You are on page 1of 199

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274048785

Energy Efficiency and Certification of Central
Air Conditioners

Technical Report · April 2003

CITATIONS READS

6 452

24 authors, including:

Philippe Rivière Luis Pérez-Lombard
Mines ParisTech, PSL Research University Universidad de Sevilla
63 PUBLICATIONS 158 CITATIONS 17 PUBLICATIONS 1,668 CITATIONS

SEE PROFILE SEE PROFILE

José Ortiz Roger Hitchin
Building Research Establishment BRE Building Research Establishment BRE
9 PUBLICATIONS 1,632 CITATIONS 39 PUBLICATIONS 173 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The inspection of building services through continuous monitoring and benchmarking –The
iSERVcmb project View project

IAQ & VENTILATION IN HEALTHCARE FACILITIES View project

All content following this page was uploaded by Philippe Rivière on 25 March 2015.

The user has requested enhancement of the downloaded file.

Study for the D.G. Transportation-Energy (DGTREN) of the Commission of the
E.U.

Energy Efficiency and Certification of Central
Air Conditioners (EECCAC)
FINAL REPORT - APRIL 2003

Contract DGXVII-4.1031/P/00-009
CO-ORDINATOR: Jérôme ADNOT, ARMINES, France
Assisted by Paul WAIDE
PW Consulting, UK

PARTICIPANTS
Jérôme ADNOT, Philippe RIVIERE, Dominique MARCHIO,
Martin HOLMSTROM, Johan NAESLUND, Julie SABA
Centre d’Energétique, Ecole des Mines de Paris, France
Sule BECIRSPAHIC
Eurovent Certification
Carlos LOPES
ADENE-CCE, Portugal
Isabel BLANCO
IDAE, Spain
Luis PEREZ-LOMBARD, Jose ORTIZ
AICIA, Spain
Nantia PAPAKONSTANTINOU, Paris DOUKAS
University of Athens, Greece
Cesare M. JOPPOLO
Politecnico di Milano, Italy
Carmine CASALE
AICARR, Italy
Georg BENKE
EVA, Austria
Dominique GIRAUD
INESTENE, France
Nicolas HOUDANT
Energie Demain, France
Philippe RIVIERE, Frank COLOMINES
Electricité de France
Robert GAVRILIUC, Razvan POPESCU, Sorin BURCHIU
UTCB, Bucharest
Bruno GEORGES
ITF, France
Roger HITCHIN
BRE, UK
With the additional participation of experts from Eurovent Cecomaf

© 2003 ARMINES

ARMINES
60, bd St Michel
75272 Paris Cedex 06
France

Tel: (+33) 1 40 51 91 74 Fax: (+33) 1 46 34 24 91
E-mail: jerome.adnot@ensmp.fr

All rights reserved, including that of translation into other languages. No part of this publication may be reproduced or
transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or any
information storage and retrieval system, without permission in writing from ARMINES.

Editorial content: Although great care has been taken in compiling and checking the information given in this
publication to ensure that it is accurate, ARMINES shall not be held responsible for the continued currency of the
information or for any errors, omissions or inaccuracies in this publication.

................................ ............................... Importance of AC for human health and productivity performance..... 37 2...................................................................................................................................................................................................... 44 Terminal units and other peripheral equipment used............................................................................................................... 26 Impact of load reduction on the efficiency – a reporting format proposed to Eurovent .......................................... ............................................................................................................................................. 23 Packaged units can also be improved a lot ............. 38 2................................... 11 SUMMARY OF RESULTS ........................................... 13 CAC market and stock data have been assembled for the first time............ 31 Scenarios for energy efficiency have been established and quantified ... 23 Optimisation of a chiller for its least LCC........ 36 Selection of technical experts ......................................................................2. 44 General classification of systems based on chillers ............ CENTRAL AIR-CONDITIONERS IN EUROPE: DEFINITIONS AND BASIC DATA ............. 16 All the elements of a possible grading of Cooling market have been assembled ......................................................... INTRODUCTION ..................... 11 Definitions of all CAC systems found on the EU market have been given.....4 Description of systems not using chillers ..................................... 30 The newly proposed ESEER method allows grading and ranking of chillers by order of merit .................. Ventilation: our assumptions for the study .............................................................. 49 .................. 39 RAC and CAC in competition ................................................................................................................................. 39 2....... 19 Splits and Packages are grouped in one single category ( ...................................................................... 47 Number of water loops connected with the chiller .. Description of other aspects of systems .................. 38 What is "comfort"? .........................................................................................................1................................ 49 2........................................................................................... Basic definitions ................ 19 The impacts of BAU have been assessed ................................................................................................................... 28 EECCAC final figures for a European SEER method (ESEER) ...................................................................................................................................3............................................................. 33 1.............. 29 EER alone is a poor selection tool ...CONTENTS ABSTRACT ................................................................................................ 39 Basic Thermodynamics at one instant ...................... link with ventilation ....................................................... 25 Part load performance has been quantified for the first time and the methods have been tested ........................................................................................... manufacturers and national experts ................................................................ 12 All CAC equipment test standards have been reviewed and studied to assess their suitability to represent energy efficiency under real operating conditions................................ 43 Classification of the systems ... 13 The present Energy Efficiency efforts have been reviewed ................ 29 IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for comparison of chillers ....................................................................................................................................................................................................................... 27 The simulations leading to the reference values of SEER (HSEER) .................................................. 27 Magnitude of gain/losses due to part load ....................................................... 39 Main technologies for cold production ............................................................. utilities................................................................... 36 Participation of energy agencies....................................................................................................... 43 2.............. 24 System optimisation : all air systems.............................................................................................................................................................................................................. 32 All the elements for an action plan on Air Conditioning are available in the full report ..................................................... 40 CAC systems types based on distribution ..................... 38 Comfort level............... 31 Energy efficiency options have been defined for each system configuration ................ 19 Optimisation of a chiller to improve its EER on the basis of capacity cost only ............................................................

............................................................................................ SELECTION..... 50 Sizing issues ............. 51 Peripheral equipment of chiller based systems: testing and classification .................................................................................. 50 Free cooling ....................... 72 3......................................................................................................................................... 62 A general growth ............................................... 71 Other stakeholders ......................................................................................................................5.......................................................... 60 National surveys .......... 77 4...........................7............................ 61 3............................. 60 2................................................................................................................................................ 51 Chillers: the CEN and ARI approaches (at full load and IPLV) .....................................................................6............... 53 A proposal for a better characterisation of AHU .......................................................................................... 49 Inclusion of RAC in the present study..................................................................................................4....................................... 75 EER for chillers over 750 kW ... 73 Potential for efficiency gains of the selection of higher efficiency equipment........ Statistics on present Energy Efficiency on the EU market ................................................................. 49 Water Loop Heat Pump CAC systems based on local packaged AC systems............................ 58 Full system efficiency........................................... 55 2.................................................. 54 Testing and performance setting for packaged systems ...... 58 Year round thermodynamic balance .............................................................................................................. 60 Correction and treatment of data ................................................. 49 Local package CAC systems: roof-tops and close control cabinets ...............................................................1 Actors involved with CAC systems ................ 54 Ventilation efficiency and Air Conditionning .................................................................................................................................................................................................................... VRF (Variable Refrigerant Flow) CAC systems ........................................................... Statistical databases used and information gathered ......................................................................................................................................................................................................... Technical response to the demand................................................................................................ 49 Summary of choices in terms of local versus central systems ........................................................................2........................... 77 4................. Overall view of energy performance ........................................................... 76 4....... 77 Guidelines for the design of CAC systems .....1........................................................ 58 Definitions ............................................................................................................................................................................. 77 The main barriers to efficiency ..................................... 62 National differences in demand ................................................................................................................................. 66 3................ MAIN FIGURES OF AIR-CONDITIONING IN EUROPE ...................... 51 2............................................................................................................... INSTALLATION AND OPERATION OF CAC SYSTEMS ....................................... 68 The share between distribution systems in chiller based CAC ......................2 Practices and procedures adopted in CAC system design ....... Testing standards and performance standards ....................................................................................... 73 EER as a function of capacity and cooling medium for a chiller under 750 kW .......................................... 62 3............................. The demand for AC in Europe ............................................................................................................3.......................................... 60 Data from manufacturers associations .............. 77 .......................... 69 The choice between chiller-based systems and packages ............. FACTORS GOVERNING THE DESIGN............................................................................ 68 Reversible use of Air Conditioning ............................................................... 64 Evolution of market shares of techniques ............... 64 Comparisons with US market ......................................... 69 The value and nature of the European CAC market ................ 64 Market share of each technology .................... 62 3................................................................................................................................................................... A few technical trends on the market ......................

.............................................................................. 105 Coverage of situations with the DOE software .......................................................................... 2010 and 2020............ 90 Additional prescriptive HVAC requirements ...................................................................................................... 86 The Energy Performance of Buildings Directive (to be transposed nationally) ....... 112 ..................................................................................... 82 The status of regulations in other EU Member States ..................................................................................... 92 Australia..................................................... 1998................. 107 Preliminary results for the cooling consumption of the office building ..4.... 99 Seventh type: envelope and ventilation......................................... 81 Portugal: An example of a national scheme to promote energy-efficient AC through building thermal regulations .................................................................. 94 Technical measures which could increase the efficiency of CAC systems ...... 101 Some national results............................... 108 5............. 94 Measures which could increase globally the efficiency of CAC .......................... 100 Some global results .................................................................................. 98 Third type: improvement of the detailed structure of the system and control options................... 104 Real buildings for the simulation of CAC systems with DOE ................................................... 103 The share between technical systems . 87 The draft Directive on Energy Demand Management (to be defined).. 78 market-transformation efforts within the EU (equipment) .............. Korea and Taiwan ..................................6 Choices and measures which could increase the efficiency of CAC systems ....................... 99 5...................................................................................................................................... 88 4......................... 100 5...............................................................................................................3Previous ................................................... 88 ASHRAE 90...............1 ......................................................... 107 Extension to all economic sectors.................... 108 Preliminary results for the cooling consumption of the office building ..................... 2010 and 2020.......................... 81 Existing national regulations within the EU (which apply at the system level).................................................................................... 86 The draft Framework Directive for “Eco-design of End-Use Equipment” (to be adopted)..............................................................................................................5 Regulatory structure and market transformation at the wider international level ..............................................................................................................................................3 Energy consumption in 1990................... 95 Synthesis of policy measures to raise the efficiency of CAC systems ................ 78 An example of a utility-led energy-efficient AC promotional campaign ... 88 Minimum efficiency standards and energy labelling in the USA ............................. 81 Summary of UK building regulations for space cooling and ventilation.................................... 78 The Eurovent Certification programme ............................................................. 88 Mandatory HVAC Provisions in ASHRAE 90................ 100 Evolution of the market .... 104 5............................................................................................4 ...........................................................................2 Computation of energy consumption in European conditions ............................................ 1998................. Japan................................................................... 80 4...................1 AC Stock and market in 1990.............................................................. 98 First type: selection of more efficient components by whoever decides .............................................................................................. 98 Second type: choice of the best general structure of the system........................ PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO) ................................................................ 105 Adjustment for chiller quality and options not covered in DOE software................................................................. 98 Fifth type: maintenance and operation improved ............ 98 Sixth type: energy and power control ....................................................................................................................... 91 Continuous maintenance of the ASHRAE standard ...................................................................................................................1: a comprehensive approach to raise CAC energy efficiency ........................................................................................................ system types and EU climates ........................................................ 102 Sectoral market .......... 79 The UK Market Transformation Program .............................................. 93 4.................................................... 98 Fourth type: reversible use of the system ...... other measures ...................................................................................................................................... 92 Links between an ASHRAE standard and the US Energy Codes... 87 Practices and procedures adopted in CAC system operation.......

............................ 118 6................................................................... 135 7........................................ 126 6.............................................................................................................................. 136 Motors and fans efficiency ....... 122 Chiller prices as a function of the refrigerating fluid and EER .................................................................................................................................................. 134 7........................................................................................... 124 Optimisation of the chiller used as baseline without any system consideration .................................................................................................................................... 128 Life cycle cost analysis ............................................... 138 Water-side efficiency by sizing and control ................................................................................................................................................................................................. 135 Heat recovery on primary air ........................................................ 115 TEWI (Total Equivalent Warming Impact) and leak rates of CAC systems ...................... 112 Energy by economic sector ..... 134 Energy consumption for a given comfort level ..........................................................................3 Other cost & efficiency trade-offs ............ 122 Role of condensing medium ......................................................................................................................................1 Energy-engineering analysis of chillers ...... 131 Heat recovery section of AHU .............................................................. 138 Design of flow in water circulation .......................................................................................................................................................................2 Engineering approach of the performance of Packaged units .................................. 138 AHU improvement .......................... 135 Primary Air and ventilation .................................................................................................................... TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY EQUIPMENT USED IN CAC ................................................................. 139 Simultaneous demand of heating and cooling ......................................................................................... 137 Quality of Air Diffusion ................................. 127 The US energy engineering analysis ................................................................................................................ 139 Influence of terminal equipment.......................................................................................................................................................................... 118 5............ 141 ..........................................3 Energy Efficiency of Air Handling Units seen as tradable goods ...... 136 Variable air flow and lower head losses .........................2 The improvement of the efficiency of air handling systems in CAC ................................................................................................................... 117 Use of water..................1 Comparison of different CAC systems ............................................................................. 138 7......... 124 Incremental costs as a function of efficiency............................................... 134 Comparison of costs and sensitivities ................................................................................... Overall values ..............................................................................................5 Heating............................ 122 Additional costs for reversibility ....................................................................................................... 116 Numerical results about CO2 emissions for cooling in Europe .............................................................................. 125 Water cooled chillers .................................................................................. 137 Air Side Free Cooling (Economiser) ............................................................................................................... 123 Available data and simulation tools................................................................ reversible or not .............................. 131 Fans integrated in AHU .............................................................. 129 6...................... 115 Atmospheric pollution reduced to CO2............................................ 140 Heat rejection ................................................................................................................................................. TECHNICAL & ECONOMIC EVALUATION OF CAC SYSTEM PERFORMANCE AS A FUNCTION OF THE DESIGN OF THE AC SYSTEM.......................... 132 7...................................... 122 6........................................................................................... 115 5...........................................................................................................................................................................................................................................................4 The possible strength of regulatory efforts and the minimum LCC solutions ...................................... 141 7................................................................................................. 122 Defining chiller part-load efficiency ...................................... 124 Optimisation of a chiller in a system ...................................................................................4 Global warming and other environmental impacts .......... 136 Terminal reheat issues .....

........................................................................................................................... Seville................................................................................................. Concentration of efforts on Air based systems ................................................................................................................................. 171 Perspective of the proposed ESEER ....................................................................................................... 156 Reduction of European hourly load curves to a set of four conditions (based on the example of Milano) . 171 9.......................2 Is the IPLV approach directly applicable to European conditions? ............. 143 Importance of establishing a EU method about part load ................................................ 164 EECCAC final figures -Simplification of the figures and uncertainty estimate ..................................................................................... 173 Scenario 2 THE BEST CHOICE AMONG EXISTING COOL GENERATORS BASED ON FULL LOAD INFO.................................................................. 156 Sizing issues for chillers rating as shown by the simulation of the buildings.. 143 8... ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY CAC SCENARIOS ...........THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO .............................................................................. 148 Building cooling load calculation in US-IPLV.............................................................................................................................. 167 Second way to realise the testing needed for the ESEER proposed certification method ................................................................ 152 Impact of load reduction on the efficiency – a reporting format proposed to Eurovent ............................................................................. 155 8.......................................................... 165 IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for comparison of chillers .................... 148 Buildings used in deriving the US-IPLV .............. 152 Testing conditions and available testing results ...................................................................... 151 8......................................................................... 143 Staging of Part capacity (control issues) ....................................... 160 Extrapolating to the European stock of chillers in use ........................... Construction of a data base of EU chillers at part load –understanding part load .......................................................... 170 Final choice of the ESEER testing methodology........................... 156 The simulations leading to the reference values of SEER (HSEER) .......... 153 Air cooled chillers –experimental results .... 173 .................................................................................. 173 9 ..................................................................................................... 167 First way to realise the testing needed for the ESEER proposed certification method ............... 166 The proposed ESEER method allows grading and ranking of chillers by order of merit ................ 149 EMPE: an answer to a need for a European weighting with IPLV-like testing .................... 146 8........ ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY CAC SCENARIOS ................................. 145 High pressure control at part load...................... 173 Scenario 3 BAT.......................3..................................................................................................................................... 162 8........................................................................................................................................................................ 149 Calculating US weighing coefficients ............................. EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE ......... 150 Reduction of EMPE or IPLV to 2 points with extrapolation ........................... 143 How to reduce the capacity of a chiller? ........................................ 173 9............................... 149 Interpolation scheme needed to reduce testing time ................................................................................................................................................................5 Is there a method good enough for classification of products by order of merit? ....................1 The importance and nature of part-load management measures ....................................... 165 EER is a poor selection tool .......................................................................... 158 Results for more extreme weather conditions (London............ 148 Climate used in IPLV derivation .......................... 164 Classification : who is right? ...................................4 Derivation of a new SEER method (ESEER) ................................................................... 153 Water cooled chillers –experimental results ........................... different distribution systems) .................................................................................................................. 173 Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE ................................................................................................................................. 141 8.................... 141 The result of optimisation .........

.......................... 175 Scenario 3 BAT............................................ 175 Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON FULL LOAD INFO ........................................................................ 177 10........................ 178 10........ 176 Scenario 5 Variable Air Flow ................................ 178 A to G efficiency grading of central air conditioner components...........................................................................................................2 Results of scenarios.............................................. 176 Scenario 6 British regulation on AC – heating.............................................THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO ......................................... 178 10........................................................... 194 Legal basis for policy measures targeting O&M ............................................................................................................................3 Policies and measures to encourage the adoption of more efficient system structures .................. 192 10......................................................................................... 197 REFERENCES ................................................................................................................................. 176 Scenario 4 FREE COOLING ......................................... 195 Definitions and general terms used in the study ....................... 188 Encouraging the selective acquisition of more efficient equipment by other means .............. 174 Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE ............ 174 General Evolution ................................ 179 Market mixed statistics based on the scheme (splits and packages mixed) ....................... POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC ENERGY PERFORMANCE ......................... 195 Specific recommendations................... 197 List of abbreviations .................... 191 Specific recommendations................................................................................................................................................................................................................. 174 Scenario 6 British regulation on AC – heating........................................................................................... 174 9...................... 178 Measures to provide information to end-users and equipment procurers .............................................................................4 Policies and measures to improve system maintenance and operation ................................. 194 Policy aims and potential measures targeting improved O&M .................................................. Scenario 4 FREE COOLING ................................... 174 Scenario 5 VAV .....................................................................adapted for each EU climate ........................................... 190 Legal basis for policy measures targeting more efficient system structures .................... 189 10.................................adapted for each EU climate .........1 Some fundamental considerations regarding policy measures ............................. 190 Policy aims and potential measures targeting the adoption of more efficient system structures ............................................................. 187 Removing less efficient equipment from the market (MEPS and voluntary agreements) ........... 199 ....................................... cooling and air movement................................................................................... cooling and air movement..................................2 Policies and measures to encourage the selection of more efficient equipment ....................................................................... 194 Broadening the application of existing policy measures addressing O&M .....................................................................................................................................................

Impact of load reduction on the efficiency of a chiller may be positive but has to be certified by Eurovent : a reporting format has been proposed to Eurovent as well as a European SEER method (ESEER) for quantification. which are defined as air conditioning systems with more than 12kW of cooling capacity in the EU. Energy Efficiency and Certification of Central Air Conditioners ABSTRACT Air-conditioning constitutes a rapidly growing electrical end-use in the European Union (EU). Scenarios for energy efficiency have been established and quantified. 2200 Mm2 in 2010 (5m2/inhabitant). Definitions of all CAC systems found on the EU market have been given. there are certainly other ways to reach 3. We show that the life cycle cost minimum occurs for large packaged units with an EER of 3. As opposed to room air- conditioners (RAC) central air conditioning (CAC) systems. who choose the system technology without any direct influence from the customer. EER is a poor selection tool . Is there a margin for further improvement ? Optimisation of a chiller for its least LCC shows a large possibility.25 SEER than the ones investigated. yet the possibilities for improving its energy efficiency have not been fully investigated. but our objective was to find out if there is a margin for improvement. The optimal level of performance for the chiller considered is about 40% more efficient than the present « bottom » of the market : it has an SEER between 3. except for the specification of the desired 11 . yet the possibilities for improving its energy efficiency have not been fully investigated. In terms of market transformation. Eurovent. We can keep a few figures in mind : 1200 Mm2 cooled in year 2000 (3m2/inhabitant). in a situation that may be called BAU and leads to electricity consumption around 51 TWh for all AC in 2000 (18 MtCO2) becoming 95 TWh in 2010 (33 MtCO2). engaged in identifying the most suitable measures to improve the energy efficiency of commercial chillers and AC systems. The proposed ESEER method allows perfect grading and ranking of chillers by order of merit . All CAC equipment test standards have been reviewed and studied to assess their suitability to represent energy efficiency under real operating conditions. European CAC systems are commercial AC systems usually specified by engineers or technicians.50 and an initial overcost of +12% paying for itself rapidly. a piece of equipment that we would call a RAC in Europe. They may be selected by an installer of packaged units. European CAC market and stock data have been assembled for the first time.22 W/W. Within the EECAC study twelve participants from eight countries including the EU manufacturers' association. They play a negligible role. The definition of CAC applied in the EU does not correspond to the definition used in the USA where a “Central Air Conditioner” is a ducted package AC system. based on full load behaviour. with a share of reversibility around 25%. For manufacturers. Energy efficiency options have been defined for each system configuration and for the components outside the chiller. have been assembled. the US IPLV and the Italian EMPE are more accurate than EER for classification but do not give enough accuracy for comparison of chillers. less expensive. All the elements for an action plan on Air Conditioning are available in the full report. They are usually designed by an AC engineer and the components selected following the engineers’ recommendations. Packaged units can also be improved a lot. namely thanks to part load control. are not bought or selected in a shop. The present Energy Efficiency efforts have been reviewed. One thing can be done rapidly : all the elements of a possible grading of chillers on the market. which are relatively infrequent in Europe and sized often under 12kW. SUMMARY OF RESULTS Air-conditioning constitutes a rapidly growing electrical end-use in the European Union (EU).00 and 3.

The structure of a CAC system and consequently its name results from the accumulation of a number of decisions on the choice of essential components. but also in the nature of comfort) obtained. There are a large variety of systems and technical options (regarding system structure and control) in use while there is also a large variation in comfort conditions (not only in terms of the set-point. The energy efficiency options list has also to cover secondary systems (distribution) which were found of equal importance for reaching the minimum cost of service. This study benefited from the co-operation between laboratories. Air Handling Units (AHU) are used. which are a particular assembly of small “split systems” and were originally investigated under the EERAC study. engaged in identifying the most suitable measures to improve the energy efficiency of commercial chillers and AC systems.e.. Other AC systems are applicable to a series of rooms or spaces such that their applicability is dependent on the number of rooms and the general situation of the building. Even in this predominant CAC system a choice must be made on how to transfer “cold” to the air of the room. In many cases large Unitary Air Conditioners (or Packaged). which are self-contained direct-expansion (i. The first choice determining a system is the type of the fluid being centrally refrigerated and circulated. Definitions of all CAC systems found on the EU market have been given. like space heating. To transfer the “cold” to air. 12 . It was made easier by the existing information scheme by a subsidiary of the manufacturers’ association called Eurovent-Certification. which was established during the conduct of the SAVE sponsored EERAC study concerned with room air conditioners (EERAC 1999). chilled water is circulated up to the rooms and the air of the room comes directly in contact with it. However the existing information scheme in Europe is based on testing at nominal operating conditions and lags behind the information available in some foreign countries for the same type of equipment (such as the ARI certification programme in the USA). energy efficiency options have to be defined not on the basis of nominal operating conditions but at a variety of part load conditions. Eurovent. Within the EECAC study twelve participants from eight countries including the EU manufacturers' association. These system descriptions lead to the idea of a CAC system description tree of which a non- exhaustive set of branches is presented in Figure 1. which better reflects the CAC operating modes that occur in real use. The two most common approaches are an “induction system” and a “Fan-Coil Unit” (FCU) and these both of these systems can operate with a water distribution network having two or three or four pipes. In the majority of situations however. etc. which generates cold water (typically at 7°C) and which is used to transfer "cold" to the building space in part via a water distribution network and in part via a centrally treated airflow.environment and other conditions such as maximum overall price. is capable of significant energy savings and has occupied a market segment. consultants and Eurovent. but more generally as a modulated capacity system. The most frequent (and really dominant option) is the use of a chiller. without using water as an intermediate heat transfer vector) apparatus can be applied as can Multi-Split systems. To be really effective. In addition a new variant of the “split system” concept most commonly known as the VRF (Variable Refrigerant Flow) system.

