLF to
conventional wood frame construction. That led to inclusion of one type of wall (wood frame), two types
of ceilings (ceiling/roof and ceiling/attic combinations), and three types of floors (exposed, crawlspace,
and slab). Surface orientation was not a variable (all wall orientations are combined) and solar
absorptance was treated as a variable only for roofs. It is believed that additional surface types could be
added via straightforward extension of the current procedures.
Fenestration
A goal for the fenestration model was to separation of latitude-dependent exterior effects from building-
dependent effects. It was determined by experimentation that this is achieved by factoring out peak hour
irradiance incident on the fenestration exterior, leading the PXI formulation shown in Equation (6). As
with the opaque surface models, this form is similar to prior residential and non-residential methods.
Many combinations of effective window aperture were included in the regression data set used to find the
FFs coefficients. The final adjusted R2 value was over .995.
An attempt was made to eliminate exposure-specific FFs values, leaving PXI as the only exposure-
dependent input. This produced significantly worse regression results. The FFs coefficients (Table 8)
show a physically reasonable relationship with exposure. East surfaces produce less cooling load per unit
irradiance than do west surfaces, as is expected. Prior methods that relied on averaging show a less
plausible E/W and SE/SW symmetry.
Distribution losses
Duct losses can be calculated using models specified in ANSI/ASHRAE 152-2004, Method of Test for
Determining the Design and Seasonal Efficiencies of Residential Thermal Distribution Systems and
Palmiter and Francisco 1997. These models are fully implemented in the ResHB. Using typical input
values, ResHB was exercised to produce Table 12 suitable for use with RLF hand calculations
Heating Loads
The RLF procedure for heating loads calculation is identical, in most respects, to previously published
ASHRAE 2001 residential heating load calculation procedures. The heating load calculation is based on
a steady-state UA∆T calculation, with no solar radiation and no internal heat gains. Infiltration leakage
rate is based on Equation (10). The calculation procedure for heat losses from surfaces in contact with the
ground has been revised as described in the following sections.
Basement wall heat losses

For basement wall and floor heat transfer, the ASHRAE Handbook of Fundamentals has incorporated a
procedure described by Latta and Boileau 1969 for a number of years. The Latta and Boileau method has
the advantage of simplicity. One check of its accuracy was described by Sobotka, et al. 1994, who
showed that the Latta and Boileau method underpredicted the peak heating load of one basement by 16%.
Correlation-based methods (Krarti and Choi 1996, Beausoleil-Morrison and Mitalas 1997) have been
developed that offer significantly improved accuracy. However these later methods have a large number
of coefficients; this complicates presentation in a handbook. Therefore, the RLF procedure has
incorporated a revised version of the Latta and Boileau procedure with the suggestion that for buildings
where the heating loads are significantly impacted by ground heat transfer should be analyzed with one of
the more accurate methods.

The Latta and Boileau method is based on the assumptions that the surface temperature of the ground is at
a calculable winter design temperature and that the heat flow paths may be approximated as
circumferential with radial isotherms. See Figure 26. It also assumes that the thermal resistance of the
ground may be estimated based on the path length of the heat transfer. In the original Latta and Boileau
1969 paper, tabulated U-factors included inside thermal resistance, thermal resistance of a concrete wall,
thermal resistance of insulation (if any), and thermal resistance of the soil. The tabulated values were
based on a coarse numerical integration and were specific to single combinations of soil thermal
conductivity and insulation thermal resistance. The approach taken in the tables also depends on the
interval value for the numeric integration being one foot. This presentation has been, at times, somewhat
confusing. In fact, the values in the 2001 SI version of the Handbook of Fundamentals are wrong,
apparently having been misconverted due to the dependence on the interval. In addition, at some point,
the original Latta and Boileau recommendation to use a ground temperature calculated as the mean
ground temperature minus the annual amplitude, A, was re-expressed to use a ground temperature
calculated as the average winter air temperature minus the annual amplitude. This results in significant
overprediction of the ground heat loss.
Heat flow
lines
Radial isotherms
(from intersection
of grade and Radial isotherms
basement wall) (from intersection
of basement wall
and floor)
Figure 26. Heat flow from basements (2001 ASHRAE Handbook of Fundamentals)
In the RLF procedure, the original Latta and Boileau work was revisited and reformulated in a more
flexible manner. The revised procedure allows for variation of the soil thermal conductivity and, if
desired, partial wall insulation with any thermal resistance. Furthermore, an analytical expression for the
average U-factor has been developed, along with new tables. This may be summarized as follows.
In cases where the basement wall is partially insulated, it will be desirable to calculate the heat loss
separately for portions of the wall with differing amounts of insulation. Consider the region between
depth z1 and z2 in Figure 27. (Here z1 and z2 can be any region of the wall, including the entire wall.

However, if partial insulation is used, it might be desirable to treat the wall as two or more separate
regions.)
z1
zf z2
Figure 27. Definition of basement wall and floor dimensions
For the region of interest, in steady-state heat transfer, there are several thermal resistances of interest –
the soil, the concrete wall, the insulation (if any) and the inside surface resistance. If all thermal
resistances besides the soil are lumped into a single value, Rother, the average U-factor between the
basement air and the ground temperature is:
2k soil ⎡ ⎛ 2k soil Rother ⎞ ⎛ 2ksoil Rother ⎞ ⎤
U avg ,bw = ⎢ ln ⎜ z2 + ⎟ − ln ⎜ z1 + ⎟⎥ (27)
π ( z2 − z1 ) ⎣ ⎝ π ⎠ ⎝ π ⎠⎦
where
Uavg,bw = average U-factor between basement air and ground temperature over region of interest
shown in Figure 27, W/m2-K (Btu/h-ft2-F)
ksoil = soil thermal conductivity, W/m-K (Btu/h-ft-F)
z1 = depth of upper bound of region of interest (see Figure 27), m (ft)
z2 = depth of lower bound of region of interest (see Figure 27), m (ft)
Rother = combined resistance of wall, insulation and surface conductance, m2-K/W (ft2-h-F/Btu)
While values of soil conductivity vary widely with soil type and moisture content, a typical value of 1.4
W/m-K (0.8 Btu/h-ft-F) was used in past editions of the Handbook of Fundamentals to tabulate U-factors.
Rother is the sum of the resistance of the concrete wall, insulation (if any) and the inside surface resistance.
In past editions of the Handbook of Fundamentals, Rother was approximately 0.25 m2-K/W (1.47 ft2-h-
F/Btu) for uninsulated concrete walls. For these parameters, Uavg,bw is tabulated for a range of depths and
insulation levels in Table 18.
Table 18a. Average U-factor for basement walls with uniform insulation (SI units)

Uavg,bw from grade to depth, W/m2-Ka
Depth (m) Uninsulated R-0.88 R-1.76 R-2.64
0.3 2.468 0.769 0.458 0.326
0.6 1.898 0.689 0.427 0.310
0.9 1.571 0.628 0.401 0.296
1.2 1.353 0.579 0.379 0.283
1.5 1.195 0.539 0.360 0.272
1.8 1.075 0.505 0.343 0.262
2.1 0.980 0.476 0.328 0.252
2.4 0.902 0.450 0.315 0.244
a
Soil conductivity is 1.4 W/m-K; insulation is over entire depth.
For other soil conductivities and partial insulation, use equations.
Table 18b. Average U-factor for basement walls with uniform insulation (I-P units)
Uavg,bw from grade to depth, Btu/h-ft2-Fa
Depth (ft) Uninsulated R-5 R-10 R-15
1 0.432 0.135 0.080 0.057
2 0.331 0.121 0.075 0.054
3 0.273 0.110 0.070 0.052
4 0.235 0.101 0.066 0.050
5 0.208 0.094 0.063 0.048
6 0.187 0.088 0.060 0.046
7 0.170 0.083 0.057 0.044
8 0.157 0.078 0.055 0.043
a
Soil conductivity is 0.8 Btu/h-ft-F; insulation is over entire depth.
For other soil conductivities and partial insulation, use
equations.
Basement floor heat losses

The RLF procedure uses an analogously updated version of the Latta and Boileau procedure for basement
floors. For cases where the entire basement floor is uninsulated, or has uniform insulation, the average
U-factor is:
2k soil ⎡ ⎛ Wb z f k soil Rother ⎞ ⎛ z f k soil Rother ⎞ ⎤
U avg ,bf = ⎢ ln ⎜ + + ⎟ − ln ⎜ + ⎟⎥ (28)
π Wb ⎣ ⎝ 2 2 π ⎠ ⎝ 2 π ⎠⎦
where
Uavg,bf = average U-factor between basement air and ground temperature for entire basement floor,
W/m2-K (Btu/h-ft2-F)
ksoil = soil thermal conductivity, W/m-K (Btu/h-ft-F)
Wb = basement width, which should be taken to be the shortest dimension, m (ft)
zf = depth of slab bottom (see Figure 27), m (ft)
Rother = combined resistance of floor, insulation and surface conductance, m2-K/W (ft2-h-F/Btu)
For a soil conductivity of 1.4 W/m-K (0.8 Btu/h-ft-F), Uavg,bf for uninsulated basement floors are tabulated
in Table 19.
Table 19a. Average U-factor for basement floors (SI units)

Uavg,bf, W/m2Ka
Wb (shortest width of basement), m
zf (depth of foundation wall 6 7 8 9
below grade), m
0.3 0.370 0.335 0.307 0.283
0.6 0.310 0.283 0.261 0.242
0.9 0.271 0.249 0.230 0.215
1.2 0.242 0.224 0.208 0.195
1.5 0.220 0.204 0.190 0.179
1.8 0.202 0.188 0.176 0.166
2.1 0.187 0.175 0.164 0.155
a
Soil conductivity is 1.4 W/m-K; floor is uninsulated.
Table 19b. Average U-factor for basement floors (I-P units)

Uavg,bf, Btu/h-ft2-Fa
Wb (shortest width of basement), ft
zf (depth of foundation wall 20 24 28 32
below grade), ft
1 0.064 0.057 0.052 0.047
2 0.054 0.048 0.044 0.040
3 0.047 0.042 0.039 0.036
4 0.042 0.038 0.035 0.033
5 0.038 0.035 0.032 0.030
6 0.035 0.032 0.030 0.028
7 0.032 0.030 0.028 0.026
a
Soil conductivity is 0.8 Btu/h-ft-F; floor is uninsulated.
Slab-on-grade floor heat losses

Concrete slab floors have been previously approximated as having heat losses solely proportional to the
perimeter length by Wang 1979 and Bligh et al. 1978. More recent work (Bahnfleth and Pedersen 1990)
has shown a significant effect of the area-to-perimeter ratio. The correlation-based methods for basement
wall and floor heat transfer described above (Krarti and Choi 1996; Beausoleil-Morrison and Mitalas
1997) also have procedures for dealing with a wide range of slab-on-grade configurations. Again, if slab
heat loss is a significant factor in the building heating load, one of these procedures should be used.
However, for handbook presentation, the previous approach was retained, with the exception that the
table which showed some dependence of the perimeter heat loss factor on the number of degree-days was
simplified by eliminating the degree day dependence.
The simplified approach gives heat loss for both unheated and heated slab floors with the following
equation:
q = F2 ⋅ P ⋅ ∆T (29)

where
q = heat loss through perimeter, W (Btu/h)
F2 = heat loss coefficient per unit length of perimeter, W/m-K (Btu/h-ft-F) (see Table 20)
P = perimeter or exposed edge of floor, m (ft)
∆T = heating design temperature difference, °K (°F)
Noting that the degree-day dependence previously given in the Handbook of Fundamentals is relatively
small, the table of F2 factors was simplified by only giving the value for 2970 Kelvin degree-day (5350
Fahrenheit degree-day) climates.
Table 20a. Heat loss coefficient F2 of slab floor construction (SI units)
Construction Insulation F2 (W/K-m)
200 mm. block wall, Uninsulated 1.17
2
brick facing R-0.95 K-m /W from
0.86
edge to footer
200 mm. block wall, Uninsulated 1.45
brick facing R-0.95 K-m2/W from
0.85
edge to footer
Metal stud wall, Uninsulated 2.07
stucco R-0.95 K-m2/W from
0.92
edge to footer
Poured concrete wall Uninsulated 3.67
with duct near R-0.95 K-m2/W from
a 1.24
perimeter edge to footer
a
Weighted average temperature of the heating duct was assumed at 43ºC during the
heating season (outdoor air temperature less than 18ºC).
Table 20b. Heat loss coefficient F2 of slab floor construction (I-P units)
Construction Insulation F2 (Btu/h-ft-
F)
8 in. block wall, Uninsulated 0.68
brick facing R-5.4 from
0.50
edge to footer
4 in. block wall, Uninsulated 0.84
brick facing R-5.4 from
0.49
edge to footer
Metal stud wall, Uninsulated 1.20
stucco R-5.4 from
0.53
edge to footer
Poured concrete wall Uninsulated 2.12
with duct near R-5.4 from
a 0.72
perimeter edge to footer
a
Weighted average temperature of the heating duct was assumed at 110ºF during the
heating season (outdoor air temperature less than 65ºF).
Verification of Results
The RLF method was added to the ResHB application, allowing RLF vs RHB cooling load calculations to
be performed on test cases. Figure 28 shows typical results of such a comparison for a building not
involved in the regression process and using design weather data for 20 diverse U. S. locations. As can be
seen, there is generally good agreement but a trend remains in that RLF predicts too high for low loads
and too low for high loads. This is being investigated and may lead to model refinement.

RHB (kBtu/h)
0 10 20 30 40 50 60 70
70
20
60
15
50
RLF (kBtu/h)
40
RLF (kW)
10
30
20
5
10
MAE: 3.3% = 548 W (1870 Btu/h)
RMSE: 4.1% = 679 W (2318 Btu/h)
MBE: 0.51% = 84.8 W (289 Btu/h)
Cor: 0.966
0
0
0 5 10 15 20
RHB (kW)
Figure 28. RLF vs. RHB sensible cooling load comparison. Test building calculated for representative
range of climate and construction conditions (1280 cases)

5. Conclusions
The development of RHB shows the ASHRAE heat balance method to be the procedure of choice for
residential load calculations. In particular, the ability of RHB to directly handle floating temperature
conditions allows straightforward and rigorous treatment of temperature swing, master / slave control, and
buffer spaces. The first-principles basis of RHB means that the method can be applied with few
limitations to virtually any residential problem. Extreme configurations, such as large or concentrated
fenestration areas, have been problematic for prior methods but are readily modeled with RHB.
The heat balance approach has two primary drawbacks. First, it is computationally intensive. Current
generation interactive load calculation applications require “instant” calculations (a small fraction of a
second). Based on experience with ResHB, it appears that such speed is achievable, but not without
clever implementation. Second, the occasional HB convergence failure remains troublesome; additional
algorithm development is needed before the method can robustly handle all cases encountered in practice.
In addition to providing accurate peak loads, the hourly, multi-room, varying temperature capabilities of
RHB offer many opportunities for improved residential system design procedures. An immediate
possibility is automated identification of poor zoning configurations, replacing the user judgment required
in prior procedures. More ambitious, but certainly possible, are applications such as distribution system
optimization and automated determination of zoning.
The largest advantage of heat balance is how it transforms the continuing improvement of loads
calculation methods. Direct empirical comparisons can be made between experiments and heat balance
models, leading to model validation and/or refinement. Further, sensitivity studies can guide research
efforts to areas where refinement is particularly important. As is discussed in Section 6 (Future
Research), experience during RHB development indicates that high priority should be placed on
improving models of convective heat transfer in air-cooled residential spaces, ground heat transfer during
cooling season, clear-sky solar radiation, and interior shading. Additional research on residential
occupancy patterns, appliance use, and interior shade operation is also needed. Some of these topics are
addressed by current or proposed ASHRAE research projects. The significant point, however, is that as
results become available, they can be immediately integrated into RHB and other HB methods.
The ASHRAE Load Toolkit proved to be an invaluable starting point for this work. Ultimately much of
the code was reworked. However, we had a running heat balance application “right out of the box”. This
accelerated the work and provided reference cases for testing purposes. Also, certain complex pieces of
code, such as CTF generation and psychrometric calculations, were used with little modification.
Intermodel comparison proved to be very useful for evaluating models. The flexibility of ESP-r allowed
testing, comparison, and error detection. The frequent comparison to ESP-r during development gives
confidence in the accuracy of ResHB and was instrumental in identifying key areas of sensitivity, such as
exterior surface long wave radiant transfer.
Linear regression was shown to be an effective tool for RLF development and could be usefully applied
to other situations. The regression approach necessarily means that a broad array of cases are involved in
simplified method formulation, giving confidence that the intended range of applicability is covered. In
addition, having RHB available as a reference method compatible with RLF allows uncertainties to be
checked at any time.
Finally, it should be noted that all simplified models having the RLF form (including RLF) have the
distinct disadvantage that they do not give any indication of when they become inapplicable. For

example, the peak cooling hour cannot be identified from the RLF procedure; if the peak is shifted, the
internal gain load component could be significantly in error.