Eurovent-Certification. usage and regulatory environment were conducted by the EECCAC study participants for their country. except these involve a mixture of testing at full and part-load conditions so that the results can be extrapolated to provide the average annual performance of CAC equipment. a branch of Eurovent has defined test conditions at which equipment energy performance is to be reported by European industry. CAC market and stock data have been assembled for the first time. although it has been established in this study that they are not suitable for use in European conditions without modification. in accordance with CEN standards. It received a very significant help from the manufacturers 13 . based upon performance testing at full load. There are ISO efforts for the testing aspect but the respective standards are not all available. Thus in some way. CAC system description tree showing the most common CAC systems.U. LOCAL OR CENTRAL LOCAL CENTRAL ROOM BY ROOM SERIES OF ROOMS BUILDING RAC Roof Top-Splits Other CAC FLUID: AIR ONLY FLUID: AIR AND WATER FLUID: REFRIGERANT A. the US standards have progressed futher than the European ones.H.s and INDUCTION FAN-COIL DUCTWOR UNITS UNITS 2 pipes 3 pipes 4 pipes MULTI. National surveys of the CAC market. This has enabled US legislative bodies to readily establish minimum performance criteria which are based upon comparative performance under representative operating conditions. VRF SPLIT ‘Fluid’ refers to the primary heat transfer fluid from the building to the refrigeration system All CAC equipment test standards have been reviewed and studied to assess their suitability to represent energy efficiency under real operating conditions. This took advantage of each participants national contacts including assembling and synthesising rough data supplied by local manufacturers’ or importers’ associations or even involved subcontracting national consultants. Refrigerating and Air Conditioning Engineers (ASHRAE) have defined US-national test standards in a similar way.Figure 1. The American Refrigeration Institute (ARI) and American Society of Heating.

the relative share of the total EU cooled floor-area of countries such as France or Germany. Germany.associations. Annual addition of building cooled-floor area by CAC in the EU (either really added or replaced) AREA COOLED EU-15 added (or replaced.00 100.00 40. Greece. National shares of installed CAC-cooled floor area in EU buildings in 1998 Others Germany 13% 11% Greece 5% UK 8% Spain 24% Italy 25% Portugal France 2% 12% The market shares for all competing AC systems.00 0. see Figure 4 for example based on year 1998. The resulting set of country reports for: Austria. as can be seen from the additional cooled building floor- area installed from 1980 to 2000 for the EU-15.00 1975 1980 1985 1990 1995 2000 2005 Because of a strongly differential growth rate across EU Member States.00 80. Figure 2.00 20. France. which was large in the 1980’s has become small in the 1990’s. Figure 2 (including new systems and refurbishment). Portugal. United Kingdom (with special thanks to the BRE) provides a unique set of data at the national level. Spain.) m2 (YEARLY MARKET) 120. Figure 3. Italy. have been determined in all usage sectors and for all years between 1990 and 2020. 14 .00 M m2 60. The CAC market is expanding rapidly in Europe. The high growth in CAC installed in Italy and Spain means that these countries now account for more than 50% of the EU market. as is apparent from the CAC floor-area installation figures for 1998 by country shown in Figure 3.

the choice of AC equipment is very different. The figures cover the same years (1999-2000) and the same type of building stock (non residential buildings in use).Figure 4. The share of cooled-floor area by AC type in non-residential buildings in the EU for 1999-2000 15 . The share of cooled-floor area by AC type in non-residential buildings in the USA for 1999-2000 US A (E IA ) chillers pack ages all RA C Figure 5b. Figure 5a. Figure 5. however. from the CBECS programme of the US Department of Energy’s Energy Information Administration. Packaged AC accounts for the majority of cooled floor area in the USA while chiller based CAC systems dominate in Europe. the USA. AC market share by AC type expressed in terms of newly installed cooled-area in EU buildings in 1998 Splits >12kW 7% RAC< 12 kW 36% chillers 45% VRF 2% Rooftops 5% Packages 5% Similar data to the European data is available for the world’s largest market.

5 3. but rather represents the temperature regime of cooling towers.57 with water whereas it is 2.0 1. Nevertheless. Figure 6. Chiller energy efficiency (EER) at full load as a function of cooling capacity for chillers available on the EU market in 1999.52 for the systems with air as a rejection medium under the conditions of the testing Standard. and will 1 Energy efficiency ratio. which is taken from the Eurovent directory.0 air cooled water cooled 0.5 2 R = 0.0 2. see Table 1. however.5 1. the distribution of chiller EER1 as a function of cooling capacity and condensation cooling fluid has been determined. for which an arbitrary estimate is made in the standard. rather the efficiency improvements that do occur tend to happen haphazardly. The present Energy Efficiency efforts have been reviewed In the EU.5 Regression (air cooled) Regression (water cooled) 0.5 4. Figure 6. that cannot be compared) 4.0 2 R = 0. units which use water as the condensing medium are significantly more efficient than those that use air.0073 3. with distinct testing conditions (water cooled and air cooled.0 0 100 200 300 400 500 600 700 800 900 1000 Capacity kW The average EER is indeed 3. Through an analysis of data on chiller energy performance at full load test conditions. There are two groups of chillers. the US market is so large in absolute terms that for every CAC type there are more square metres of cooled floor space in the USA than in the EU. From this data it is clear that there is no statistical significant dependence of chiller efficiency on the chiller cooling capacity. In fact this apparent difference is not internal to the chiller. One could conclude in favour of a clear superiority of Water cooled systems over Air cooled systems. water-cooled systems are expensive (for both when a cooling tower or natural ground water is used). which is measured under full load conditions 16 . the energy efficiency of the AC system is not presently a criterion that plays any major role in the AC design and installation process. E UR (E E CCA C) c hillers pack ages all RA C Despite this difference in technical preferences.0003 EER 2.

The operating conditions on the field may be very different from the testing conditions and reverse premature conclusions.therefore only be common among large capacity systems. 17 .

the reversible systems. have an almost identical energy efficiency to the cooling-only systems which have an average EER of 2.96 Floor 3.9 3.Table 1.53 3.48 W/W. the US IPLV approach. the only attempt to address the specific issues of chiller energy performance.09 Condenserless Cooling only water conditioning 2. A number of countries outside the EU have implemented market transformation policy measures to raise the energy efficiency of CAC systems installed in their markets.48 2. which have an average EER of 2.9 3.39 Cooling only water conditioning 2. 18 .9 2. Mexico.34 3.53 W/W. including: the USA.29 reversible air conditioning 1. Canada. was -before this study and the AICARR’s EMPE proposal. Korea. Most policy measures have been aimed at packaged units. Australia and New Zealand. which are not so important on the European market.31 3. which has been found to not be directly applicable in the EU.9 2.76 3. however.73 4.09 reversible water conditioning 2.69 Interestingly. The range of chiller energy efficiency (EER) for different types of chiller systems found on the EU market EER Categories Type Condenser Application min ave max Complete unit Cooling only air conditoining 1. Japan.21 3.57 4.

7) D (>2. with a share of reversibility around 25%.0% 35.00 > EER > 2.40 > EER > 2.0% 30. while taking into account typical over sizing.3) F (>2. Table 2 Example of proposed grading in Cooling function -mixed statistics (splits and packages mixed) . allowing a fair comparison of equipment despite of testing conditions and technical differences.0% R22 HFC 15.1) B (>2.0% 5. Cooling. The BAU definition is the absence of large regulatory actions and of significant changes in consumers choice.0% 10.20 26% F 26% 2.0% A (>3. We have estimated the areas cooled in a way compatible both with manufacturers statistics (capacities.1) G (<2. Figure 7 Air cooled chillers.All the elements of a possible grading of Cooling market have been assembled Tables have been produced on the basis of the technical findings and of the market statistics that allow to define a grading scale for each segment of the market.80 > EER > 2.20 < EER 2% A 2% 3.60 15% D 15% 2.1) Splits and Packages are grouped in one single category (Air-cooled air conditioners) for presentation of results in table 2.20 > EER > 3.5) E (>2.80 7% C 7% 2.60 > EER > 2.).00 5% B 5% 3. The penetration of AC can be expressed in a variety of standardised ways such as the unit cooled-area per inhabitant (in m2/hab. statistics with proposed grading and for each refrigerant 40. table 3.0% R407C R134a 20. below 750kW .40 22% E 22% 2.46 Class definition % on market Grade % with equal class width 3. present market average = 2. Figure 8. Statistics show for each segment (like figure 7 for the largest segment) which part of the market falls in each grade.20 > EER 23% G 11% The impacts of BAU have been assessed Projections of energy consumption have been made. 19 . We can keep a few figures in mind : 1200 Mm2 cooled in year 2000 (3m2/inhabitant).0% 25. 2200 Mm2 in 2010 (5m2/inhabitant). numbers of pieces) and with national statistics (square meters cooled).0% 0.9) C (>2.

17 14.35 0.24 102.27 BE Mm2 cooling 4.39 20.22 3.07 54.83 3.74 414.06 26.45 2.07 1214.19 50.78 6.19 326.78 2.68 16.35 2.36 294.41 IT Mm2 cooling 130.84 64. Only trade and offices really grow in relative terms and they may reach 70% of stock by 2020.01 7.04 23.84 52.74 14.04 Total Mm² reverse 136.23 87.51 365.71 342.90 15.73 57.12 72.01 33.70 1.08 1.07 295.61 56.81 9.41 42.24 502.03 6.71 213.52 171.38 78.64 216.63 400.21 8.49 8.07 2.11 136.35 2.37 13.68 IR Mm2 cooling 5.62 Mm² reverse 1.55 6.73 34.57 42.23 773.28 534.87 61.29 814.69 248.28 50.74 8.92 UK Mm2 cooling 94.53 36.61 41.43 NE Mm2 cooling 22.22 43.88 20 .68 172.63 258.29 33.20 Mm² reverse 34.26 69.47 108.86 120.24 70.83 2558.20 Mm² reverse 0.10 0.59 2693.17 23.21 Mm² reverse 4.38 138.23 80.18 2 LU Mm cooling 0.01 792.38 15.09 54.24 37.10 350.28 110.50 12.30 2.15 248.28 Mm² reverse 14.43 43.33 186.71 101.28 3.84 1.99 Mm² reverse 1.17 83.28 88.67 7.06 36.37 Mm² reverse 0.08 6.26 106.89 PO Mm2 cooling 8.17 0.57 SW Mm2 cooling 38.84 17.74 298.41 78.79 45.61 GE Mm2 cooling 34.88 16.33 467.27 Mm² reverse 4.23 1793.89 FR Mm2 cooling 93.88 145.57 6.62 11.50 4.01 15.07 0.88 450.59 141.29 127.06 43.41 53.46 8.69 FI Mm2 cooling 15.17 20.74 4.03 15.85 175.71 7.41 31.47 27.93 132.24 182.88 24.07 19.39 180.06 2.57 472.76 368.29 11.81 73. Total A/C in 2000 7 6 5 m2 /in 4 ha bit ant 3 2 1 0 B DK D GR E F IRL I L NL A P FIN S UK EU-15 The evolution of the various economic sectors and their demand for comfort vary a lot.06 48.26 0.68 20.65 40.76 2.09 56.46 12.73 DE Mm2 cooling 3.29 127.23 GR Mm2 cooling 11.27 11.31 SP Mm2 cooling 64.25 18.40 0.66 97.11 41.84 4.34 1.02 161.Figure 8.08 6.43 0.80 340.95 Mm² reverse 1. Table 3 Area conditioned in each country and year (such areas can be compared with national statistics) Years Country 1990 1995 2000 2005 2010 2015 2020 AU Mm2 cooling 12.51 18.42 2207.13 Mm² reverse 4.03 8.96 660.29 Mm² reverse 0.30 Mm² reverse 0.08 68.65 51.98 20.03 6.81 30.99 Mm² reverse 5.12 6.98 106.77 52.37 293.30 19.14 8.39 Mm² reverse 32.75 1.85 Mm² reverse 29.72 8.43 10.87 1.49 113.24 10.63 173.07 Total Mm² cooling 538.84 3.24 390.54 18.20 352.40 129.81 51.88 87.36 32. Average cooled-floor area per inhabitant for EU countries and the EU as a whole in 2000.01 191.97 140.25 0.83 5.07 66.02 66.29 30.92 29.28 47.25 39.88 9.

Figure 9 Consumption of the 18 systems in three climates as simulated with DOE software Reference office building unitary cooling consumptions 140. Figure 10 shows the first two values (cooling and associated heating consumption by technical type) for the BAU.0 0. In other words. the winter demand of the cooled areas with the reversible use presently estimated. We computed the electricity consumption of a square meter for AC depending on its location.0 Seville 40.0 S F e ps s l st um R ts op t n n l ) g) ) al tro is g ng lit rg VR io tio M LE uc di in in m rd to lo Sp ut +h la n i bu l ol l Ks D er R IL co oo oo e ai SP rib co on at tri H ir le C . the winter demand of the cooled areas if no reversible use took place.0 100.0 20.( is ng PA di on st um rd id er is + + Si C di er m er rd at PS ai s PA +h r hu at AC at er e w ai h at O w w at ir r+ t de i R w ith LO +a w w de ith si ai de w d + si w ut d O le ith ed si le ut d O TW d oo le ut oo w O le l oo oo rC O oo d C le C C rC Ai er oo er er at Ai C at at W W r W Ai CAC Systems The three main sections of our BAU scenario predictions relate with : the actual cooling demand.0 London Milan 60.0 kWh/m2 80.Then we had to move from area statistics to energy use statistics. its economic sector (typical use) and on the AC system. Figure 10 Energy for cooling consumption split by technical type of cooling and related conventional heating Total cooling consumption by subtype 300 000 250 000 200 000 RAC PACK GWh 150 000 FCU CAV total conventionnal heating 100 000 50 000 0 1990 1995 2000 2005 2010 2015 2020 21 .0 120.(c + st ity +a C K& w t.(c . we have obtained (through DOE simulation and physical extrapolation) energy consumption figures for each system . each building use and each climate as shown in figure 9 under the form of a specific value : consumption per square meter.

330 236.894 28.727 109.875 42.495 18.374 11.579 (Electricity only) Heating function 51.The tables 4.084 164.879 33. (El.) Table 5 Cooling only energy consumption by country and year (for comparison with national projections) Total Cooling Year GWh/ year Country 2000 2005 2010 2015 2020 AU 469 549 633 689 707 BE 274 422 559 681 708 DE 71 122 180 232 260 FI 206 210 229 242 246 FR 5 010 8 213 10 954 13 240 14 071 GE 2 286 4 012 5 542 6 785 7 415 GR 2 909 5 365 7 269 9 399 9 734 IR 127 180 222 252 264 IT 16 209 24 336 27 445 29 795 30 890 LU 11 18 23 27 29 NE 605 690 797 869 892 PO 1 020 2 049 3 072 4 039 4 621 SP 19 689 28 333 33 573 38 719 39 915 SW 391 378 403 421 425 UK 2 359 3 227 3 826 4 241 4 401 Total 51 636 78 103 94 727 109 631 114 579 Table 6 Numerical results about CO2 emissions due to cooling in Europe Kt CO2 2000 2005 2010 2015 2020 AU 164 192 221 241 248 BE 96 148 196 238 248 DE 25 43 63 81 91 FI 72 73 80 85 86 FR 1 754 2 874 3 834 4 634 4 925 GE 800 1 404 1 940 2 375 2 595 GR 1 018 1 878 2 544 3 289 3 407 IR 44 63 78 88 93 IT 5 673 8 518 9 606 10 428 10 812 LU 4 6 8 9 10 NE 212 242 279 304 312 PO 357 717 1 075 1 414 1 618 SP 6 891 9 916 11 751 13 552 13 970 SW 137 132 141 148 149 22 . Table 4 Total energy demand generated by AC (TWh either electric or gas or added) Energy demand (TWh) 1990 1995 2000 2005 2010 2015 2020 Cooling function 22.333 45.103 94.683 51.844 Without REV.442 111. with the same value as electricity.598 74.913 35.517 203.040 With present REV. 5 and 6 give the main values (EUR15) for the three functions. Note that gas is accounted for as a secondary energy.636 78.631 114. Heating function 7.765 250.

). better lighting. packages). but limited if we compare them with other uses in buildings (heating.4 2. electricity prices ranging from 6 2 Seasonal energy efficiency ratio. the cold generating plant (depending on its number of hours of operation. and that there is no significant EE measure on that market. etc.) and we see how the price of the service rendered (the kW of cooling capacity) varies.7 2.3 2. Optimisation of a chiller to improve its EER on the basis of capacity cost only We have performed some engineer economic calculation and compared the technical improvements proposed in the study with the diversity found on the market. through successive additions. The range from 2. the compressor can be reduced when EER increases. etc. Optimisation of a chiller for its least LCC The energy consumption of equipment will be more and more considered in the equipment design process. including part load options (Figure 12). The search for the optimum has been done in the same way as previously. We introduce one by one the possible improvements (better compressor. The part load benefits will then be optimised and the optimisation can then be made on the basis of energy consumption.80 shows reasonable prices for a chiller judged only on capacity.2 2.00 to 2. UK 826 1 129 1 339 1 484 1 540 Total 18 073 27 336 33 154 38 371 40 103 We can keep in mind a 51 TWh consumption estimate for all AC in 2000 (18 MtCO2) becoming 95 TWh in 2010 (33 MtCO2). So the next question is : how far can we improve the balance? What is the potential of improvements paying for themselves but not realised by the present market structure? This question can be tackled at three levels : the most frequent cold generating equipments (chillers. So the cost per kW decreases with the first steps of performance and only increases later (see figure 11). the energy efficiency ratio which reflects the real usage conditions of the equipnent over the year 23 . It corresponds exactly to the present market. For a given electrical power the capacity varies proportionally to EER. So it is interesting to define the “optimum” taking into account consumption. a definition of chillers performance based on SEER2 and SCOP will be substituted to the ones given as EER and COP. including distribution. Such impacts are not small. home electronics. for a given capacity.8 2.50). with a 6% discount rate.80. One day. Remember the figures given correspond to BAU.6 2. The minimum cost chiller according to our analysis has the same EER as the average market (EER 2. better evaporator. which may be considered as a validation of our cost reconstruction.5 2. Figure 11 The cost of a chiller at nominal capacity according to its EER Optimisation of Cost/kW final 110 100 Euro/kW 90 80 2 2.9 EER Conclusion : the best chiller having the same cost (assumed here 100 Euros/kW) as the present “worst performer” has an EER around 2. and climate). and the full system.1 2.

Packaged units can also be improved a lot For 26kWc packaged units.80 4.to 17 cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe). less expensive.80 3.40 2. there are other ways to reach 3. Figure 13 LCC curve of a 26kW package in Europe presented as an annual cost €1.22 W/W.60 3. but our objective was to find out if there is a margin for improvement. improved compressor) and a capacity split between 3 or 4 compressors. It may correspond to a chiller with a correct EER around 2.22 W/W when a 6% real discount rate is applied. They show that the life cycle cost minimum occurs for large packaged units with an EER of 3. and equivalent usage durations (at full load) taken as 400 or 800 hours/year .2 3.200 €1.25 SEER.00 SEER The optimal level of performance for the screw chiller considered is about 40% more efficient than the present « bottom » of the market : it has an SEER between 3. Although the overall life cycle cost per kW are lower for the 52 kW unit the minimum still occurs for an EER of 3.40 3.6 EER (W/W) 24 . The results of the analysis taking these factors into account is shown in Figure 13 for the 26kWc unit. analyses used for US regulations imply an average equivalent of 2097 hours of full load operation per year while 800 hours per year is deemed more likely for the EU.000 Life cycle cost (€/kWc €800 €600 €400 €200 €0 2.400 €1.4 3.0 3.60 2.8 3.50.6 2.20 2.46 (enhanced evaporator and condenser. 12 The annual cost of the service rendered by a chiller in terms of SEER 10 9 8 Total cost (Euros/m2) ALCC17-800h 7 ALCC10-800h ALCC17-400h ALCC6-800h ALCC10-400h 6 ALCC6-400h 5 4 3 2.00 and 3.00 2.00 3.20 3. which is most representative of the EU market. Figure. For manufacturers.

10 and 17 cEuro/kWh a full all air system in Seville for lowest ALCC 39 37 35 33 ALCC17 Euros/m2 ALCC10 31 ALCC6 29 27 25 0. Figure 15 Optimising with 6. the optimal trajectory of improvement of the annualised cost of ownership (ALCC) is given in figure 15.00% 40. After sorting options and combinations.00% 60.00% 20. including part load options (Figure 12). specially if we consider the highest cost of electricity. The regulatory measure could be taken anywhere between a 0% and a 60% reduction without generating overcosts (in the LCC definition) in Seville.00% 100. electricity prices ranging from 6 to 17 cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe). 25 . The designers need the whole range of AC solutions to cover the domain of geometries and air quality requirements.00% 120.System optimisation : all air systems We have concentrated our efforts on air based distribution systems which show presently the most consumption and the highest cost. and equivalent usage durations (at full load) taken as 400 or 800 hours/year. through successive additions.00% % of reference The optimum is very flat. So the bottleneck to the expression of a global reduction in consumption will be the point (shown on figure 14 by an array) where the improved air based solutions start not to pay for themselves : the designers will find it is too heavy a constraint to get under this value. Figure 14 The key point of energy efficiency : the best attainable air based system ALCC Euros/m2/year air water rac packages SPECIFIC CONSUMPTION kWh/m2/YEAR The search for the optimum has been done in the same way as previously.00% 80. with a 6% discount rate.

We have been able to define a new method called ESEER that enables to calculate the seasonal efficiency for all European chillers (centrifugal units are not treated completely in this document by lack of specific information but seem likely to be covered by the proposed method. the part capacity performance can be better than full load at the same temperature. understanding better part load. one for part load reporting and certification. The issue is : how to represent. certify and translate in a single figure those improvements. split by condensation type. The new ESEER method has been compared with the US-IPLV and EMPE proposal under both respects : time spent and accuracy. EMPE was not based on a sufficiently large climatic and technical investigation. There is much progress being made in the control of these part load phenomena. The main tool used was actual testing of EU equipment but a number of group meetings allowed to build a common thinking frame. Its strength (being very close to the existing US method. We had the opportunity to go further by constructing a data base of EU chillers at part load. and proposing two separate methods. Tested air-cooled chillers Name Type Circuits Compressors Available Stages N° 5 Scroll 1 2 3 N° 7 Scroll 2 4 4 N° 8 Herm rec 2 2 2 N° 9 Scroll 2 4 4 N° 2 Screw 2 2 Partially continuous Table 8. which aggregated many factors) was also its weakness. in particular. it cannot be said that Italy would need air conditioning at 12. The first remark in the Italian proposal is that the operating conditions are rather different from Southern Europe conditions. if Northern Europe countries may need air conditioning in summer. an analysis of the method shows that the ARI part-load temperature testing points are "sized" to be "representative" of US buildings (cooling until in negative Celsius temperatures. Original knowledge has been generated during the “Joint project” of EDF R&D facility and manufacturers from Eurovent wanting to promote part load performance. Further to this. The AICARR proposal. The technical description of the chillers tested follows on tables 7 and 8. with different energy weights and. the other one for the computation of SEER. Therefore. Table 7. The percentage of operating hours assigned at each part load condition (in the US-IPLV) is intended to be representative of the US climate and buildings but not of the European ones. named EMPE (Average Weighed Efficiency in Summer regime in Italian) directly deriving from IPLV.Part load performance has been quantified for the first time and the methods have been tested The report shows how in the case of most chillers. AICARR proposed a new energy index.as can be shown by drawing the loads in terms of outside temperatures). it is to say that the error coming from the reduction of the data to single points should be inferior to the testing uncertainty. The compression ratio is decreased so that compressor isentropic efficiency increases. more suitable for the European climate and requirements in the air conditioning field. There was on the table the original US-IPLV method and a European version called EMPE. Tested water-cooled chillers Name Type Circuits Compressor Available Stages N° 1 Screw 2 3 8 N° 3 Screw 2 2 4 N° 4 Scroll 2 4 4 N° 6 Screw 1 1 Continuous 26 . due to the Us experience). And even. for instance. This results from the reduction in refrigerant flow –and consequent improvement in heat transfer efficiency at part load.8°C as normal operating conditions. with different temperatures at the condenser inlet. The constraint was to minimize the testing time while ensuring maximum precision.

4 and 8. We arrived also at applicable conclusions on the way to report the SEER in the Eurovent directory. type of secondary system) can be reduced to four or five external conditions without loss of accuracy. based on a SEER. A significant market share of chillers have various compressors and a complex circuiting. the manufacturer has only to declare which piece of its equipment is operating and to indicate CC .92 Magnitude of gain/losses due to part load The overall performance improvement (or degradation) at part load (temperature effects being substracted) is given on figure 16 for the four air cooled tested units. plus the final Eurovent grading when it is available. No need to test every condition reported in the “map”: the benefit of Eurovent is the fair and independent choice of a few points on the map. For each stage. as usual. It is so when there is one single compressor per chiller. and the associated independent testing. type of chiller. For discrete stages chillers. like in table 9 and following. the DOE reference that we generated. For the others as many testing points as desirable have been obtained.47 4. there is no need for combined testing & weighting (like IPLV). country. The ESEER index proposed here is a set of 4 conditions given for E. For all chillers. namely “low noise” chillers with the possibility of reduced fan speed. For the very few continuously controlled chillers. only IPLV or EMPE points plus the CEN one were available. But a given part load regime has to be defined by the actual status of each piece of equipment. We have introduced a format for the description of the stages of a chiller.12 3.00 EER 4. or various identical circuits. Table 9 : Part load performance of water cooled scroll chiller N°4. the electric power absorbed. as could be reported in Eurovent part load certification scheme N° 4 // WT : 30°C STAGES 1 2 3 4 Compressor 1 0 0 0 1 Circuit 1 Compressor 2 0 1 1 1 Compressor 3 0 0 1 1 Circuit 2 Compressor 4 1 1 1 1 EP (kW) 8.60 27. suitable for Eurovent specification. It is proven that each set of outside conditions (for each sector. climate. leading to improved part load performance. The certifying body has only to check a few of the values. selected in the same conditions as usual. We started from HSEER.27 4. some common testing points were made according to either the US-IPLV or the EMPE conditions depending on the manufacturer will. Impact of load reduction on the efficiency – a reporting format proposed to Eurovent One important finding is that a percentage (like 50%) is not enough to report the part load behaviour of a chiller.17 38. but there can be as many similar indices as specific demands: sector. For all the tested chillers.27 CC (kW) 37. Note that this procedure is in fact already used for some chillers with various speeds. as a whole. 3. fours stages can be defined in terms of input. 27 . not only values at the four arbitrary percentages and temperatures. it would be easier to describe performance at a given stage not at a given percentage. the cooling capacity and EP. Since temperature and load can be tested independently and recombined. a supplementary point was added to fulfil the CEN EnV requirement : nominal inlet condensing temperature (35°C for air and 30°C for water) and 50% load ratio referred at this nominal inlet condensing temperature For chillers n° 2.50 78.80 17.50 150. In all circumstances a simple model has been used to draw the performance maps from existing testing points. etc. About certifying Part Load : what the manufacturers give to their customers is a « map » of performance.00 112.U. The customer can rely on the Eurovent SEER computed from this map … or compute its specific SEER for its specific demand.

8 N° 2 0. CAC air and water distribution equipments have been simulated using the European average efficiency values. three climates have been simulated. the cycling formula enables to calculate the electric power. including the high pressure control impact on each stage.6 0.2 0.1 N° 5 1 N° 7 0. the weighting of electric power of each stage is found by the expression of the weighted average. Then the load is compared to each stage capacity.7 0. Figure 17 Reduced efficiency while decreasing part load ratio (same source temperatures) for the water cooled chillers Reduced efficiency of the part load stages for air cooled chillers 1.2 Part load ratio The simulations leading to the reference values of SEER (HSEER) Two buildings were simulated on computer. Otherwise.3 1. If the load is lower than the smallest available capacity step. but buildings that do exist : an office and a commercial mall.2 0.7 0.2 1.Figure 16 Reduced efficiency while decreasing part load ratio (same source temperatures) for the testedwater cooled chillers Reduced efficiency of the part load stages for water cooled chillers 1.9 N° 1 N° 3 0. 28 . used then as a reference for other methods.9 N° 8 N° 9 0.2 1.4 0. At each hour the outside enable to calculate all known stage capacities and respective electric powers. adopting different envelope characteristics when moving the building around Europe.6 0.6 0.4 0.2 Part load ratio The overall performance improvement (or degradation) at part load (temperature effects being substracted) is given on figure 17 for the five water cooled tested units.6 0. Hour after hour.5 0 0.5 0 0.8 N° 6 0. The different systems identified in the stock and market study have been simulated.8 1 1.8 1 1. For each one. the simulation uses then the characteristics of the real chillers modelled to compute the exact yearly performance index : the HSEER (Hourly SEER).1 1 N° 4 0.

EER alone is a poor selection tool We shall compare now four classifications : according to EER. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers ESEER ARI EMPE Weighting Weighting Weighting Part load ratio Temperatures Temperatures Temperatures coefficients coefficients coefficients 100 35 3% 35 1% 35 10 % 75 30 33% 26. For 50 and 75%.9 42% 26. am I really selecting a better chiller? This is completely false if we base ourselves only on EER (figure 17). 29 . included single compressor units.3 30 % 50 25 41% 18. ESEER weighting coefficients give more weight to the 25% point load than both index. coefficient are nearer to the EMPE index. US-IPLV.8 20 % Temperatures of the ESEER are comprised between EMPE temperatures above and ARI temperature under.3 45% 23. We take the point of view of a user of the Eurovent certification system : by selecting a “better” chiller.7 42 % 31. It mainly comes from the uncertainty measurement on the temperature difference at the evaporator. Table 10.EECCAC final figures for a European SEER method (ESEER) Our work clearly shows also that the methodology for air and water cooled chillers enabled to extract seasonal operating temperature conditions with errors on the seasonal efficiencies that are inferior to the experimental uncertainties.5 40 % 25 19 23% 12.9 20% Temperatures of the ESEER are embedded by the EMPE ones above and ARI temperature beneath except for the 25% point. nearer from the IPLV one. For 50 and 75%. ESEER. The 100% coefficient is 3%.3 12% 21. A comparison of the conditions of the 3 available indexes is proposed Table 11 for water cooled chillers. coefficients are nearer to the EMPE index.4 10% 75 26 33% 23.5 40% 25 18 23% 18.3 45 % 27. for all chillers.9 30% 50 22 41% 18.4 1% 29. EMPE. largely under the experimental uncertainty. In order to simplify the application of the index. it also shows that the experimental uncertainty is quite high. The ESEER weighting coefficients give more weight to the 25% point load than both index. some rounding can be done without modifying noticeably the ESEER figures obtained. However. The ESEER 100% weighting coefficient is nearer from the IPLV one. Comparison of the ESEER conditions with the EMPE and IPLV indexes for water cooled chillers ESEER ARI EMPE Weighting Weighting Weighting Part load ratio Temperatures (°C) Temperatures Temperatures coefficients coefficients coefficients 100 30 3% 29. Table 11.8 12 % 23. using as a reference the actual EU values obtained by simulation in the three locations and properly weighted. A comparison of the conditions of the 3 available indexes is proposed Table 10 for air cooled chillers.

5 IPLV 30 .5 4 4.5 2 2.5 3 3.5 1 1.5 0 0 0.5 2 2.5 1 0. Figure 18 comparison of US-IPLV with HSEER for the tested chillers HSEER versus IPLV 4 3. Comparison of HSEER with EER on the tested chillers HSEER versus EER 4 3.5 EER IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for comparison of chillers A classification based on US-IPLV or EMPE would be largely false but would not distort completely the market (figures 18 and 19).5 0 0 0.5 3 2.5 3 3.5 HSEER 2 1.5 1 1.5 1 0.Figure 17.5 3 2.5 HSEER 2 1.

the two methods. the literature 31 . More than 20 basic systems and 100 variations were considered on a qualitative basis and then "filtered".5 0 0 0.5 3 2. The newly proposed ESEER method is more accurate in a noticeable manner and satisfies the needs of Eurovent certification process as well as the expectations of the DGTREN in a market transformation. IPLV and EMPE have at the end the same advantages and disadvantages. The newly proposed ESEER method allows grading and ranking of chillers by order of merit We see on figure 20 that the new method has the most important characteristic expected from a market transformation tool : almos no misclassification. and not always in the same direction.5 3 2. a piece of equipment graded better than an other one is better or equivalent. A number of systems and energy efficiency options do not need detailed quantitative simulation because one or more of the following apply: they are infrequent.5 1 0. comparison of ESEER with HSEER for the tested chillers HSEER versus ESEER 4 3.5 2 2. Figure 20 .5 4 EMPE Based on similar assumptions.5 0 0 0.5 1 1. Energy efficiency options have been defined for each system configuration Th possibilities are so numerous.5 1 0.5 2 2.Figure 19 comparison of EMPE with HSEER for the tested chillers HSEER versus EMPE 4 3.5 4 ESEER Conclusion : the differences are relatively large between existing methods and reality.5 1 1. and so system dependent that the process of “filtering” the most promising was difficult.5 HSEER 2 1.5 3 3.5 3 3.5 HSEER 2 1. their use is are not expanding.

nobody wanting to buy a « poor » image equipment. consumers and utilities) would find a benefit in the deployment of efficient CAC. The policy measure associated is to ban many classes of equipment or a negotiated agreement on average full load performance like ACEA agreement for cars. but should also address all the factors which have to reshape and activate the chain relating the energy used to the final service: the conditioned square meter. At the European level a draft Directive on Energy Efficiency of Buildings is currently under transposition. or simulation is impossible. the imposition of limits on cooling demand). On the basis of this inspection.e. the competent authorities shall provide advice to the users on the possible improvement or replacement of the air- conditioning system and on alternative solutions. The policy measure associated is to ban some classes of equipment either directly (Directive ) or by voluntary agreement. at a net negative cost. The US ASHRAE 90. The Portuguese measures cover a number of sizing obligations. We can also expect that a certain number of years of labelling and communication by energy agencies reaches the same point. and require the use of a central AC system with a number of energy saving features such as ’heat recovery’. which itself demands knowledge of Air Conditioning (and other) system efficiencies. information (through a national release of the Eurovent directory database and the later elimination of the least efficient equipment) and voluntary "best practice " initiatives. So there will be a movement towards improvement also in the existing buildings. and ’free cooling’ over a certain cooling capacity threshold (usually 25kW). For the others a detailed simulation has been organised. between the building renter and the CAC operator. This requires the calculation of building energy performance. however part load is not taken into account. design guidelines and specific system requirements. the absence or inadequacy of building codes addressing this new and rather complex equipment segment. however part load is not taken into account in Eurovent grading and so the corresponding improvement is not obtained. which in fact seems to be the smaller difficulty. High energy savings are possible. Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON FULL LOAD INFO On average packages and chillers reach in 2005 the EER level corresponding to the minimum LCC (Best Available Technologies with present information). This will help the adoption of the best technical solution in new buildings. such that all parties (manufacturers. 32 . the lack of incentives for operators to optimise efficiency. the problem of certifying something built on site and only once. Article 8 requires that central air-conditioning systems of greater than 12 kW cooling capacity shall be regularly inspected.is already sufficient to enable the savings potentials and costs to be defined. different incentives caused by the separation between the plant owner and the building renter. etc.1 standard is a fully integrated set of policy measures which include minimum energy performance thresholds. It is fully described in the report. As a result the policy measures to be proposed in the EECCAC study should not only address the offer of more efficient CAC equipment. as well as significant CO2 emissions reductions. The French measure is basically a minimum efficiency threshold above which AC systems can benefit from EDF’s marketing and financial support. Scenarios for energy efficiency have been established and quantified Scenario 1 MOVING ALL COLD GENERATORS TO AVERAGE PERFORMANCE All packaged AC and chillers presently below the market average EER reach that level by 2005. However the chain going from the manufacturer (most have already improved equipment in their directories together with simpler one) to the final consumer is distorted by a number of factors including: the consideration of initial cost as the only decision criterion by most designers and installers. which is very difficult to model. ’part load staging’. The UK Market Transformation Programme is a comprehensive set of possible policy measures including: obligations written into building codes (i. Monitoring and maintenance are also included in the Portuguese RSECE the local regulation applicable to CAC). which shall include an assessment of the air- conditioning efficiency and the sizing compared to the cooling requirements of the building.

COOLING AND AIR MOVEMENT- ADAPTED FOR EACH EU CLIMATE Introduction of a MEPS on total electricity used for Heating ventilating and AC in kWh/ m2. There is a reduction in cooling demand which is climate dependant and has been expressed here by country and system. to know the cost we have to evaluate the less costly options.Scenario 3 BAT. this policy induces almost no extra cost for any stakeholder. national values are different and have been derived from UK with corrections for DD and fitted to each country. The policy measure associated is to ban many classes of equipment or a negociated agreement on average part load performance like ACEA agreement for cars Scenario 4 FREE COOLING IS USED AT ITS MAXIMUM POTENTIAL Obligation of introducing free cooling on air side of air based distribution systems at a certain value of air flow (Portuguese regulation and Ashrae) even for primary air (which is the case of our simulations. at comfort level TC).THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO All packages and chillers reach in 2005 the SEER level with the minimum LCC (BAT with upcoming information given by part load testing). All the elements for an action plan on Air Conditioning are available in the full report The analysis presented in this study has shown that there is a significant variation in energy efficiency for all types of CAC equipment that have been investigated when tested under standard test conditions. To obtain this “free” market transformation a prescriptive minimum should be applied to local manufacturers and importers at the same time. which may be on either side. c) can be used in building thermal regulations to demonstrate compliance with requirements 33 .1 and which like ASHRAE 90. grading. Scenario 6 BRITISH REGULATION ON AC – HEATING. primary or secondary. fleet-average efficiency targets) • Financial and/or fiscal incentives for higher efficiency equipment • Public procurement and general market transformation programmes The European Commission and/or a coalition of willing Member States should consider: • the development of an EU model building code that addresses air conditioning amongst other energy end-uses.g. Starting from their present averages of EER and SEER. Part load IS taken into account in Eurovent grading and so the corresponding improvement is obtained. The policy instrument would be a strong and harmonised implementation of EPB directive. The measures which can be considered to encourage the higher energy efficiency levels for new CAC equipment are: • Provision of information (labelling. The less expensive way of attaining the objective is the improvement of chillers. b) can enable to compare of the relative benefits of different system designs. and absolutely no cost provided it’s applied to all manufacturers (and so that they all pass on the costs to the customer). (an EU equivalent to ASHRAE 90. Scenario 5 VARIABLE FLOW COMPULSORY IN ALL AIR BASED SYSTEMS There is a reduction in cooling demand which is climate dependant but has been applied here on auxiliaries consumption in Air based systems with national values.1 is subject to continuous improvement) • The development of practical public domain CAC system design tools which: a) can aid system designers to develop energy efficient CAC designs. efficiency ratings) • Removing less efficient models from the market (MEPS or voluntary agreements) • Encouraging higher sales-weighed average efficiency levels through negotiated agreements (e.

Member States could also consider the development of low cost mechanisms to encourage the adoption of good practice for CAC operation and maintenance (namely by ESCOs). occupancy and purpose. The European Commission and/or a coalition of willing Member States should also consider: • Making efforts to define best practices in operation and maintenance • Making efforts to define best practices in operation and maintenance performance contracting With an aim of informing national building thermal regulations and the implementation of the Energy Performance in Building Directive. cooling and heating degree days) Further to this. quality of comfort provision and climate (e. 34 . • The development of EU benchmarks for CAC system efficiency expressed in terms of: building function and size.g. Member States should undertake a revision of their building thermal regulations to address a number of specific issues aimed at reducing CAC energy consumption which are described in the report.

35 .

Energie Demain are leading consultancies on demand-side management (DSM) in France. • Eurovent Certification was established by the Eurovent/Cecomaf manufacturers’ association for the certification of performance of air-conditioning and ventilation equipment. • to make appropriate recommendations. • The University of Athens. the famous HVAC engineering school in Sevilla. • to estimate potential energy savings deriving from the use of more efficient CAC. and namely the Department for Energy studies is the main supporting laboratory for the HVAC engineers gathered in AICARR. and dissemination of information. • UTCB is the Romanian Technical University for building sciences. on the basis of a cost–benefit analysis. education.1. • ITF is an HVAC consultancy in the region of Chambery. The EECCAC study follows-on from the EERAC SAVE study which investigated the potential for measures to raise the energy efficiency of room air-conditioners (RAC) and which is available from the same co-ordinator (EERAC 1999). The EECCAC working group began a study in April 2001 to investigate the technical and economic potential of measures to raise the energy efficiency of Central Air-Conditioners (CAC). the group carries out research. with activities ranging from technological development to socio- economic investigations. INTRODUCTION SAVE II is an EU programme to promote the rational use of energy within the European Community. application projects. • INESTENE and –later. The work has been organised in tasks for which the best experts have been chosen as task leaders. • PW Consulting is a UK-based consultancy specialising in equipment energy-efficiency initiatives and programmes around the world. specialised studies. • Politecnico di Milano. • AICIA supports the research by ETSIS. utilities. • to investigate ways in which these savings can be realised. The objectives of the study are: • to estimate the electric power consumption of CAC. France 36 . Selection of technical experts The following technical experts are participating in the EECCAC study: • Armines is a research association supported by the Ecole des Mines de Paris and is especially active in the field of energy efficiency. Since the EERAC study covered all types of AC of under 12 kW cooling capacity. is very active in the field of solar cooling and energy conservation in buildings. the present study is concerned with air-conditioning systems over 12 kW and will eventually integrate the two segments in a common picture of the European industry and market. A broad coverage of energy agencies. The working party has been gathered and co-ordinated by Jérôme Adnot from Armines. technical experts and manufacturers’ representatives are involved in the work in order to be sure that state of the art knowledge is available for each aspect of the study. in particular the Group of Building Environmental Studies.

Participation of energy agencies, utilities, manufacturers and national experts
The coverage of national and industrial expertise in the study was ensured through the participation of
the following bodies:
• EdF (Electricité de France), the French electricity utility, who brings an important contribution to
the study of this growing electrical end-use. EDF are represented by Pascal Dalicieux and Frank
Colomines
• ADENE-CCE, the Portuguese energy-conservation agency, has a significant experience;
• AICARR, the Italian association of Air Conditioning, Heating and Refrigerating engineers, who
includes professionals working on international standards development, maintains special
Observatories on Hospital air conditioning technology and Refrigerant fluids and cycles.
• EVA, the Austrian energy research and policy institution in which the federal and provincial
administrations (‘Bund’ and ‘Länder’, respectively) and some 30 important institutions and
corporations from a variety of economic sectors co-operate.

• IDAE, the Spanish energy-conservation agency, has close relationship with all bodies having an
influence on CAC in Spain,

• BRE is the leading UK centre of expertise on buildings and construction. Its energy-related activities
include technical research and consultancy, managing the Government's buildings energy-efficiency
information programme, providing strategic analysis of energy-efficiency policy options and
modelling the energy performance of the UK building stock.

• Eurovent/Cecomaf and Eurovent/Certification is the manufacturers’ associations for refrigeration,
air-conditioning and ventilation equipment, represented by a number of members, namely Mr
Sormani from Climaveneta, Mr Van de Velde and Ms Jacques from Daikin, Mr Coates from
Airedale, Mr Zucchi from AERMEC, Mrs Huguet and Ferrand from Carrier, Mrs Goral and Legin
from Trane.
Not only did Eurovent actively participate in the plenary meetings of the EECCAC working groups, but
they have also held a number of specific working meetings with their members (European
manufacturers), including those of WG6A (May, 28 and November, 23, 2001 and October, 3, 2002),
with the following companies in attendance: ACE-Airwell, AERMEC, Airedale, Carrier, Climaveneta,
Daikin-Europe, Ferroli, Galletti, Lennox-Europe, Multiclima, Teba, Trane, York. The manufacturers
have opened their laboratories for visits by the EECCAC co-ordinator and some of them (Carrier &
Trane) have shared some valuable data bases for use in the study. Through the course of the EECCAC
study an agreement to conduct a common part load testing programme for chillers on the European
market has been reached between EDF and Eurovent on the basis of a shared costs procedure.
During each of the EECCAC working meetings held in Paris, Athens, London, Lisbon3, Madrid, Milano and
Vienna, a number of national AC experts were invited to attend and express national views and policies,
which has been a valuable input to the work. On total 50 professionals or representatives of associations
attended our dissemination and exchange meetings.
Further contact concerning present EECCAC study or EERAC results or paper copies of the reports can be obtained
through:
Prof. Jérôme ADNOT
Centre d'Energétique-Armines-ENSMP
60, Bd St Michel - F 75272 Paris Cedex 06
Tel 33 1 40 51 91 74 Fax 33 1 46 34 24 91 Mail jerome.adnot@ensmp.fr

37

2. CENTRAL AIR-CONDITIONERS IN EUROPE: DEFINITIONS AND BASIC
DATA

2.1. Importance of AC for human health and productivity performance, link with
ventilation
The use of air conditioning is increasing rapidly in Europe, as a result of an increasing trend towards control
of the indoor environment and a wider diffusion of air-conditioning systems as a consequence of economic
growth, which has made them more affordable. However, this is also partly a consequence of a movement
towards higher economic productivity. Accurate figures have shown that a better indoor working
environment leads to less quality problems, higher productivity and less accidents in the workplace, provided
it does not create too much noise. We are not in a situation where air conditioning specialists and companies
generate an artificial need, but are rather in a situation where they offer new ways to answer existing needs or
decrease total costs. However there is a need for them to be able to prove case by case that their techniques
are cost effective for the intended purpose and that they have optimised their proposed solution, which is a
major reason why industry has been so active in co-operating with the EECCAC study.

What is "comfort"?
In any case, the comfort level to be reached should reflect the nature and quality of the activity which takes
place in the conditioned space. There is no value in “cooling buildings”, but there is in being able to establish
desired comfort levels in the internal spaces where people work or perform other tertiary activities.
To give a rough presentation of the range of comfort conditioning requirements and circumstances that can
be encountered, the following main cases are listed:
NC- Natural Cooling which is obtained, day or night time, by forced-ventilation, when outdoor conditions
permit, or by any other ways of transferring heat to the outside, provided they are not based on the operation
of a compressor. Natural cooling is usually insufficient to attain always and everywhere the required comfort
levels but can be found sufficient in most circumstances in climates like the UK or Northern Europe.
PC- Partial Cooling which is obtained with air conditioning equipment that provides partial control of the
temperature. For instance, the rooms are cooled but fresh air may be introduced without cooling, or the
installed cooling capacity of the AC system may be insufficient for all circumstances, as a result the internal
air cannot be kept at a constant temperature. This may be felt as comfortable in France or Germany.
TC- Total Cooling, wherein the AC system provides full temperature control and includes the provision of
the minimum rate of ventilation air changes required for hygienic purposes at an adequate temperature. This
type of equipment allows a degree of dehumidification consequent to the cooling effect– it is a very frequent
level of comfort today.
TAC- Total Air Conditioning, which includes full control of temperature and humidity as well as provision
of the minimum ventilation rate required for hygienic purposes but is not capable of attaining air purity
conditions for specific IAQ (Indoor Air Quality) levels.
AAC- Advanced Air Conditioning, same as TAC but with a full control of IAQ. These systems are
particularly applied in hospitals or clean rooms in the electronic industry.
The variation in comfort level changes the consumption of energy. If one wanted to make a complete
comparison, it would be necessary to give a monetary value or some other proxy quantity to discomfort and
to balance it with the total cost of the service. The question was less difficult for the RAC studied in the
EERAC study which were deemed to provide " equal service" at the PC or TC level of comfort. The
differentiation of comfort has substantial consequences for the systems’ energy consumption, but one should
not regard a decrease in the quality of the indoor environment as a means of saving energy without being
conscious of the trade-offs involved.

38

Comfort level, Ventilation: our assumptions for the study
From one extreme to the other both the cost of installing the initial AC equipment purchase and of operating
it can vary by a factor 10, but consideration of the choice of who or what really needs a certain level of
comfort is outside the scope of the present project. Most comparisons between systems are thus to be done
internally to a given assumed comfort level even if it is possible to gain an idea of the relative cost of the
various possible comfort levels. Total Cooling will be used as the appropriate value for benchmarking. It is
more difficult to predict the future level of comfort expectations and even more the speed of change.
There is a strong interaction between two functions: ventilation (i.e. air changes) and cooling. The technical
systems used differ from one country to another depending on the basic philosophy embedded in the
regulations and building codes. Two philosophies of ventilation seem to exist in Europe : in the first one
(adopted by Northern countries), ventilation comes first as an hygienic necessity and then a further decision
leads to cool the space or not. In the second one (apparently Southern States), the decision of A/C comes first
and leads to more air changes with the outside, and to controlled ventilation. Central ventilation (with
cooled “primary air”) is the base of our technical study together with the TC comfort level.

2.2. Basic definitions
Air conditioning is a technology, supported by thermodynamic and physical science, intended and
designed to change and improve the conditions (mainly temperature and moisture content) of the
outdoor air to be supplied in an enclosed space in order to make it possible to fulfill an industrial process
(Industrial Air Conditioning ) or maintain specific conditions needed by equipment installed in the space
(Control Air Conditioning) or for the well-being of the human presence (Tertiary1 and Residential Air
Conditioning).
The EECCAC study is concerned with tertiary air conditioning and is focused on the important and
practical issues that surround the design of "conventional" air-conditioning systems of types that are
already well established in the market-place and thus does not include the so-called "low energy cooling
options" : passive buildings, dessicant or evaporative cooling. Despite this the range of equipment and
technical issues included in the EECCAC study is very wide and therefore the related energy
conservation topics of lowering cooling loads, and deploying innovative but, as yet, little-used systems
are outside its scope.

RAC and CAC in competition
Central Air Conditioning systems (CAC), the subject of our study, are characterised by a central
refrigerating unit operating together with an air treatment unit and make use of a fluid (air and/or water)
to transport cold to the air conditioned space. They perform other functions than just refrigerating, like
controlling air change, air quality and humidity. Their specifications are determined by engineers or
technicians, who usually design the system and its associated energy performance without any direct
influence from the final customer or user (except for the preliminary limitations on cost).
A ‘Room Air-Conditioner’ (RAC), as opposed to an ‘air-conditioning system’ (CAC), is an appliance
that can be bought by a household, with a direct link between the customer and the selection of the
purchased good – either direct purchase by the household or through an installer with whom negotiation
and specification of the appliance takes place. The existing results of EERAC, the previous study on
RAC, can be used as a basis for the present study, when we come to such RACs. We have excluded
from our research absorption machines (running on gas or waste heat), and the use of any other fuel than
electricity, which are still very infrequent solutions.

Basic Thermodynamics at one instant
The simplest AC system, as illustrated in Figure 2.1, cools the space around the people or the process in
summer by rejecting the heat outside the room, with a limited or complete control of the room humidity

1
Tertiary is a European word indicating all human activities and related buildings other than industry or households

39

or humidity control) extract heat in summer from inside the room (Pc). or flooded evaporators consisting of a tabular shell in which the refrigerant expands and cools a fluid circulating in a bundle of tubes inserted in the shell. From an energy perspective.and air quality. or air quality. A similar index to the EER.2). 40 . seen from an energetic perspective Pr heat rejected air conditioned space AC Pc cooling effect system control of: Temperature Humidity IAQ Pe energy input The accepted energy performance index is called the ‘energy efficiency ratio’ (EER) and is defined as: EER = Pc / Pe Cooling only systems (not including ventilation. which is usually obtained through ventilation. The characteristics of the evaporator technology depend primarily on the required application and the type of cold source. Two broad categories exist: • air-cooled evaporators. and filtration components. There are also some cooling systems which offer the possibility to produce heat instead of cold by reversing their refrigeration cycle: such systems are called ‘reversible’. the coefficient of performance (COP). R407C. In fact the desired comfort conditions may include thermal comfort (which is expressed in terms of a mix of convective and radiative temperatures). this situation is summarised by a “load” to be extracted or cooling effect Pc (the minimum thermodynamic quantity necessary to maintain the defined comfort conditions). etc.) creates the "cold" in the evaporator. Figure 2.1 Essential quantities in the process of air-conditioning in summer. Usually the heat rejected outside (Pr) has an energetic value equivalent to Pe + Pc.e. which subsequently absorbs heat from the refrigerated space. humidity control and indoor air quality (IAQ). • liquid-cooled evaporators. is applied to indicate the performance of reversible AC in the heating mode. by the change of indoor air. Main technologies for cold production Evaporation of the liquid "refrigerant" (R22. After its full evaporation the refrigerant vapour is compressed using a compressor for which the following main technologies are used: • reciprocating compressor • screw compressor • scroll compressor • centrifugal compressor. through the use of electricity (Pe). R134a. It is the ratio of the heat input into the conditioned space and the electric power consumed to transfer it. the largest single equipment (figure 2. or direct expansion evaporators consisting of a pack of finned tubes through which the air is forced. i. We shall describe the steps of the technology and give the specific names for the chillers. approximately equivalent to the value of the “load”.

• water-cooled condensers consisting of finned tubes with internal grooves to increase the heat transfer surface area and the overall heat transfer coefficient. presenting detacheable parts to access the compressor’s main components and coupled to separate electric or thermal engines. which have their body directly coupled to an electric motor cooled by the refrigerant and enclosed in a totally sealed shell. that would occur only at non-nominal or bad designed points.3). The water used as the coolant may be from a natural water source (such as a river or aquifer) or from re-circulated water that’s been cooled in a cooling tower.2 An aircooled chiller (courtesy Climaveneta) After its compression the refrigerant vapour is condensed while evacuating the heat corresponding to the one absorbed at evaporator level and the thermal equivalent of the work of the compressor. • "Semi-hermetic" compressors that are similar to the open type compressors but have a common casing with the electric motor. The ratio of pressure producing work to kinetic energy output is known as the impeller reaction and ranges from 0. Figure 2. They can be used with any refrigerant but are generally employed in systems with medium to high cooling capacity. is the outside air temperature. one can find a diffuser that ends converting kinetic energy into pressure lift. the speed of the refrigerant would be consequently reduced as related to the speed of the impeller.00 in terms of EER and 5. The temperature and flow rate of water have the greatest influence on the condensing temperature. 2 They enable a high pressure ratio because of the absence of alternative compression. The condenser technology depends primarily on the required application and the heat source. The primary factor which influences the performance of the condenser. after the impeller. between 50 and 100 % better than the chillers of smaller size. They are relatively infrequent industrial compressors of a large size and very efficient2 and we can relate them more with “district cooling” or “block cooling” that really with “Air Conditioning”. That’s why. they are generally used for systems with medium cooling capacity.7 in terms of SEER. Hermetic compressors. Condensers used in CAC systems are divided into three categories: • air-cooled condensers consisting of a finned tube heat exchanger (figure 2.4 to 0.The centrifugal compressors will not be studied in details in this report. these are generally used for systems with a small to medium cooling capacity. If the fluid enters the impeller with a tangential component or swirl. 41 . The turbine is called the impeller. The observed performance is around 6.7. A wide range of technologies are used to couple the compressor to the electric motor: • open type or accessible compressors. The compression ratios can vary between 2 and 30.