6. Future Research
Needed research for load calculations falls into at least two categories, which, for lack of better
terminology might be called:
• Fundamental research – new data, models or theory are needed that are currently not available.
• Applied research – data or models are available that cover the topic. However, the research has
not been “brought to practice” -- it is not in a form that is suitable for design engineers to use.
The following research topics are divided into the two categories, though there may necessarily be some
overlap in any one project.
Fundamental Research
• Interior convection correlations. 1199-RP has shown significant sensitivity to interior convection
coefficients for some cases. Yet, the ASHRAE Handbook still uses single values determined for
natural convection conditions over 60 years ago. More recent research has been useful to some
degree, but is not well aligned to common building practice. The suggestion is to develop
convection correlations that are harmonized to standard diffuser design practices. There are five
outlet categories in Chapter 32 of the 2001 HOF. Convection correlations ought to be developed
that match these categories for realistic ranges of ADPI.
• Exterior convection correlations. There has been some recent work applicable to both residential
scale buildings (Yazdanian and Klems 1994) and high rise buildings (Loveday and Taki 1998).
However, in both cases, there are still gaps in applicability – neither covers roofs and the latter
does not give any guidance for still air conditions.
• Longwave radiation – building to sky. The Berdahl and Martin model seems to be well-accepted
for predicting clear sky temperature seen by horizontal surfaces. However, sky temperature seen
by vertical surfaces, which “see” the sky through longer path lengths through the atmosphere, has
only been handled heuristically (Walton 1983, McQuiston et. al. 2000).
• Longwave radiation – building to surroundings. As shown by 1199-RP, peak cooling loads for
some buildings are surprisingly sensitive to longwave radiation between the building and
surroundings. Two sets of seemingly reasonable assumptions give differences on the order of
10% of the peak cooling load.
Applied Research
• Sky models – solar radiation. The current ASHRAE ABC model has some questionable features,
e.g. how, precisely, to treat the clearness number and it is only applicable for the continental U.S.
More detailed models are available but usually require parameters that are unknown to design
engineers.
• Ground heat transfer – A whole range of slab-on-grade and basement heat transfer models have
been developed. However, relatively few are suitable for design load calculation procedures.
The best prospects are two correlation-based methods (Beausoleil-Morrison and Mitalas 1997,
Krarti and Choi 1996). However, both methods have large numbers of coefficients and users of
either method would face some difficulties.

• Multi-dimensional heat conduction. What is the best (quickest with reasonable accuracy) way for
design engineers to represent walls with wood studs, steel studs, etc.? The Handbook treats
steady-state heat transfer in some detail, but gives no guidance for design load calculations, where
walls are entered in a layer-by-layer fashion.
• Interior shading (shades, blinds, etc). This topic is being addressed by 1311-RP.

7. References
ACCA. 1986. Load Calculation for Residential Winter and Summer Air Conditioning – Manual J 7th
Edition, Air Conditioning Contractors of America, Arlington, VA.
ACCA. 2003. Manual J Residential Load Calculations, 8th Edition. Fourth Printing, February, 2003.
Air Conditioning Contractors of America, Arlington, VA.
Alamdari, F. and G.P. Hammond. 1983. Improved data correlation for buoyancy-driven convection in
rooms. Rep. SME/J/83/01 (Cranfield: Applied Energy Group: Cranfield Institute of Technology)
Alexander D.K. 1997. HTB2 (A Model for the Thermal Environment of Buildings in Operation) User
Manual 2.0b. UWIST, Cardiff.
Andersen, B. R., L. Kosmina, E. Panzhauser, J. Lechleitner, J. Achtziger, P. I. Sandberg, B. Johnsson, C.
Pompeo, T. Frank, H. Muhlebach, and B. Torroja. 1999. Analysis, selection, and statistical treatment
of thermal properties of building materials for the preparation of harmonised design values. Glasgow:
BRE Scottish Laboratory.
ASHRAE. 1972. Handbook of Fundamentals.
ASHRAE. 1994. Air Leakage Performance of Detached Single-Family Residential Buildings.
ANSI/ASHRAE Standard 119-1988 (RA 94).
ASHRAE. 2004. ANSI/ASHRAE 152-2004, Method of Test for Determining the Design and Seasonal
Efficiencies of Residential Thermal Distribution Systems.
Bahnfleth, W. P. and C. O. Pedersen. 1990. A Three-Dimensional Numerical Study of Slab-on-Grade
Heat Transfer. ASHRAE Transactions 96(2): 61-72.
Barnaby, C. S. and J. D. Spitler. 2005. Development of the Residential Load Factor Method for Heating
and Cooling Load Calculations (1199-RP). ASHRAE Transactions (in preparation).
Barnaby, C. S., J. D. Spitler, and D. Xiao. 2005. The Residential Heat Balance Method for Heating and
Cooling Load Calculations (1199-RP). ASHRAE Transactions (in preparation).
Beausoleil-Morrison, I. and G. Mitalas. 1997. BASESIMP: A Residential-Foundation Heat-Loss
Algorithm for Incorporating Into Whole-Building Energy-Analysis Programs. Building Simulation
'97, Prague.
Beausoleil-Morrison, I. 2000. The adaptive coupling of heat and air flow modeling within Dynamic
whole-building simulation. Ph.D. Thesis, University of Strathclyde.
Berkeley Solar Group. 1986. Energy Calculations and Data. Concrete Masonry Association of California
and Nevada.
Berdahl, P. and M. Martin. 1984. Emissivities of Clear Skies. Solar Energy, 23, 5, 663-664.
BLAST Support Office 1995. BLAST Users' Manual. United States Army Construction Engineering
Research Laboratory, Champaign, IL.

Bloomfield, D., R. Brewer, B. Stringer, I. Cooper, and S. Lera. 1988. IEA Task 8, the UK Experience:
Improving the design of low energy passive solar houses. Building Research & Practice, the Journal
of CIB, no. 6.
BRE. 1988. Final Report of BRE/SERC Collaboration, An investigation into analytical and empirical
validation techniques for dynamic thermal models of buildings. Building Research Establishment,
U.K.
Brunger, A. and F. Dubrous, S. Harrison. 1999. Measurement of the Solar Heat Gain Coefficient and U
Value of Windows with Insect Screens. ASHRAE Transactions 105(2).
Building America. 2003. Building America Research Benchmark Definition, Version 3.1. Available at
http://www.eere.energy.gov/buildings/building_america/pdfs/benchmark_def_ver3.pdf
CAN/CSA-F280-M90. 1990. Determining the Required Capacity of Residential Space Heating and
Cooling Applicances. Canadian Standards Association, Rexdale (Toronto), ON, Canada.
Carpenter, S.C., J. Kosny, E. Kossecka. 2003. Modeling Transient Performance of Two-Dimensional
and Three-Dimensional Building Assemblies. ASHRAE Transactions 109(1).
CEN. 1999. Building materials and products – Hygrothermal properties – Tabulated design values. prEN
12524. Brussels: European Committee for Standardization.
Clarke, J.A. 1977. Environmental Systems Performance. Ph. D. Thesis, University of Strathclyde,
Glasgow UK.
Clear, R.D., L. Gartland, and F.C. Winkelmann. 2001. An Empirical Correlation for the Outside
Convective Air Film Coefficient for Horizontal Roofs. LBNL-47275, Lawrence Berkeley National
Laboratory, Berkeley, CA.
EnergyPlus. 2004. http://www.eere.energy.gov/buildings/energyplus/
Enermodal Engineering Limited. Modeling Two- and Three-Dimensional Heat Transfer Through
Composite Wall and Roof Assemblies in Transient Energy Simulation Programs (1145-RP). Final
Report. Available from ASHRAE.
ESRU. 2000. The ESP-r System for Building Energy Simulation: User Guide Version 9 Series. ESRU
Manual U00/1, University of Strathclyde, Glasgow UK.
Forest Products Laboratory. 1999. Wood handbook--Wood as an engineering material. Gen. Tech. Rep.
FPL-GTR-113. Madison, WI: U.S. Department of Agriculture, Forest Service, Forest Products
Laboratory. (Available at http://www.fpl.fs.fed.us/documnts/FPLGTR/fplgtr113/fplgtr113.htm)
HRAI. 1996. Residential Heat Loss and Gain Calculations, Student Reference Guide. Heating,
Refrigerating and Air Conditioning Institute of Canada. Mississauga, ON, Canada.
IDA ICE. 2004. http://www.equa.se/eng.ice.html
Judkoff, R., J. Neymark. 1995. Building Energy Simulation Test (BESTEST) and Diagnostic Method.
NREL/TP-472-6231. Golden, CO: National Renewable Energy Laboratory.
Judkoff, R., D. Wortman, B. O’Doherty, J. Burch. 1983. A Methodology for Validating Building Energy
Analysis Simulations. SERI/TR-254-1508. Golden, CO: Solar Energy Research Institute (now
NREL).
Krarti, M. and S. Choi. 1996. Simplified Method for Foundation Heat Loss Calculation. ASHRAE
Transactions 102(1): 140-152.
Kusuda, T., and P.R. Achenbach. 1965. Earth temperature and thermal diffusivities at selected stations in
the United States. ASHRAE Transactions 71(1): 61-75.

Latta, J.K. and G.G. Boileau. 1969. Heat losses from house basements. Canadian Building 19(10):39.
LBL and Ayres Sowell Associates. 1997. SPARK User’s Manual. Lawrence Berkeley National
Laboratory and Ayres Sowell Associates, Inc.
Liesen, R.J. and C.O.Pedersen. 1997. An Evaluation of Inside Surface Heat Balance Models for Cooling
Load Calculations, ASHRAE Transactions 103(2): 485-502.
Loveday, D. and A. Taki. 1998. Outside surface resistance: Proposed new value for building design.
Building Services Engineering Research and Technology, Vol. 19, No. 1, pp. 23-29.
Machler, M. A. and Iqbal, M. 1985. A Modification of the ASHRAE Clear Sky Irradiation Model.
ASHRAE Transactions 91(1A): 106-115.
Martin, M. and P. Berdahl. 1984. Characteristics of infrared sky radiation in the United States. Solar
Energy 33(3/4): 321-336.
McQuiston, F. C. 1984. A study and review of existing data to develop a standard methodology for
residential heating and cooling load calculations. RP-342. ASHRAE Transactions 90(2A):102-36.
McQuiston, F.C., J.D. Parker, J.D. Spitler. 2000. Heating, Ventilating, and Air Conditioning Analysis
and Design, Fifth Edition. John Wiley and Sons, New York.
McQuiston, F. C. and J. D. Spitler. 1992. Cooling and Heating Load Calculation Manual, Second
Edition. ASHRAE.
NIST. 2000. NIST Standard Reference Database 81: NIST Heat Transmission Properties of Insulating
and Building Materials. Accessible via http://srdata.nist.gov/insulation/.
Palmiter, L. and T. Bond. 1991. Interaction of Mechanical Systems and Natural Infiltration. Proceedings
of the 12th AIVC Conference on Air Movement and Ventilation Control Within Buildings, Air
Infiltration and Ventilation Centre, Coventry, Great Britain.
Palmiter, L. and P. Francisco. 1997. Development of a Practical Method of Estimating the Thermal
Efficiency of Residential Forced-Air Distribution Systems. Electric Power Research Institute report
TR-107744. Palo Alto, CA.
Park, C., D. Clark, G. Kelly 1986. HVACSIM+ building systems and equipment simulation program:
building loads calculation. National Bureau of Standards, NBSIR 86-3331.
Pedersen, C. O., D. E. Fisher, and R. J. Liesen. 1997. A heat balance based cooling load calculation
procedure. ASHRAE Transactions 103(2): 459-468.
Pedersen, C. O., D. E. Fisher, J. D. Spitler, and R. J. Liesen.. 1998. Cooling and heating load calculation
principles. ASHRAE, Atlanta, GA.
Pedersen, C. O., R. J. Liesen, R. K. Strand, D. E. Fisher, L. Dong, and P. G. Ellis. 2001. A Toolkit for
Building Load Calculations. ASHRAE, Atlanta, GA.
Perez, R., P. Ineichen, R. Seals, J. Michalsky and R. Stewart. 1990. Modeling daylight availability and
irradiance components from direct and global irradiance. Solar Energy 44(5): 271-289.
Perez, R., R. Seals, P. Ineichen, R. Stewart and D. Menicucci. 1987. A new simplified version of the
Perez diffuse irradiance model for tilted surfaces. Solar Energy 39: 221-231.
R. 2004. Open source software available at http://www.r-project.org/
Rees, S.J., J.D. Spitler, and P. Haves. 1998. Quantitative comparison of North America and U.K. cooling
load calculation procedures – Results. ASHRAE Transactions 104(2): 47-57.

Rees, S.J. and J.D. Spitler. 1999. Proposals for a building loads diagnostic test procedure. ASHRAE
Transactions 105(2): 514-526.
Sherman, M.H. 1992. Superposition in infiltration modeling. Indoor Air 2:101-14.
Sobotka, P., H. Yoshino, et al. 1994. Thermal performance of three deep basements: a comparison of
measurements with ASHRAE Fundamentals and the Mitalas method, the European Standard and the
two-dimensional FEM program. Energy and Buildings 21: 23-34.
Spitler, J.D., D.E. Fisher, and C.O. Pedersen. 1997. The Radiant Time Series Cooling Load Calculation
Procedure. ASHRAE Transactions. 103(2):503-515.
Spitler, J.D. and S.J. Rees. 1998. Quantitative comparison of North America and U.K. cooling load
calculation procedures – Methodology. ASHRAE Transactions 104(2): 36-46.
Spitler, J.D., S.J. Rees, and D. Xiao. 2001. Development of an analytical verification test suite for whole
building energy simulation programs – building fabric (1052-RP): Final Report, ASHRAE, Atlanta
GA.
Walker, I. S. and D. J. Wilson. 1998. Field validation of equations for stand and wind driven air
infiltration calculations. International Journal of HVAC&R Research 4(2).
Walker, I. S. and D. J. Wilson. 1990. The Alberta Air Infiltration Model. The University of Alberta,
Department of Mechanical Engineering, Technical Report 71.
Walton, G. N. 1980. A New Algorithm for Radiant Interchange in Room Loads Calculations, ASHRAE
Transactions. 86(2): 190-208.
Walton, G. N. 1983. Thermal Analysis Research Program Reference Manual. NBSIR 83-2655.
National Bureau of Standards (now National Institute of Standards and Technology), Washington,
DC.
Walton, G.N. 1993. Computer programs for simulation of lighting/HVAC interactions. NISTIR 5322.
National Institute of Standards and Technology, Gaithersburg, MD.
Xiao D. 2001. Developing Updated Method of Residential Heating and Cooling Load Calculations:
Research Proposal for Ph.D Preliminary Examination, Oklahoma State University. School of
Mechanical and Aerospace Engineering.
Xiao, D., J.D. Spitler, and S.J. Rees. 2002. An Analytical Verification Test Suite for Multizone Building
Fabric and Control Models in Whole Building Energy Simulation Programs. Proceedings of the
eSim 2002 Conference, Montreal, Canada. September 11-13, 2002. pp. 260-267.
Yazdanian, M. and J. Klems. 1994. Measurement of the Exterior Convective Film Coefficient for
Windows in Low-Rise Buildings. ASHRAE Transactions, Vol. 100, Part 1, pp. 1087–1096.

Appendix A. The 1199-RP Source Code and Data CD
Development files, source code, and other project materials have been archived on the 1199-RP Source
Code and Data CD and delivered to ASHRAE with this report. Copies of the CD are available from the
ASHRAE Manager of Research.
The materials on the CD are freely usable for personal, research, and teaching purposes. However,
commercial use requires an additional license from ASHRAE. Please review the file
“1199RP_TermsOfUse.txt” found in the root directory of the CD.
The 1199-RP Source Code and Data CD is organized into several folders, each of which contains a file
“ReadMe.txt” or “ReadMe.doc” that provides pertinent procedures and details.
• InterModelValidation. Materials related to the comparison of ResHB and ESP-r results.
Includes ESP-r source code, PERL scripts, and various supporting files. Note that this folder is
large.
• Regression. Regression source code (for R) and data sets used for RLF development and other
data fitting.
• Report. Contains this report in PDF format.
• ResHB. Source and executables for ResHB and RHBGen, plus associated test input files.
With the exception of inter-model comparisons, all 1199-RP work was conducted on Windows-based
PCs. Batch files and build procedures in ResHB assume a root folder “\1199RP”. Thus the following
installation procedure is recommended:
1. Create a folder “X:\1199RP”, where X is the target hard drive.
2. Copy selected folders from the CD into that folder.
Unless the reader is interested in developing new correlations or running inter-model comparisons, there
is no reason to copy the InterModelValidation or Regression directories to a hard drive. See
“InterModelValidation\ReadMe.doc” for additional information on the inter-model validation files. See
"Regression\ReadMe.doc" for additional information on the RLF regressions.

Appendix B. Algorithms and Assumptions
Fenestration
The Loads Toolkit fenestration model requires input of angular SHGC and absorption values. This is not
practical for a routine use.
ResHB implements an idea designated “fenestration class.” A fenestration class can be thought of as a
family of fenestration types that exhibit common behavior. The fenestration class embodies the ratio of
transmission to absorption and the angular characteristics of the fenestration system. An actual
fenestration is specified by its NFRC-rated U-factor and SHGC plus its fenestration class. The required
angular characteristics are taken from the fenestration class and are scaled by the ratio of rated SHGC to
nominal (fenestration class) SHGC.
As shown in Figure 29, 6 classes are sufficient represent the overwhelming majority of residential
glazings (although there is some error for reflective glazing) --
• Clear single
• Heat absorbing single (for heat absorbing and reflective single glazings)
• Clear double (for clear double plus low-E double)
• Heat absorbing double (for heat absorbing and reflective double)
• Clear triple (for clear and low-E triple)
• Heat absorbing triple (for heat absorbing and reflective triple)
Fen Classes, Sacramento South Glass, Stud Walls

4 Bars from top: Exact, 6 Classes, 3 classes, 1 class
Peak Hourly Sensible Load, W

0 500 1000 1500 2000 2500 3000 3500
"1C "
1H
1R
2C
Glass Type
2EH
2EL
2H
2R
3C
3H
Figure 29. Load grouping as function of fenestration classification

Internal mass
The presence of internal mass (such as furniture) in a room has two effects on cooling load.
• The load is increased due to enhanced surface area available for convective exchange.
• The load is decreased by the heat storage in the mass (which in turn depends on its construction).
The ultimate impact of internal mass depends on which of these effects dominates in a particular case.
For the 1199-RP procedure, the intent is to provide fixed assumptions for internal mass. Preliminary
parametric studies have been performed to investigate what assumptions might be appropriate.
One important note is that residential buildings typically have relatively small rooms and thus have a
significant amount of partition area per unit floor area. For example, a 15 ft square room (225 ft2) with an
8 ft ceiling and one exterior wall has 360 ft2 of exposed interior partition area. If the room is not on the
top floor, an additional 225 ft2 of ceiling area is exposed. Thus the ratio of partition area to floor area can
easily exceed 1 and often exceed 2. All this surface area functions as internal mass via the mechanisms
described above.
Another important interaction is that between temperature swing and internal mass. The larger the
temperature swing, the more the heat storage effects of the mass tend to reduce cooling load.
Finally, there is a strong interaction between internal mass and other mass elements in the space,
especially if solar gain is targeted to those surfaces.
To investigate internal mass impact on typical residential cooling loads, several ResHB cooling load
calculations have been performed for a Washington, DC design day. Some of the assumptions:
• Prototype 1: large (162.6 m2) house without partitions
• Prototype 2: single 25 m2 room with one exposed wall
• Floor construction: wood frame or slab on grade
• Temperature swing: varies 0 – 2.5 °C
• Internal mass construction: 12 mm wood exposed 1 side, area varies 0 – 2 x floor area.
Figure 30 - Figure 33 show results for these runs. The conclusion from this study is to assume 12 mm
wood as internal mass and fix its area at 1 x floor area.