Figure 2. Figure 2. which is used to throttle the refrigerant fluid back to the evaporator pressure and to control the refrigerant flow. Among cooling towers there are three principal systems: • Indirect contact (Dry) cooling towers where there is no contact between the cooling fluid (air) and the fluid to be cooled (water) • Direct contact (Wet) cooling towers where there is a direct contact between the two fluids thus providing better heat transfer • Wet-dry towers.• Evaporative condensers. A wet cooling tower (which displays better energy performance) is more at risk of cultivating the legionella bacillus and consumes water. They are especially used to reduce water vapour plumes and hence water consumption.4 A water cooled chiller (Courtesy Carrier) After condensation the refrigerant is expanded by an expansion valve. which are used in industrial applications and combine a condenser and a cooling tower in a single apparatus. which contain a conventional wet type tower in combination with an air-cooled heat exchanger.3 Inside an air cooled chiller (Courtesy Airedale) A water cooled chiller (figure 2.4) is generally used with a cooling tower. Three systems are used: 42 .

of which some significant branches are presented in Figure 2. so as to create a common vocabulary and terminology in further EU regulatory work on CAC. • thermostatic expansion valves that are controlled via the superheating. The structure of a CAC system (and consequently its name) results from the accumulation of a number of decisions on the essential components. CAC systems types based on distribution The system is integrated by a number of interrelated elements controlling the total comfort. The consequences in terms of energy consumption of all these technical choices will be investigated in the rest of the study. • Terminal room units for local air treatment (most frequent). • Assembly of automatic controls to keep the requested conditions and general safety. occupation scenarios. “induction systems” or “fan-coil systems” and these can be used with a water distribution network including two or three or four pipe assemblies. central or local re-heat . such as air filtering. There are. The VRF (Variable Refrigerant Flow) system (also sometimes called a “modulated capacity” system) is a relatively new development on the CAC market that is based on the “split system” and has the potential to produce some interesting energy savings. humidity treatment. each of them addressing different parts of the space having different loads. • Air treatment central unit if we want to provide refrigerated air. etc. The number of possible systems that can be obtained by the combination of these elements is very large: the EECCAC study has developed a set of documents (additional to this report) to cover the systems that can be defined in an exhaustive manner. The most frequent (and really dominant option) is the use of a chiller that generates cold water (typically at 7°C). Even in this predominant CAC system a choice must be made regarding how to transfer “cold” to the air. The first choice determining a system is the choice of the fluid being refrigerated centrally and circulated.. Essential components of a CAC system are: • Water chiller with an electrically driven vapour compressor. Among the many systems to consider (50 or so). direct-expansion (without water) apparatus can be applied as well as Multi Split systems. for instance. ducts and pumps for refrigerated air and water circulation. In many cases large Unitary Air Conditioners (or Packages). load levels. Other systems are applicable to a series of rooms and their application depends on the number of rooms and the general situation of the building. 43 . some are infrequent and costly and only half a dozen deserve real interest for their low initial cost or for their comfort or adequacy to the needs. These descriptions lead to the idea of a descriptive systems tree. a particular assembly of residential “split systems”. which is is used to transfer "cold" to the building space partly through a centrally treated flow of air and partly through a water distribution network.• expansion devices with a constant pressure difference. • Distribution structure including networks of fans. • electronic expansion valves that are also controlled via the superheating. which are self-contained. originally covered in the previous EERAC study. some are obviously too complicated. Classification of the systems Most large plants have to combine a number of systems.5. In many cases the same system has to also take care of the heating mode. In this study we have considered only generic systems (one system for one zone) and not the combinations of various systems in such larger plants.

5.U.6. controllers.s and INDUCTION FAN-COIL DUCTWORK UNITS UNITS 2 pipes 3 pipes 4 pipes MULTI. Figure 2.H. induction units. humidifiers and de-humidifiers. etc.3.7). which transfer heat from the air in the locally cooled zone to a cold water circuit (Figure 2. Description of other aspects of systems Terminal units and other peripheral equipment used The optimisation of many pieces of equipment is required to attain the optimal operation of the system: constant or variable flow mixing boxes (air). VRF SPLIT 2. air handling units.6) and Air Handling Units which cool air more centrally before its distribution and diffusion into the rooms (Figure 2. Plan of a fan coil unit 44 . fan coil units. The main devices which provide comfort are Fan Coil Units.Figure 2. balancing valves and dampers. CAC system description tree showing the most common CAC systems LOCAL OR CENTRAL LOCAL CENTRAL ROOM BY ROOM SERIES OF ROOMS BUILDING RAC Roof Top-Splits Other CAC FLUID: AIR ONLY FLUID: AIR AND WATER FLUID: REFRIGERANT A.

• 2 pipe fan coil with change-over (2PR). the heat exchanger comprises one supply and one return pipe only for refrigerated water.Water to air Heat exchanger Circulation fan Air filter Outdoor Indoor air supply air return There are various kinds of FCU: • a 2 pipe fan coil (2P) for cooling only.8) used for the remote preparation of cold air. This system can heat or cool simultaneously in two different rooms. The same system is used in all zones and comprises one supply and one return pipe as the previous one but circulation of which can be of either hot or cold water. Figure 2.7 and 2. A 3 pipes system as well as a 4 pipes system with only one coil existed and disappeared. cooling and variable air flow distribution 45 . This system may be reversible or not. • 2 pipe fan coil with electric heating (2PE). There are also a wide-variety of air handling units (AHU –see figures 2. • 4 pipe fan coil (4P) with two coils frequently assembled together The same system can be used in all zones. An additional electric resistance heater can be complementary to the reversible heating mode or can be the main heating source for weak loads during the winter period. according to needs. A reversible chiller supplies either cooling or heating and changeover from one mode to the other is centrally regulated according to season.7 Air Handling Unit with heating. This system cannot heat and cool simultaneously in two different rooms and hence is used when the summer-winter transition is easily distinguishable. and comprises a supply and return for both hot and cold water and can thus heat or cool simultaneously different rooms of the same building.

Figure 2. less frequent nowadays. Return air Outdoor air variable flow boxes S Figure 2.9. are used when centrally distributed air is further cooled at the local level through thermal contact in the IU with refrigerated water circulated in a central water loop.8 The « coil » providing heating and cooling inside an AHU (Courtesy Trane) Induction units (IU). 46 .

systems using water chillers are the most frequent type of central air conditioning system used in the EU. cools water. air handling units CONDITIONED AIR AIR AIR A. or employ a heat recovery system. ALL AIR SYSTEMS refrigeration: chiller (air or water cooled with/out cooling tower) air treatment: central station type. have terminal reheat or not. comprising a group of equipment to refrigerate water. which may be: single duct or dual duct. Table 2. A water chiller. ZONES DISTRIBUTION VOLUME TEMPERATURE CONFIGURATION SINGLE-ZONE SINGLE DUCT FIXED VARIABLE CAV with terminal re-heat with by-pass on re- 47 .1).H.9 Induction Unit Diffusion Filter Coil Indoor air Nozzles Air from central system General classification of systems based on chillers These can be divided into single-zone or multi-zone systems. AHU systems can also be distinguished between those that have a variable flow and those which have constant flow. which in turn is used to cool off centrally supplied air in an air handling unit and/or is circulated to room terminal units (RTU) for local air treatment. Due to the complexity of descriptions involved..1 Classification of central systems based on chillers SYSTEM CLASSIFICATION SUMMARY 1.Figure 2. Terminal reheating carried out with an electric resistance heater or a gas furnace heater but never with a water coil. etc. The EECCAC study has developed a specific terminology to describe this variety of systems. use fresh air only. As mentioned before..U. it is useful to present all the possible systems which could work with chillers in a synthetic table (2...

induction units. circulation with by-pass on mixed air with control on coil capacity MULTI-ZONE SINGLE DUCT FIXED VARIABLE Multi-zone CAV PER EACH ZONE PER ZONE PER ZONE VARIABLE CONSTANT VAV systems VARIABLE VARIABLE VVT systems DUAL DUCT FIXED VARIABLE with terminal re heat first duct . AIR-AND-WATER SYSTEMS refrigeration: chiller (air or water cooled with/out cooling tower) Air treatment: primary air – central station type. air handling units secondary air – room treatment TYPE OF AIR AIR VOLUME AIR AIR DISTRIBUTION TEMPERATURE PRIMARY AIR FIXED VARIABLE SINGLE DUCT std velocity for systems with fan coils high velocity for systems with induction units SECONDARY AIR Treatment by room terminals: fan coils.cold air second duct – hot air DUAL CONDUIT High pressure systems PRIMARY AIR FIXED VARIABLE SECONDARY AIR VARIABLE CONSTANT 2. radiant panels 48 .

they don’t use water pipes to distribute cold and have no. are still considered to be “central AC systems” because they don’t work on a room-by-room basis and their cooling capacity is often much higher than 12 kW. Inclusion of RAC in the present study 49 . They are frequently used in super markets and in telephone central technical rooms although in other economies (i. or alternatively to each side of the chiller (reversible heating and cooling). their sizing can be minimised. Although they are similar to multi-split systems (a residential split system serving several rooms). because when the outdoor condensing unit is selected in accordance to the indoor units needed for the entire system. the installation has to be done and adapted to the site. the USA) they are commonly used in much wider applications. VRF units are available in three versions: cooling only. in fact. Roof Top Units and Close Control Cabinets. “Free cooling” can be available in many types of Roof Tops and two operating modes are possible for most of the available technologies: cooling only and reverse heat pump cycle. The advantage of the system is that the closed water circuit can transfer heat rejected from the units operating in cooling mode to the others which are operating in heating mode and thereby minimise energy consumption. A more efficient version of this system makes use of thermal storage and in some cases of ground water sources. the installation needs to be adapted to the site. Although the use of a central chiller and a central boiler is often also necessary. VRF systems are capable of being controlled to simultaneously satisfy different building zones requiring different thermal conditions. It can be used for cooling when connected to a water chiller and for heating. to transfer “heat” and “cold” according to the local need with a very low energy consumption. Local package CAC systems: roof-tops and close control cabinets Although these systems are not installed far from the rooms that they have to cool.Number of water loops connected with the chiller The simplest solution is to have one single water loop in the conditioned space. air ducts. This system is particularly viable when there are simultaneous cooling and heating needs in the building. Roof top units are always air cooled whereas close control cabinets may be water or air cooled. or very limited use of. they have not been accounted for in the previous EERAC study due to their higher cooling capacity.e. For more comfort. VRF systems are classified as built-on-site systems. Water Loop Heat Pump CAC systems based on local packaged AC systems This system is based on a closed loop of water-cooled packaged reversible heat pumps which can potentially operate independently with some in cooling and some in heating mode. in all cases they are self-contained units which are completely assembled in a factory. They are able. the transition being called “change over”. VRF indoor units are equipped with electronic expansion valves which continually adjust the flow of refrigerant to match their specific cooling capacity to the local “load“ requirements. In addition. The same loop can be served by a reversible chiller provided refrigerated waterin cooling mode or hot water in the heat pump mode. The two commonly used systems. two water loops are installed (hot and cold) with distinct generators. heat pump and heat recovery.4 Description of systems not using chillers VRF (Variable Refrigerant Flow) CAC systems VRF or modulated capacity systems are based on the “residential split system” technology although in this case a large series of rooms potentially up to the level of an entire building can be served. 2. As for VRF systems. when connected to a boiler. The chiller is still only connected to one loop at the same time. In a further refinement the chiller is installed between the cold and the hot loop. taking complete advantage of COP and providing at each instant a reversible solution.

and reject hot air from the condenser to the exterior space through a duct. consisting of several indoor units (usually four or more) connected to one outdoor unit. Water-cooled units of any type under 12 kW cooling capacity were part of the previous EERAC study. There is a range of “large split systems” over 12 kW of cooling capacity and therefore are usually classified separately from the residential splits. Table 2. consisting of two sections (one indoor and one outdoor unit) connected only by two pipes that transfers the refrigerant and a cable for the electric power. VRF (Variable Refrigerant Flow) systems can be considered as a version of multi-split systems but are always over 12 kW in cooling capacity. To sum up. RAC are used for 78% in the economic sectors and for 22% in households.2 Summary of choices having in mind internal factors of Energy Efficiency ST AH SITUATION COLD SOURCE CAV Central Hydronic VAV Central Hydronic RT Central Package FC4P Zonal Hydronic FC2P Zonal Hydronic WLHP Zonal Package PTAC Zonal Package Sizing issues In general all the related issues within the borders of CAC energy efficiency have been treated as diligently as possible. leads to the choice of a carrier but generates positive scale effects. Since a CAC system is designed and sized by a professional engineer who is usually concerned to minimise the 50 . the main use of water-cooled RAC is therefore limited to closed-loop heat pumps as previously described in this section as one of the CAC systems with a comparatively high system efficiency. This classification based on the EE view is shown in table 2.2 . while the other provides direct cooling to indoor space by means of an air circulation fan. while the outdoor unit has a compressor and a condenser. figures about the market and level of cooling provided by RACs drawn from the EERAC study have been utilised in the present study. Heating. The water used in RAC could in principle be drawn from a natural water source. Summary of choices in terms of local versus central systems A growing distance between the “centre” and the rooms increases losses and auxiliaries. but this is seldom available. Four types of air-cooled RAC are widely used: • Split-packaged units. This family of AC equipment is partly under 12 kW of cooling capacity and partly over. The indoor unit includes the evaporator and a fan.Should we neglect RAC or include them here? To answer this question. most system types can be classified according to the air handling situation (central versus zonal) and cold source (hydronic versus package) criteria. but it can be generated far from the room or close to it. and should thereby avoid potentially distorting impacts from isolated policy actions. It is better to compare the shares of all competing AC types in a single scenario for a given economic sector and country. the potential problem of equipment over-sizing had to be treated explicitly. • Single-packaged units. In particular. This also helps to understand the real basis of competition between RAC and CAC systems. which are packaged AC appliances that are kept inside the room while cooling the space. are commonly known as ‘window’ or ‘wall’ RACs wherein one side of them is in contact with the outdoor air for condensation. ventilation and air conditioning systems can been split in secondary (air-side) and primary (water side) systems. There is always an air side. • Multi-split-packaged units. • Single-duct units.

Free cooling At some time during the year. but ISO standards are not currently available for all the required equipment types. etc. the US standards have progressed more than the European ones. although they are not directly applicable in European conditions. we can expect that on average there is no over-sizing in this case. it is assumed that the average European RAC is over-sized by a factor of 2 (i. being used twice in opposed ways. In the same way the American ARI (American Refrigeration Institute) and ASHRAE (American Society of Heating. The ratios of consumption per square meter and the cooled areas are all presented as “standardised” area (based on 120 W/m2) but on one occasion figures have been produced with a variable sizing depending on location. In the case of RAC. There are however many examples of oversizing. which should be related to some specific factor (type of contractor.initial investment cost. system type to allow national comparison. conventions may vary from one country to another. which is being developed by ISO TC 86. Testing standards and performance standards A branch of the trade association Eurovent has defined the set points for which performance is reported by industry. Refrigerating and Air Conditioning Engineers) set US standards. 100% over sizing). building. testing conditions and extrapolation to yearly behaviour of equipment in the US in order to facilitate energy performance requirements to be set by a legislative body.5. An international test standard is being developed as ISO PWD 19298-2001-Draft 5: ‘Liquid-chilling packages using the vapour compression cycle – testing and rating for performance’ This ISO standard. This value. Efforts have also been made by ISO to define internationally applicable testing standards.e. has no influence on our statistics. nature of control : based on temperature only or on enthalpy). We made the assumption of a fair and economic sizing. outside air can be used directly to cool the space without any special thermal treatment. level of expertise. a fact that gives interest to our repeatable “standardised” area. Thus in some ways. thus allowing performance comparison at full load. not the strict area of activity rooms. For consistency reasons we have chosen a single value for sizing all CAC systems and converting capacities into areas and later areas into capacities (120 W/m2) and another one for all RAC systems (240 W/m2). is not yet ready and is not expected to be for some time. in accordance with CEN standards. There are control issues associated (flow rate. The US test method is defined in ARI 550/590-1998 ‘Water-Chilling Packages Using the Vapor Compression Cycle’. Note that the notion of conditioned area is uncertain in national statistics : not the gross built area. 2. However the ARI documents often mix pure testing standards. Chillers: the CEN and ARI approaches (at full load and IPLV) The US and European full-load chiller test conditions (with a built in condenser) are as follows: -.4°C (85°F) (water 51 . The European method to test chiller cooling capacity is defined in the proposed European standard prEN 12055 and in the equivalent Eurovent Certification standard 6/C/003-2001. which are bought directly without any advice or even from a retailer interested achieving the highest purchase value or from an installer who enjoys a margin on the equipment sold. movement of dampers. time allocated for sizing. etc).7°C (44°F) Entering chiller water temperature 12°C About the same (the water flow is fixed by the standard) Entering condenser water temperature 30°C (water cooled) 29.Cooling operation Eurovent Certification (at ARI Standard 550/590 – full load) 98 Leaving chiller water temperature 7°C 6.

3 Parameters used in the US IPLV % load Air entering condenser Water entering Operating hours % (DB) condenser 100 35 29.7 23.9 42 50 18. which is the normal operating condition for AC systems and the one where very significant energy efficiency improvements appear to be possible.45C + 0. The Integrated Part Load Value is thus calculated using the following equation: IPLV = 0. and by 3% water cooled chillers EER. which overestimates by 1% air cooled chillers EER. The weighing of part load points in ARI 550/590 are given in Table 2.3. ARI Standard 550/590 – 98 testing at part load conditions The current European test standards do not include part-load ratings. cooled) Entering condenser air temperature 35°C (air cooled) 35°C (95°F) (air cooled) -.01A + 0.3 18.3 45 25 12.8 18. The intention of part load rating is to enable the energy and cooling performance at part-load to be assessed over a wide range of typical operating conditions. For this reason energy efficiency measurements at full load are directly comparable between the two systems. Table 2.12D Where A = EER at 100% of full load 52 . the seasonal average efficiency of a chiller.42B + 0. However the situation is not the same for part-load operation.3 12 This load profile is the basis used to calculate the integrated part load value (IPLV). whereas the chiller certification programme operated by ARI in the USA does include part-load performance ratings. The only difference is that the pumping power is “forgotten” in ARI values.4 1 75 26. ISO harmonisation on full load is easy.Heating operation Eurovent Certification (at full load) Leaving condenser hot water 45°C temperature Entering condenser hot water 40°C temperature Evaporator inlet air temperature 7°C dry-bulb 6°C wet-bulb ( air cooled) Evaporator inlet water temperature 10°C ( water cooled) There is no great difference in the full load chiller test conditions under ARI (US) or Eurovent (European) specifications.

An ISO Committee Draft on testing of ‘Air Conditioning Condensing Units’ has been produced as CD 13258. Performance testing requirements are stipulated in prEN 13053 for fans. The thermal performance of the AHU casing is also tested and classified under prEN 1886-1997 such that the AHU thermal transmittance is measured under a 20 to 25 K steady-state temperature difference. which must always be less than 3.5 W. Heat recovery sections are tested according to EN 305 ‘Heat exchangers.Air handling units .96 l. called EMPE. damper sections. The following international standards are also under development by ISO TC 86/SC 6: ISO NP 17524 ‘Testing and Rating of Air Terminals’ and ISO NP 17515 ‘Testing and Rating of Air Diffusers’ Under prEN 1886-1997 the casing air leakage of the assembled AHU is tested and the system is graded according to the measured leakage rate. heat recovery sections with intermediate heat transfer medium. Mixing sections are tested for air mixing efficiency. mixing sections. An additional test is conducted on the filter bypass leakage and the filter leakage performance is classified according to the results. An ISO test standard for air terminal units.K-1 are classified as T1 while those with a transmittance of greater than 2. As of August 2001 a DIS was being prepared for ballot which would include refrigeration condensing 53 .K-1 are classified as T5.m-2. Before our study a first proposal by the Italian AICARR was made. and regenerators (heat recovery sections containing thermal accumulating mass). components and sections’ prEN 1886: 1997 ‘Ventilation for buildings . Three types of heat recovery section are recognised: recuperators.m-2. Definitions of performance of heat exchangers and the general test procedure for establishing the performance of all heat exchangers’. Damper sections are tested according to EN 1751-1998 ‘Ventilation for buildings – Air distribution and diffusion – aerodynamic testing of dampers and valves’ and must satisfy maximum permissible air leakage requirements. Units with a transmittance of less than 0. filter sections and sound attenuation sections. is under development by ISO TC 86/SC 6.0 W. The following test standards are in use by Eurovent: Eurovent 6/3 ‘Thermal Test method for Fan Coil Units’ and Eurovent 8/2 ‘Acoustic testing of Fan Coil Units in Reverberation Room’. The T2 to T4 classes occur at intermediate values. In addition individual coils must be sealed within the AHU casing such that the resulting air gaps do not exceed maximum permissible levels. heat recovery sections.s-1m-2 at 700 Pa negative test pressure for the positive pressure sections of units that operate under both positive and negative test pressure.Air handling units – Mechanical performance’ In the USA the ARI 430-1999 ‘Central station air-handling units’ test standard is used. B = EER at 75% of full load C = EER at 50% of full load D = EER at 25% of full load The suitability of this index for European operating conditions is considered in Chapter 9. known as ISO NP 17524.s-1. humidifiers. Peripheral equipment of chiller based systems: testing and classification The following European test standards are in use for Air Handling Units (AHUs): prEN 13053: 1999 ‘Ventilation for buildings .m-2 at 400 Pa negative test pressure for units that always operate under negative test pressure and should be less than 5. Additional testing of thermal bridging in the casing is required and the casing’s thermal bridging performance is classified on a non-linear one to five scale. coils.Ratings and performance for units. Room fan coil units are the subject of a proposed working draft of a new ISO standard (ISO/PWD 5) oinISO/TC 86/SC 6. Coils are tested and rated according to ENV 1216-1998 ‘Heat exchangers – forced circulation air-cooling and air-heating coils – test procedures for establishing the performance’.70 l.

A proposal for a better characterisation of AHU An AHU is characterised by the following information: 1. Diffuser typology. The listed information is necessary and sufficient to exhaustively define the functions and performances of an AHU. The comparative method is based on the principle that each AHU can be evaluated at every working condition comparing it with a reference AHU with the same functional characteristic. Possibility of partial or/and total air recirculation 6. Cooling. Diffuser arrangement. Humidification Dehumidification 4. Partial load air efficiency 9. Available static head. Location of the AHU (indoor/outdoor) 8. Expected treatment typology: Heating. The Cooling Tower Institute in the USA has also issued a test standard to rate the performance of Cooling Towers. The conventional nominal values would be reported in tables and settled on the basis of criterion that allow for economics and technical factors. The nominal efficiency value of an air diffusion system depends on the following factors: 1. there are some directions. Ventilation efficiency and Air Conditionning The capability of an air diffusion system to usefully introduce air in a room is described by the ventilation efficiency conventional parameter ev defined by the ratio between the air flow ideally needed to keep the required air quality level in the room with the hypothesis of perfect air mixing and the real air flow required in the real air diffusion systems. intended as the difference between the static pressure measured upstream and downstream the AHU working at nominal flow without air recirculation. The reference machine is built keeping the layout of the AHU being evaluated and adopting conventional values for losses and efficiencies at nominal load and partial load values obtained by simplified methods. Yet. Filtration class. These three factors are combined together and they can not be considered independently from the performances standpoint. Nominal power consumption. 2. published by each manufacturer. ISO TC 86/SC 6/WG 8 has developed a proposed working draft PWD 16345 for ‘Cooling Towers – Testing and Rating for Performance’. subdivided in: a. 54 . 3. In the USA the following test standard is used: ARI 365 – 1987 ‘Commericial and industrial unitary air conditioning condensing units’ . Nominal flow. expressed in kW. intended as the working flow recommended by the manufacturer expressed in m3/h. Partial load exchange efficiency b. Presence of sensible or total heat recovery 7. The approach will be tested as a possible UNI (Italian) standard. Supply temperature. 3. expressed according to the EN 779 standard. which recommend the right installation conditions and the maximum operating temperatures at nominal load. 2. A proposal of a comparative method has been made. Partial load efficiency. 5. CTI 201-1986 ‘Certification Standard for Commercial Water Cooling Towers’.units.

There is much to gain by this approach. the occupied conventional space.3 0.75 0. The test conditions are always at full- 55 .5 indicate the ventilation efficiency values for each system category.3 Displacement diffuser 1.70 0. .60 Table 2. or inside. The European Standards EN 814 and EN 255 are fully consistent with the ISO standards although they classify AC equipment by whether it is operating in the cooling or heating mode and not whether it is ducted or not. The standard test conditions for the cooling capacity test are shown in Table 2.4 and 2.75 Linear slot diffuser 0. Testing and performance setting for packaged systems ISO 5151-1994(E) – 'Non-ducted air conditioners and heat pumps -. Table 2. and air-cooled or water-cooled units. realized and installed in conformity with the manufacturer directions and that we remain at nominal conditions.4 – ventilation efficiency for mixing system with the air introduced above the occupied zone.3 1.5: ventilation efficiency for Displacer systems or mixing systems with the air introduced below. The following tables 2.00 Cones diffuser 0.In particular.00 1. two different categories of air diffusion system can be identified: . the occupied conventional space. Diffuser typology ev ev Dt < 0°C Dt ≥ °C Helicoidal effect diffuser 1.1 Underseat diffuser or similar 1. contributing to cut the heat loads and the air pollutants down.2 1.Mixing system with the air introduced above the occupied zone. in the hypothesis that every system is projected.90 0. or inside. Diffuser typology ev ev Dt < 0°C Dt ≥ 0 °C Floor helicoidal diffuser 1. embedded into a UNI standard but not yet quantified here.6. Testing for these systems was discussed in the previous EERAC study. single-phase and three-phase.Displacement systems or mixing systems with the air introduced below.60 Outlet with single or double fin rows 0.8 The efficiency value characterizes the share of the flow rate introduced that actually reaches the occupied zone. An equivalent standard exists for ducted equipment: ISO 13253- 1995 – Ducted air conditioners. The ISO test procedure applies to packaged air conditioners of any capacity and type provided they are non- ducted including cooling-only and reversible.Testing and rating for performance' is applicable to all packaged air conditioners. A new version is being prepared. Water-cooled heat-pumps are not included and neither are part-load test conditions thus in practice it is not possible to use the test procedure to properly rate the performance of variable or multiple speed drive air conditioners.

thus the part-load performance of variable or multiple speed drive units is not reflected.load and with a single set of stable environmental conditions. 56 .

57 . The US efficiency regulations classify a central air conditioner depending on its cooling (or heating capacity). Unitary air conditioners or heat pumps of below 65000 Btu/h capacity and which are not single-packaged room air conditioners or heat pumps nor packaged terminal air conditioners or heat pumps are classed as ‘central’ systems and have a different test procedure and set of efficiency requirements.05 kW) and 135000 Btu/h (40 kW) are tested in the same way as large systems but have different efficiency requirements. SEER. Central air conditioners in the USA include both ducted systems and ductless split systems (i. All central air conditioners. The definition of “central air conditioners” used in the USA and Canada is much narrower than that applied in the EECCAC study because they are limited by the maximum cooling (or heating) capacity and they must be packaged units. split- packaged units) although ducted systems are predominant on the market. including split-packaged systems. that includes the following product types: • central air conditioners and heat pumps • small commercial package air conditioners and heat pumps • large commercial package air-conditioners and heat pumps With some minor differences. All ‘commercial’ units with capacities below 135 000 Btu/h (40 kW) are classed as ‘small’.e. with capacities up to 65000 Btu/h. For equipment designed for other uses. are rated using a seasonal energy efficiency ratio. that is based on the amalgamated results of testing cooling capacity at four different test conditions. Small ducted systems are either split systems or single package systems. 2) Representative of equipment working with cooling towers. Small commercial units with capacities between 65000 Btu/h (19. Units with cooling or heating capacities above 135000 Btu/h (40 kW) are classed as ‘large commercial systems’ and have different efficiency requirements from other types. Note that performance degradation due to real part load is determined in the standard and that variable speed or capacity systems are tested in a way which allows their advantages to be apparent. ISO Parameter Standard test conditions T1 T2 T3 Temperature of air entering indoor side (°C) dry-bulb 27 21 29 wet-bulb 19 15 19 Temperature of air entering outdoor side (°C) dry-bulb 35 27 46 1) wet-bulb 24 19 24 Condenser water temperature (°C) 2) inlet 30 22 30 outlet 35 27 35 T1 = Standard cooling capacity rating conditions for moderate climates T2 = Standard cooling capacity rating conditions for cold climates T3 = Standard cooling capacity rating conditions for hot climates 1) The wet-bulb temperature is not required when testing air-cooled condensers which do not evaporate the condensate. but mostly the latter. the manufacturer shall designate the condenser water inlet and outlet temperatures or the water flow rates and the inlet temperature in the ratings The USA has an extensive program for air conditioners and heat pumps.6: Test conditions for the determination of cooling capacity.Table 2. Canada and Mexico have adopted the US approach.

In many situations there is a possibility to take advantage of “free cooling”. water or evaporatively cooled. In addition complex buildings often experience simultaneous cooling and heating demands within different parts of the building because the heating and cooling seasons are not cleanly separated and the internal spaces have different uses and loads and different sun exposures. HSPF. there are thus possibilities of using the AC system to simultaneously heat and cool different building zones without relying on thermal energy drawn from outside the building. are classified as unitary air conditioners and are subject to energy efficiency targets. defined as those with capacities above 135 000 Btu/h (40 kW). To apply this strict definition. energy labelling and are also subject to aspirational efficiency targets. that we consider being only two : heating with ventilation. Both the fixed and variable-speed air conditioners (either room or unitary) must satisfy MEPS (Minimum Efficiency Performance Standards). mostly the ventilation auxiliaries at that time : do they act in the 58 . furthermore there are plenty of ways of recovering cold from some parts of the system in order to cool other parts. The performance of the equipment at part load is significantly different from what it is at full load. of a packaged type. Definitions To facilitate understanding AC system energy performance. an air cooled heat pump or a water/evaporatively cooled air conditioner. 2. which is available whenever the outside air is cooler than inside air. a common vocabulary has been defined based upon the specification of seasonal or annual quantities. we should come back to each time step and make some sophisticated computation on the effects of the auxiliaries. The commercial (small and large) air conditioners and heat pumps can be air. For large commercial air conditioners and heat pumps. Larger room air conditioners. The existing targets differentiate depending on whether an appliance is integral (windows) or split-type and whether it is cooling-only or reversible. In Japan a central air conditioner would be classified as a unitary air conditioner. The variable speed units are tested and rated using a seasonal energy efficiency ratio (SEER).Similarly all central heat pumps with capacities up to 65000 Btu/h are rated using a similar approach to produce a heating seasonal performance factor. which is a measure designed to reflect performance under part-load conditions. COP and IPLV. In establishing this balance. For small-commercial air conditioners and heat pumps with capacities above 65000 Btu/h but below 135 000Btu/h. however these would not be representative of the year round energy performance of the component nor of the energy performance of the on-site AC system for many reasons that are now described. Free cooling increases the time during which the building is only ventilated. not MEPS and labelling. Reversible units are not subject to COP requirements but are required to satisfy the cooling-mode performance requirements. due to come into effect between 2004 and 2007. Korea is unique in having devised AC efficiency standards and targets that treat constant speed air conditioners (those using a single-speed compressor) differently from those using a variable speed. a distinction is made depending on whether the capacity is above or below 760 000 Btu/h (205 kW) and whether the unit is an air cooled air conditioner. you separate the auxiliaries according to the function being performed. Multi-split systems are currently excluded but are about to be included in new energy efficiency target regulations. Efficiency is measured and regulated in terms of the EER. the cooling performance is regulated for the EER (static test conditions) and Integrated Part-Load Value (IPLV). This implies that either the draft international test procedure for multi-split units is to be adopted or that a new unique Japanese test procedure will be created. The heating performance is measured and regulated using a static COP test. but it’s not related with the ventilation function but with cooling. The targets for reversible units are a combination of EER and COP targets. Overall view of energy performance Year round thermodynamic balance EER and COP figures can be measured for AC equipment submitted to testing under specified steady- state conditions.6. cooling with ventilation.

package. 3.g. addressing the family of “free cooling” options). SCL: the Summer Cooling Loads or "cooling energy". SSCOP the System Seasonal COP for winter: the ratio of WHL to WED.direction of heating or in the direction of cooling? We have done something far simpler : when there is heating only or cooling only.e. The quantities of interest are : 1. this is not the subject of the present study. outdoor air load. etc). This parameter could be miscalculated due to the large preponderance of electricity consumption by auxiliary applications in most systems. that of the: Fans + Pumps + Primary equipment. SSEER. SEC = the electricity consumption of the compressor of the cooling equipment (e. 10. including the energy for cooling and humidity treatment. SSEER: System Seasonal EER for summer: the ratio of SCL to SSEC. SCOP. ventilation is integrated in the demand . SSCOP and partly WHL. SSEC: the System Summer Electric Consumption for cooling. the equivalent of SCL in winter season. etc). It is well known that with a suitable budget it is possible to deploy passive measures which can provide an elevated degree of comfort without the use of AC. WEC: the Summer Electricity Consumption for reversible cooling. WEC = the electricity consumption of the compressor of the cooling equipment used to heat (e. SEER: Seasonal energy efficiency ratio during the summer: the ratio of SCL to SEC. 2. The SCL takes into account zone loads. WED: the Winter Electricity Demand (Consumption) for reversible use of air-conditioning. however. WHL: “Winter Heat Loads" represent the heat demand of the building .e. the heating of air that is passed through fans and the real system operating schedule and thermostatic control. 7. 6. the equivalent of SSEER for winter. WHG: “Winter Heat Gains" represent the heat generated by the equipment and effectively used (used in substitution of normal heating mode in winter for reversible equipment). at the end of the year the unaffected auxiliaries (floating or simultaneous heating and cooling) are allocated to heating and to cooling in proportion of the total yearly demand. 4. This study is concerned with identifying beneficial means to increase the SEER. is completely satisfied by the AC system. SSEC = electricity consumption of the whole system i. including all auxiliaries. 11. reversible heating has some limitations and not all the load can be satisfied by the cooling equipment. To give a clear definition of quantities we have called Winter the total of time periods during which the heating demand is higher than the cooling demand and Summer the sum of all times where the opposite occurs. the real index of performance. As far as the cooling loads are concerned this study is confined to an investigation of the ways of decreasing the SCL that are related with equipment and control choices (i. the chiller. the chiller. package. The study does not address issues concerning the total heating load or the nature of the back-up heating system because it is only concerned with that part of the heating which is provided by the AC equipment. SEC: the Summer Electricity Consumption for cooling. it is assumed that the total summer cooling load. the same may be said about the SCOP (the seasonal coefficient of performance during the winter for reversible systems. Note that the use of the terms Summer and Winter in relation to AC comfort loads refers to the typical seasonal period during which they apply and in some cases there will be parts of the building that have to be cooled all year round.g. 9. 8. It is also sometimes called the ”coil cooling energy” (kWh) or ”coil cooling load” (kW). 59 . 5. SCOP the Seasonal COP for winter: the ratio of WHG to WEC.

usage and regulatory environment were conducted by the EECCAC study participants for their country. have to rely in many cases on air conditioning to attain desired comfort levels in the existing building stock. could also be extended to the whole plant. Portugal. Simulation results. have targeted their CAC energy efficiency efforts purely on the reduction of cooling loads through measures to improve the thermal performance of the building shell in conjunction with electro-mechanical ventilation. and the hot/cold water generators. France. however. Germany. In some other countries. Data from manufacturers associations Here are some details on the four databases used in the EECCAC study. Italy. Thanks to the simulation the weights related to the partial load working conditions for the space of a year are obtained. temperature and humidity trend graphs. National surveys The national surveys of the CAC market. the results of simulations using computer models and from national survey data gathered for each of the countries that are directly represented in the study. Greece. this is made possible mainly because of the milder climate.Some countries. Full system efficiency The total efficiency value results from a combination of all the system components efficiencies at different load conditions. effective for the chiller. Spain. can be analysed to directly get the weights associated to the different part load conditions. expressed by airflow. This took advantage of each participants national contacts including assembling and syntheting rough data supplied by local manufacturers’ or importers’ associations or even involved subcontracting national consultants. such as France and Portugal. The results depend on the application context intended as the whole of the following factors: Climatic factors Building type Type of the activity carried out in the air-conditioned environments Plant type For all plants the estimation can be carried out by means of a simulation. such as Switzerland. 60 . The Mediterranean countries. because of their geographical position. We need a computer program to do this accurately.7. The simplified IPLV approach. The resulting set of country reports for: Austria. United Kingdom (with special thanks to the BRE) provides a unique set of data at the national level. a combination of policy measures have been deployed that combine passive measures to increase comfort in unconditioned spaces (and thereby resulting in a lower demand for AC) with those that are intended to raise the efficiency of AC equipment used in conditioned spaces. by evaluating with similar procedure air water system and terminals. 2. with an evaluation of the weight that each part load condition takes in comparison to the whole yearly functioning cycle. Statistical databases used and information gathered The conclusions of this study are based upon information gathered from four technical and market databases.

It should be stressed that all percentages are based on weighed statistics. except the particular case of the AHUs which were for 1999. both in terms of numbers and cooling power. under the condition that any data pertaining to a specific manufacturer is strictly anonymous. The statistics are intended to give the right weight to the country. climate.e. Manufacturers EUROVENT directories Directories National EUROVENT confidential confidential sales data sales data Two database are permanently maintained by Eurovent at EU level. with no figure being an arithmetic average. Checks on the consistency of various sources of information have found them to be reasonably high (i. etc. 61 . At the time of the EECCAC study Eurovent had also assembled a temporary data base on the numbers of AC units sold in 2001 that included their efficiency although this was not made available for use in the study. A data base similar to the Eurovent sales database exists in some countries within the national associations and some of these have been made available to the study. the Eurovent market data provide an extraordinary means to reconstitute CAC stocks. Correction and treatment of data This being said. which is a data base on total units sold without reference to performance. We used some estimated ratios to correct for these issues when no better data were available. is confidential although some segments have been made available for use in this study. Obtaining good market figures for the year 1998 was not the real objective but rather to gather enough data that would allow reasonable projections of the AC stock to be assessed. The data is generally based on the year 1998. There are also directories and technical literature of some manufacturers that constitute by themselves a data base that can be considered a good representation of the market. Projections on the size of the stock going back in time were made using rates based on measured data from 1996 to 2000. but approximately 90% of the total sales • Part of the chillers which were sold are not used in air-conditioning applications • There is a renewal rate (how many pieces replace identical worn or obsolete equipment?). with the additional benefit of giving an idea of public prices. and estimated data from prior to 1996. with discrepancies of a few tens of percent only). type of AC system. The first one is public: the directory of certified products. The second one ("sales"). The Eurovent market figures need to be corrected for the following three reasons: • Eurovent does not incorporate all the market. at least in relative terms hence allow an indication of the relationship between cost and efficiency to be established. Obviously real national data from country reports have replaced these ratios whenever they became available from the countries with a national participant.

including RACs. as shown in Figure 3.00 Mm2 60.00 ed Germany France 10.00 Italy M m Spain 2 ad 20.00 Others UK 5.00 80. and also by country Figure 3. Apparent annual additional building floor area conditioned by CAC from 1980 to 2000.00 20.00 1975 1980 1985 1990 1995 2000 2005 Source: EECCAC from Eurovent Experts and country reports 62 .1.00 0.00 40. by EU Member State (apparent means additions and replacements) 30.1. for the EU (apparent means inclusive of additions and replacements) EU-15 added (or repl.1. The demand for AC in Europe A general growth The CAC market is expanding rapidly in Europe.00 de d or re pl ac 15.3.2).00 1975 1980 1985 1990 1995 2000 2005 Source: EECCAC. ) have experienced larger rates of growth in AC than some more Southern countries such as Portugal or France (Figure 3. Figure 3.) m2 120. MAIN FIGURES OF AIR-CONDITIONING IN EUROPE 3.00 25.2. Germany.00 100. In fact a number of central European countries (Belgium. etc. Economic growth in the South is resulting in AC levels rising in regions where tertiary sector is important. Country reports National differences in demand The growth of AC is partly related to the differences in climatic conditions but also to the development of the tertiary sector especially offices. which are also mostly used in the workplace). Apparent annual additional building floor area conditioned by CAC from 1980 to 2000. The figures are given here for the total market.00 Greece Portugal 0.

Today just two countries.3.4. Figure 3. Share of CAC installed by tertiary sector for six European countries 100 90 80 70 60 Others 50 Trade 40 Offices & work places 30 Hotels / restaurant / bar 20 Hospitals 10 0 default France Portugal UK Austria Italy However CAC is also in competition with RAC so it is relevant to examine the type of building where each type of system are installed. by country Others Germany 13% 11% Greece 5% UK 8% Spain 24% Italy 25% Portugal France 2% 12% It is also pertinent to consider what type of buildings AC is being installed in. Spain and Italy.As a result of different growth. Apparent additional building floor area conditioned by CAC in 1998. which was large in the 1980s has become small in the 1990s. the relative weight of some countries like France or Germany as a proportion of total installed AC within the EU. account for more than 50% of the entire EU market Figure 3.3.4 shows the share of conditioned floor area by type of tertiary activity and country for CAC systems alone. Figure 3. (Figure 3. Figure 3.5). 63 .

00% 30. corresponding to hotels.6.00% education 50.00% trade 20.00% hotels&bar Rooftops RAC VRF chillers packages&splits hospitals This market is centred on offices and trade.00% 90. 64 .5. bar and existing medium office buildings. maybe due to their flexibility of use and installation.Figure 3.00% 80.6 (these fugures exclude room air conditioners). Technical response to the demand Market share of each technology The relative importance of each CAC technology in the European market is shown in Figure 3. Figure 3.2.7 indicates which systems and segments are experiencing the largest growth (from 1996 to 2000). The only exception are VRF type systems. 3.00% houses 40.00% 70. Share of conditioned floor space by building type for each AC system type across the EU 100. Share of installed conditioned space by CAC system in the EU in 1998 VRF Rooftops 3% Splits >12kW 7% 11% Packages 8% chillers 71% Evolution of market shares of techniques Figure 3.00% 60.00% offices 10. It is shared between CAC and RAC technique for economical reason (compared price) but also because various building sizes lead to the choice of distinct solutions.00% 0.

5% +2. The competition is focused on the "new" segment of smaller buildings (trade.5% are very different from the overall average growth rate of 9%.14 0. Average annual rate of growth in conditioned floor area by type of CAC for the period 1996-2000 0.8 illustrates for instance the importance of decentralised AC solutions in the trade sector while over the longer term the increased share of RAC sales within the total AC market corresponds to the same phenomenon.00% packages&splits chillers 20.9 shows the growth in conditioned floor area by each type of CAC&RAC system from 1980 to 2000 across the EU.00% Rooftops 40.1 0.7.5% The average growth rate for large splits of 14%.00% RAC 0.00% offices trade education hotels&bar houses hospitals Figure 3. which have correspondingly smaller average loads.8. etc).00% VRF 60.12 0. It should be remembered that "Splits" refers to large split systems of over 12 kW in cooling capacity and that smaller ones are included in the term "RAC".04 0.00% 80.00% 100. The percentage of AC supplied by each AC type by user sector for the EU in 1998 120.08 0. small offices.06 0.02 0 Large Chillers Packages Rooftops VRV Small Splits +8. for VRF of 13% and small AC for 10.16 0. Figures 3. We see a growing competition of RAC against chiller based solutions and an adaptation of solutions for the treatment of smaller sites. 65 .Figure 3.5% +9% +13% A/C +14% +10. Figure 3.

and central chillers dominating in Europe. The US and EU figures cover the same years (1999-2000) and the same building stock (non-residential buildings in use). as shown by comparison of the data in Figures 3. the preferred technical solutions are very different. Water 58% Nat. Water Classic(FCU) 1% Comparisons with US market Similar data supplied by the CBECS programme of the US DOE’s Energy Information Administration has been gathered for the US market.Figure 3. however.12. Figure 3.11 and 3.10. Figure 3.The share of chiller CAC systems (based on installed conditioned floor area) by sub-system type across the EU for 1998 Subsystems with chillers Two loops 2% Two loops Air (AHU) Air (AHU) Classic(FC 39% U) Nat. with packages dominating in the US. Total conditioned floor area provided by each type of AC in the EU tertiary and industrial sectors from 1980 to 2000 Market shares on TOTAL A/C market 160 140 120 100 RAC<12 kW VRF M m2 80 Rooftops Packages Splits >12kW chillers 60 40 20 0 1980 1985 1990 1995 1998 2000 CAC systems based on chillers account for the majority of the CAC market. 66 .10.9. which is the world’s largest. but among these there are two dominant subsystems with market shares of the same order of magnitude: chiller systems using AHU and those using FCU.

The share of non-residential conditioned building floor area cooled by each primary AC type in the USA in 1999-2000 US A (E IA ) chillers packages all RA C Figure 3.12. The share of non-residential conditioned building floor area cooled by each primary AC type in the EU in 2000 E UR (E E CCA C) chillers packages all RA C However the US market is so large in absolute terms that for every segment there is more conditioned floor area in the USA than in Europe. Figure 3.Figures 3. 67 .11.13.

and to have data on the growth rates of each.3. uses air distribution systems. The share of chiller systems using water distribution systems (based on installed conditioned area) in the EU % Water distribution/ Total 120% 100% 80% 60% 40% 20% 0% ly n y l ria e e s m ga an nd ai nc ec Ita do .14. A few technical trends on the market In order to use the data provided by European manufacturers to estimate total AC stock sizes by type of AC system hence to project the associated energy consumption.Figure 3. The share between distribution systems in chiller based CAC Based on adjusted numbers of AHU and FCU sales and applying some other scaling ratios obtained in the study. it has been possible to estimate the share of water distribution systems.14 shows the large variation in the share of chillers using water distribution systems by EU country. Figure 3. the installed area of installations with water distribution divided by the total area of installed AC.. It is necessary to have an image of the global industry and the main stakeholders. Conditioned non-residential building floor area by AC type in the EU and in the USA in 2000 Mm2 9000 8000 7000 6000 all RAC 5000 packages 4000 chillers 3000 2000 1000 0 USA (EIA) EUR (EECCAC) 3. In fact this value is equivalent to the installed area of installations with FCU divided by the total area of installed AC. Figure 3.13.e.. i. uses water distribution systems. it is necessary to be able to determine the proportion of conditioned floor area which: is treated by reversible AC systems. st Sp rtu m a re la m Au ng Fr er er Po G iu ki h G lg U et Be N 68 .

it's easy to see if it is reversible (the owner may use the reversibility feature or not) but for chillers tighter definitions are required.15. The share of conditioned non-residential building area provided by reversible CAC (for chillers only and for all CAC) and by water-based (using FCUs) distribution systems for four EU countries in 1998. 100% 90% 80% 70% Spain 60% France 50% 40% Italy 30% UK 20% 10% 0% Reversibility of Total reversibility %age of water chillers systems The choice between chiller-based systems and packages The share of different AC types in non-residential buildings (based on conditioned floor area) varies appreciably from one EU country to another. Figure 3. 69 . It is further assumed that the pure air cooled and the centrifugal chillers are none reversible.15. Figure 3. Figure 3.Reversible use of Air Conditioning One important aspect of this study is the reversible use of the cooling equipment for heating. The statistics on chiller reversibility are derived from data on a number of system sub-types: ƒ water cooled chillers including water-to-water heat pumps ƒ air-cooled chillers including condenserless water-cooled systems ƒ air-to-water heat pumps with reverse cycle ƒ water-to-air heat pumps with reverse cycle on a water loop ƒ centrifugal chillers either hermetic or open type according to connection between the motor and the compressor. It has been assumed here that reversibility is a feature of 10% of the water cooled chillers and all the air-to- water heat pumps.16. On a packaged or split product.

Market shares of AC technical solutions in four European countries (based on installed conditioned floor area in non-residential buildings) in 1998. Average cooling capacity of chillers in four EU countries. RAC VRV UK Rooftops Italy Packages France Spain chillers Large Splits 0% 20% 40% 60% Chillers are predominant in France while RAC Italy is equally divided between RACs and chillers. Figure 3.17.16. based on 1998 data 250 Cooling capacity (kW) 200 Size of large Splits 150 (kW ) Average size of 100 chillers (kW ) 50 0 UK Spain France Italy A number of "mini-chillers" with a small capacity are more popular solution in Italy than in other European countries (Figure 3.18).Figure 3.17. which corresponds to the importance in the AC market of small trading enterprises and small offices compared with the larger tertiary building complexes found in the UK. Figure 3. the Italian market of chillers when expressed in terms of the number of chillers sold is growing rapidly whereas some other national markets have risen smoothly or have even stagnated. 70 . Packages have a comparatively large market share in Spain as do VRV systems in the UK. Competition between “local” systems. VRV and mini-chillers for the medium-size building market is the dominant issue for the future. The average size (cooling capacity) of chillers is smaller in Italy and the other Mediterranean countries. As a result.