Effect of Internal Mass
(162.6 m 2 house, no partitions, w ood floor)
1.300
1.250
1.200
Swing = 0
1.150
Swing = .83 C
1.100
Swing = 1.67 C
1.050
Swing = 2.5 C
1.000
0.950
0.900
0 0.5 1 1.5 2 2.5
IM area / floor area
Figure 30. Effect of internal mass, wood floor house

(162.6 m 2 house, no partitions, slab floor)
1.300
1.250
1.200
Swing = 0
1.150
Swing = .83 C
1.100
Swing = 1.67 C
1.050
1.000 Swing = 2.5 C
0.950
0.900
0 0.5 1 1.5 2 2.5
Figure 31. Effect of internal mass, slab floor house

(25 m 2 room , 1 exposed w all, w ood floor)
1.300
1.250
1.200
Swing = 0
1.150
Swing = .83 C
1.100
Swing = 1.67 C
1.050
1.000 Swing = 2.5 C
0.950
0.900
0 0.5 1 1.5 2 2.5
Figure 32. Effect of internal mass, single-exposure room, wood floor

(25 m 2 room , 1 exposed w all, slab floor)
1.300
1.250
1.200
Swing = 0
1.150
Swing = .83 C
1.100
Swing = 1.67 C
1.050
1.000 Swing = 2.5 C
0.950
0.900
0 0.5 1 1.5 2 2.5
Figure 33. Effect of internal mass, single exposure room, slab floor

Framed constructions
A pervasive feature of building walls and roofs are multidimensional heat transfer elements – studs,
concrete blocks, and the like that create parallel heat transfer paths. Although such features might be
modeled with numerical methods such as finite difference and finite volume, this is infeasible for load
calculation purposes. Instead, it is highly desirable to have a method that works within the existing load
calculation methodology framework, which expects a layer-by-layer description of homogenous layers.
One possible approach is simply to have users create separate layer-by-layer wall descriptions for the
different parallel paths. That is, replace a single stud wall with two walls – one that is siding, sheathing,
insulation, drywall, and another that is siding, sheathing, stud, drywall. This method should work very
well for cases where the parallel path method is acceptable for calculating U-factors, e.g. wood stud walls.
However, it is unlikely to work well for cases where the parallel path method is unacceptable, e.g. steel
stud walls.
ASHRAE attempted to address this problem with 1145-RP, “Modeling Two- and Three-Dimensional
Heat Transfer Through Composite Wall and Roof Assemblies in Transient Energy Simulation Programs”
(Carpenter et al. 2003). The intent of this project was to develop a procedure that would allow a
composite wall or roof assembly to be represented with a layer-by-layer description of homogenous
layers. The resulting procedure does allow this. Unfortunately, the first step is to model the wall in a
three-dimensional numerical model! Users of the procedure are expected to obtain their own numerical
model and use it to develop response factors. We do not believe this is feasible for residential load
calculations, or commercial load calculations, for that matter, unless the entire procedure is packaged into
a “black box” with an easy-to-use user interface.
Therefore, a simpler method is sought. The proposed method involves substitution of the parallel layer
with a homogenous layer that preserves the thermal capacity of the parallel layer and the overall
resistance of the wall, as determined by the appropriate method recommended by the ASHRAE HOF.
Note that the equivalent layer does not have the resistance that would be calculated by just combining the
two parallel elements – this would be equivalent to the isothermal planes method.
The proposed method, which might be called the “equivalent homogenous layer method” may be
summarized as follows:
1. Determine the surface-to-surface or air-to-air resistance with the method recommended by the
ASHRAE HOF. The 2001 HOF suggests: for wood frame walls -either isothermal planes or
parallel path method (p. 25.2-25.3), masonry walls -isothermal planes (p. 25.3); metal stud walls -
modified zone method (pp. 25.11-25.12).
2. Determine the resistance for all layers except the parallel layer.
3. Subtract the resistance determined in (2) from that determined in (1). This gives the required
equivalent layer resistance.
4. The conductivity for the equivalent homogenous layer is the parallel layer thickness divided by
the equivalent layer resistance determined in (3).
5. The density for the equivalent homogenous layer is the sum of the individual material densities
multiplied by their volume fraction.
6. The specific heat for the equivalent homogenous layer is the sum of the individual ρcp multiplied
by their volume fractions, divided by the density determined in (5).
This method has been tested against results published in the 1145-RP final report (Enermodal 2001). A
total of 20 building assemblies were analyzed as part of 1145-RP. Of these, eight represent features such
as window headers and corners. The remaining twelve are composite walls of various forms. For each
wall, the R-value, response factors and conduction transfer functions computed with the detailed three-

dimensional numerical model are reported. The twelve walls are summarized in Table 21. Unfortunately,
some problems were detected in some of the CTF coefficients and a consistency check was run for each
of the twelve walls. For each wall, the stated U-factor (U1145) was compared to U-factors computed
from the CTF coefficients. For a number of the walls, the errors are higher than expected.
Table 21. Summary of 1145-RP walls, stated U-factors, and U-factors calculated from CTF coefficients
1145 RP U factor comparison W/m2K
W all No Wall Type U1145 UY-CTF UZ-CTF Error
1 2x4 wood stud with R-11 insulation 0.4988 0.4988 0.4988 0.00%
2 2x6 wood stud with R-19 insulation 0.3249 0.3090 0.3249 4.91%
7 Insulated Concrete Form (ICF) Wall: Polysteel ties with finish 0.5056 0.5054 0.5056 0.04%
8 Concrete/foam/concrete sandwich with metal connectors 0.7418 0.7403 0.7418 0.21%
9 Concrete/foam/concrete sandwich with plastic connectors 0.5369 0.5369 0.5369 0.00%
10 2x4 steel stud 0.6460 0.6460 0.6460 0.00%
15 2x4 steel stud wall + 1”EPS + Brick 0.4442 0.4396 0.4442 1.03%
16 2x6 steel stud wall + R-19 insulation 0.5022 0.5022 0.5022 0.00%
17 2x6 s teel stud wall + R-19 insulation + 1-in. EPS + stucco 0.3749 0.3707 0.3749 1.10%
18 2x6 steel stud wall + R-19 insulation + 1-in. EPS + brick 0.3674 0.3665 0.3674 0.25%
19 Two-core concrete blocks + empty cores + stucco 4.1917 3.9909 4.1142 4.79%
20 Two-core concrete blocks + insulated cores + stucco 2.4797 2.4796 2.4797 0.00%
For several of the walls, a comparison has been made for a design day where the exterior surface
temperature varied cyclically between 30 C and 15 C and the interior surface temperature was held at 20
C. Interior heat fluxes calculated with the 1145-RP CTFs are compared to interior heat fluxes calculated
with the equivalent homogeneous layer method. (After using the method to determine the properties of
the equivalent homogeneous layer, CTFs are calculated using procedures from the Loads Toolkit. These
CTFs are used to calculate the interior heat fluxes.) Sample results for walls 1 and 18 are shown in Figure
34 and Figure 35. As can be seen, the approximate method gives approximately correct results. It is not
perfect, but may be the best that can be done at present. (Also note that wall 18 has some error in the
CTF coefficients – it may be that some of the discrepancy is due to this problem.)
Wall Assembly -- 1
5
4.5
4
3.5
3
Flux (W/m2)
2.5
2
1.5 Flux – 1145RP CTF
1 Flux-Homogenous
0.5 Approx.
0
-0.5
-1
-1.5
-2
-2.5
0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25
Time
Figure 34. Wall 1 of interior heat flux calculated with 1145-RP CTF and CTF generated with equivalent
homogeneous layer method

Wall Assembly -- 18
3.5
3
2.5
2
Flux (W/m2)
1.5
1 Flux – 1145RP CTF
0.5 Flux-Homogenous
Approx.
0
-0.5
-1
-1.5
-2
0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25
Time
Figure 35. Wall 18 interior heat flux calculated with 1145-RP CTF and CTF generated with equivalent
homogeneous layer method
Material properties
In spite of the fact that the building simulation field is approaching 40 years old, it remains difficult to
find a coherent and well-documented list of the properties of common building materials suitable for use
in detailed computer models.
In order to assemble material property assumptions for this project, a number of sources have been
reviewed and best-estimate values selected and built into ResHB. Sources included Andersen et. al. 1999,
CEN 1999, Forest Products Laboratory 1999, ASHRAE 2001 (Table 4, Chapter 25 and Table 22, Chapter
29), Berkeley Solar Group 1986, NIST 2000, Pedersen et. al. 1998, and ESRU 2000.

Appendix C. ResHB Documentation
Command Line and Files

ResHB is a 32 bit Windows console application. It invoked from the command line as follows --
ResHB runID [options]
runID is the base name for input and output files for the run (see below)
Options can be any combination of the following --
Option Effect
-C<filepath> Case file. If present, requests that a 1-line, comma-separated result summary be appended to the
specified file. This file is useful for import into Excel.
-G<filepath> Regression export file. If present, requests that a 1-line comma-separated result summary be appended to
the specified file. This results summary is similar to the case file summary but includes additional building
characteristics data.
-I<dirList> Include directories. Specifies a ;-delimited list of directories that ResHB searches for include files (the
current directory is searched first). Example: -Iinc1;c:\inc2
-M<xxx> Flags selecting alternative models, consisting of one or more upper case plus lower case pairs. The upper
case letter identifies the model, the lower case selects the specific alternative, as described below. For
example –MSvIt selects the visibility solar model and the Threlkeld incident solar model. The choice with *
is the default.
Model Descripton
C Inside surface convective coefficient
x* Runtime weighted. System on: 5 W/m2-K on all surfaces. System off: ASHRAE
still air values with flow-direction enhancement; linear transition of value from ∆T =
-1 °K - +1 °K.
s Same as ‘x’ except no linear transition.
y Same as ‘x’ except system on = 4.5 W/m2-K
a Fixed coefficients: 1.25 ceiling, 4.67945 wall, 4.37 floor (all W/m2-K)
d TARP detailed model (∆T^.333 and tilt dependent) (from Toolkit)
v Fisher ACH model (from Toolkit)
E Elevation adjustment of inside and outside convective coefficients
1* Enabled
0 Disabled
I Sloped surface incident diffuse solar (i.e. short wave sky model)
t* ASHRAE Threlkeld-Jordan “Y” model
i Isotropic sky model: view factor = 0.5 · (1 + cos( tilt))
p Perez
K Sky temperature
b* Berdahl/Martin
6 tAir – 6 °K
S Clear sky solar
a ASHRAE HOF
r* ASHRAE with Machler/Iqbal revised coeffiients
v Machler/Iqbal visibility
0 No incident solar (for testing)
X Outside surface convective coefficient
c* Constant 12.5 W/m2-K
m MoWiTT (ignoring wind direction)

Command line options (continued)
-R<xxx> Report options. Flag characters that specify what is included in reports. By default, ResHB writes
runID.rep that includes only a loads summary. –R causes additional information to be added to this file
(except –D, which generates an additional file).
flag effect
+ Append report to pre-existing runID.rep (if it exists). Default action is to overwrite prior file.
A building geometry (surface areas)
B room energy balance
C construction CTFs (for all constructions found in input file)
c construction CTFs (for constructions used in current run)
D generate additional report runID.dbg containing detailed debugging information; the
contents of this file changes as needed
E input echo with all built-in and include file text
e input echo, primary input file only
G Room internal gains with GAINITEM breakdown
g Room internal gains, totals only
S hourly surface info
s surface input summary
U building UA
W weather data summary
-X<xxx> Results file options. Flag characters which specify what is include in the results file. The results file,
runID.res, is a comma-separated file intended for import into Excel and other applications. By default, the
results file includes only a loads summary.
flag effect
B room energy balance
D room details (numerous 24 hour values for plotting)
S Solar model(s) output (re inter-model comparison)
W weather data in format suitable for plotting
Files
ResHB reads or writes the following files.
File name Use Description
<runID>.rhb Input, must exist Input file.
ResHB.log Output, append (created if does not Log file. Information is appended to this file
exist) documenting each run, showing date/time of run,
general information about the run, and any error
messages.
ResHB.err Output, append (created if does not Error file. Only written if errors are detected. Can serve
exist) as an error flag – delete before a group of runs, if exists
at end, there has been at least one error.
-C<filepath> Output, append (created if does not Case results file. One line of data appended to this file
exist). Written only if valid file per run in comma separated format. Useful for import
specified and if no errors into Excel for analysis and plotting of run sets (see
below).
-G<filepath> Output, append (created if does not Regression export file. One line of data appended to
exist). Written only if valid file this file per run in comma separated format. Useful for
specified and if no errors import into regression applications for post processing
(see below)
<runID>.rep Output, overwritten. Report file. Loads results plus optional info selected via
the –R command line switch.
<runID>.res Output, deleted, then created if run Results file. Room, zone, and system level results in a
successful format for import into post-processing programs (format
described below). Additional info can specified via the
–X command line switch

Case results file format
If –C<filename> is specified on the command line, each ResHB run adds one line <filename> (creating
the file if needed). The line consists of comma-separated items to support simple import into Excel and
other applications, as follows:
runID, timeStamp, caseTag, loadHtg, loadClgSen, loadClgLat, <24 hour loadClgSen>
where:
• runID runID from command line (root file name of input, report, and results files)
• timeStamp date and time of execution
• loadHtg calculated heating load (W) (peak for all systems)
• loadClgSens peak sensible cooling load (W) = MAX( 24 hour loadClgSen)
• loadClgLat peak latent load
• 24 hour loadClgSen 24 hour load profile (W) (sum of all systems)
Note that the case file loads are building-wide totals derived by summing all systems. These results
meaningful for single-system buildings but must be used with care in multi-system situations.
Results file format
The results file, runID.res, always contains a loads summary, documented below. In addition, other
information can be added via the –X command line switch. The format of these additional summaries is
self-documenting.
The entire results file is comma-separated for convenient import into Excel.
runID
! timeStamp
B, , <# systems>, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
S, <systemID>, <# zones>, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
Z, <zoneID>, <# rooms>, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
R, <roomID>, 0, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
R, <roomID>, 0, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
R [repeats for each room in zone]
Z, <zoneID>, <# zones>, <QLH>, <Ta>, <Ts>, <QLCSen>, <QLCLat>, <AVFClgDes>
Z [repeats for each zone in system]
S [repeats for each system in building
Where
AVFClgDes Design air volumetric flow (l/s) (generally = peak of hourly AVF)
QLCLat Latent cooling load (W)
QLCSen Sensible cooling load (W)
QLH Heating load (W)
Ta Indoor air temp (C)
Ts Supply air temperature (C). For rooms = at register. Average values for systems and zones

General Information
Format
An ResHB input file is a free-format text file, prepared in NotePad or suitable alternative text editor.
On each line, all text beginning with an exclamation point is treated as a comment and is dropped.
Commands consist of a list of comma-separated items and are terminated by a semi-colon.
Items are either numeric (N) or alphanumeric (A). Items may be enclosed in quotes (allowing inclusion
of commas and/or semi-colons). The maximum length of A items varies but is generally at least 40 chars.
Some items may be either N or A. ResHB examines the input text and treats it as N if it can be
successfully decoded as a number or A if not. This feature is used, for example, to allow 24 hourly values
to be input as single value to be used for all 24 hours or a schedule reference.
In some cases, items in a command command are optional. That is, they can be omitted (blank(s)
between commas) or, if all remaining items are optional, the command can be terminated with a semi-
colon after the last non-optional item. Optional items take default values as indicated in the detailed
documentation.
Command and ID matching are done case-insensitive. “SCHEDULE” and “Schedule” are both
recognized as a SCHEDULE command. A schedule with ID = “MySched” can be referenced as
“mysched”.
Error Checking
ResHB performs some amount of input validation and checking. Warning or error messages are issued
when anomalies are detected and the loads calculation is suppressed.
Caveat user: ResHB is a research program. Input data is not checked to the same level of detail in all
areas of the application. Results should be checked carefully. In addition, certain kinds of errors are not
detected – for example, file access problems are not handled.
Command Order
With the exception of ROOMs, SURFACEs, and GAINs (see below), ResHB commands can be in any
order. The entire input file is read into memory at startup. Input processing is done from the in-memory
image, allowing any command to be located when needed.
Exception: SURFACEs and GAINs belong to ROOMs and this relationship is determined from input file
order. All SURFACEs and GAINs found in the file belong to the ROOM that they follow.
Note also that INCLUDE commands, PARAMETER commands, and parameter resolution are handled as
the file is initially read, making order significant for these.
Include Files
Input files may include other input files via the INCLUDE command. INCLUDE is a powerful feature
that allows single-copy maintenance of standard constructions, materials, and other common items.