Figure 3.00 Index relative to 1996 140. Growth rates are high in countries with both small and large systems and smaller in countries with only large hydronic systems (UK) 180.00 Italy 40. In terms of market share the manufacturers can be categorised into three broad groups. Market share of the "Top Ten" European chiller manufacturers Name and main country grouped % of market average size A (3) Trane (FR) 35 % 12 % Carrier (FR-IT) York (UK-DK-FR) +Daikin B (7) Climaveneta (IT) 40 % 6% Clivet (IT) Mc Quay (IT) RC (IT) Lennox (FR) CIAT (FR) AERMEC (IT) C (30) Others (Same countries) 25 % <1% Source: EECCAC co-ordinator + Eurovent Experts Even if the most significant European manufacturers belong to American groups. Japan and the USA) which results in a transfer of technology and experience to their European branches.00 20. The largest are usually foreign owned companies.00 UK 80.00 0.1. it appears that many manufacturers operate on an EU- wide level.00 90 91 92 93 94 95 96 97 98 99 00 19 19 19 19 19 19 19 19 19 19 20 The value and nature of the European CAC market According to the information gathered for this study. The European CAC equipment industry is self-sufficient within Europe and is fairly concentrated although less than the car industry.00 France 120. which is 71 .00 100.18.00 160.00 60.g. Table 3. their centres of development in Europe constitute a rather autonomous technical base of the European AC industry.1. but it partly explains the operation of the market. Table 3. resulting from the fact that a number of the countries where they originate have a large and mature internal market (e. This does not mean that these local companies of foreign corporations have no technical autonomy.

ensures a great autonomy of supply for Europe. There is almost no freedom for installers and designers to be rewarded for the extra energy efficiency of the systems they may promote although operators can be reimbursed through performance contracting. 84. designers and operators all have to adapt to the customer demands.25 1700 &CT splits 90% 5% 85% 1. compared with its American equivalent. An indication of the problems and opportunities of CAC in the EU Accession States has been gained through a detailed study of the situation in 72 . Estimated value of the European CAC market by CAC equipment type (source: coordinator) Segments % production % import for Balance net Capacity Value MEuros/ Estimated EUR production MW/1998 Income (MEuros) /MW chillers 95% 5% 90% 6.1 0.15 100 VRF 50% 10% 40% 1. Extension to EU accession states of the CAC market is already a reality. Summer peaking may be a problem for some Southern European utilities but is often seen as a market opportunity for Northern European utilities. but nobody appears to be happy with their current regulations. however. It results in the independence and equality of the European manufacturers’ association.2% 21.the third largest in the world after the USA and Asia. They have to display a competitive cost. or be able to guarantee a high reliability (better servicing. Table 3. Thermal insulation. without considering the value of components imported from outside Europe compared with those produced in Europe). better contracts) in order to compete.6 . Building thermal regulations usually aim to minimise AC energy demand but often "don't know how” it should be done.2 1000 Total 87. Governmental agencies and ministries are responsible for the development of building codes. Other stakeholders Installers.7 0. very often also helps lower cooling needs.8 0. This base. Utilities are important stakeholders.2 200 rooftops 95% 5% 90% 0.25 1 250 FCU/AHU 100% 0% 100% 6.1 650 RAC 75% 0% 75% 5. which is often introduced into building codes to limit heating requirements.9 0.e. joined to the existence of average-sized manufacturers which are 100% European.2.8 0. Eurovent. leaving the designer free to choose the elements and to assemble them to reach the target (the obligation of results). European countries cope with this problem in different ways. Table 3. The problem arises from the lack of energy consumption calculation methods that are applicable to a wide range of systems. ARI. There is a hesitation between a pure prescription on some elements (an obligation of means) and a global limitation of demand. in some cases increased insulation can aggravate summer discomfort and increase the need for AC.0 0. 4100 weighted in income These figures do not include the value of installation but do include the profit margin of the equipment suppliers.2 200 > 12kW Packages 90% 10% 80% 0.2 shows the estimated value of the European CAC market without taking into account the imported contents of the equipment (i.7% . The US and Japanese markets are worth about 10 000 MEuros per annum on the same basis.

our recomendations on a European IPLV are not yet put in practice and we have based the study of present efficiency on the existing information : nominal EER. that cannot be compared) 4. Figure 3.5 3. 3. EER as a function of capacity and cooling medium for a chiller under 750 kW Figure 3.0 0 100 200 300 400 500 600 700 800 900 1000 Puissance frigorifique kW Source: Eurovent directory 73 . with distinct testing conditions (water cooled and air cooled. Note that in terms of chillers the directory is limited to 750 kW capacity which practically leaves uncovered the centrifugal chillers only.Romania.0 R2 = 0. In a first moment information was limited to electric power and cooling capacity. The Directory is meant as an instrument to direct the buyers by giving certified performance information.0073 3. There are two groups of chillers. To be perfectly representative the study should be based on a proxy of the SEER (like the American IPLV) because this is the figure having an influence on the electricity consumption.4. Now EER and COP are highlighted to promote Energy Efficiency. Statistics on present Energy Efficiency on the EU market Eurovent – Certification runs a directory of products on the EU market which gives good information of product performance. Chiller EER as a function of cooling capacity for 1998.0 1. however this type is sold always on specific request.0003 EER 2.5 1.19. Using data drawn from the Eurovent directory as well as a few individual manufacturers product directories. The methodology applied regarding the creation of national CAC stock statistics from an analysis of export and import figures can be applied in other CEEC countries in the same way as for the EU countries and be used to project CAC energy consumption and identify nationally specific issues. either for chillers or packaged air conditioners. Over this capacity the companies are very few and we have used directly data provided by some of them.19 shows the EER as a function of capacity for chillers on the EU market according to their mode of condensation. This has given some insights into how the findings may be applicable in the rest of the CEEC.0 2. a complete data-set of CAC capacity and nominal energy performance (at full load) has been assembled and analysed for chillers under 750 kW. The intention is to go even further by making use of part load information for a more appropriate selection of products. However.5 4.5 R2 = 0.5 Linéaire (air cooled) Linéaire (water cooled) 0.0 air cooled water cooled 0.

16 3.5 water cooled.55 3.48 2.04 100kW<CC≤300kW 99 2.86 Packaged.73 300kW<CC≤500kW 5 2. ≤50kW 8 2. reverse cycle. Summary of average and extreme EER values by chiller category on the EU market EER Categories Type Condenser Application min ave max Complete unit cooling air conditioning 1.09 Packaged. conditioning 50kW<CC≤100kW 10 3.99 2. cooling only.Tables 3.96 100kW<CC≤300kW 31 2.39 cooling water conditioning 2.4 indicate the average and range of EER values for CAC systems found on the European market.12 2. Average and extreme EER values for chillers on the EU market. split according to Eurovent internal categories.31 3.7 300kW<CC≤500kW 6 2. reverse cycle.56 2.04 >500kW 15 2. cooling only.43 2.34 3.48 2. air ≤50kW 6 3.52 2.62 4.49 2.98 4. for year 1998 Categories Cooling Capacity CC Number of EER EER EER in kW models min ave max Packaged.72 4.18 3.29 Source: Eurovent statistics Table 3.5 100kW<CC≤300kW 3 2.76 3.16 3.9 3.69 Statistically there is no relationship between chiller EER and its cooling capacity.39 cooled.21 3.79 4. conditioning 50kW<CC≤100kW 5 2.9 2.3.32 3.75 4. reverse cycle.18 3.9 3.06 water cooled.53 3.3 cooled.84 3.09 reversible water conditioning 2.73 4.56 2.55 3. on average there is significantly higher EER for chillers which are cooled with water compared with those that use air.46 >500kW 7 2.57 4.9 2.84 100kW<CC≤300kW 14 2. conditioning 50kW<CC≤100kW 35 1.31 3.3 and 3.27 3.09 Remote condenser. Table 3.93 2. ≤50kW 8 3.53 2. air ≤50kW 105 1.05 300kW<CC≤500kW 20 3.85 300kW<CC≤500kW 4 2.66 Packaged.4.25 conditioning 100kW<CC≤300kW 14 2. cooling ≤50kW 6 3.99 3. conditioning 50kW<CC≤100kW 102 1.77 3.28 3.41 2.03 3.9 3.31 3.76 3.34 3.57 2. 50kW<CC≤100kW 3 3. but rather represents the temperature regime found in cooling towers.96 3. The values used for testing the two types of system are somehow arbitrary and it may be that the 74 .8 300kW<CC≤500kW 5 3.09 Condenserless cooling water conditioning 2. water cooled.59 >500kW 6 2.9 3.73 2.13 3.1 2.85 3. In fact this improvement is not inherent to the chiller.29 reversible air conditioning 1.98 >500kW 7 3.9 2.53 only.49 3.45 3. floor cooling/heating Packaged.9 3.9 2.87 3.96 Floor 3.2 3. air ≤50kW 174 1.9 3.94 3.96 cooled. however.

3 and 3. Potential for efficiency gains of the selection of higher efficiency equipment It is clear that the apparent variations chiller EER seen in the product directories are partly explained by differences in the standard testing conditions.47 13. Interestingly.7 gives additional data that enable a comparison of chiller efficiency as a function of compressor type and condensing medium.5 W/W for those using scroll or reciprocating compressors.3 5 2.e. water-cooled systems are relatively expensive (because of the additional cost of using either a cooling tower or of accessing a natural water supply) and will therefore only tend to found for larger capacity systems. The following compressor types can be distinguished: . Based on the standard test data the average EER for water-cooled chillers is 3. then reciprocating. From a total of 698 chillers in the Eurovent database the type of compressor was known for 304 of them. as shown in Figure 3. not the best on average) are the ones with scroll.6 shows the proportion of chillers by compressor type within this subset of models as a function of the condensing medium.06 20.34 20 15 % 13. The only significant performance difference on overall averages is seen for the chillers with water-cooled condensers that use screw compressors who have an average EER of ~3.57 W/W whereas for air-cooled chillers it is 2. then screw.52 W/W.4) there is still a wide spread in product efficiency. when a piece of equipment is compared with its direct peers (as expressed through the product categories described in Tables 3.17 10 5.9 W/W compared to ~3. Figure 3. We note also that the best air cooled chillers (the top runner."reciprocating" (i."scroll" (orbital) .5. 75 . the average EER of the reversible systems is almost the same as for the cooling-only systems .5 W/W.apparent benefit from water cooling is not fully realised in practice. The average performance of chillers with air-cooled condensers is almost independent of the compressor type at ~2.44 1.5.58 0 75-80 81-85 86-90 91-95 96-100 101-105 106-110 111-120 121-130 EER/EER aver It is interesting to consider to what extent this difference can be attributed to differences in the type of chiller compressor used."screw" .6 11. with pistons) Table 3. Table 3. however.04 of models 11. Nevertheless. Distribution of EER/EERave (where the average EER is average for the same product category) for chillers on the EU market in 1998 25 21.

Table 3.6. partial statistics (not from Eurovent) on compressor type

Type Number % Air Water
Scroll 202 66,45 58,48 7,97
Screw 50 16,45 4,61 11,84
Reciprocating 52 17,1 8,55 8,55

TOTAL 304 100 71,64 28,36

Table 3.7. Comparison of chiller full-load nominal performance values depending on the type of compressor
and type of condensing medium

Type Cond. number Min.kW ave.kW Max.kW Min EER ave EER Max EER
Scroll air 178 12.2 49.8 158 1.9 2.5 3.39
wat 24 13.7 64 163.6 3.11 3.51 4
Screw air 14 196.1 451.3 789.1 2.35 2.5 2.66
wat 36 132 478.6 920.7 3.65 3.91 4.09
Reciprocating air 26 24.2 163.5 350 2.16 2.52 2.74
wat 26 136 407.2 847 2.99 3.54 4.06

It is also interesting to consider to what extent chiller EER depends on the choice of refrigerant. The data
shown in Table 3.8 suggests there is a small variation but perhaps not as much as had been expected.

Table 3.8. Chiller energy efficiency at full-load as a function of the type of refrigerant used

capacity EER
Refrigerant Min. Ave. Max. Min. Ave. Max.
R22 12.1 143.03 921 1.9 2.79 4.09
R407C 12.5 106.125 782.28 2.1 2.68 4.06
R22/R407C 12.2 62.68 163.6 1.9 2.6 4

Reciprocating compressors were prevalent for small cooling capacity systems but tend to give way to orbital
(scroll) compressors at medium cooling capacities, (because of multiple advantages: lower noise, wear, etc.),
and to screw compressors for the larger capacities.

EER for chillers over 750 kW
We have analysed a few chillers over 750 kW. Over this capacity the manufacturers are very few and we
have used directly data provided by some of them. The EER are very high. It’s a niche market (large malls,
airports, some district cooling) which is completely separate from the rest.

76

4. FACTORS GOVERNING THE DESIGN, SELECTION, INSTALLATION AND
OPERATION OF CAC SYSTEMS
4.1 Actors involved with CAC systems
The main barriers to efficiency
As opposed to RAC, CAC are usually designed and specified by a chain of engineers or technicians, who
define the system assembly for a given building without the direct influence of the customer. Everybody
could gain something from the marketing of efficient CAC, but each actor has a limited vision of the chain:

¾ the consideration of initial cost as the only decision criterion by most designers and installers,
because this is almost the only way the customer judges them ;
¾ the separation between the plant owner and the renter, between the renter and the operator, etc. ,
nobody being ready to pay for the other’s benefit ;
¾ the customer only judges the initial cost because he is not aware of the other aspects (no coded
information on other aspects) or not interested (owner and occupant having distinct interests);
¾ the competition between manufacturers is only expressed in terms of Euro/kW, not Euro/EER or
Euro/kWh consumed later ;
¾ the absence and/or intrinsic difficulty of developing building codes for this relatively new source of
energy consumption which is very complex to model and characterise;
¾ the problem for consultants of completely specifying and certifying the quality of something so
complex, built on site and only once,
¾ the lack of incentives for energy efficiency for most system operators, except in case of a good EPC
(Energy Performance Contract),
As a result of these complex factors the measures to be proposed in EECCAC cannot only address the
efficiency of the CAC equipment supplied to the market, which in many ways is the smallest problem, but
should also aim to reshape and activate the chain leading to the final service: the conditioned square meter.
Hence the necessity to develop a better description of the chain.

4.2 Practices and procedures adopted in CAC system design
Guidelines for the design of CAC systems
Standards and guidelines for the design of CAC systems are often developed nationally within EU Member
States. Typically they take the form of technical documents released by national associations as follows:
UK: CIBSE guides
France: AICVF guides
Italy: AICARR guides
There is also an unpublished REHVA guide (REHVA, 1997) based on the national guides of Italy
(AICARR), France (AICVF) and the UK (CIBSE), which makes a synthetic presentation of the following
issues:
• comfort conditions
• air quality, noise
• cooling loads
• system types description and naming
• sizing
• reference meteorological data.
CEN and other standardisation bodies have not yet addressed this subject, probably because system sizing
brings a responsibility that the solution "will work", which is not a responsibility normally taken by a

77

standardisation body. The technical options related with sizing have to be defined in such a way as to leave
degrees of freedom to the designer that will bear this responsibility.

4.3Previous market-transformation efforts within the EU (equipment)
In most EU countries air conditioning has been viewed for years as a very marginal electricity use
compared with the largest end uses and has therefore not received much attention. As a result the market
has developed spontaneously and has established its current terms of reference without having
embedded energy efficiency within the decision and design criteria, with the exception of a few local
voluntary efforts.
Utilities, national energy agencies, ministries, manufacturers’ associations, etc. have made a few
attempts in the past to advise the public or the professionals about energy efficiency. But in terms of the
impact on the market, little value has been given to energy efficiency. As shown in section 3.4, the
market doesn't pay for improved EER; it gives some value to the brand name, a little to reversibility, but
nothing to improved EER. If observed the other way round, this has resulted in the current
circumstances where purchasers can acquire more efficient equipment at no extra cost. Since the
operating costs for such equipment are lower, this results in a negative incremental cost for
purchaser/users of energy efficient AC. In fact this ‘free ride’ is partly illusory because manufacturers
with an interest in energy efficiency are not recouping the investments they make in producing more
efficient equipment and this naturally holds back the rate at which even higher efficiency equipment is
developed. Even if the implementation of new energy efficiency policy measures eventually causes the
average cost of AC equipment to rise, the changes will still be profitable for the customer.
In the EU, Energy Efficiency improvement is not presently a criterion playing any major role in the design
and installation process (see 3.5); scattered performance improvements happen spontaneously, even if some
actors give them some importance locally. As examples of such local or national efforts, we present
hereafter a few indications.

The Eurovent Certification programme
The Eurovent-Certification programme is a trans-national AC energy performance-certification programme.
The managing body, Eurovent Certification, is a business association created specifically for the purpose. By
participating in the scheme and allowing their products to be independently tested, manufacturers have the
right to include their products in the annual Eurovent product directory, which is circulated to around 20 000
consultants and installers. They are also allowed to use the Eurovent Certification label (Figure 4.1). For a
cost of less than 0.2 % of their total turnover manufacturers can, depending on the number of models, have
all their models listed in the directory.
The equipment to be tested is independently selected by Eurovent Certification (not by the manufacturer)
and then tested according to international standards and the specific requirements of Eurovent Certification.
In order to ensure true comparability and reproducibility of the test results all equipment are tested in a single
designated test centre. There is always a risk that some manufacturers certify only their best equipment.
Eurovent has moved, following a similar move by their American counterpart, ARI, to a "Certify All" policy
wherein a manufacturer can only report performance data of equipment BEING ALL certified if it wants to
claim the benefits of its participation to the Eurovent scheme.
About 10% of all models on the European market are tested by Eurovent Certification each year. The fear of
a test result that could contradict a manufacturer’s self-determined values has led a number of manufacturers
to readjust the declared EER values printed in their commercial literature.
A high percentage of the European AC market is already included in the Eurovent Certification scheme
(~80-90% of the total market) and all the most important manufacturers participate in the scheme. This is
very useful to ensure the reliability of the efficiency data quoted in the EECCAC study and should provide a
strong basis for the establishment of voluntary and transparent agreements between the EU and the
manufacturers.

78

test conditions (temperatures. In order to be able to compare products from different manufacturers. Therefore in order to avoid possible discrepancies. . Sometimes. For other programmes. this principle could not be applied for all programmes for various reasons. Acceptable deviations from test conditions and required accuracy of measuring devices are also specified.1 Label stamped on Eurovent certified appliances. There are always subtle differences between laboratories and human factor plays an important role.Figure 4. Eurovent Certification is based on a few principles: . a progressive implementation of this principle has been necessary. humidity.regular testing by independent bodies must be carried out on randomly selected units. When all products presented by a manufacturer in the given scope are certified. An example of a utility-led energy-efficient AC promotional campaign EdF aims to establish reversible air-conditioners as the principal heating source in new French households and is directing its promotional efforts in that way.all products in the defined scope must be certified (“certify-all principle”) . This is applied at ARI which served as a model for Eurovent procedures. any laboratory applying the same test method should in principle obtain the same results when testing the same units. the image of certification is clear and there is no possibilities for misunderstanding. In practice. Air conditioning and refrigeration equipment are complex by their nature and it is not possible to determine their performances with a sufficient accuracy without testing. the testing capacity (expressed in number of tests to be performed during a year) of available laboratories was not sufficient. incorporation of unit to be tested. it is not possible to achieve such an ideal work. However. Finally Eurovent Certification now uses nine independent laboratories from five countries for its thirteen programmes. Certify-all principle is not yet fully applied in all certification programmes but the goal will be achieved in a few years. etc. testing must be performed according to precisely defined procedures. Test standards contain specifications for test installation. 79 .Test results must be very close to claimed performance characteristics – otherwise the catalogues data must be re-rated. These procedures are generally described in test standards which now exist on European or International level for almost all products in the scope of Eurovent/Cecomaf. Everything being well defined.) and method of calculation of performance characteristics. the principle of single laboratory for a given product was introduced in Eurovent Certification. However for some equipment. flow. historically several laboratories had been used by participants and it was difficult to select a single test house.

this provides a means of introducing indicative performance levels ("good practice".2 mode Heating 7 6 40 45 COP>2. perhaps at European level –e. and identified a range of market transformation options that can be expected to constrain the expected increase. Information on this programme is accessible at www. Minimum energy performance requirements for reversible heat pumps (or chillers) of the air /water type with a capacity higher than 30 kW (Application in fan coil or hybrid systems (FC or radiators + h&cfloor)) that are included in EdF’s promotion Generator Outdoor Unit Water Coefficient of performance air / water minimum Dry Bulb Wet Bulb Inlet temperature Outlet temperature temperature(°C) temperature(°C) (°C) (°C) Cooling 35 ** 12 7 EER>2. Others are highly desirable but require legislation. revisions to building regulations.com 80 . For air-conditioning.5 * temperature function of the flow and identical to the one in heating mode at à +7°C outside ** non-controlled The commercial importance of being included in EdF’s scheme has resulted in these efficiency requirements having a substantial impact on the efficiency of the market in France. All values used are Eurovent certified values. Some of the policy options are measures that have already been agreed upon.mtprog. before being used by EdF. A key feature of the programme is that the policy development is open to public comment. An example of these requirements are given in Table 4. EdF also obliges products promoted within the directory to attain a minimum energy efficiency value under each test condition. Supplying the data to Eurovent means that it can be independently verified. One specific activity is to place the Eurovent performance data on an (existing) interactive product performance database.Manufacturers who participate in the EdF scheme must supply Eurovent with more test-point data than the minimum which is required to be quoted in the Eurovent directory (which is currently the T3 test results in the heating mode and the T1 test results in the air-conditioning mode). through their certification process. the analysis and debate has identified measures that can be implemented without legislation. The UK Market Transformation Program The primary purpose of the UK Market Transformation Programme is to develop quantified policy options for the government regarding measures to improve the energy and environmental performance of products.5 mode -7 -8 * 45 COP>1. In addition. In addition to providing system-specifiers with easier access to the data. Table 4.1. but there are many sets of requirements for the various equipment types. subject to independent testing. for example.g.1. "best practice") that can form the basis of voluntary procurement programmes. and the involvement of key stakeholders is actively sought. the programme has thus far modelled the carbon emissions consequences of three scenarios. mandatory minimum efficiency standards for AC equipment. EdF and Eurovent have an agreement wherein Euroevent will do additional certification testing for the RACs that are in the EdF programme.

The electric heating power provided by the Joule effect cannot exceed 5% of the total heating power installed. Free cooling is mandatory in "all-air" systems with a ventilation air flow greater than 10000 m³/h. ƒ Energy recovery is promoted. according to the Table 4. ƒ Power stages should allow adaptation to the demand. i. There is a hesitation between a pure prescription on some elements (obligation of means) and a global limitation of demand. the improved thermal characteristics and solar factors set out in the RSECE are a design option and are not mandatory. better lighting systems) to compensate for not attaining the 20% improvement in envelope performance. when the rejection air has a thermal power greater than 80 kW. The problem arises from the lack of consumption calculation methods applicable to a wide range of systems. Terminal re-heating is allowed for cooling-only systems but cannot exceed 10% of the installed cooling power. Portugal appears to be the only country in the EU following an alternative approach and has imposed limiting values on both summer and winter energy needs through the building code known as the RSECE (Decree-Law 118/98). Among EU Member States only Portugal and the UK have significant measures in their building regulations designed to limit the energy consumption of air conditioning systems. The reference thermal characteristics of the building envelope for buildings using HVAC under the RSECE are more thermally efficient than those required by the general regulation on the thermal characteristics of buildings. 81 . and (iii) the respect for environment. Rather they are one of the aspects taken into account in the calculation of the nominal power of the system to be installed. leaving the designer free to choose the elements and to assemble them to reach the target (obligation of results). In these circumstances it obliges the use of central AC systems rather than distributed RACs. The reason is that the RCCTE was been designed mainly for application in the residential sector while the use of the larger HVAC systems addressed in the RSECE justifies stronger requirements. The RSECE defines regulations that need to be complied with when HVAC systems are sized and installed.2. ƒ Limits to individual AC appliances. Portugal: An example of a national scheme to promote energy-efficient AC through building thermal regulations National building thermal regulations in Europe usually only address the minimisation of winter energy consumption and are not intended to influence energy consumed in air-conditioning. Energy recovery from the exhaust air is mandatory during the heating season. Thus. the RCCTE. (ii) quality and safety of the facilities. these equipment are required to have a number of stages depending on the power. It applies to all HVAC systems with an installed cooling power of more than 25 kW or when the aggregate heating and cooling thermal power is more than 40 kW. However. The RSECE sets out a number of limits to be applied to the equipment energy use: ƒ Limits on the Joule effect. Individual air conditioning appliances for heating or cooling are only allowed in spaces with special internal conditions. so as to ensure (i) a minimum energy efficiency level of systems and equipment used to meet indoor thermal comfort and air quality requirements. To minimise the heating and cooling generators from working at partial load. nor 25kW by independent zone. conditions which are different from the rest of the independent zone.4Existing national regulations within the EU (which apply at the system level) In general building thermal regulations "want" to reduce the energy demand associated with air conditioning but usually they "don't know how”.g.4. ƒ Limits on terminal re-heating. These are described in detail below. Different countries cope with this problem in different ways. the designer is free to choose other options (e.e.

Summary of UK building regulations for space cooling and ventilation The legal requirement specifically refers to air conditioning and mechanical ventilation systems that serve more than 200 m2 of floor area. This also considers the performance of the building as a whole. These are: The Elemental Method. which considers the performance of each aspect of the building individually (e. cooling towers. offering increasing design flexibility in return for greater demands in terms of calculations. It requires that the proposed building should cause carbon emissions that are no worse than a notional building that satisfies the requirements of the Elemental Method. hydraulic tests. insulation levels and heating system efficiency. providing adequate shading. The regulation imposes the adoption of a maintenance plan and a monitoring system. or designing for night cooling operation. heat exchangers. chillers (power and efficiency). by imposing minimum u-values). Some flexibility is provided for trading off. noise levels and overall functionality. For spaces with glazing facing only one orientation.0. Compression machines must have a COP (EER) greater than 2. the EMS should allow the centralised optimisation of the parameters. Commissioning tests are required for boilers. air conditioning and lighting systems be capable of being operated in such a way as to limit the carbon emissions per square metre below a given benchmark. The energy consumption of all equipment with an electric power greater than 12. minimum number of stages required in the RSECE Heating Cooling Power Nº Stages Power Nº Stages <100 kW 1 <40 kW 1 100 to 1000 kW 2 40 to 200 kW 2 1001 to 4000 4 201 to 500 4 >4000 kW 6 >500 kW 6 ƒ The attainment of minimum efficiency levels for AC equipment is compulsory.g.Table 4. For systems with a thermal power greater than 500 kW. Energy Management Systems (EMS) are required for systems with a thermal power greater than 250 kW.5 kW should be independently metered. The UK employs three alternative methods for demonstration of compliance with the national building thermal regulations for tertiary buildings. and requires that the heating. pumps. controllers. The Carbon Emissions Calculation Method. but is applicable to any building type. Elemental Method: Avoiding solar overheating The general guidance is that naturally ventilated spaces should not overheat and that cooled spaces should not require excessive cooling plant capacity. There are less detailed whole-building methods for schools and hospitals. ventilation. This mainly applies to offices.2. The Whole-Building Method. the requirement will be satisfied by limiting glazing area to a percentage of the internal area: North 50% NE/NW/S 40% E/SE/W/SW 32% 82 . for example. This may be satisfied by limiting glazing area.

the space will not overheat when subjected to an internal heat gain of 10 W/m2 Elemental Method: Heating efficiencies The carbon intensity of the heat generating equipment at maximum output and 30% (system) output should be no higher than: Maximum output 30% output Natural gas 0. The maximum allowable ratings (all in kgC/m2/year) for new installations are: New building Existing building Air conditioning 10.113 In effect. Notionally. in the absence of mechanical cooling or mechanical ventilation.showing by (acceptable) calculation that.35 For substantial modifications to existing systems. it operates as a limit on the installed cooling power and fan power per m2 of floor area. air conditioning systems should have a satisfactory "Carbon Performance Rating" (CPR). In practice. For offices.088 (all figures in kgC/kWh) However.074 electricity 0. Key calculation parameters: Equivalent hours of full load operation: 83 . Carbon emission factors for different fuels are tabulated: significant figures are natural gas 0.091 0. the performance must be the least demanding of either these values or a 10% improvement on the original value.the solar heat load per unit floor area averaged between the hours of 7:30 and 17:30 would not be greater than 25 W/m2 with the solar irradiances for the location that are not exceeded more than 2.065 Other fuels 0. this is a limit on the carbon emissions per m2 of floor area from air conditioning or mechanical ventilation systems under standard operating conditions. since the calculation procedure prescribes standard figures for equivalent hours of full load operation. Benefit can be claimed for a number of control and other features. it is acceptable to show either that: .053 oil 0. these figures taken together define the minimum CoP that is required for heating by reverse- cycle operation of air-conditioning Elemental Method: Air conditioning efficiency. these figures may be exceeded if building insulation levels are increased beyond the minima.5 7.068 0.3 11.35 Mechanical ventilation 6.Horizontal 12% Alternatively.5% of occasions for July (between 1976 and 1995) .

column B when any of: energy metering. including modular plant 0. this is "partial ice storage".cooling plant 1000 hours per year The calculated CPR may be reduced by applying factors to the fan or cooling capacity to reflect energy- saving design features.90 0. This causes some difficulties when applying the Carbon Emissions Calculation Method (see later) and guidance on the calculation of SSEER is being prepared.90 0. there are no explicit requirements for air conditioning system efficiency.95 "Efficient control of capacity" requires good part load efficiency (without defining this): hot gas bypass does not qualify.93 0.95 Controls which restrict the hours of operation of the system 0. It is only permissible when the perimeter zone of the space accounts for more than 80% of the floor area. internal zone temperature monitoring are provided.95 Efficient means of controlling air flow rate 0. there is a general requirement that components such as fans.95 Controls which prevent simultaneous heating and cooling in the same zone 0.93 0.90 0.90 0. run-hour metering.95 "Mixed mode" requires the provision of sufficient opening windows and an interlock to prevent air- conditioning operating when windows are open.95 Partial ice storage 1.. column A when. "Full ice storage" requires chillers to operate only at night: if day and night chiller operation is intended.93 0.85 0.75 0. in addition.95 Efficient control of plant capacity. 84 .90 0. For buildings other than offices. They are multiplicative.90 Full ice thermal storage 0. However. Refrigeration plant: A B C Free cooling from cooling tower 0. "Air flow rate control" requires variable speed drives or variable pitch fan blades: throttling or inlet guide vanes do not qualify for the allowance. the monitoring system has the ability to draw attention to "out of range" values.86 1.85 0.90 0. so there is scope to introduce performance requirements. Air distribution systems: A B C Operation in mixed mode with natural ventilation 0.90 0.95 Variation of fresh air using economy cycle or mixed mode 0. Column C figures apply if no plant monitoring is provided.80 1.93 0. pumps and refrigeration equipment are reasonably efficient and appropriately sized.90 0.mechanical ventilation or air conditioning fans 3700 hours per year .95 Controls which restrict the hours of operation of the system 0. These factors depend on the level of plant monitoring provided.85 0.93 0.85 0.