Note that include files are located and read along with all other input, before command-by-command
input processing begins. Thus, the contents of include files are treated as part of the main input file.
The format of the INCLUDE command is:
INCLUDE, “fileName”;
Note that the -I command-line option can be used to simplify access to include files.
Built-in Commands
ResHB has a number of so-called “built-in” items, such as MATERIALS and FENCLASSs. These are
discussed in the detailed command description. This capability is implemented as an implicit (invisible)
INCLUDE file that is read prior to all other input. Thus, all input rules apply with respect to built-in
items – for example, a built-in MATERIAL cannot be redefined.
Parameters
The PARAMETER command allows simple string substitution at the item level, as shown in this
example:
PARAMETER, area, 20;
(other input)
SURFACE, Ceiling, Wood, @area, ...
The PARAMETER command associates a string with a name (in this case “20” with “area”). When an
item is encountered that begins with “@”, the string associated with the name is substituted.
Rules:
• Parameters are defined and used in the order they are encountered in the file. Parameter values
can be changed by subsequent PARAMETER commands referring to the same name.
• Parameters are resolved at the whole item level. That is, parameter references are identified by
the initial @, so they cannot be embedded within items.
Example
The following example calculates loads for a residence in Atlanta, GA. For simplicity, the house is
treated as two rooms corresponding to living and sleeping areas. The main file references two include
files, containing location-specific data (LocGAAT.rhi) and standard definitions (RHBLib.rhi). The
contents of those files is shown after the listing of main file Example.rhb.
File Example.rhb
! Example.rhb
! Location and design conditions

INCLUDE, "locGAAT.rhi";
BUILDING, 0, 3;
INFIL, Std, AIM2_LCLS, G, 2.4, 1;
SYSTEM, Sys01, *ALL*;
ZONE, Zone01, 20., 24.,, *ALL*;
ROOM, Living, 48.00, 2.40, 115.20, conditioned, 1.67;

SURFACE, LC, CEIL-WM, 48.00, 0, 0, AMB;
SURFACE, LW-F, WALL-WL, 14.40, 0, 90, AMB;
SURFACE, LW-L, IWALL-W, 19.20, 90, 90, ADB;
SURFACE, LW-B, WALL-WL, 14.40, 180, 90, AMB;
SURFACE, LW-R, WALL-WL, 12.00, 270, 90, AMB;
SURFACE, WN-LW-R, FEN-WOX2, 7.20, 270, 90, AMB;
SURFACE, LF, FLOR-WM, 48.00, 0,180, AMB;
SURFACE, IMASS, IMASS-W, 48.00, 0, 0, IM;
GAIN, GZLiv;
ROOM, Sleeping, 48.00, 2.40, 115.20, conditioned, 1.67;

SURFACE, SC, CEIL-WM, 48.00, 0, 0, AMB;
SURFACE, SW-F, WALL-WL, 7.20, 0, 90, AMB;
SURFACE, WN-SW-F, FEN-WOX2, 7.20, 0, 90, AMB;
SURFACE, SW-L, WALL-WL, 19.20, 90, 90, AMB;
SURFACE, SW-B, WALL-WL, 14.40, 180, 90, AMB;
SURFACE, SW-R, IWALL-W, 19.20, 270, 90, ADB;
SURFACE, SF, FLOR-WM, 48.00, 0,180, AMB;
SURFACE, IMASS, IMASS-W, 48.00, 0, 0, IM;
GAIN, GZSlp;
! schedules, constructions, fenestrations, and materials

INCLUDE, "RHBLib.rhi";
! end of file
File LocGAAT.rhi
! lat long tzn elv
Location, "Atlanta, GA", 33.65, -84.42, -5.0, 315;
! dst db dr wb tgnd
DesCondCooling, JUL21, 7, 21, D, 32.6, 9.6, 23.4, 0;
DesCondHeating, -7.9, 6.7;
File RHBLib.rhi
!=============================================================================
! RHBLib.rhi: common definitions for ResHB
!=============================================================================
!=============================================================================
! SCHEDULES
!-----------------------------------------------------------------------------
! California Energy Commission internal gain schedule (fraction of daily total)
SCHEDULE, CECGain, 0.024, 0.022, 0.021, 0.021, 0.021, 0.026, 0.038, 0.059,
0.056, 0.060, 0.059, 0.046, 0.045, 0.030, 0.028, 0.031,
0.057, 0.064, 0.064, 0.052, 0.050, 0.055, 0.044, 0.027;
! Building America appliance/plug schedule (fraction of daily total)

SCHEDULE, AppPlug, 0.034, 0.029, 0.029, 0.027, 0.027, 0.034, 0.045, 0.052,
0.052, 0.048, 0.042, 0.042, 0.038, 0.034, 0.034, 0.034,
0.043, 0.035, 0.059, 0.059, 0.056, 0.056, 0.050, 0.041;
! Building America lighting scheds: consolidated, living, sleeping (frac of day total)
SCHEDULE, LightAll, 0.0085, 0.0085, 0.0084, 0.0084, 0.0237, 0.0500, 0.0558, 0.0499,
0.0219, 0.0151, 0.0151, 0.0151, 0.0151, 0.0151, 0.0151, 0.0261,
0.0557, 0.0783, 0.1052, 0.1252, 0.1270, 0.0878, 0.0487, 0.0203;
SCHEDULE, LightLiv, 0.0081, 0.0081, 0.0081, 0.0080, 0.0161, 0.0322, 0.0374, 0.0361,
0.0186, 0.0186, 0.0186, 0.0186, 0.0186, 0.0186, 0.0186, 0.0348,
0.0736, 0.1027, 0.1268, 0.1341, 0.1208, 0.0695, 0.0373, 0.0161;
SCHEDULE, LightSlp, 0.0091, 0.0091, 0.0091, 0.0091, 0.0365, 0.0801, 0.0871, 0.0733,
0.0275, 0.0092, 0.0092, 0.0092, 0.0092, 0.0092, 0.0092, 0.0115,
0.0252, 0.0366, 0.0686, 0.1098, 0.1374, 0.1191, 0.0682, 0.0275;
! Building America occupancy schedules: consolidated, living, sleeping (peak multiplier)

SCHEDULE, OccAll, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.8831,

0.4086, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419,
0.2950, 0.5531, 0.8969, 0.8969, 0.8969, 1.0000, 1.0000, 1.0000;
SCHEDULE, OccLiv, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.3438, 0.5000,
0.4086, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419, 0.2419,
0.2950, 0.5531, 0.8969, 0.8969, 0.8969, 1.0000, 0.5000, 0.0000;
SCHEDULE, OccSlp, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.6562, 0.3831,
0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,
0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.5000, 1.0000;
! Gain items (Wh/day)

! Appliance latent and exhaust fractions per Building America
! Appliance sensible assumed 60% radiant (except refrig)
! Lighting 90% incandescent, radiant all LW (avoids confusing SW "leakage")
! Wh Wh/m2 A Wh/m2 CFA Wh/BR Wh/Occ mult FConv FSW FLW FLat FExh
GAINITEM, Refrig, 1833, 0, 0, 0, 0, 1, 1;
GAINITEM, RangeEl, 1655, 0, 0, 0, 0, 1, .16, 0, .24, .30, .30;
GAINITEM, Dishwasher, 282, 0, 0, 94, 0, 1, .34, 0, .51, .15;
GAINITEM, Washer, 144, 0, 0, 48, 0, 1, .40, 0, .60, 0;

GAINITEM, DryerEl, 1145, 0, 0, 381, 0, 1, .06, 0, .09, .05, .80;
GAINITEM, Plug, 0, 49.25, 0, 0, 0, 1, .36, 0, .54, .1;
GAINITEM, Lighting, 1247, 0, 23.59, 0, 0, 1, .21, 0, .79, 0;
! Occupants: Building America data

! Living: 230 Btuh sens, 190 lat = 420 Btuh (123 W), 45% lat
! Sleeping: 210 Btuh sens, 140 lat = 350 Btuh (103 W), 40% lat
GAINITEM, PrsnLiv, 0, 0, 0, 0, 123, 1, .22, 0, .33, .45;
GAINITEM, PrsnSlp, 0, 0, 0, 0, 103, 1, .30, 0, .30, .40;
GAINITEM, PrsnAvg, 0, 0, 0, 0, 113, 1, .26, 0, .31, .43;
GAINITEMSET, Kitchen, Refrig,, RangeEl,, Dishwasher;

GAINITEMSET, Laundry, Washer,, DryerEl;
! Living / sleeping
GAINZONE, GZLiv,
Kitchen, AppPlug, Laundry, AppPlug, Plug, AppPlug,
Lighting, .63 LightLiv,
PrsnLiv, OccLiv;
GAINZONE, GZSlp,
Plug, AppPlug,
Lighting, .37 LightSlp, PrsnSlp, OccSlp;
! consolidated
GAINZONE, GZAvg,
Kitchen, AppPlug, Laundry, AppPlug, Plug, AppPlug,
Lighting, LightAll, PrsnAvg, OccAll;
!=============================================================================
! FENESTRATIONS
!-----------------------------------------------------------------------------
! wood operable frames, default SHGC and U-factor
FENESTRATION, FEN-WOC1, OpWood, Clear, 1;
FENESTRATION, FEN-WOE2, OpWood, LoEHS, 2;
FENESTRATION, FEN-WOX2, OpWood, LoELS, 2;
! custom frames, glass only SHGC, center-of-glass U-factor

FENESTRATION, FEN-GXC1, Custom, Clear, 1, .86, 5.91;
FENESTRATION, FEN-GXE2, Custom, LoEHS, 2, .70, 1.99;
FENESTRATION, FEN-GXX2, Custom, LoELS, 2, .41, 1.70;
!==============================================================================
! FENCLASS
!------------------------------------------------------------------------------

! used in test cases, from OSU, origin unknowns
FENCLASS, FCTEST-C2, 2, 10,
0, 10, 20, 30, 40, 50, 60, 70, 80, 90, !diff angles
.701, .701, .698, .691, .678, .648, .585, .456, .237, 0, .606, ! SHGC
.607, .606, .601, .593, .577, .546, .483, .362, .165, 0, .510, ! TRANS
.167, .168, .170, .175, .182, .190, .200, .209, .202, 0, .185, ! ABS1
.113, .113, .115, .116, .118, .119, .115, .101, .067, 0, .111; ! ABS2
!=============================================================================
! CONSTRUCTIONS
!-----------------------------------------------------------------------------
! wood frame sub-constructions (10% and 25% framing factor)
! thickness adjusted when ref'd
Construction, WF10Empty, PP, $F 90 .10 FrameWood Empty;
Construction, WF10FbrGl, PP, $F 90 .10 FrameWood FbrGlsBatt;
Construction, CONC-L, PP, ConcHvy 200;

Construction, CONC-M, PP, FoamBoard 25, ConcHvy 200;
Construction, CONC-H, PP, FoamBoard 200, ConcHvy 200;
! massless constructions (same overall conductance as CONC)

Construction, MASSLESS-L, PP, .0909;
Construction, MASSLESS-M, PP, .953;
Construction, MASSLESS-H, PP, 6.98767;
Construction, CEIL-WL, PP, WF10Empty 190, GypBoard;

Construction, CEIL-WM, PP, WF10FbrGl 190, GypBoard;
Construction, CEIL-WH, PP, FbrGlsBatt 140, WF10FbrGl 190, GypBoard;
Construction, CLRF-WL, PP, AsphaltShingle 9, Plywood 20, CEIL-WL;

Construction, CLRF-WM, PP, AsphaltShingle 9, Plywood 20, CEIL-WM;
Construction, CLRF-WH, PP, AsphaltShingle 9, Plywood 20, CEIL-WH;
Construction, ROOF-ASPH, PP, AsphaltShingle 9, Plywood 20;

CONSTRUCTION, ROOF-TILE, PP, ConcHvy 50, .2, Plywood 20;
Construction, WALL-WL, PP, Plywood 20, WF25Empty 90, Gypboard;

Construction, WALL-WM, PP, Plywood 20, WF25FbrGl 90, Gypboard;
Construction, WALL-WH, PP, Plywood 20, FoamBoard 25, WF25FbrGl 140, Gypboard;
! Interior ceil, wall, floor (1/2 thickness)

Construction, ICEIL-W, PP, WF10Empty 95, GypBoard;
Construction, IWALL-W, PP, WF25Empty 45, Gypboard;
Construction, IFLOR-W, PP, WF25Empty 95, Plywood 19, OakFloor 8;
! Internal mass
Construction, IMASS-W, PP, FrameWood 12;
! Floors
Construction, FLOR-WL, PP, WF10Empty 190, Plywood 19, OakFloor 8;
Construction, FLOR-WM, PP, WF10FbrGl 190, Plywood 19, OakFloor 8;
Construction, FLOR-WH, PP, FoamBoard 50, WF10FbrGl 190, Plywood 19, OakFloor 8;
Construction, SLAB-L, PP, Soil 300, ConcHvy 100;

Construction, SLAB-M, PP, SLAB-L, .185; ! carpet: approx R-1.05 IP
Construction, SLAB-H, PP, SLAB-L, .370; ! carpet: approx R-2.10 IP
Construction, CRAWL-FLOR, PP, Soil 400;
! Test constructions -- retained to allow exact match of test case results

Construction, KARNSEXTWALL, PP, HARDBOARD_SIDING 9.5, FELT_SHEATHING .8,
REGULAR_DENSITY_SHEATHING 12.7, KARNSSTUDR11 88.9, GypBoard 12.7;
Construction, KARNSCEILING, PP, KARNSTRUSSR19 88.9, GypBoard 12.7;
Construction, KARNSFLOOR, PP, PVC 12.7, AIRGAP 0, AIRGAP 0,
PARTICLE_BOARD_UNDERLAYMENT 15.9, FELT_SHEATHING .8,
KARNSJOISTR19 304.8, PLYWOOD_SHEATHING 12.7;
Construction, KARNSINTPART, PP, Gypboard1 12, airgap 0, GypBoard1 12;
Construction, KARNS-DBL-GLZ, PP, PLATE_GLASS 6, AIRGAP 0, PLATE_GLASS 6;

!==============================================================================
! MATERIALS (most materials built-in)
!------------------------------------------------------------------------------
! k rho Cp R
! W/m-K kg/m3 kJ/kg-K m2-K/W
Material, AIRGAP, 0, 0, 0, .17;
Material, HARDBOARD_SIDING, .787, 1900, 1.17, 0;
Material, FELT_SHEATHING, .19, 960, .95, 0;
Material, REGULAR_DENSITY_SHEATHING, .055, 290, 1.3, 0;
Material, KARNSSTUDR11, .055, 159, 1.2, 0;
Material, PVC, .16, 1379, 1.004, 0;
Material, PARTICLE_BOARD_UNDERLAYMENT, .11, 640, 1.21, 0;
Material, KARNSJOISTR19, .14, 185, 1.2, 0;
Material, PLYWOOD_SHEATHING, .12, 540, 1.21, 0;
Material, KARNSTRUSSR19, .034, 133, 1.2, 0;
Material, PLATE_GLASS, .76, 2710, .837, 0;
Material, ASPHALT_SHINGLES, 1.29, 1100, 1.26, 0;
Material, KARNSTRUSSAIR, .222, 43, 1.63, 0;
Material, CONCRETE, 1.95, 2240, 0.84, 0;
Material, GypBoard1, .19, 950, .84, 0;
! end RHBLib.rhi

Commands
SIMCONTROL
Specifies convergence criteria for calculations.

SIMCONTROL, dayMax, dayConv, hourMax, hourConv, swingMax, swingConv;
# Item Format Default Description

1 dayMax N 300 Max number of day iterations (if day convergence not reached)
2 dayConv N 1E-5 Day convergence fraction
3 hourMax N 100 Max number of iterations for a single hour (if hour convergence not
reached)
4 hourConv N .01 Hour convergence fraction
5 swingMax N 30 Max number of temperature swing iterations
6 swingConv N (°C) .01 Temperature tolerance for swing search
DOCUMENTATION
Provides run documentation passed to output files to aid parametric studies. Optional.
DOCUMENTATION, runDoc, caseTag;

1 runDoc A (blank) Run documentation, included in reports
2 caseTag A (blank) User-defined text string that identifies the run.
LOCATION
Provides information about the building site.

LOCATION, name, lat, long, tzn, elev;

1 name A (req) Location name, documentation only. Note use of quotes allows
inclusion of comma: “Knoxville, TN”
2 lat N (° N) (req) Latitude. > 0 for northern hemisphere, < 0 for southern.
Valid range is -90 - +90
3 long N (° E) 0 Longitude. >0 for east of Greenwich, <0 for west.
Valid range is -180 - +180
4 tzn N (hr) 0 Time zone, hours after UCT. <0 for western hemisphere locations
(e.g. -5 for EST)
5 elev N (m) 0 Site elevation above sea level.

DESCONDCOOLING
Specifies outdoor conditions used for cooling loads calculation.

ResHB requires 24 hour sequences of outdoor conditions as input for its cooling load calculations
simulation. These values can be specified hour-by-hour, via schedules, or generated from single design
values plus daily range. If dry bulb temperature is scheduled, then web bulb must be as well.
Support for multiple DESCONDCOOLINGs planned; currently only 1 is allowed.
DESCONDCOOLING, ID, mon, day, dst, tdb, dr, twb, tg, ws, vis, clr, grRflct;

1 ID A (ID) (req) ID of cooling design condition
2 mon N (1-12) (req) Month
3 day N (1-31) (req) Day of month
4 dst A (code) (req) Time convention: S=Std, D=Daylight savings. If D, the profiles of
generated outdoor conditions is shifted later by 1 hour.
Generated outdoor conditions include solar irradiance and
defaulted temperature profiles (not scheduled temperature
profiles).
5 tdb N (° C) (req) Outdoor maximum dry bulb temperature. Tdb and dr (see below)
are used to generate an hourly temperature profile that peaks at
1500 (standard time) or 1600 (daylight time).
A (Schedule ID) ID of schedule specifying dry bulb temperatures (° C) by hour.
Note: Scheduled values are used as input, no daylight savings
shift is applied by ResHB.
6 dr N (° C) (req) Daily range of dry bulb (ignored if temperatures are specified by
schedule)
7 twb N (° C) (req) Wet bulb temperature coincident with tdb.
A (Schedule ID) ID of schedule specifying wet bulb temperatures (° C) by hour.
Note: Scheduled values are used as input, no daylight savings
shift is applied by ResHB.
8 tg N (° C) 0 Ground temperature (° C) used for all hours
A (Schedule ID) ID of schedule specifying ground temperature (° C) by hour
9 ws N (m/s) 3.4 Design wind speed
10 vis N (km) 30 Visibility (used for visibility solar model)
11 clr N (.8 – 1.2) 1 Clearness number (used for ASHRAE solar model)
12 grRflct N (0 – 1) .2 Default ground reflectivity (.2 is a typical value)

DESCONDHEATING
Specifies outdoor conditions used for heating loads calculation.

ResHB performs a single-point heat loss calculation.
DESCONDHEATING, tdb, ws;

1 tdb N (° C) (req) Heating outdoor design dry bulb temperature
2 ws N (m/s) 6.7 Heating wind speed
SCHEDULE
Specifies a named sequence of 24 values. Schedules are widely used in ResHB to input data by hour.
SCHEDULE, ID, 1 or 24 value(s);

1 ID A (ID) (req) ID of schedule. Must be unique across SCHEDULEs and
OHFINs.
2 (-25) value(s) N (units (req) Schedule value(s). If 1 value is present, it is used for all 24
depend on hours. Otherwise, 24 values must be specified.
use of
schedule)
Note: The time base of schedules is local time (standard or daylight, per DESCONDCOOLING). This is
generally appropriate for modifying occupancy-related values such as internal gain. However, care must
be exercised when applying a schedule to items that involve solar time, such as surface shading or
outdoor temperature profiles.

GAINITEM
Defines a type of internal gain. GAINITEMs have no effect until they are included in a room via
reference in GAIN commands. See example with GAIN (below)
GAINITEM, ID, gain, gainA, gainCFA, gainBR, gainBR, schedMult, fConv, fSWRad, fLWRad,
fLat, fExh;
Note: fConv + fSWRad + fLWRad + fLat + fExh must equal 1.

1 ID A (ID) (req) ID of gain item. Must be unique across GAINITEMs,
GAINITEMSETs, and GAINZONEs.
2 gain N (W) (req) Gain value
3 gainA N (W/m2) 0 Gain per unit of applied conditioned floor area (room area or
GAINZONE area)
4 gainCFA N (W/m2) 0 Gain per unit of building conditioned floor area
5 gainBR N (W/BR) 0 Gain per bedroom (per BUILDING NBR)
6 gainOcc N (W/Occ) 0 Gain per occupant (per BUILDING NOcc)
7 schedMult N and/or 1 Gain schedule / multiplier (see below)
schedule ID
8 fConv N (0 – 1) 0 Fraction convective
9 fSWRad N (0 – 1) 0 Fraction short wave radiant (i.e. light)
10 fLWRad N (0 – 1) 0 Fraction long wave radiant
11 fLat N (0 – 1) 0 Fraction latent
12 fExh N (0 – 1) 0 Fraction exhaust (not contributing to internal gain)
The schedMult value consists of any number of space-separated numeric constants or schedule IDs that
are multiplied together to yield a 24-hour profile for the gain. A numeric constant is treated as a 24 hour
schedule with the constant value for all 24 hours.
Examples
schedMult Result
.3 all hours = .3
.3 .4 all hours = .12
.3 AllDay each hour = 0.3 · hour value of schedule AllDay
AllDay OccProf each hour = (AllDay hour value) · (OccProf hour
value)

GAINITEMSET
Combines a group of GAINITEMs, with optional schedule / multiplier modification. GAINITEMSETs

allow convenient groupings of GAINITEMs. As with GAINITEMs, GAINITEMSETs are definitions of
possible gains and have no effect unless referenced via a GAIN command.
GAINITEMSET, ID, gainItemID, schedMult [, gainItemID, schedMult,...];

1 ID A (ID) (req) ID of GAINITEMSET. Must be unique across GAINITEMs,
2 gainItemID ID (blank) ID of GAINITEM to be included in this GAINITEMSET.
3 schedMult N and/or ID 1 schedule / multiplier (see GAINITEM). GAINITEMSET
schedMult is multiplied by GAINITEM schedMult to yield
combined 24-hour profile.
3- (additional a maximum of 10 GAINITEMs can be included in a
gainItemID / GAINITEMSET.
schedMults)
GAINZONE
Combines a group of GAINITEMs and/or GAINITEMSETs, with optional schedule / multiplier

modification. A GAINZONE allows area weighted application of gains to a group of rooms, see GAIN.
GAINZONE, ID, gainItemID, schedMult [, gainItemID, schedMult,...];

1 ID A (ID) (req) ID of GAINZONE. Must be unique across GAINITEMs,
2 gainItemID ID (blank) ID of GAINITEM or GAINITEMSET to be included in this
GAINZONE
3 schedMult N and/or ID 1 schedule / multiplier (see GAINITEM). GAINZONE schedMult
is multiplied by GAINITEM schedMult to yield combined 24-
hour profile. If gainItemID references a GAINITEMSET, all
schedMults are combined.
3- (additional a maximum of 10 GAINITEMs/GAINITEMSETs can be
gainItemID / included in a GAINZONE.
schedMults)

SURFACEDEFAULTS
Provides default values for several SURFACE arguments, optional command. Note that input is not
provided for all SURFACE arguments, since only some of them commonly vary. Any value can be
specifically set on a surface-by-surface basis (see SURFACE).
SURFACEDEFAULTS, SWAbsO-RF, SWAbsO-WL, SWAbsI-CW, SWAbsI-FL;

1 SWAbsO-RF N (0 – 1) .8 Default roof outside SW (solar) absorptance
2 SWAbsO-WL N (0 – 1) .6 Default wall outside SW (solar) absorptance
3 SWAbsI-CW N (0 – 1) .45 Default ceiling and wall inside SW (solar) absorptance
4 SWAbsI-FL N (0 – 1) .6 Default floor inside SW (solar) absorptance
Summary of surface absorptance defaults:

Surface Short wave (solar) Long wave (thermal)
Outside Inside Outside Inside
Roof / ceiling (tilt < SWAbsO-RF or .8 SWAbsI-CW or .45 .9 .9
60 °)
Wall (180° < tilt ≤ SWAbsO-WL or .6 SWAbsI-CW or .45 .9 .9
60°)
Fenestration per FenClass per FenClass .84 .84
Floor (tilt = 180°) 0 SWAbsI-FL or .6 .9 .9
ROOMDEFAULTS
Provides default values for ROOM arguments, optional command.

ROOMDEFAULTS, MCSwing, SlvSwing;

1 MCSwing N (°C) 0 Default temperature swing for master
and conditioned rooms.
2 SWAbsO-WL N (0 – 1) 0 Default temperature swing for slave
rooms

OHFIN
Specifies a geometric shading configuration. Geometric shading can be applied to any vertical surface,
whether opaque or fenestration.
Left and right in OHFIN are defined as looking from the outside of the surface.
Note that srfWidth and srfHeight are used for geometric calculations only. The area implied by these
dimensions need not equal the area(s) of referencing surface(s).
OHFIN, ID, srfWidth, srfHeight, srfReveal, ohDepth, ohSrfDist, ohLeftExt, ohRightExt,
ohFlapHeight, lfDepth, lfSrfDist, lfTopExt, lfBotShort, rfDepth, rfSrfDist,
rfTopExt, rfBotShort;

1 ID A (ID) (req) ID of this Ohfin. Must be unique across OHFINs and
SCHEDULEs
2 srfWidth N (m) (req) surface width
3 srfHeight N (m) (req) surface height
4 srfReveal N (m) 0 surface reveal (setback of target area from nominal surface
face)
5 ohDepth N (m) 0 depth of overhang (from surface face)
6 ohSrfDist N (m) 0 distance from top of surface to overhang
7 ohLeftExt N (m) 0 overhang extension beyond left edge of surface
8 ohRightExt N (m) 0 overhang extension beyond right edge of surface
9 ohFlapHeight N (m) 0 height of vertical flap (down) at outer edge of overhang
10 lfDepth N (m) 0 depth of left fin
11 lfSrfDist N (m) 0 distance from left edge of surface to left fin
12 lfTopExt N (m) 0 distance left fin extends above top of surface
13 lfBotShort N (m) 0 distance left fin stops short above bottom of surface
14 rfDepth N (m) 0 depth of right fin
15 rfSrfDist N (m) 0 distance from right edge of surface to right fin
16 rfTopExt N (m) 0 distance right fin extends above top of surface
17 rfBotShort N (m) 0 distance right fin stops short above bottom of surface

MATERIAL
Specifies a material and its associated thermal properties.

MATERIAL, ID, cond, dens, spHt, res [, thk];

1 ID A (ID) (req) ID of material. Must be unique across MATERIALs and
CONSTRUCTIONs.
2 cond N (W/m-K) (req) Conductivity (if 0, material is resistive)
3 dens N (kg/m3) (req) Density (ignored if resistive)
4 spHt N (kJ/kg-K) (req) Specific heat (ignored if resistive)
5 res N (m2-K/W) (req) Resistance. If conductivity is 0, material has a fixed
resistance (independent of layer thickness). Ignored if
conductivity is non-zero.
6 thk N (mm) 0 if resistive, Optional default layer thickness. If present, material
else –1 = can be referenced in a CONSTRUCTION layer without
unspecified specification of thickness. This is useful for materials
that have a common typical thickness, such as carpet.
CONSTRUCTION
Specifies a multi-layer construction for use in surface(s).

Construction layers are defined outside to inside. Thus, layer 1 is the outside surface. Up to 10 layers are
supported.
CONSTRUCTION, ID, Utarg, lyrDesc1 [, lyrDesc2, …, lyrDesc3];

1 ID A (ID) (req) ID of construction. Must be unique across
CONTRUCTIONs, MATERIALs, and
FENESTRATIONs.
2 Utarg N (W/m2-K) (req) Target U-factor for construction, used only for framed
constructions.
A (code) Specifies that Utarg should be derived from layer data
• PP: Utarg = Upp (parallel path)
• IP: Utarg = Uip (isothermal planes)
• MPI: Utarg = (Upp + Uip) / 2
3 – 12 lyrDesc A (at least 1 req) Layer description(s), see table below.
Layer descriptions are multipart specifications of layer thermal properties. Several options provide
flexibility in how layers are described. A layer description consists of 1, 2, or 5 values, as follows (note
that these values are delimited by whitespace, not commas):
rmc [thk [fFrm matFrm matCav]]

Form Effect
Numeric R (m2-K/W) Massless layer with specified resistance R
matID [thk (mm)] Layer having properties of MATERIAL matID and thickness thk.
• If matID is resistive or has a default thickness, thk may be
omitted
• If matID is resistive, thk is ignored even if specified
consID [thk (mm)] Specifies a “sub-construction”. The layer(s) of previously-defined
CONSTRUCTION consID are added to the current construction. This is
useful for defining common framed layers.
• If consID has exactly 1 layer, thk may be specified.
• If thk is not specified, the thickness of consID is used.
$F thk fFrm matFrm Specifies a framed layer, consisting of a framing material and a cavity
matCav material. ResHB creates an homogenized fictitious material such that the
overall U-factor of the construction matches Utarg.
$F code indicates framed layer
thk N (mm) layer thickness
fFrm N (0 – 1) framing fraction (fraction of layer face area
consisting of framing material); typical
values for wood frame construction are in
the .10 - .25 range.
matFrm ID Framing MATERIAL ID
N (m2-K/W) Alternative: framing resistance
matCav ID Cavity MATERIAL ID
N (m2-K/W) Alternative: cavity resistance
CONSTRUCTION example (referenced materials are assumed to be suitably defined):

CONSTRUCTION, WF25FbrGls, PP, $F 200 .25 Wood FbrGlsBatt;
CONSTRUCTION, WFWALL, PP, Siding 13, Plywood 20, .1, WF25FbrGls 100, GypBoard;
These statements specify two constructions:

• WF25FbrGls consists of one 200 mm framed layer, 25% wood and 75% fiberglass batt insulation.
• WFWALL consists of 13 mm of Siding, 20 mm Plywood, a resistance of .1 m2-K/W, 100 mm of
insulated stud framing, and gypsum board.
Note that the thickness of the stud framing layer is not the same as the original value specified in the
WF25FbrGls definition. Also note that GypBoard must have a default thickness, allowing it to be
referenced without specifying a thickness.
The “PP” code indicates that the target U-factor for the construction is calculated from layer properties
using the parallel path method. The properties of the framed layer are adjusted so the resulting U-factor
equals Utarg.
Report options “c” and “C” cause construction details to be printed for checking.
Also note that the $REV option in SURFACE allows using constructions in reverse layer order. See
SURFACE.

FENESTRATION
Defines a fenestration type for use in surface(s).

Note that while the formal definition of “fenestration” includes doors, in ResHB, the term is applied to
assemblies that transmit solar radiation (e.g. windows and skylights). Doors should be entered as opaque
surfaces.
The ResHB fenestration model is based on NFRC ratings augmented by FENCLASS (“fenestration
class”) data that describes heat gain characteristics as a function of incident angle. The fenestration class
data is normalized to the SHGC of the fenestration for application in the model.
FENESTRATION, ID, frameTy, glzTy, nGlz, SHGC, UorCons, UHtg;
# Item Format Defaults Description

1 ID A (ID) (req) ID of fenestration. Must be unique across
FENESTRATIONs and CONTRUCTIONs.
2 frameTy A (ID) (req) Frame type code (see below)
3 glzTy A (req) Glazing type code (see below)
A (Fenclass ID) ID of FENCLASS (see above)
4 nGlz N (1 – 3) (req) # of glazings
5 SHGC N (0 – 1) per frameTy and NFRC rated Solar Heat Gain Coefficient. Note this
glzTy, see below is the overall SHGC, including the effects of the
frame.
6 UorCons N (W/m2-K) per frameTy and NFRC rated heating U-factor (used for cooling
glzTy, see below calculations).
A (Construction Alternative. If specified, conduction is modeled
ID) using detailed construction layers as opposed to the
default massless U-factor model (provides
compatibility with prior ASHRAE Toolkit methods,
generally not used.)
7 UHtg N (W/m2-K) same as cooling U- Optional U-factor to be used for heat loss calculation
factor (heating U-factor may differ due to storm windows).
Frame code (for frameTy)

Codes
Frame construction
Operable Fixed
OpAlum FxAlum Aluminum (or other metal)
OpAlumTB FxAlumTB Aluminum with thermal break
OpRfVnl FxRfVnl Reinforced vinyl
OpACWood FxACWood Aluminum clad wood
OpVnl FxVnl Vinyl
OpWood FxWood Wood
OpInsFbg FxInsFbg Insulated fiberglass
OpInsVnl FxInsVnl Insulated vinyl
(any other) Custom frame, fenestration characteristics determined from other input

Glazing type codes (for glzTy)
Code Description Default FENCLASS by NGlz
1 2 3
Clear Clear FCA-1a FCA-5a FCA-29a
LoEHS Low-e, high solar n/a FCA-17c FCA-32c
LoELS Low-e, low solar n/a FCA-25a FCA-40c
HtAbs Heat absorbing FCA-1c FCA-5c FCA-29c
Rflct Reflective FCA-1l FCA-5p FCA-29c
Default U-factor: If the cooling U-factor (UorCons) is not specified, a default value is derived from
glzTy and frameTy using values from Table 4, Chapter 30 (Fenestration), 2001 HOF, using values shown
below.
Default SHGC: If SHGC is not specified, a default value is derived from glzTy and frameTy. The SHGC
is calculated using frame fraction, frame SHGC, and glazing only SHGC, using the values given on p.
30.39 (Fenestration), HOF 2001, as follows:
Frame Frame
FrameTy codes
fraction SHGC
OpAlum, OpAlumTB .15 .14

OpRfVnl, OpACWood, OpVnl, OpWood, OpInsFbg, .27 .04
OpInsVnl
FxAlum, FxAlumTB .11 .11
FxRfVnl, FxACWood, FxVnl, FxWood, FxInsFbg, FxInsVnl .13 .03