5 20.for example. commissioning details. and for new systems in existing buildings or substantial alterations to existing systems should not exceed 3. rather than the principle. the design assessment for CPR or other benchmarks. These relate to its implementation. and details of all meters provided. Carbon Emissions Calculation Method This route permits the greatest design flexibility. designed to comply with the Elemental Method. the electricity demand (of motor rating) per airflow unit (expressed as W/litre/second).5. Of particular relevance to the EECCAC study is the need to make explicit assumptions about . lighting and air-conditioning systems. A recommended template for this has been published by CIBSE. There is also a requirement that sub-metering be provided.or calculations of .some design parameters (notably air-conditioning system efficiency) are not defined in the elemental method but are required to enable the comparison to be carried out. There are. There are exceptions for non-comfort applications Whole-building method For offices. The main problems are: . and preferably 1. It requires the designer to show that the carbon emissions from the proposed building are not greater than those for a notional building of the same shape and size. but a completed copy of Appendix B of CIBSE AM11 "Building Energy and Environmental Modelling" is an acceptable demonstration. the SPF should not exceed 2. Specific Fan Power. air leakage of the building.8 Mechanically ventilated 10. probably high-profile.the designer is required to design two buildings and their systems in order to demonstrate compliance . There is no prescriptive list of acceptable calculation methods.2004). buildings. but the building envelope requirements of the elemental method must also be complied with. still some constraints on acceptable values for some parameters . amongst other things.0 Air-conditioned 18.0. it has already generated some criticism. For ACMV systems in new buildings.4 Calculation procedures are explained in BRE Digest 457. Guidance on both these points.1 7. however. operating instructions. proportioning of cooling may be 85 . and encouragement of variable flow control. in the form of a simplified calculation for the notional building. as an alternative to the air-conditioning CPR.0 11. Information Provision: metering and logbooks The Regulations require that the owner and/or occupier of the building be provided with a logbook that contains.about the seasonal system energy efficiency ratio (and seasonal system coefficient of performance). The required values of whole-building CPR are: Building type New office refurbished office Naturally ventilated 7. a similar calculation may be calculated for the combined emissions from the heating. Although this route is expected to be used only for a few. This includes separate metering for tenancies of more than 500 m2 (though for tenancies below 2500 m2. This provides greater building services design flexibility.0. has been prepared (CIBSE.There are requirements for SPF.

that the cooling load should be as low as possible according the state of the art. The objective of the Directive is to promote the improvement of the energy performance of buildings within the European Community. any chiller installation (which may include more than one chiller) of greater than 20 kW input power should be separately metered. and any motor control centre providing power to fans and pumps of more than 10 kW input power. as well as indoor climate requirements and cost effectiveness. The approach under consideration is based on the standardisation of algorithms used in a simulation tool known as Consoclim. Law N. has a general theme of the “rational use of energy in buildings” and is intended to limit “energy consumption” as much as possible and encourage improvements in energy efficiency. which should include maximum limits for the energy consumption of air conditioning systems and which may ultimately be integrated into the German energy saving regulation. DPR 412/93) addressing the heating systems of buildings has been established but the same has not yet been done for the cooling system. including the heating and cooling systems. taking into account outdoor climatic and local conditions.g. UNI standards and similar) which give the physical parameters of applications. are preparing guidelines for the application of the European Energy Performance of Buildings Directive (see below) that shall subsequently be transposed into Italian Law. The Italian Engineering Association AICARR and CTI. laws are followed by Law-Decrees which establish ways and means to fulfil law dictates. Implementation of the Directive is subject to the principles of subsidiarity and as such the Directive is not always precise and does not specify detailed implementation mechanisms. This all encompassing law. The status of regulations in other EU Member States France France is in the process of developing building thermal regulations. it is intended. Spain also has a voluntary building energy labelling scheme which includes specifications for cooling system energy performance. member states have some flexibility concerning the exact manner in which it is to be implemented. Italian Thermal Committee. Studies. The Energy Performance of Buildings Directive (to be transposed nationally) European Directive 2002/91/EC on the Energy Performance of Buildings was published in December 2002. a number of stopping devices. Spain There are no requirements in the national building regulations regarding the energy performance of central air conditioning systems although there are requirements for maximum U-values. The new Energy Saving regulation of 2001 requires. Generally. however. but does not specify how this is to be determined. Some specific demands which are already certain to be included in the new regulation are applied and have been integrated in the existing code: a bonus for more efficient terminal units. and finally official Standards (e. 10 of 1991. reduction of solar gains. A law decree (no. Germany Germany has recently initiated a process to develop national regulations to limit the energy consumption of building active cooling systems. Under Italian legislation. not even the one on comfort without air conditioning. reduction of lighting and ventilation loads. proposals and recommendations are still under examination. Transposition of the Directive into national law is required by January 2006. to generate within the DIN 4601 a new part (Part 11). The full 86 . no regulation is applied. Accordingly. which will address the cooling mode and more immediately a means of determining building energy performance although the results are not expected for this year.acceptable). Italy In Italy. In the absence of a full calculation method. separation of air renewal of different buildings. yet none has been transformed into official documents so far. thermal regulations affecting the building envelope and internal energy-using systems. and a number of local control obligations. fall within the framework of a general law.

through the following: A) Establishment of a general framework and common methodology for calculating the integrated energy performance of buildings. The Directive aims at reducing the consumption of energy in new and renovated buildings. The draft Directive on Energy Demand Management (to be defined) The objective of this proposed Directive is to complete the internal market for energy by developing and encouraging energy efficiency on the demand side. The determination of this is to be based on the results of a technical-economical analysis. This includes a supportive framework for implementation and financing of energy services. The draft Framework Directive for “Eco-design of End-Use Equipment” (to be adopted) The European Commission has developed a draft proposal for a new Directive. adapted to each Member States’ liberalised market. It also creates an EU-wide technical committee comprised of representatives from Member States that will be responsible for the development and maintenance of the inspection rules. Article 8 requires central air-conditioning systems over 12 kW to be regularly inspected. C) Certification schemes for new and existing buildings on the basis of the above standards and public display of energy performance certificates and recommended indoor temperatures and other relevant climatic factors in public buildings and buildings frequented by the public. Article 9 requires Member States to put in place a system that ensures that certification of buildings and inspection of equipment are carried out by qualified and independent personnel.e. This proposal is in lieu of additional public service obligations in the Amended Internal Market Directives and the Commission’s 87 . excluding industrial buildings. Buildings of more 1000 m2 in floor area are to addressed in A. The annex of the Directive stipulates that the level of energy efficiency used in the standards will be set aiming at the least life cycle cost for the final users using a real discount rate of 5% and realistic assumptions about product lifetime. (e. The measures (B) and (C) envisaged in Article (6) may require the modification of the insulation.. lighting and ventilation requirements existing in some countries and thus have an indirect influence on air-conditioning. especially as it is provided by utilities and service companies in the form of energy services. air conditioning system inspection and building certification (see below) may not be delayed beyond January 2009. B) The development and application of minimum energy performance standards to new buildings and to certain existing buildings when they are renovated. The field of application of (D) for CAC i. even after the harmonisation. It is envisaged that Member States will set targets to promote and support energy efficiency services. D) Specific inspection and assessment of boilers and heating/cooling installations.g. As yet there is no clear time line regarding when this draft will be submitted to the council of ministers and parliament for approval.4 2004. A minimum energy efficiency target to be reached through energy services each year is proposed for Member States that corresponds to 1% of the total electricity and gas sales. namely in the very technical field of CAC.application of requirements for boiler inspection. The method of calculation to be established in (A) is the responsibility of each Member State and thus is not unified across Member States. An Annex to the proposal contains the main aspects to be taken into account when calculating the energy performance of buildings and requirements for inspection of boilers and central air conditioning systems. There could be obvious advantages in harmonising such national building codes. All these measures are to be taken nationally. heating remains the essential concern of EU building codes. Indeed. B and C. especially for smaller energy consumers such as households and SMEs. > 12 kW is exactly that of this study. which amongst other measures would give the Commission the right to establish mandatory minimum energy performance standards (MEPS) for end-use equipment. before January. third party financing) and programmes. It should include the energy consumption of AC and ventilation.

It is a little bit out of fashion due to the existence of other typical contracts. the company gives its estimate on operational budgets. as it is needed.5 Regulatory structure and market transformation at the wider international level Minimum efficiency standards and energy labelling in the USA The US Environmental Protection Agency (EPA) has implemented its ‘Energy Star’ voluntary award-style energy label for central air-conditioners and heat pumps that satisfy minimum energy performance criteria. Indeed the guarantee of the results implies a perfect knowledge of the installations but also. A contract of results is incontestably the form which it is advisable to give to a technical management contract when there are. It does not apply to low-rise residential buildings.2 standard. Whereas a contract of means can be of low duration. As described in Chapter 2. most AC equipment must attain a minimum EER and/or SEER level prescribed by the USDOE to be allowed for sale on the US market. Minimum energy performance standards (MEPS) are the main energy efficiency policy option presently being implemented in the US for AC systems. Contracts of "results" The contracts with obligation of results strongly engage the responsibility for the company which must fulfill successfully the mission which is defined by the contract. by nature. very often. commissioning and adjustment of the installations.1: a comprehensive approach to raise CAC energy efficiency The objective of the ASHRAE 90.1 standard (ANSI/ASHRAE/IESNA 90. It implements the means that it judges necessary. until obtaining the contracted result. expensive and complex air conditioning installations. 4. However. the contract of results can be only a contract of long duration. the building owner entrusts the execution of specific tasks to a company.1-1999) on the ‘Energy efficient design of new buildings except low-rise residential buildings’ is to “provide minimum requirements for the energy-efficient design” of commercial buildings.Amended Rational Planning Techniques Directive proposal from March 1997. Presently labelling is not the main means of action on the market of central systems because it has a low impact and that Minimum performance Standards and building codes are more efficient in influencing CAC efficiency in the US . which are covered under the ASHRAE 90.1 standard provides: 88 . Thus. its guarantee on the quality of air conditioning and well-being in the buildings. Their importance derives from the importance of the public markets. Practices and procedures adopted in CAC system operation The operation and maintenance of CAC systems is usually contracted out to a specialist company. This type of contract in general defines only frequencies of visits and the nature of the services to be carried out as well as labour and material means. as described below. ASHRAE 90. as central air-conditioning systems are designed and installed on-site by professionals policy measures which address the overall quality and energy efficiency of the system design are also required. The 90. significant investments in time for the knowledge. on the maintenance of the materials which are entrusted to them and the compliance with the code of practice. Two broad types of contract are used: Contracts of “means” Within the framework of a contract with obligation of means.

1—1999. New systems and equipment in existing buildings. “In general. For alterations and additions. beginning with the simplest and most common design obligations.1 which was issued in 1999 has several differences from the previous 1989 version. which comprises mandatory provisions and prescriptive requirements 2. (b) the following systems and equipment used in conjunction with buildings: 1. ventilating. Energy-conscious comfort in ASHRAE 90. too. The technical sections of the standard.1 sets minimum requirements for the simulation software. the 90. heating. Standard 90. and TRACE™. Using the ECB approach requires simulation software that can analyse building energy consumption and model the energy features of the proposed design. New buildings and their systems. has been substantially reorganised compared with the 1989 edition. It has been reorganised for ease of use. or commercial processes. New portions of buildings and their systems.” A life-cycle-cost analysis was used to define the criteria in the 1999 edition and thereby balance energy efficiency with economic reality. or 2. More importantly. Suitable programs include BLAST. DOE-2. and 5. Note. such that the new standard clarifies requirements and provides a simplified compliance path for small commercial buildings.4 Btu//h*ft2 (10W/m2). The latest version of Standard 90. The ECB method permits tradeoffs between building systems (lighting and fenestration. ventilating. service water heating.1 focuses on comfort conditioning rather than industrial.(a) minimum energy-efficiency requirements for the design and construction of. Section 6. lighting. 2. that the stated purpose of the standard is to provide minimum requirements.1 User's Manual notes that. But that's not all. HVAC-related requirements are presented in order of complexity. An Energy Cost Budget method. manufacturing. electric power distribution and metering provisions. 3. cooled by a cooling system whose sensible output capacity is greater than or equal to 5 Btu/h * ft-2 (15 W/m2). for example) if the annual energy cost estimated for the proposed design does not exceed the annual energy cost of a base design that fulfils the prescriptive requirements. A Prescriptive Path. Standard 90. 4. specifically address components of the building envelope. Moreover. 1. (b) criteria for determining compliance with these requirements. electric motors and belt drives. and 3. the 1999 edition expands the standard's scope to include both new and existing buildings and building systems. service water heating. lighting. the prospective design must first satisfy the general requirements and mandatory provisions of each technical section.1—1999 addresses building components and systems that affect energy usage. Standard 90. The provisions of the standard apply to: (a) the envelope of buildings provided that the enclosed spaces are: 1.1. Because the HVAC section is 21 pages only the key requirements are summarised here.1 The present HVAC section of Standard 90. and motors. Section 6 of ASHRAE/IESNA 90. some sections also include prescriptive and performance requirements.1—1999 describes mandatory and prescriptive requirements for commercial heating. and air conditioning. Sections 5 through 10. heated by a heating system whose output capacity is greater than or equal to 3. The design must either (a) fulfil additional prescriptive and performance requirements described in each technical section or (b) satisfy the energy cost budget (ECB) method. To comply with Standard 90. the Standard only applies to building systems and equipment…that are being replaced. It also defines three methods for compliance: 1. power. 2. Each technical section contains general requirements and mandatory provisions. a designer or owner can always exceed these basic conditions for compliance. which combines mandatory provisions and a computerised methodology that permits tradeoffs between various building systems and components 89 . HVAC systems and equipment. and air-conditioning systems.

0 EER (86°FEWT) ARI320d(ARI/ISO-13256- 4tons (cooling mode) 1after29-Oct-2001) Water-cooled screw chiller. that is 0.) 2.7EER ARI340/360 15tons Water-source heat pump. Another prerequisite (there are others) is that each air- cooled or evaporatively-cooled HVAC system serves only one zone.80 COP 3. the “simplified approach” consolidates the provisions on roughly two pages so that design professionals can quickly locate all applicable requirements. Drawings. Controls. and a narrative of the system operation must be supplied to the building owner. The 1999 standard upgrades the minimum efficiency requirements for many types of HVAC equipment and adds efficiency requirements for heat-rejection equipment. which makes a lot of sense.3. Mechanical equipment efficiency. and completion.50 IPLV ARI590 125tons Water-cooled centrifugal 5. Shutoff damper controls that automatically close when the systems or spaces served are not in use (these dampers must also satisfy a maximum allowable leakage rate. it's unlikely that energy savings will accrue if the operator doesn't know how the system should work. ground- source heat pumps. 2001.10 IPLV ARI550 chiller. Summary of the revised of ASHRAE 90. entering condenser water temperatures. Climate and placement dictate insulation requirements for ducts. 8.5EER 9. The 1999 standard also contains extensive HVAC control requirements regarding deadbands. 3. restrictions for set-point overlap.3EER (85°FEWT) 12. 90 .1—1989 After29-Oct-2001 Test Procedure Rooftop air conditioner. 9. and absorption chillers. 300tons In the case of centrifugal chillers. Even if an engineer designs a great system. controls.10 SI or better. Standard 90. For equipment covered under the previous edition. Table 4. Table 4.516/COP. For piping. the 1999 standard allowed the 1989 efficiencies to apply until October 29.3.1—1999 also provides tables for centrifugal chillers selected at non-standard conditions (leaving chilled water temperatures. A Simplified Approach option.1 energy performance requirements Equipment Type Per90. manuals. Setback controls 5. leakage tests) and insulation of ducts and piping.10 COP 6.45 COP 4. Zone isolation capabilities that permit areas of the building to continue operating while others are shut down 3. with COP in W/W]. Mandatory HVAC requirements also address construction (duct sealing. insulation. Stipulations for off-hour controls include all of the following: 1.3. Optimum start controls after the system airflow exceeds 10000 cfm Construction. These requirements are an integral part of every compliance path. Mandatory HVAC Provisions in ASHRAE 90.1 Mandatory requirements for HVAC systems include mechanical equipment efficiencies. construction. both the full-load COP and IPLV must be 6.30 IPLV 6. insulation.576 kW/ton or less [kW/ton figure of merit = 3. Automatic shutdown 4. or condenser water flow rates).20 COP 5. and completion.90 IPLV 4. Eligibility for this approach requires that the building occupy less than 25000 sq ft of gross floor area and not more than two stories. which consists of a subset of all mandatory provisions and prescriptive requirements For small buildings. The difference lies in ease of use and the degree of flexibility allowed. and off-hour controls. the requirements depend on pipe location and the operating temperature range of the fluid.

Climate and equipment size dictate the prescriptive requirements for airside and waterside economisers. static pressure set point shall be reset based on the zone requiring the most pressure. unlimited reheat is permitted if at least 75 percent of the reheat energy originates from a site-recovered or on-site solar energy source. an economiser is unnecessary. the pressure drop of the waterside economiser must be less than 15 feet of water or a secondary loop must be created to avoid its pressure drop altogether when the economiser is not in use. Fan power limitations. It further requires control elements to be calibrated. that is. Economisers (automated free cooling).The standard also addresses balancing for air systems larger than 1 hp and for hydronic systems larger than 10 hp. too—but not for variable-flow systems nor where it “…cannot be implemented without causing improper operation of heating. adjusted and in proper working condition for buildings that exceed 50000 sq ft. cooling. Exceptions provide sufficient flexibility to maintain either temperature or humidity control. If the EER of the selected rooftop air conditioner is 11. Another notable addition to this set of prescriptive requirements is the following: Set Point Reset.e. capable of operating in conjunction with mechanical cooling. which has 6921 Cooling Design Days—base 50 (CDD50). it does not ban simultaneous heating and cooling.7. This requirement will almost certainly prompt increased use of variable-speed drives or vaneaxial fans in systems of this size. 2. Unless there are three or fewer control valves in the system. the fan must be able to reduce its speed to two-thirds or less if its motor is 7. For systems with direct digital control of individual zoned boxes reporting to the central control panel. i.1 in US units.5 hp or larger. is an EER of 9. This requirement will undoubtedly prompt greater use of variable-speed drives. now expressed in terms of nameplate power. Simultaneous heating and cooling. a cooling tower with less than two cells must be equipped to reduce fan speed on all cells — perhaps with pony motors. must be met when the total fan power for the system exceeds 5 hp (about 3. a prospective HVAC design must satisfy specific prescriptive requirements in addition to the mandatory provisions reviewed above. i. Beyond this power limit. Air system design and control. In addition. For heat-rejection equipment such as cooling towers. or dehumidifying systems.e. Such provisions should increase the popularity of heat-recovery designs that salvage heat from the condenser in an applied chilled-water system or a desuperheater in a direct-expansion system. Supply-temperature reset is required. humidifying. An economiser can be omitted from unitary equipment if its performance is efficient enough. Also known as fan-pressure optimisation. two- 91 . Fans of 30 hp and larger must use less than 30 percent of design power at 50 percent of design air volume and at one-third of the total design static pressure. the 1999 standard requires that the pump in a variable-flow system draws substantially less power at part load. the requirement for a 20-ton rooftop air conditioner in Tucson. Additional prescriptive HVAC requirements Under the Prescriptive Path. For example. the basic premise of set point reset is that the static-pressure set point can be reduced dynamically.” Heat-rejection equipment.6 kW). each pump with a head greater than 100 feet and a motor larger than 50 hp must include a means for reducing electrical demand to 30 percent of design power at 50 percent of design water flow. For example. The 1999 standard increases the power allowance to compensate for pressure increases imposed by specific filtration or heat-recovery devices and when the supply-air temperature is less than 55°F. Although the 1999 standard limits this practice. Hydronic system design and control.8 SI or better. The economiser must also be integrated. Like the fan on the air side of the system. the set point is reset lower until one zone damper is nearly wide open. which lets energy savings accrue rapidly.

” In fact. in a report entitled ‘Commercial Buildings Determinations — Explanation of the Analysis and Spreadsheet’.2% 7.1. Rather than periodic updates (every five years. Aggregation of the energy use intensities produced by the simulations was done as follows: 1) extract zone- based energy use intensities from simulations. and where the service-water heating load exceeds 1 million Btu/h. 8) weight energy use intensity results by building construction floor area estimates. Links between an ASHRAE standard and the US Energy Codes The US Energy Policy Act or EPAct (P.1—1999 Building Type Floor Area Electricity Gas Site EUI Source EUI $UI (USD) Weighting Assembly 0. particularly in retrofit applications in which ventilation airflow is brought into compliance with ANSI/ASHRAE 62. committee members can request changes to the standard at any time. Heat recovery for service water heating is required in facilities that operate 24 hours a day. If the cooling tower has three cells. both the report and SSPC 90. the ASHRAE committee responsible for maintaining the standard.1. source energy. 5) weight results for frame and mass wall construction .S. 2) aggregate results by required economiser usage in each climate. and energy cost intensities.eren. Percentage Change in Whole-Building Energy Use Intensity (EUI) and Dollars Use Intensity ($UI) through application of ASHRAE/IESNA 90. Exceptions to this airside requirement include (but are not limited to) series-style energy recovery and systems in which the largest exhaust air stream is less than 75 percent of design outdoor airflow. Continuous maintenance of the ASHRAE standard As a continuous-maintenance standard. Department of Energy (DOE) evaluate subsequent revisions of Standard 90. 102-486) requires states to certify that their energy codes meet or exceed the requirements of ASHRAE Standard 90.5 92 . This proviso will probably lead to the increased use of energy recovery in air handlers dedicated to ventilation.5% -5. Systems larger than 5000 cfm that bring in lots of outdoor air (at least 70 percent of design airflow) must include energy recovery.3% 4.gov. EPAct also requires that the U.1—1999 Table 4. at least two of them must be equipped with speed control.1 to determine whether they improve energy efficiency in commercial buildings. acknowledge that application of the revised standard will not necessarily increase efficiency for all building types or for all components and systems compared with the 1989 standard.4% 7. Energy recovery. When consensus is reached. ASHRAE/IESNA 90. In some instances. Estimates of the aggregate impact of the new standard at the national level are derived from energy use intensities (EUI) developed through simulations of the building stock under each edition of the standard.speed motors. it issues an addendum for public review and comment. Note that the simulation is done for the entire change of standard from one version to another and that it is not possible to separate out the impacts which are solely due to changes in the AC requirements.1 remains a dynamic document. the means of recovery must be at least 50-percent effective. 6) weight results for heating fuel 7) convert energy use intensities by fuel type to site energy. or variable-speed drives. considering those differences that can be reasonably quantified. 3) map simulation results by climate 4) scale simulation results to existing building stock floor area by building type and region.1—1989. the 1999 edition [of ASHRAE/IESNA Standard 90. The U. If the committee sees merit in a proposed change. Table 4. Public proposals submitted by February 20 are considered at the ASHRAE annual meeting (usually held in June). the 1989 standard was either unjustifiably stringent (in the case of metal roofs.doe. where the heat rejection capacity exceeds 6 million Btu/h.068 9. the addendum is incorporated in the standard. www.4 shows the estimated energy savings from application of the revised standards ASHRAE/IESNA 90.4. The report observes that “Overall. for example). DOE posted the results of its quantitative analysis on its Web site.1] will increase the energy efficiency of commercial buildings.L.S. for example) from a cost standpoint or did not adequately reflect real-world conditions (in the case of warehouse lighting).

2% -6.1-25.1-12.8 Retail 0.4% -0.1% -58.5 Office 0.9 Warehouse 0.0 2.2 Australia.246 15. Minimum energy performance requirements for packaged air conditioners with a cooling capacity between 7.Education 0.7% 9.7% -0.40 18.5 2.173 -71.6% 9.06 = (EER+COP)/2 2004 93 .8% -0. At the same time the Australian government assessed the requirements.218 11.027 -1. see Table 4.50 30.5.6% -11.0 2.88 2004 Unitary AC (heating and cooling) 3.60 45.35 15.0 2.1-30.7% -30.2% 1.55 37.5 2.1-65.6% 3.6% -12. Table 4.3% -8.2% 8. Thus far Japan has developed the following requirements for central AC systems: Table 4.1-1989 requirements and the current Australian requirement are loosely based on these.2% 9.079 0. Japan.6.7% 8.4% -6.6-45.7% 14.65 Japan Japan has adopted the “Top Runner” policy under which quasi-mandatory minimum energy performance requirements are set at a level corresponding to the most efficient equipment on the market at the time the requirements are developed.9 Lodging 0.30 12.7% 14.7 National 1.7% -59. Australia currently has no MEPS for chillers and has no energy labelling requirements for packaged air conditioners.3% -45.25 10.9% 6.5 kW and 65 kW.6-18.45 25.5 2.190 10.0 Food 0. They have introduced MEPS for packaged air conditioners with cooling capacity between 7.9% 5.0 2.3% 5.7% -0.1-37.7% 12.5% -1.6-10.6% -0. Minimum energy performance requirements for unitary air conditioners with a cooling capacity between 7 and 28 kW in Japan AC type Minimum EER or COP (W/W) Date of application Unitary AC (cooling only) 2.6-15.5.0 2.000 7. The most stringent MEPS being applied internationally were found to be the US ASHRAE 90.0 2. Korea and Taiwan All of these countries have adopted minimum energy efficiency requirements for central air conditioning equipment as follows: Australia Australia has adopted a policy of applying the world’s most stringent MEPS as their national requirements.5 and 65 kW in Australia Cooling Capacity Minimum cooling (kW) COP (W/W) 7. Australian test standards for packaged AC units are compatible with ISO and European test standards.