Resulting default U-factor and SHGC values --
Operable Fixed
Glazing Glass OpRfVnl OpWood OpInsFbg FxRfVnl FxWood FxInsFbg

type N glz ID only OpAlum OpAlumTB OpACWood OpVnl / OpInsVnl FxAlum FxAlumTB FxACWood FxVnl FxInsVnl
Clear 1 1a U 5.91 7.24 6.12 5.14 5.05 4.61 6.42 6.07 5.55 5.55 5.35
SHGC 0.86 0.75 0.75 0.64 0.64 0.64 0.78 0.78 0.75 0.75 0.75
2 5a U 2.73 4.62 3.42 3.00 2.87 2.53 3.61 3.22 2.86 2.84 2.72
SHGC 0.76 0.67 0.67 0.57 0.57 0.57 0.69 0.69 0.67 0.67 0.67
3 29a U 1.76 3.80 2.60 2.25 2.19 1.91 2.76 2.39 2.05 2.01 1.93
SHGC 0.68 0.60 0.60 0.51 0.51 0.51 0.62 0.62 0.60 0.60 0.60
LoELS 2 25a U 1.70 3.83 2.68 2.33 2.21 1.89 2.75 2.36 2.03 2.01 1.90
SHGC 0.41 0.37 0.37 0.31 0.31 0.31 0.38 0.38 0.36 0.36 0.36
3 40c U 1.02 3.22 2.07 1.76 1.71 1.45 2.13 1.76 1.44 1.40 1.33
SHGC 0.27 0.25 0.25 0.21 0.21 0.21 0.25 0.25 0.24 0.24 0.24
LoEHS 2 17c U 1.99 4.05 2.89 2.52 2.39 2.07 2.99 2.60 2.26 2.24 2.13
SHGC 0.70 0.62 0.62 0.52 0.52 0.52 0.64 0.64 0.61 0.61 0.61
3 32c U 1.42 3.54 2.36 2.02 1.97 1.70 2.47 2.10 1.77 1.73 1.66
SHGC 0.62 0.55 0.55 0.46 0.46 0.46 0.56 0.56 0.54 0.54 0.54
HtAbs 1 1c U 5.91 7.24 6.12 5.14 5.05 4.61 6.42 6.07 5.55 5.55 5.35
SHGC 0.73 0.64 0.64 0.54 0.54 0.54 0.66 0.66 0.64 0.64 0.64
2 5c U 2.73 4.62 3.42 3.00 2.87 2.53 3.61 3.22 2.86 2.84 2.72
SHGC 0.62 0.55 0.55 0.46 0.46 0.46 0.56 0.56 0.54 0.54 0.54
3 29c U 1.76 3.80 2.60 2.25 2.19 1.91 2.76 2.39 2.05 2.01 1.93
SHGC 0.34 0.31 0.31 0.26 0.26 0.26 0.31 0.31 0.30 0.30 0.30
Rflct 1 1l U 5.91 7.24 6.12 5.14 5.05 4.61 6.42 6.07 5.55 5.55 5.35
SHGC 0.31 0.28 0.28 0.24 0.24 0.24 0.29 0.29 0.27 0.27 0.27
2 5p U 2.73 4.62 3.42 3.00 2.87 2.53 3.61 3.22 2.86 2.84 2.72
SHGC 0.29 0.27 0.27 0.22 0.22 0.22 0.27 0.27 0.26 0.26 0.26
3 29c U 1.76 3.80 2.60 2.25 2.19 1.91 2.76 2.39 2.05 2.01 1.93
SHGC 0.34 0.31 0.31 0.26 0.26 0.26 0.31 0.31 0.30 0.30 0.30
FENCLASS
Defines a fenestration class (a collection of data that defines the angular heat gain characteristics of a
fenestration system).
ResHB has a number of built-in FENCLASSs that support default characteristics (see table below).
Additional FENCLASSs can be input to handle special cases.
FENCLASS layers are defined outside to inside. That is, layer 1 is the outside (or only) layer.
The number of values required for a FENCLASS depends on the number of glazing layers and the
number of angles for which data is available. See example under General Information (above).

FENCLASS, ID, nLayers, nAngles, angles, SHGC, trans, abs1, abs2, abs3;

1 ID A (ID) (req) ID of fenestration class. Must be unique across FENCLASSes
2 nLayers N (1 – 3) (req) number of glazing layers
3 nAngles N (2 – 10) (req) number of angles for which data is provided
angles N (°, 0 - 90) (req) list of angles in ascending order (nAngles values)
SHGC N (0 – 1) (req) list of SHGC values at each angle plus hemispherical diffuse
(nAngles + 1 values)
trans N (0 – 1) (req) list of transmitted solar gain at each angle plus hemispherical
diffuse (nAngles + 1 values)
abs1 N (0 – 1) (req) list of layer 1 (outside) absorbed fraction for each angle plus
hemispherical diffuse ( nAngles + 1 values)
abs2, N (0 – 1) (req) absorbed fraction data for layer 2 and layer 3 if required.
abs3
ResHB includes a number of built-in FENCLASSs that are used to provide defaults for typical
fenestration configurations; they can also be referenced as needed for custom situations. The associated
data is taken from Table 13, Chapter 30 (Fenestration), ASHRAE HOF (2001), refer to that table for
complete information.
ID NGlz SHGC (glazing Description
only)
FCA-1a 1 .86 clear
FCA-1c 1 .73 bronze heat absorbing
FCA-1l 1 .31 reflective
FCA-5a 2 .76 clear / clear
FCA-5c 2 .62 bronze heat absorbing / clear
FCA-5p 2 .29 reflective / clear
FCA-17c 2 .70 clear / low-E (“high solar”)
FCA-25a 2 .41 low-e / clear (“low solar”)
FCA-29a 3 .68 clear / clear / clear
FCA-29c 3 .34 green heat absorbing / clear / clear
FCA-32c 3 .62 clear / clear / low-e (“high solar”)
FCA-40c 3 .27 low-e / low-e / clear (“low solar”)
FENCLASS Example:
FENCLASS, FCA-25a, 2, 7,', & ! double LE CLR (low-e low solar)
0, 40, 50, 60, 70, 80, 90,
.41, .40, .38, .34, .27, .14, 0, .36, ! SHGC
.37, .35, .33, .29, .22, .11, 0, .31, ! TRANS
.24, .26, .26, .27, .28, .23, 0, .26, ! ABS1 (outer)
.04, .04, .04, .04, .03, .03, 0, .04; ! ABS2

INFIL
Specifies infiltration parameters. ResHB includes a full implementation of the AIM-2 infiltration model
as well as a simpler air change model.
INFIL, ID, meth, L, Hstack, Heave, FCrawl, ShelClass, PRef, N, R, X, Y, ShelClassFlue;

1 ID A (ID) (req) Infiltration ID
2 meth A (code) (req) Code specifying infiltration model and input form. See L, next
argument
3 L N or A (req) Air exchange rate, leakage area, or leakage class
meth L
ACH air exchange rate or ID of 24 hr ACH
schedule; no additional arguments.
AIM2_C Total leakage coefficient (L/s-Pan)
AIM2_ALEAK Total leakage area (cm2)
AIM2_LNORM Normalized leakage as defined in Standard
119 (value range is .1 for very tight to 1.6 for
leaky)
AIM2_LCLS Leakage class, A-J (see table below)
4 Hstack N (m) (req) Average stack height = height difference between floor (or grade if
lowermost floor is below grade) and ceiling averaged over
conditioned footprint of building
5 Heave N (m) Hstack + Eave height above grade, determines wind modifier.
1
6 FCrawl N 0 Fraction floor area above crawl space; 1 – FCrawl is assumed
above slab or basement.
7 ShelClass N (1-5) 4 Building shelter class (see below)
8 Pref N (Pa) 4 Reference pressure for leakage areas. Unused if meth = AIM2_C
9 N N .67 Pressure exponent
10 R N .5 Leakage distribution – (floor + ceiling) / total
11 X N 0 Leakage distribution – (floor – ceiling) / total
12 Y N 0 Leakage distribution – flue / total
13 ShelClassFlue N 2 Flue shelter class (see below)
AIM-2 shelter classes

Class Description
1 No obstructions or local shielding
2 Light local shielding with few obstructions within two house heights
3 Heavy shielding, may large obstructions within two house heights
4 Very heavy shielding, many large obstructions within one house height
5 Complete shielding, with large buildings immediately adjacent
AIM-2 leakage classes

Leakage class Normalized leakage
A .09
B .12
C .17
D .24
E .34
F .48
G .68
H .96
I 1.36
J 1.60

DUCTSYS
Specifies a duct system. Note: as of version 1.00, the duct model is not integrated with the loads model.
These inputs are experimental.
DUCTSYS, ID, config, supF, supLoc, supLF, supR, supMat, retF, retLoc, retLF, retR,
retMat;

1 ID A (ID) (req) Duct system ID
2 config A (req) Duct system configuration. Only current choice is ‘radial’ (to be
expanded)
3 supF N (0 – 1) 1 Fraction of supply duct located outside of conditioned space
4 supLoc A (ID) (req) Supply duct location: room ID or ‘outdoors’
5 supLF N (req) Supply duct leakage fraction
6 supMat A (req) Supply duct material (“sheet metal”, “flex vinyl”, or “duct board”
7 supR N (m2-K/W) (req) Supply duct insulation resistance (added to material resistance)
8 retF N (0 – 1) 1 Fraction of return duct located outside of conditioned space
9 retLoc A (ID) (req) Return duct location: room ID or ‘outdoors’
10 retLF N (req) Return duct leakage fraction
11 retMat A (req) Return duct material (“sheet metal”, “flex vinyl”, or “duct board”
12 retR N (m2-K/W) (req) Return duct insulation resistance (added to material resistance)
BUILDING
Specifies building-wide information.

BUILDING, azm0, NBR, Nocc;

1 azm0 N (°, -180 - 0 Actual azimuth of surfaces having azm = 0. azm0 is added to all
+180) azimuth values, effectively rotating the building clockwise.
2 NBR N 2 Number of bedrooms in building (used re determination of internal
gains, see GAINITEM)
3 NOcc N NBR+1 Number of occupants in building (used re determination of internal
gains, see GAINITEM)
SYSTEM
Specifies a system (collection of zones). Not that ResHB is does not perform any system modeling. A
system is simply an accounting structure to which loads are accumulated for reporting purposes.
SYSTEM, ID, zoneList;

1 ID A (ID) (req) system ID
2 zone A (at least 1 req) list of IDs of zones that are associated with the zone. May be
list ‘*ALL*’ to select all zones.

ZONE
Specifies a building zone (a group of rooms). The rooms in a zone share HVAC controls.
ZONE, ID, tspHeat, tspCool, rhClg, roomList;

1 ID A (ID) (req) zone ID
2 tspHeat N (° C) (req) heating setpoint
3 tspCool N (° C) (req) cooling setpoint
A (Schedule ID) schedule cooling set point
4 rhClg N (0 – 1) .5 design indoor relative humidity
5 roomList A (at least 1 req) list of IDs of rooms that are associated with the
zone. May be ‘*ALL*’ to select all rooms.
ROOM
Specifies a single space for which loads are calculated.

ROOM, ID, cfa, height, vol, type, swing;

1 ID A (ID) (req) room ID
2 cfa N (m2) (req) conditioned floor area. Nominal room area, generally the sum of the
areas of all floor surfaces.
3 height N (m) 2.4 nominal room height
4 vol N (m3) cfa * height nominal room volume used for e.g. air change calculations.
5 type A (code) conditioned ‘master’ = master room within zone (thermostat site)
‘slave’ = slave room in master / slave zone
‘conditioned’ = maintains zone set point (no master-slave control)
‘uncond’ = temperature floats
‘attic’, ‘crawl/base’ = uncond, adjacent surfaces are appropriately
classified
6 swing N (° C) 0 allowed temperature swing above cooling set point

VENT
Specifies direct room ventilation. Vent air flow adds to infiltration and any other outdoor air assumed in
the model.
0 or 1 VENT statements are allowed in each room.
VENT, vFlow [, vACH[, vMult]];

1 vFlow N (l/s) 0 Volumetric flow rate of outdoor air introduced to room
A (sched 24-hr schedule flow
ID)
2 vACH N (ACH) 0 Additional air volume of outdoor air introduced to room. If omitted, taken as 0
A (sched 24-hr scheduled ACH
ID)
3 mult N 0 Additional multiplier. If omitted, taken as 1.
A (sched 24-hr scheduled additional multiplier
ID)
Vent air volume flow rate is calculated for each hour as follows:
Total vent flow (l/s) = (vFlow + vACH • RoomVol • 1000 / 3600) • mult
Note that air volumes are stated in terms of indoor conditions. ResHB uses constant (i.e. approximate) air
properties evaluated at indoor conditions.
GAIN
Specifies an internal gain to a room. References, scales, and schedules a GAINITEM, GAINITEMSET,
or GAINZONE (see above). An example showing the use of GAIN is below.
GAIN, gainItemID, schedMult;

1 gainItemID A (ID) (req) Reference to GAINITEM, GAINITEMSET, or
GAINZONE
2 schedMult N and/or 1 see schedMult under GAINITEM. If gainItemID
schedule ID references a GAINZONE, all 24 schedMult hours must
have the same value (that is, GAINZONEs can be
scaled but not scheduled in the referencing gain).
Any number of GAINs can be included in a ROOM. All gains are combined. Negative gains are
permitted.
If the referenced gainItemID is a GAINITEM, then the total possible gain = (gain + gainA · RoomCFA +
gainCFA · bldgCFA + gainBR · bldgNBR + gainOcc · bldgNOcc). The fractions fConv, fSWRad,
fLWRad, fLat, and fExh are then applied to the total. The fExh fraction is discarded and the other
fractions are applied as indicated in the heat balance model.
If the referenced gainItemID is a GAINITEMSET, each GAINITEM in the GAINITEMSET is treated as
above, with the application of the additional GAINITEMSET schedMult.
If the referenced gainItemID is a GAINZONE, then the treatment is identical to GAINITEMSET except

gainA multiplies by the total area of all rooms in the GAINZONE (that is, rooms containing GAINs that
reference the same GAINZONE).
The resulting total gain is then allocated in proportion to room floor area to all rooms in the GAINZONE.
GAINZONE schedMult is applied after gains are allocated to rooms. Generally the default value (1) is
used so the total gain in all rooms in the gain zone reconciles with the expected total. Also, if a
GAINZONE is referenced more than once in a room, the room will receive multiple gains.
Note that care must be taken to avoid incomprehensible multiple schedMult combinations. schedMult is
allowed at all levels (GAINITEM, GAINITEMSET, GAINZONE, and GAIN), but should be used
selectively.
Example –
SCHEDULE, OccSched, 1, 1, 1, 1, 1, 1, 1, 1,.2,.2,.2,.4,
.2,.2,.2,.5,.5, 1, 1, 1, 1, 1, 1, 1;
SCHEDULE, LightSched, 0, 0, 0, 0, 0, 0, 0,.3, 0, 0, 0, 0,
0, 0, 0, 0, 0,.2,.5,.8, 1,.7,.4,.2;
GAINITEM, Person, 130, 0, 0, 0, 0, 1, .25, 0, .33, .42;

GAINITEM, Lights, 0, 10, 0, 0, 0, 1, .2, .1, .7, 0;
GAINITEMSET, Standard, People, 4 PeopleSched, Lights, LightSched
ROOM, Test, 20, 2.4,, master, 0;

GAIN, Standard;
In this example, two GAINITEMs have been defined.

• Person, representing a single person with a gain of 130 W
• Lights, representing incandescent lighting at 10 W/m2
These GAINITEMs are referenced in ROOM “Test” via GAINITEMSET “Standard.” The resulting heat
gains are—
• People = 4 “Person” = 520 W. Hourly gains are scheduled per OccSched. The gains are applied
as 25% sensible convective, 33% long wave radiant, and 42% latent.
• Lights = 20 m2 • 10 W/m2 = 200 W, scheduled per LightsSched. The gains are 20% convective,
10% short wave radiant, and 70% long wave radiant.

SURFACE
Defines and characterizes a room surface.

SURFACE, ID, cons, area, azm, tilt, bcCode, bcAdj, shade, grflct, SWAbsO/EAC,
SWAbsI/IAC, LWAbsO, LWAbsI;
Note that only the first 5 arguments are required (ID – tilt). All remaining arguments take default values
as specified in their descriptions.
1 ID A (ID) (req) Identifies surface for error messages. Must be unique within room.
2 cons N (W/m2-K) (req) Surface U-factor (W/m2-K). If cons is a valid numeric value, the
surface is treated as having massless opaque construction. U-factor
value should include surface conductances (34 W/m2-K outside and
8.3 W/m2-K inside)
A Ref to defined Construction (surface is opaque)
(Construction
ID)
‘$REV’ A Ref to defined Construction (surface is opaque). Construction is
(Construction reversed, allowing the same construction to be referenced in adjacent
ID) rooms (such as the ceiling of conditioned space and floor of attic
above). Without this feature, the ceiling construction would have to be
defined twice with opposite layer order.
A Ref to defined Fenestration (surface is fenestration)
(Fenestration
ID)
3 area N (m2) (req) Net (actual) heat transfer area (ResHB makes no adjustment for sub-
surfaces such as fenestration).
• Opaque: generally based on outside-to-outside dimensions
(measured on building exterior, less sub-surface area)
• Fenestration: NFRC projected area (approximately equal to
rough opening)
4 azm N (deg) (req) Surface azimuth (0 = north, 90 = east, 180 = south, 270 = west). Note
actual azimuth depends on BUILDINGNORTH.
5 tilt N (deg) (req) Surface tilt (0 = horiz facing up (roof), 90 = vertical, 180 = horiz facing
down (floor)).
6 bcCode A (code) AMB Boundary condition code (see below)
7 bcAdj N or sched ID (blank) Additional data re boundary conditions (see below).
8 shade ‘none’ none Shaded fraction = 0
‘shaded’ Shaded fraction = 1
N (0 – 1) Shaded fraction for surface (beam solar multiplied by this value)
A (sched ID) Schedule specifying shaded fraction by hour
A (Ohfin ID) Ref to defined Ohfin specifying geometric shading
9 grRflct N (0 – 1) Note 1 Ground reflectivity adjacent to surface.
10 Opq: N (0 – 1) Note 2 Outside short wave absorptance.
SWAbsO
Fen: EAC N (0 – 1) 1 Exterior attenuation coefficient = incident solar reduction factor re
exterior shade screens
11 Opq: N (0 – 1) Note 2 Inside short wave absorptance. Ignored for fenestration.
SWAbsI
Fen: IAC N (0 – 1) 1 Interior attenuation coefficient (re interior shading)
12 LWAbsO N (0 – 1) Note 2 Outside long wave absorptance.
13 LWAbsI N (0 – 1) Note 2 Inside long wave absorptance.

Note 1: default is DESCONDCOOLING grRflct
Note 2: see SURFACEDEFAULTS re this default

ResHB can model a number of different surface boundary conditions, as described in the following table.
BcCode determines the air temperature, radiant temperature, solar exposure, and convective conditions
for the outside surface. Note the following:
• Only AMB surfaces receive solar radiation.
• Most of the bcCodes allow a temperature adjustment via an additional numeric or schedule input
in bcAdj. Examples:
o ..., AMB, 5, ... means “outdoor air temp = design conditions + 5” (plus solar exposure).
o ..., TARA, <schedID>, ... means “outside air temp is the same as the current room +
schedule value”.
• The codes TARA and TSRS allow reference to other rooms and surfaces. This allows, for
example, the outside temperature of a ceiling to be coupled to the inside surface of an attic.
ResHB calculates each room independently, so TARA/TSRS references do not result in true
simultaneous solutions. However, the methods are generally adequate for loads calculation
purposes. Some examples:
o “TARA” means “outside air temp is same as the current room.”
o “TARA=Living” means “outside air temp is same as in room Living.”
o “TSRS” means “outside surface temp is the same as the inside surface temp.”
o “TSRS=Ceiling” means “outside surface temp same as the inside temp of surface Ceiling
in the current room.”
o “TSRS=Attic:Floor” means “outside surface temp same as the inside temp of surface
Floor in room Attic.”
Note that all the above can be further adjusted via bcAdj.
• BcCode may be prefixed by ‘X’ to force classification of the surface as exterior in certain reports
(geometry and regression export file). For example, by default an ADB surface is included in
partition totals in those reports. ‘X ADB’ would change that classification to exterior. Note that
the X prefix has no effect on building modeling, just reporting.
Inside surface conditions. All surface inside faces are coupled convectively to room air and radiantly to
all visible surfaces (having different tilt and/or azimuth). The sole exception is bcCode = IM, which is
coupled radiantly to all surfaces.

bcCode Description bcAdj Outside conditions
AMB Ambient N (° C) or • Air temp per design conditions + bcAdj
sched ID • Convective coeff = outside
• Radiant surround = combined sky temp and air
temp
• Incident solar per exposure
AMS Sheltered ambient N (° C) or • Air temp per design conditions + bcAdj
sched ID • Convective coeff = inside (no wind)
• Radiant surround at air temp
• Incident solar = 0
TASC Air temp scheduled N (° C) or • Air temp = bcAdj
sched ID • Convective coeff = inside
TARA Air temp same as this N (° C) or • Air temp same as ref’d room + adj
or TARA=<roomID> or ref’d room sched ID • Convective coeff = inside
Blank roomID = this room
TSG Surface temp = N (° C) or • Surf temp = design conditions ground temp +
ground temp sched ID adj
TSSC Surface temp N (° C) or • Surf temp = bcAdj
scheduled sched ID
TSRS Surf temp same as N (° C) or • Surf temp = inside temp of ref’d surface + adj
or TSRS= ref’d surf in another sched ID Blank roomID = this room
<roomID>:<surfID> room Blank roomID and blank surfID = this surface
QSC Scheduled heat flow N (W/m2) or • Specified heat flow (for testing)
sched ID
ADB Adiabatic (unused) No heat transfer
ADG Adiabatic ground (unused) Equivalent to ADB except causes surface to be
categorized as adjacent to ground
IM Adiabatic internal (unused) No heat transfer at outside surface. Inside face
mass coupled radiantly to all surfaces (see above).

ATTIC
Generates an unconditioned room of type ‘attic’, including all required surfaces and internal mass.
Surface defaults are used for all values for which there is no input. If finer control is required, the ROOM
and SURFACE commands should be used.
ATTIC, ID, config, l, w, ht, azm, ACH, TRBelow, ceilCons, rfCons, rfSWAbsO, rfLWAbsI;

1 ID A (ID) (req) ID of attic
2 config A (req) attic configuration. Only option is ‘hip’, which
generates a hip roof with a rectangular plan.
3 l N (m) (req) length
4 w N (m) (req) width
5 ht N (m) 1.5 ridge height
6 azm N (°, 0 = north) 0 azimuth of attic ‘front’ = long side
7 ACH N (ach) 4 air change rate
8 TRBelow N (° C) or 24 assumed air temperature in conditioned room(s) below
schedule ID ceiling
9 ceilCons N or A (req) ceiling construction, see SURFACE cons. Note that
$REV is useful here
10 rfCons N or A (req) roof construction, see SURFACE cons.
11 rfSWAbsO N Note 1 roof outside surface short wave (solar) absorptance
12 rfLwAbsI N Note 1 roof inside surface long wave absorptance. This value
may be altered to represent radiant barriers.
Note 1: see SURFACEDEFAULTS re this default

The floor surface (that is, the house ceiling) generated by ATTIC has ID ‘Flr’. Thus, to couple to an attic,
the boundary conditions for the ceiling(s) of adjacent room(s) should be set to TSRS=<AtticID>:Flr.

CRAWLBASE
Generates an unconditioned room of type ‘crawlbase’ (crawlspace or basement) including all required
surfaces. Surface defaults are used for all values for which there is no input. If finer control is required,
the ROOM and SURFACE commands should be used.
CRAWLBASE, ID, config, l, w, ht, azm, ACH, TRAbove, HsFlrCons, WallCons, CBFlrCons;

1 ID A (ID) (req) ID of crawlspace / basement
2 config A (req) Configuration. Only option is ‘CS_vnt’, which generates
a ventilated crawlspace.
3 l N (m) (req) length
4 w N (m) (req) width
5 ht N (m) 1.5 height
6 azm N (°, 0 = north) 0 azimuth of ‘front’ surface = long side
7 ACH N (ach) 4 air change rate
8 TRAbove N (° C) or 24 assumed air temperature in conditioned room(s) below
schedule ID ceiling
9 HsFlrCons N or A (req) house floor construction (= crawlspace / basement
ceiling construction), see SURFACE cons. Note that
$REV is useful here
10 WallCons N or A (req) crawlspace / basement above grade wall construction
11 CBFlrCons N or A (req) crawlspace / basement floor construction
The ceiling surface (that is, the house floor) generated by CRAWLBASE has ID ‘HsFlr’. Thus, to couple
to an crawlspace / basement, the boundary conditions for the floor(s) of adjacent room(s) should be set to
TSRS=<cbID>:HsFlr.

Appendix D. RHBGen Documentation
General Input Info

See ResHB documentation
Command line and files

RHBGen is a 32 bit Windows console application. It invoked from the command line as follows --
RHBGen runID [options]
runID is the base name for input and output files for the run (see below).
Options can be any combination of the following --
Effect
Option
-I<dirList> Include directories. Specifies a ;-delimited list of directories that RHBGen searches for include
files (the current directory is searched first).
Example: -Iinc1;c:\inc2
Exit code
RHBGen returns an exit code indicating the outcome of processing --
Exit code Meaning
0 Success, no errors: all cases generated and run without error
1 Internal program error
2 – 19 Input or file generation error(s), review RHBGen.log.
20 All cases successfully generated, but ResHB detected errors and at least one case was not
successfully run. Review RHBGen.log and/or ResHB.log.
Files
RHBGen reads or writes the following files.
File name Use Description
<runID>.rhg Input, must exist Input file.
RHBGen.log Output, append (created if does not exist) Log file. Information is appended to this file
documenting each run, showing date/time of run,
general information about the run, and any error
messages.
RHBGen.err Output, append (created if does not exist) Error file. Only written if errors are detected.
Can serve as an error flag – delete before a
group of runs, if exists at end, there has been at
least one error.
<outDir>\<runID>.scl Output, replace (overwritten if exists) Successful case list. Lists parametric codes for
all cases successfully generated and run (or just
successfully generated if no RUNCASE)

Example
The following example runs 72 cases based on prototype ‘S’ (3 cases in case set, 6 locations, and 4
orientations).
File “Example.rhg” –
! Example.rhg
GENDIR, Gen, "*.rhb;*.bad;*.rep;*.res;*.cas;*.err;*.gxf";
RUNCASE, "c:\1199rp\resHB\run\ResHBD", "-Ic:\1199rp\resHB\run\hbIncludes";
RUNSET, 0, TestSet,,,
L=DCWJ|AZPH|FLKW|CAPR|TNME|CASD,
O=000|090|180|270;
!================================================================
! CaseSets
CASESET, TestSet,
R2HSLCOALO000QmEm00AmmmmCwmhamWwmmGx110llllPwxMwmFwmax,
R2HSLCOALO000QmEm00AmmmmCwmhtmWwmmGx110mmmmPwxMwmFwmax,
R2HSLCOALO000QmEm00AmmmmCwmhamWwmmGx11mhhhhPwxMwmFwmax;
! Prototypes
INCLUDE, "Prototypes.rhg";
! Example.rhg end
File “Prototypes.rhg”
! Prototypes.rhg -- RHBGen prototype geometry definitions
! PROTOTYPE: 5 x 10 m single room, 1 exposed side w/5 m2 fen + 5 m2 skylight, 1 story

PROTOTYPE, PROTOTYPE-M, MFUnit, 2.5;
PRROOM, ROOM1, 50, 2.5;
PRSURF, Ceil-01, IC, 50, 0, 0, 5;
PRSURF, Wall-01, XW, 12.5, 0, 90, 5;
PRSURF, Wall-02, IW, 25, 90, 90, 0;
PRSURF, Wall-03, IW, 12.5, 180, 90, 0;
PRSURF, Wall-04, IW, 25, 270, 90, 0;
PRSURF, Flor-01, IF, 50, 0, 180, 0;
! PROTOTYPE: 5 x 10 m single room, 5 m2 fen on all facades, 1 story

PROTOTYPE, PROTOTYPE-B, Box, 2.5;
PRSURF, Ceil-01, XC, 50, 0, 0, 5;

PRSURF, Wall-01, XW, 25, 0, 90, 5;
PRSURF, Wall-02, XW, 12.5, 90, 90, 5;
PRSURF, Wall-03, XW, 25, 180, 90, 5;
PRSURF, Wall-04, XW, 12.5, 270, 90, 5;
PRSURF, Flor-01, XF, 50, 0, 180, 0;
!--------------------------------------------------
! PROTOTYPE: 5 x 10 m single room, 5 m2 fen on all facades, 3 stories

PROTOTYPE, PROTOTYPE-C, Box, 7.5;
PRROOM, ROOM1, 50, 2.5, , GZAvg;

PRSURF, Ceil-01, XC, 50, 0, 0, 5;
PRSURF, Wall-01, XW, 25, 0, 90, 5;
PRSURF, Wall-02, XW, 12.5, 90, 90, 5;
PRSURF, Wall-03, XW, 25, 180, 90, 5;
PRSURF, Wall-04, XW, 12.5, 270, 90, 5;
PRSURF, Flor-01, XF, 50, 0, 180, 0;
!--------------------------------------------------
! PROTOTYPE: 300 m2 single room, 15 m2 fen on all facades, 2 stories

PROTOTYPE, PROTOTYPE-D, Box, 5;
PRROOM, ROOM1, 300, 5, , GZAvg;
PRSURF, Ceil-01, XC,150, 0, 0, 2;

PRSURF, Wall-01, XW, 50, 0, 90,15;
PRSURF, Wall-02, XW, 75, 90, 90,15;
PRSURF, Wall-03, XW, 50, 180, 90,15;
PRSURF, Wall-04, XW, 75, 270, 90,15;
PRSURF, Flor-01, XF,150, 0, 180, 0;
!--------------------------------------------------
! PROTOTYPE: 5 x 5 m single room, 1 wall exposed, 5 m2 fen

PROTOTYPE, PROTOTYPE-R, Room;
PRSURF, Ceil-01, XC, 25, 0, 0, 0;

PRSURF, Wall-01, XW, 12.5, 0, 90, 5;
PRSURF, Wall-02, IW, 12.5, 90, 90, 0;
PRSURF, Wall-03, IW, 12.5, 180, 90, 0;
PRSURF, Wall-04, IW, 12.5, 270, 90, 0;
PRSURF, Flor-01, XF, 50, 0, 180, 0;
!--------------------------------------------------
! PROTOTYPE: 2 room "shoebox"

PROTOTYPE, PROTOTYPE-S, ShoeBox;
PRROOM, Great, 48, 2.4, , GZAvg;

PRSURF, GC, XC, 48.0, 0, 0, 24.0;
PRSURF, GW-F, XW, 14.4, 0, 90, 0;
PRSURF, GW-L, IW, 19.2, 90, 90, 0;
PRSURF, GW-B, XW, 14.4, 180, 90, 7.20;
PRSURF, GW-R, XW, 19.2, 270, 90, 9.60;
PRSURF, GF, XF, 48.0, 0, 180, 0;
PRROOM, Living, 48, 2.4, , GZAvg;

PRSURF, LC, XC, 48.0, 0, 0, 0;
PRSURF, LW-F, XW, 14.4, 0, 90, 7.2;
PRSURF, LW-L, XW, 19.2, 90, 90, 9.6;
PRSURF, LW-B, XW, 14.4, 180, 90, 0;
PRSURF, LW-R, IW, 19.2, 270, 90, 0;
PRSURF, LF, XF, 48.0, 0, 180, 0;
!---------------------
! PROTOTYPE: Manual J comparison

PROTOTYPE, PROTOTYPE-J, "Manual-J";
PRROOM, Room1, 162.58, 2.44, 396.44;
PRSURF, CLG, XC, 162.58, 0, 0, 0;
PRSURF, WF, XW, 37.16, 0, 90, 10;
PRSURF, WL, XW, 26.01, 90, 90, 10;
PRSURF, WB, XW, 37.16, 180, 90, 10;
PRSURF, WR, XW, 26.01, 270, 90, 10;
PRSURF, FLR, XF, 162.58, 0, 180, 0;
! PRSURF, "DOOR-01", 11A0, 1.86
!-----------------------------------------------------

! PROTOTYPE: 4 room square for RLF regressions
! 100 m2, fen = 16% of floor area evenly distributed
PROTOTYPE, PROTOTYPE-Q, Square, 2.4;
PRROOM, FL, 25, 2.4,, GZAvg;

PRSURF, GC, XC, 25, 0, 0, 2;
PRSURF, GW-F, XW, 12, 0, 90, 2;
PRSURF, GW-L, XW, 12, 90, 90, 2;
PRSURF, GW-B, IW, 12, 180, 90, 0;
PRSURF, GW-R, IW, 12, 270, 90, 0;
PRSURF, GF, XF, 25, 0, 180, 0;
PRROOM, BL, 25, 2.4,, GZAvg;

PRSURF, GC, XC, 25, 0, 0, 2;
PRSURF, GW-F, IW, 12, 0, 90, 0;
PRSURF, GW-L, XW, 12, 90, 90, 2;
PRSURF, GW-B, XW, 12, 180, 90, 2;
PRSURF, GW-R, IW, 12, 270, 90, 0;
PRSURF, GF, XF, 25, 0, 180, 0;
PRROOM, BR, 25, 2.4,, GZAvg;

PRSURF, GC, XC, 25, 0, 0, 2;
PRSURF, GW-F, IW, 12, 0, 90, 0;
PRSURF, GW-L, IW, 12, 90, 90, 0;
PRSURF, GW-B, XW, 12, 180, 90, 2;
PRSURF, GW-R, XW, 12, 270, 90, 2;
PRSURF, GF, XF, 25, 0, 180, 0;
PRROOM, FR, 25, 2.4,, GZAvg;

PRSURF, GC, XC, 25, 0, 0, 2;
PRSURF, GW-F, XW, 12, 0, 90, 2;
PRSURF, GW-L, IW, 12, 90, 90, 0;
PRSURF, GW-B, IW, 12, 180, 90, 0;
PRSURF, GW-R, XW, 12, 270, 90, 2;
PRSURF, GF, XF, 25, 0, 180, 0;
!------------------------------------------------------
! PROTOTYPE: 4 room square for RLF regressions (revised)

! Conditioned floor area = 168 m2 (1808 ft2)
! Fen = ~16% of floor area evenly distributed
! Skylight = ~2% of floor area
! Exterior wall same all facades but based on rectangular plan
! width = 8.5 m (28 ft)
! length = 168 / 8.5 = 20 m (approx)
! perim = 2 x (20 + 8.5) = 57 m
! ext wall area = 57 x 2.5 = 142.5 m2
! ext wall per room facade = 142.5 / 8 = 17.8 m2
! Interior wall area = 83% of floor area (estimate based on discussion with
! Joe Huang)
! 168 x .83 = 140 m2 (approx)
! must be entered 1 sided -> total = 280 m2
! int wall per room facade = 280 / 8 = 35 m2
! 3 bedrooms, 4 occupants; standard gains allocated to all rooms
! Infil height = 2.5 m
PROTOTYPE, PROTOTYPE-X, RLF base, 2.5, 3, 4;
PRROOM, FL, 42, 2.5,, GZAvg;

PRSURF, GC, XC, 42, 0, 0, 0.84;
PRSURF, GW-F, XW, 17.8, 0, 90, 3.4;
PRSURF, GW-L, XW, 17.8, 90, 90, 3.4;
PRSURF, GW-B, IW, 35, 180, 90, 0;
PRSURF, GW-R, IW, 35, 270, 90, 0;

PRSURF, GF, XF, 42, 0, 180, 0;
PRROOM, BL, 42, 2.5,, GZAvg;

PRSURF, GC, XC, 42, 0, 0, 0.84;
PRSURF, GW-F, IW, 35, 0, 90, 0;
PRSURF, GW-L, XW, 17.8, 90, 90, 3.4;
PRSURF, GW-B, XW, 17.8, 180, 90, 3.4;
PRSURF, GW-R, IW, 35, 270, 90, 0;
PRSURF, GF, XF, 42, 0, 180, 0;
PRROOM, BR, 42, 2.5,, GZAvg;

PRSURF, GC, XC, 42, 0, 0, 0.84;
PRSURF, GW-F, IW, 35, 0, 90, 0;
PRSURF, GW-L, IW, 35, 90, 90, 0;
PRSURF, GW-B, XW, 17.8, 180, 90, 3.4;
PRSURF, GW-R, XW, 17.8, 270, 90, 3.4;
PRSURF, GF, XF, 42, 0, 180, 0;
PRROOM, FR, 42, 2.5,, GZAvg;

PRSURF, GC, XC, 42, 0, 0, 0.84;
PRSURF, GW-F, XW, 17.8, 0, 90, 3.4;
PRSURF, GW-L, IW, 35, 90, 90, 0;
PRSURF, GW-B, IW, 35, 180, 90, 0;
PRSURF, GW-R, XW, 17.8, 270, 90, 3.4;
PRSURF, GF, XF, 42, 0, 180, 0;
!---------------------
! Prototypes.rhg end

Commands
GENDIR
Specifies output location for generated files.

GENDIR, genDir, delFiles;

1 genDir A (req) Path to directory where ResHB files are to be generated and
optionally run.
2 delFiles A (blank = no files Semi-colon delimited list of files or wildcards to be deleted prior
deleted) to run. Note that since this list is delimited by semi-colons, it
must generally be quoted (“*.rhb;*.err”)
RUNCASE
Specifies application for analysis of each generated case. Optional, if omitted, no program is run.
If present, RHBGen executes the following command for each successfully generated file:
<runCmd> <runID> <runArgs> <caseArgs>
RUNCASE, runCmd, runArgs;

1 runCmd A (blank) Command to be invoked for each generated case. Should include
path to executable as required. Example: “c:\LoadsDir\ResHB”
2 runArgs A (blank) Additional arguments supplied to runCmd.
CASESET
Defines a set cases (parametric codes) for reference in RUNSET.

CASESET, ID, ParCode1, ParCode2, ...;

1 ID A (ID) (req) ID of this CASESET
2- parCode A Parametric code (up to 50 can be included)

RUNSET
Specifies cases to be generated and run. Any number of RUNSETs can be included.
RHBGen is driven by a list of parametric codes that are optionally modified via variants.
RUNSET, control, caseSetFile, caseTag, caseResPath, caseArgs [, caseVar(s)];

1 control N 0 Controls use of this RUNSET
• -1: disable, no cases run
• 0: run all variant combinations
• >0: repeat run <control> times, with randomly selected
combinations from among available variants
2 caseSetfile A (req) ID of a CASESET or path to case set file. Provides base set of
parametric run codes (see below).
3 caseTag A (blank) Text passed to the ResHB DOCUMENTATION command
4 caseArgs A (blank) Optional args passed to RunCmd (see RUNCASE). A common
option is –C to request results to be written to a case results file.
5- caseVars A (no Zero or more variants to be applied to each base parametric code.
caseVars) See below.
The base set of parametric codes used for a RUNSET are defined by the referenced CASESET or read
from a case file. Parametric codes are described in a later section of this document.
Case sets or case files contain lists of parametric codes to be run. These files are external to the main
input file to allow sets of cases to be generated using automated schemes. The format is one parametric
code per line, with the following rules –
• Blank lines are ignored
• Any characters after ‘!’ are ignored (for comments)
• A line consisting of “*STOP*” terminates processing. This is useful for testing.
Variants allow generation of multiple cases from a single base parametric code.
A variant is a parametric code prefix followed by a list of alternative values for that prefix field. For
example, O=000|090|180|270 means “generate/run each case at the 4 orientations 0, 90, 180, 270.” If
multiple caseVars are specified, all combinations are generated and run (assuming control = 0). Rules for
variants –
• Any position containing ‘*’ retains the original parametric code character from the case file.
• Any position containing ‘~’ retains the current parametric code character. This feature allows
nested combinations within the same field. For example, O=1|2, O=~10|~20 would generate/run
cases with orientation 110, 120, 210, and 220.
• The variant code may be shorter than the associated parametric code field (but not longer). Short
variants are implicitly padded with ‘*’s to the required length.
• Orientation variants have a special case: 0=999 causes an orientation value 0 – 359 to be
randomly selected.
Example 1 --
RUNSET, 0, myCases.gfl, , , , O=000|090|180|270, W=*l|*m|*h;

This command generates and runs 12 cases for each line in the case file (4 orientations times 3 wall
insulation levels). Wall construction is not modified (due to the “*”)
Example 2 --
RUNSET, 100, myCases.gfl, , , , O=999, W=*l|*m|*h;
This command generates and runs 100 cases with random orientation (1-360) and random wall insulation.
PROTOTYPE
Along with PRROOM and PRSURF, provides basic geometry for prototypes used to generate ResHB
input files. The prototype used is selected via the Hx field in the parametric code, where ‘x’ is the single
letter designation of the prototype
Any number of prototypes can be defined.
PROTOTYPE, name, comment, hInfil, NBR, NOcc;

1 ID A (ID) (req) Prototype name. Must be “Prototype-X”, where X is the single
parametric code character that selects a prototype (in the H field)
2 comment A (blank) Text written to the ResHB input file as a comment
3 hInfil N (m) 2.5 Prototype height for AIM-2 infiltration model
4 NBR N 2 Number of bedrooms in prototype
5 NOcc N NBR + 1 Number of occupants in prototype
PRROOM
Specifies the geometry of a room that belongs to the current prototype. These values are passed to
ResHB, see ResHB documentation for exact definitions.
PRROOM, name, cfa, height, vol, gainzone;

1 ID A (ID) (req) Room ID
2 cfa N (m2) (req) Conditioned floor area
3 height N (m) (req) Height
4 vol N (m3) cfa*height Volume
5 gainzone A (blank = no Internal gain zone to which this room belongs, typically ‘Living’ or
gainzone ‘Sleeping’

PRSURF
Specifies the geometry of a surface belonging to the current room.

PRSURF, ID, type, area, azm, tilt, fenArea;

1 ID A (ID) (req) Surface ID
2 type A (code) (req) Type code
• XC / XW / XF = exterior ceiling / wall / floor
• IC / IW / IF = interior ceiling / wall / floor
3 area A (m2) (req) Surface gross area
4 azm N (°) (req) Surface azimuth (0 = north or as rotated)
5 tilt N (°) (req) Surface tilt
6 fenArea N (m2) 0 Fenestration area within surface

Parametric Code
RrHmLstccOxxxQgitsxAwiflCifraWciaGtietfrblPwfMcrFcifx
Code Description Value Range

Rr Revision r Code revision. Current = 2
Hm House model m House prototype
Lstcc Location st State abbreviations
cc City code
Oxxx Orientation xxx 000 - 360: orientation of building front
Qgitsx Gains, infiltration, and g internal gain: 0: 0, l: .5 std, m: 1 std, h: 2 std; A-I: parametric
controls variation of intgain components
i Infiltration code: 0: 0, l: .1 ACH, m: .4 ACH, h: 1 ACH; A-J: AIM2
model w/ specified leakage class
t cooling setpoint. l: 21 C, m: 24 C, h: 27 C
s Room temp swing. 0: 0 C, l: 0.83 C, m: 1.67 C, h: 2.5 C
x control options: 0: sp as specified, 1: sp reduced by .5 swing, 2:
sp reduced by swing
Awifl Interior surface w Wall / partition / ceiling SW absorp: 0lmhdb1 (see note)
absorptivity i Internal mass SW absorp: 0lmhdb1 (see note)
f Floor SW absorp: 0lmhdb1 (see note)
l LW absorp (all surfaces): 0: 0, l: .2, m: .9 (fen .84), h: .95, 1: 1
Ccifra Exterior ceiling / roof c Ceiling construction: 0: massless; w: wood frame, c: solid
concrete
i Ceiling insulation (l, m, h)
f Configuration: x: ceiling only (no roof), f: flat; h: hip
r Roof construction: a: asphalt, t: tile
a Exterior surface SW abs: 0lmhdb1 (see note)
Wcia Exterior wall c Construction. 0: massless, w: wood frame, c: solid concrete
i Insulation (l, m, h)
a Exterior surface SW abs: 0lmhdb1 (see note)
Gtietfrbl Fenestration t Base fenestration type: 1, 2, e, x, 3
i IAC: 0: 0, l: .4, m: .6, h: .8, 1: 1
e EAC: 0: 0, l: .4, m: .6, h: .8, 1: 1
tflbr Fen area multiplier (adjusts prototype fen area)
0: 0, l: .5, m: 1, h: 2; A-C: area · 2, IAC = l, m, h
Pwf Interior partition w Partition construction: w: wood frame
f Interior floor/ceiling construction: w: wood frame
Mcr Interior mass c Construction type: w: wood
r Surface area ratio (fraction of room cfa)
0: 0, l: .5, m: 1, h: 2
Fcifx Exterior floor c Construction. 0: massless, w: woodframe, c: concrete, s: slab
i Insulation (l, m, h)
f Boundary conditions: 0: adiabatic, a: ambient, c: crawlspace
x Crawlspace wall insulation (l,.m, h)
Notes
• SW absorptance codes: 0: 0, l: .3, m: .6, h: .8, d: .85, b: .95, 1: 1
• All codes case sensitive, however upper and lower case variants are not used in same code position
because this would result in non-unique generated file names (file name are not case sensitive)
• Floor exterior SW abs set to same value as wall (required only for exposed floor)
• Possible additional variable = wall shading (handle with absorp?)

Roof constructions
Config code ‘f’ Attic Roof cons code ‘r’ Type Constructions
h hip attic, a Asphalt ROOF-ASPH
4/12 roof t Tile ROOF-TILE
all others (no roof)
Ceiling constructions
Config Boundary Cons code Type Insul codes ‘i’ Constructions
code ‘f’ conditions ‘c’
x AMB 0 Massless l, m, h MASSLESS-L, -M, -H
w Wood frame l, m, h CEIL-WL, -WM, -WH
c Concrete l, m, h CONC-L, -M, -H
f AMB 0 Massless l, m, h MASSLESS-L, -M, -H
w Wood frame l, m, h CLRF-WL, -WM, -WH
h Attic 0 Massless l, m, h MASSLESS-L, -M, -H
w Wood frame l, m, h CEIL-WL, -WM, -WH
Note: MASSLESS resistances match corresponding CONC
Wall constructions
Cons code ‘c’ Type Insul codes ‘i’ Constructions
0 Massless l, m, h MASSLESS-L, -M, -H
w Wood frame l, m, h WALL-WL, -WM, -WH
Fenestration
Type code ‘t’ Glazing Frame ASHRAE NFRC U-factor SHGC
ID W/m2-K
1 Single clear none 1a 5.91 .86
2 Double clear none 5a 2.73 .76
e Double low-e high none 17c 1.99 .70
solar
x Double low-e low none 25a 1.70 .41
solar
3 Triple clear none 29a 1.76 .68
Partition wall constructions

Cons code ‘w’ Type Insul codes Construction
w Wood frame -- IWALL-W (1/2 thickness, adiabatic)

Partition floor/ceiling constructions
Cons code ‘f’ Type Insul codes Construction
w Wood frame -- ICEIL-W, IFLOR-W (1/2 thickness, adiabatic)
Interior mass constructions

Cons code ‘c’ Type Insul codes ‘i’ Constructions
w Wood -- IMASS-W
Exterior floor constructions

Cons code Type Insul codes ‘i’ Constructions
0 Massless l, m, h MASSLESS-L, -M, -H
w Wood frame l, m, h FLOR-WL, -WM, -WH
s Slab -- FLOR-S (adiabatic exterior)
!=============================================================================
! CONSTRUCTIONS
!-----------------------------------------------------------------------------
! wood frame sub-constructions (10% and 25% framing factor)
! thickness adjusted when ref'd
Construction, CONC-L, PP, ConcHvy 200;

Construction, CONC-M, PP, FoamBoard 25, ConcHvy 200;
Construction, CONC-H, PP, FoamBoard 200, ConcHvy 200;
! massless constructions (same overall conductance as CONC)

Construction, MASSLESS-L, PP, .0909;
Construction, MASSLESS-M, PP, .953;
Construction, MASSLESS-H, PP, 6.98767;
Construction, CEIL-WL, PP, WF10Empty 190, GypBoard;

Construction, CEIL-WM, PP, WF10FbrGl 190, GypBoard;
Construction, CEIL-WH, PP, FbrGlsBatt 140, WF10FbrGl 190, GypBoard;
Construction, CLRF-WL, PP, AsphaltShingle 9, Plywood 20, CEIL-WL;

Construction, CLRF-WM, PP, AsphaltShingle 9, Plywood 20, CEIL-WM;
Construction, CLRF-WH, PP, AsphaltShingle 9, Plywood 20, CEIL-WH;
Construction, ROOF-ASPH, PP, AsphaltShingle 9, Plywood 20;

Construction, ROOF-TILE, PP, ConcHvy 50, .2, Plywood 20;
Construction, WALL-WL, PP, Plywood 20, WF25Empty 90, Gypboard;

Construction, WALL-WM, PP, Plywood 20, WF25FbrGl 90, Gypboard;
Construction, WALL-WH, PP, Plywood 20, FoamBoard 25, WF25FbrGl 140, Gypboard;
! Interior walls (1/2 thickness)

Construction, IWALL-W, PP, WF25Empty 45, Gypboard;
! Internal mass

Construction, IMASS-W, PP, FrameWood 12;
! Floor
Construction, FLOR-WL, PP, WF10Empty 190, Plywood 19, OakFloor 8;
Construction, FLOR-WM, PP, WF10FbrGl 190, Plywood 19, OakFloor 8;
Construction, FLOR-WH, PP, FoamBoard 50, WF10FbrGl 190, Plywood 19, OakFloor 8;
Construction, FLOR-S, PP, Soil 100, ConcHvy 100;
!=============================================================================
! MATERIALS
!-----------------------------------------------------------------------------
! k rho Cp R dflt L
! W/m-K kg/m3 kJ/kg-K m2-K/W mm
Material, Empty, 0, 0, 0, .18;
Material, R1000, 0, 0, 0, 1000;
Material, FbrGlsBatt, .039, 19, .96, 0;
Material, FoamBoard, .029, 30, 1.2, 0;
Material, FrameWood, .12, 500, 1.6, 0;
Material, Plywood, .14, 550, 1.5, 0;
Material, OakFloor, .17, 720, 1.65, 0, 8;
Material, GypBoard, .16, 800, 1.09, 0, 13;
Material, ConcHvy, 2.2, 2240, .9, 0;'
Material, Brick, 0.9, 2000, .84, 0, 90;
Material, CMUMed, 1.3, 1840, .84, 0;
Material, Soil, 1, 1900, 1, 0;
Material, AsphaltShingle, .12, 1100, 1.26, 0;
Material, Carpet, .06, 300, 1.40, 0, 13;

Executive Summary

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Executive Summary

Uploaded by

Copyright:

Available Formats

Updating the ASHRAE/ACCA

Residential Heating and Cooling Load

Charles S. Barnaby Jeffrey D. Spitler

American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

1199-RP Final Report August 20, 2004 i

1199-RP Final Report August 20, 2004 ii

1199-RP Final Report August 20, 2004 iii

1199-RP Final Report August 20, 2004 iv

1199-RP Final Report August 20, 2004 v

1199-RP Final Report August 20, 2004 vi

1199-RP Final Report August 20, 2004 vii

Planning Develop/refine RHB

Develop evaluation methods

Develop RLF method

Final deliverables: Reports, Done

Figure 1. 1199-RP project process

1199-RP Final Report August 20, 2004 1

Background and RHB Description

1199-RP Final Report August 20, 2004 2

Master / slave control, temperature swing, and cooling load

1199-RP Final Report August 20, 2004 3

Temperature swing (°F)

Figure 2. Sensible cooling load reduction due to temperature swing

Heat balance for residential applications

1199-RP Final Report August 20, 2004 4

1199-RP Final Report August 20, 2004 5

The fundamental RHB load calculation sequence is:

1199-RP Final Report August 20, 2004 6

1199-RP Final Report August 20, 2004 7

Room air temperature (°F)

Figure 3. Master / slave room temperatures with similar load profiles

1199-RP Final Report August 20, 2004 8

Room air temperature (°F)

Figure 4. Master / slave room temperatures with mismatched load profiles

Inside surface convection coefficients

1199-RP Final Report August 20, 2004 9

Elevation effects on convective heat transfer

1199-RP Final Report August 20, 2004 10

Figure 5. Effect of pressure on convection coefficient (see definitions below)

1199-RP Final Report August 20, 2004 11

Figure 6. Roof surface and attic air temperatures predicted by ResHB

1199-RP Final Report August 20, 2004 12

Fenestration and solar gain distribution

Ground heat transfer

1199-RP Final Report August 20, 2004 13

1199-RP Final Report August 20, 2004 14

1199-RP Final Report August 20, 2004 15

Figure 8. Effect of internal mass, wood floor construction

Figure 9. Effect of internal mass, slab floor construction

1199-RP Final Report August 20, 2004 16

1199-RP Final Report August 20, 2004 17

1199-RP Final Report August 20, 2004 18

1199-RP Final Report August 20, 2004 19

1199-RP Final Report August 20, 2004 20

Combined testing process

1199-RP Final Report August 20, 2004 21

Variants RHBGen RHB RHB

ESP-r input generator

Figure 10. Overall inter-model testing process

RHBGen parametric generator

ESP-r input generator

1199-RP Final Report August 20, 2004 22

1199-RP Final Report August 20, 2004 23

ESP-r test procedure