6 Choices and measures which could increase the efficiency of CAC systems Measures which could increase globally the efficiency of CAC Among the dominant types of CAC.70 4. As yet there are no measures for larger packaged units or chillers and there are no labelling requirements for this kind of AC equipment. Taiwan Taiwan has implemented the following MEPS for chillers since January 1st 2003 (Table 4. VRF units and classical packaged systems (rooftops and cabinets).50 <528 Water-cooled chillers (volumetric compressors) 3.(cooling) 6 CAC Cooling towers CAV/AHU TAC Water Cooled +air +hum.5 kW in cooling capacity.7).7.00 >1760 Water-cooled chillers (centrifugal type) 3. We have discarded such an option.40 all 4. These are required to attain a mandatory minimum EER of 2.01 3.51 5. Taiwanese test conditions for chiller units are compatible with ISO and European test standards.10 3.These requirements apply to all unitary (i. multi-splits.73 528 to <1760 Water-cooled chillers (centrifugal type) 4. Japanese test conditions for packaged AC units are mostly compatible with ISO and European test standards. Minimum energy performance requirements for chillers in Taiwan Type of chiller EER COP Cooling capacity range (kW) Water-cooled chillers (volumetric compressors) 3.30 <528 Water-cooled chillers (centrifugal type) 4.(cooling) 5 CAC Cooling towers CAV/AHU TC Water Cooled with air dist. but we have assembled the information about the performance difference among the dominant 18 types : N° Type Observation Terminal Comfort System equipment level 1 CAC Air cooled chillers FCU+1air TC Air Cooled with water distribution 2 CAC Air cooled chillers CAV/AHU TC Air Cooled with air distribution 3 CAC Air cooled chillers CAV/AHU TAC Air Cooled with air +humidity control 4 CAC Cooling towers FCU+1air TC Water Cooled + water dist.e. but in addition the government expects manufacturers to attain a minimum sales-weighted efficiency level of 2.25 W/W. Korea Korea only has MEPS in place for unitary split-packaged AC units of between 10 and 17. packaged) AC equipment within the specified cooling capacity range and hence applies to large room AC units.60 528 to <1760 Water-cooled chillers (volumetric compressors) 3. As yet there are no measures for chillers and there are no labelling requirements for this kind of AC equipment. some are known for their better Energy Efficiency.25 >1760 Water-cooled chillers (volumetric compressors) 2.93 W/W.06 2.(cooling) 7 CAC Natural Water FCU+1air TC Outside water + water dist 8 CAC Natural Water CAV/AHU TC Outside water + air dist 9 CAC Natural Water CAV/AHU TAC Outside water +air +hum 94 . Table 4.44 4. One way of improving EE is to promote some specific system types.10 4.

10 CAC Natural Water FCU+1air TC TWO LOOPS + CHILLER 11 CAC Natural Water Package TC VRF 12 CAC Natural Water Package TC PACK&SPlarge 13 CAC Natural Water Package TC Roof tops 14 CAC Water cooled RAC+1air TC RACs on one loop RAC 15 RAC Air cooled RAC RAC+1air TC Multi Splits 16 RAC Air cooled RAC RAC+1air TC Splits 17 RAC Air cooled RAC RAC+1air TC Small packages 18 RAC Air cooled RAC RAC+1air TC Single Ducts Primary air (1air) has been added here to each non air based system. 1 V3 Selection of fans characteristic curves and the pressure control V4 Improved local tangential ventilator V5 Improved fans on condensers. some under-sizing acceptable D&S3 Quality of the service of "sizing" (for example search for alternatives. 2 V2 Motor Eff. aquifer) CH4 Multi-speed strategies & swept volume control CH5 Variable speed strategies (inverters) CH6 Improved effectiveness of the tower. selection of fillings flow rates CH7 Optimal loading of stages CH8 Sharing of load among chillers & loading of chillers (when various) CH9 Optimal control of cooling tower for low auxiliaries and lowest cost effective condensing temperature CH10 Free cooling integrated into the chiller CH11 Reduction of power of cooling tower under a simpler form than CH9 Motors and fans V1 Motor Eff. Technical measures which could increase the efficiency of CAC systems There are a large range of technical measures which can lower the energy demand of CAC systems without changing the systems themselves as listed by type below. AHU. for equality of comfort. giving to electricity a weight close to its price 95 . In fact it is only installed in some situations. Chillers circuits and control CH1 Optimisation of cost/efficiency at full load (some threshold on accepted EER) CH2 Optimisation of cost/efficiency at part load (some threshold on accepted IPLV) CH3 Stable cold source at good temperature (river. Minimum LCC design) D&S4 Optimisation of multi-zone sizing & prohibition of cold & hot mixing D&S5 Sizing by full plant simulation D&S6 Regulation threshold is in primary energy. V6 Better filters Classic option in design V7 Less pressure drop in all parts V8 Variable speed In some way V9 Feasibility of local stopping of each electrical motor V10 Feasibility of central stopping of each electrical motor by BEMS Design & Sizing D&S1 Decentralised system banned over a certain limit D&S2 Over-sizing banned. sizing of fans.

RAC. etc. "decoupling". giving to electricity a weight close to its GW impact D&S8 Careful organisation of set points and control dead bands since design D&S9 Automatic adjustment of pressures planned from design Operation & Maintenance O&M1 Recording energy consumption O&M2 Monitoring energy consumption with a BEMS O&M3 Obligation to install a device allowing to measure temperature in each zone O&M4 Optimisation of the change of the filters O&M5 Cleaning of condensing coils O&M6 Cleaning of evaporating coils O&M7 Optimal scheduling of M4-M6 O&M8 Fine tuning of controls. used for homogeneous zones PACK1 Local Free cooling PACK2 Changing control set-points (T. variable speed in distribution WS7 Improved pumps WS8 Pipes insulation reinforced WS9 Less head losses.RH) PACK3 Reversibility (local heat pump) PACK4 Optimisation of cost/efficiency at full load (some threshold on accepted EER) System based on water distribution WS1 Control on the returns at 7/12 WS2 Moving to 8/14 on departure WS3 Moving to 8/16 on departure WS4 Variable temperature on departures WS5 Variable temperature on returns WS6 Circuiting of chillers.D&S7 Regulation is expressed in terms of carbon intensity. use of surfactants WS10 Better control of FCU WS11 Cold ceilings/ Beams/ Slabs WS12 Reversibility WS13 Water side free cooling System with circulation of refrigerant SCR SCR 1 Optimisation of cost/efficiency at full load (some threshold on accepted EER) SCR2 Local free cooling SCR3 Reversible MS SCR4 Moving to the VRF (compression benefits) in MS SCR5 Optimal distribution of flows by electronic unit in VRF SCR6 Reversible behaviour of the VRF 2 pipes SCR7 Reversible behaviour of the VRF 3 pipes SCR8 Changing control set-points (T. rooftops. namely through BEMS O&M9 Fault detection systematic thanks to BEMS O&M10 Contract of controlled service (which criterion?) O&M11 Performance Contracting O&M12 Balancing planned in design O&M13 Operation manual written by designer and transferred to operator Decentralised system : Packages.RH) Equipment for air handling 96 .

AS1 Central Free cooling AS2 Smaller % of outside air AS3 VAV AS4 Better fans in AHU AS5 Application of Eurovent specifications for AHUs (less leakage. more insulation) AS6 Cost effective AHU AS7 Optimised blowing temperature (10-16 °C) AS8 Quality of the moisture control system Seems critical for energy consumption AS9 Correction of the poor multi-zone efficiency of Air Systems AS10 Sensor of occupancy and other "demand controlled" ventilation AS11 Central heat/cold Recovery within the HVAC system AS12 "Displacement" strategy by use of stratification of the rooms (low inlet speeds) or by other displacement strategies AS13 The air flow follows the hygienic demand and has not a minimum value over the minimal hygienic demand AS14 Prohibition and successive cooling & reheating AS15 Ducts insulation and leakage limitation AS16 Existence of an A/C stopping & controlling possibility in each zone AS17 Ventilation should be in cascade among rooms AS18 Reversibility by use of the same chiller as a heat pump AS19 Recovery of heat for DHW Systems with water and air AWS1 Improved control of classic system AWS2 Ejector improved allowing blowing temp. like window opening C5 Quality level Alternative Equipment strategies 97 . at 18°C AWS3 Cold ceilings/ Beams/ Slabs + additional system by air AW4 "Displacement" equipment AWS5 Reversibility Systems with water loop WL1 Optimisation of cost/efficiency at full load (some threshold on accepted EER) Building envelope improvement B1 Better insulation of building for winter purposes B2 Threshold on maximum size of cooling zone in building codes B3 Access doors France : automatic closing after passage B4 Control of solar input through openings B5 Shading of facades B6 Lower electricity for lighting B7 Lower electricity for office equipment B8 Night time over-ventilation B9 Ventilation requirements closer to minimum Comfort conditions : changing the rules of the game C1 Better adaptation to occupation of zones C2 Adapted cooling. depending on outside temperature C3 Occupation sensors. like CO2 C4 Ventilation sensors.

In broad terms the measures to improve CAC energy efficiency are classifiable into seven main types as follows: First type: selection of more efficient components by whoever decides An examination of CAC energy efficiency performance data from all directories and in all product classes (not only RAC and chillers. based on rated EER although ideally it should be based on a SEER (or IPLV) rating. These gains are not completely attainable through building codes. etc. but big cabinets. Fourth type: reversible use of the system For all centralised air and water systems there are other potential gains in winter which can be relatively high since AC plants are nothing else than installed heat pumps. Condenser heat recovery is better known and documented. infrequently realised and not very well controlled. The EECCAC study has to review these options and present them in a structured way. although not prescribed. There are also savings related to efficiency of fans selected. For RAC the extremes in average equipment efficiency by equipment type are typically at -15% and +15% of the average level (for instance splits are typically more efficient than packages). controllers. Third type: improvement of the detailed structure of the system and control options For air and water centralised systems there are other potential gains in the detailed layout of the system which can be relatively high (e. literature and country reports have highlighted a number of different "philosophies" to approach CAC energy efficiency. Fifth type: maintenance and operation improved 98 .1). (see ASHRAE 90.) reveals that the more efficient products on the market are always +20-50% better than the average. in a few cases. The gains resulting from optimising the choice of the best system type can only be obtained through requirements in building codes or through equivalent voluntary agreements. there are families or types that have different average efficiency. making use of "free cooling").g. Presently the selection of equipment is made by professionals based on the initial cost per kWc or. etc. but mostly through good engineering work. A greater deployment of higher efficiency CAC components can be achieved through measures addressing tradable goods like MEPS (or equivalent VA) or information to the final customer if he/she has an influence on the chain of decisions. for CAC the potential seems larger. It is part of EECCAC terms of reference to define what can be reasonably included in building codes. Realisation of these savings could be either obtained by prescriptive way (rules of good design) or through improved methodology (apply a check list and a LCC criterion) or a mix (as in the ASHRAE standard). They are also partly related to the type of equipment. the existence of dampers. However such reversibility options are poorly known. Such structural changes are more difficult to realise than changes of the first category (individual equipment efficiency) or changes from the third category (control and detailed design) because they may affect the rules of competition. Second type: choice of the best general structure of the system Within RACs on one hand and CAC on the other.E1 District or block cooling E2 Absorption or mixed strategies E3 Cool storage E4 Use of condenser to heat DHW E5 "double dividend strategies" based on a higher efficiency of office equipment and lighting leading to lower AC loads E6 Evaporative cooling E7 Dessicant cooling E8 Natural cooling from cooling tower Synthesis of policy measures to raise the efficiency of CAC systems The study group experts. rooftops. variable speed options and the quality of control.

efforts made to improve the energy performance of the building envelope and ventilation system can avoid the need for most artificial air conditioning.The gains reported in the previous types will only be achieved over the long term therefore the maintenance or improvement of performance. Seventh type: envelope and ventilation. This could not be investigated in the EECCAC study because it was outside the scope of the study. by technical measures or contractual means (such as Energy Performance Contracting) or by periodic audit. there is a significant pressure on utilities investment to cope with peak demand up to 2010. as opposed to energy consumption. The energy consumption allowing peak power management may be a little higher. even more. building envelope and ventilation choices made to minimise winter demand (insulation and air tightness) and. 99 . Measures taken to lower the specific electricity consumption in lighting or office equipment have a very large "double dividend" in avoided air conditioning. other measures Energy efficiency measures concerning raising of the thermal efficiency of the building envelope and or of the ventilation system are outside the scope of this study. but shaped completely differently and thus can make the whole energy system more efficient. as some data show. generally speaking. Sixth type: energy and power control Gains related to the reduction of peak power. night time ventilation) have a large influence on AC energy demand. should not be overlooked especially. The regulatory approach used in Switzerland is to state that. the choices made to minimise summer demand (solar control. However. is an interesting family of options.

The year 2010 has been chosen as a reporting reference point due to the Kyoto deadline. 2010 and 2020 A base case scenario (Business As Usual) has been defined in order to analyse the technical and economical potential of single or combined policy measures in several alternative energy efficiency scenarios. The model was primed using: the market data from Eurovent for 1998. So the BAU scenario is based on an absence of any policy measure. education) are modelled as if they followed the same "learning curve". another one the obtention of data. Evolution of the market As previously described.. 100 . The air conditioned areas installed in each country.e. AC saturation coverage levels used in the EECCAC study (cooled floor area (m2)/total building floor area (m2)) Hospitals Hotels bars Offices Trade Residential Schools Average North (Others) 30 % 50 % 50 % 50 % 10 % 10 % 17 % USA=EU-South (I. 1998. hotels & bars. The overall growth in the size of the stock is the difference between the apparent market and level of renewal. The present sectoral and general AC saturation levels were compared with the current saturation levels found in the USA (from the CEBCS study from the US DOE). It was then assumed that the EU saturation levels would obtain current US levels by 2020 in the South and intermediate values were generated through extrapolation for the Northern part of the EU. which was already treated in the EERAC study. However there is not enough data to model each sector. etc. has not been remodelled here and the other economic sectors (hospitals. Validation of the results projected into the past were limited to the most reliable comparisons i.P) 81 % 75 % 83 % 70 % 20 % 62 % 32 % France 55 % 62 % 66 % 60 % 15 % 35 % 24 % Co-ordinator + USDOE CAC extrapolation is performed with an empirical function giving the yearly market in terms of the ratio: x = Stock at present time /stock at time(infinity) This choice makes it possible to represent the takeoff and the saturation of the market by one single equation. through comparison with the national market data obtained from the country reports and through comparison with national stock data when they are based on national surveys at the sectoral level. offices. Projections into the past have been mostly done for 1998 and 2000 but also back to 1990 for consistency with the terms of the Kyoto convention.5. Table 5. Table 5. The residential sector. and the historical data on national market and stock sizes for the years 1975 to 1990. EL. even if the starting point and the final saturation level (at time infinity) are different from one sector to another. trade.1. houses. Results are also reported up to the year 2020. which will be renewed 15 years later when the AC equipment is removed. the extrapolations were based on 1998 figures and national evolution trends. The distribution of AC by sector is estimated by the reconciliation of the projected data with the national data and then applying time invariant AC stock sharing coefficients by sector and by AC type. E. in Austria the hospitals are air-conditioned but the houses are not. is not an easy figure to give by country : one of the problem being the definition itself of an area effectively cooled. 1975-80..1. etc. For the past the model extrapolates back to 1970 to generate stocks for the periods 1970-75.1 AC Stock and market in 1990. namely the AC stock. PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO) 5. In fact the AC saturation levels are very different from one sector to another: for example.

101 . Some global results The total cooled area is given in Figure 5.The stocks are obtained by integration of the apparent markets extrapolated (into the past) and by simulation and integration of the annual true markets (apparent market minus renewal of existing AC) into the future.2. Evolution of the total cooled floor area in Europe from 1985 to 2020 3000 Mm2 cooled 2500 2000 1500 1000 500 0 1985 1990 1995 2000 2005 2010 2015 2020 2025 The AC stock values can be expressed in many ways. Figure 5.1. The estimated cooled area per European in the future is shown in Figure 5.3. for instance in terms of the cooled area (m2) per inhabitant as in Figure 5. Cooled area per European in 2000 by Member State and for the EU as a whole m2/inhabitant 7 6 5 4 3 2 1 0 B DK D EL E F IRL I L NL A P FIN S UK EU-15 The RAC stock figures obtained by this method differ from those given in the EERAC study except for the residential sector where the EERAC figures are used (in terms of kW not m2).2.1 and will rise from the present 1000 Million m2 to around 2000 Million m2 in 2010. Figure 5.

Evolution of cooled-floor area from1985 to 2020 at the national level Mm2 cooled 600 500 400 Spain Italy France 300 Germany Greece Portugal 200 100 0 1985 1990 1995 2000 2005 2010 2015 2020 2025 Years We propose to the reader Table 5. The values hereunder cannot be compared directly to such « national » statistics for two reasons: 1. No country has exact statistics of conditioned areas. sector. corresponding to the typical European sizing ratio (120 W/m2 for CAC. etc.4. The stock has been estimated through national surveys and by integration over time (with due replacement rates) of a few Eurovent market data available to us. Cooled area per European in 2020 per Member State and for the EU as a whole m2/inhabitant 18 16 14 12 10 8 6 4 2 0 B DK D EL E F IRL I L NL A P FIN S UK EU-15 The cooled area "per European" is projected to rise from 3 to 6 square meters over the next 20 years.the sizing ratios vary according to climate. Figure 5.2 with the most important national values. national habits.Figure 5. All conditioned areas hereunder are « standardised » areas.4. but some are close to it. Some national results We can follow national evolutions on figures like figure 5.3. 240 W/m2 for RAC). (this has been partly corrected for) 102 . because this may be a useful tool in some national frameworks. However we have to comment accounting methods. Saturation as well as differentiation between countries appear with their real importance.

01 792.61 GE Mm2 cooling 34.06 2.13 Mm² reverse 4.81 51.04 Total Mm² reverse 136.30 19.68 172.22 43.29 33.2.88 16.24 70.84 3.96 660.78 2.43 NE Mm2 cooling 22.28 88.39 180.47 27.02 66.61 41.15 248.01 15.03 15.76 2.35 2.10 350.67 7.69 248.84 64.35 2.which area is conditioned when a building is air conditioned is usually uncertain : the gross area of the building? Certainly less! The strict area of activity rooms? Certainly more! (lobbies.38 15.2 Area conditioned in each country and year (such areas can be compared with national statistics) Years Country 1990 1995 2000 2005 2010 2015 2020 AU Mm2 cooling 12.01 33.71 213.59 2693.52 171.23 87.30 2.29 127.74 14.99 Mm² reverse 5.17 0.97 140.27 11.29 11.66 97.07 1214.53 36.59 141.31 SP Mm2 cooling 64.08 1.23 773.01 191.25 0.41 78.88 87.92 UK Mm2 cooling 94.69 FI Mm2 cooling 15.19 50.68 IR Mm2 cooling 5.39 20.84 1. Sectoral market The evolution of the various economic sectors and their demand for comfort vary a lot (figure 5.89 FR Mm2 cooling 93.03 6.93 132.62 Mm² reverse 1.88 24.62 11.20 Mm² reverse 34.38 78.27 Mm² reverse 4.21 8.35 0.89 PO Mm2 cooling 8.79 45.28 110.06 26.85 175.84 52.28 50.88 145.64 216.50 12.80 340.71 7.71 342.28 534.68 16.5.26 0.07 66.49 113.75 1.08 6.92 29.88 9.03 6.30 Mm² reverse 0.25 18.57 SW Mm2 cooling 38.76 368.10 0.22 3.06 48.17 20.83 2558.47 108.29 30.46 8.98 20.81 30.68 20.20 Mm² reverse 0.40 129.70 1.38 138.90 15. etc.02 161.57 472.11 41.23 GR Mm2 cooling 11.54 18.23 1793.07 19.45 2.42 2207.11 136.29 814.12 72.74 298.36 294.84 17.41 42.88 The differences are very small on total between the two types of figures (standardised or not) but not for a specific country.41 IT Mm2 cooling 130.74 414.49 8.19 326.50 4.87 61.83 5.61 56.39 Mm² reverse 32.17 83.40 0. Table 5.81 9.28 47.33 467.73 57.07 54.24 390.25 39.46 12.43 10.01 7.07 2.21 Mm² reverse 4.63 400.86 120.88 450.24 10.51 365. The evolution of cooled floor-area by EU economic sector from 1985 to 2020 103 .17 23.09 56.07 295.73 34.26 106.73 DE Mm2 cooling 3.23 80.33 186.17 14.07 Total Mm² cooling 538.72 8.28 Mm² reverse 14.95 Mm² reverse 1.03 8.87 1.78 6.74 4.24 502.24 182.06 36.77 52.12 6.34 1.37 293.36 32.84 4.85 Mm² reverse 29.99 Mm² reverse 1.24 102.55 6.83 3.37 13. Only the “standardised” values are used in the rest of the present report.14 8.63 258. Figure 5.24 37.41 53.08 68.65 40.63 173.71 101.43 43.5).). Only trade and offices really grow in relative terms and they may reach 70% of stock by 2020.43 0.57 42.09 54.06 43.29 127.27 BE Mm2 cooling 4.07 0.26 69.18 LU Mm2 cooling 0.04 23. so both our “standardised” figures and the national figures have a margin of uncertainty and should be compared with caution.08 6.57 6.74 8.28 3.51 18.29 Mm² reverse 0.41 31.65 51.20 352.98 106.81 73.37 Mm² reverse 0. adjacent rooms.

Figure 5. 2500. 2000 RAC Rooftops 1500 PACK & Large Split VRF Chillers 1000 500 0 Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020 5. Mm2 2500 2000 1500 education residences trade offices hotelsbars 1000 hospitals 500 0 Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020 The share between technical systems In our BAU assumptions (figure 5. the care in using them and some validation on a few field results. This is also neglected in our analysis because socio-economic research on the subject is scarce.6. It may be that a factor of change comes from individual decisions of millions of citizens and that their demand for comfort (TC vs TAC) varies a lot. 104 . We have to rely on simulations that are made reliable by the years of experience gained in the US.2 Computation of energy consumption in European conditions The main problem in estimating AC energy consumption is that there is very little information available on the actual in situ use of air-conditioners in the EU. Evolution of cooled floor area provided by each AC type in the EU from 1985 to 2020 COOLED AREA IN Mm2 3000.6) no factor will influence the evolution of the share of techniques on the market and the effects of past trends disappear gradually.

It has an L floor shape and it is basically dedicated to office areas.We have chosen the American DOE model for generating consumption estimates and a summary explanation of how it is used to conduct the EECCAC system simulations is given now. Trade. restaurants. we have simulated a second building. Real buildings for the simulation of CAC systems with DOE A Spanish office building has been analysed then simplified and simulated in details (Figure 5. Figure 5. The cooled area is 12 300 m2.7a) both for envelope and equipment. Figure 5. cinemas. The cooled area is 4800 m2. a supermarket. but includes other complement uses like cafeteria. etc. The shopping mall is a real building located in Seville in the old railway station “Plaza de Armas”.. medical room and toilets. Offices account for 50% of cooled surfaces in Europe. It is composed of shops.7b The building used as a secondary reference Coverage of situations with the DOE software Main tertiary sectors are covered by the two real buildings: the Office and the Shopping center. In order to cover the second most important sector.7a The shape of the office building used as main reference. EU climate has been represented by three climates: Seville. It was rebuilt as a shopping mall after the 1992 Seville Universal Exhibition. Figure 5. London and Milan.7b shows a 3D view of this mall. 105 .

expressed in terms of specific consumption (W/m3/h) equals to 0. So there is the same number of AHU's for CAV.57 for variable volume ones (SPF = 2. Every HVAC end use is covered.Envelope has been adapted to each country construction habits. This implies that zone supply air flows are also the same for each climate. With regards to zonal systems.47 for constant volume fans (SPF = 1. low emissive in Milan and low emissive and reflective in Seville. but clear in London. namely. main remarks follow: • Chilled water loops provide water at 7 ºC to cooling coils while hot water is supplied at 60 ºC.7 W/(l/s)) and 0. • Air transport efficiencies. The final decision regarding this matter has been taking the building’s envelope from the “TRIBU” study of Building Codes (TRIBU. and. Roof-top units (RT). including air transport efficiency. and a gas standard hot water boiler (Eff = 0. 1994). expressed in terms of specific consumption (W/m3/h) is equal to 0. • An air-condensed screw chiller (EER = 2. we have used double glazing in every climate. VAV and RT.6) is used to provide chilled water to cooling coils. A variable speed motor is used to control supply flow for VAV. After an exhaustive filtering process. For VAV systems. Lighting and plug equipment. pumps. Insulation cannot be the same in the various countries. • Same ventilation level (that is same zone outdoor air flow rate) has been considered. Heat recovery is not used for this AHU. • Ventilation is guaranteed using a primary air AHU that provides neutral (22 ºC) outdoor air directly to every building zone. the minimum supply setting of each VAV box equals the design outdoor ventilation rate. variable air volume (VAV). six WS EEO have been studied and ranked. Basic AS EEO have been valued. air side economiser and exhaust air heat recovery.15 for FC and WLHP terminal units. the following issues should be pointed out: • One (or some of equal size) terminal unit is installed for each thermal zone.05 W/(l/s)). PTAC fan consumption is considered as cooling consumption since manufacturers' data include this consumption in EER. have been also considered because they generate a large share of the cooling load. outdoor flow rate is always maintained constant. Thermal comfort and indoor air quality have been guaranteed for every system to allow comparisons. package terminal air conditioners (PTAC) and water to air heat pumps connected to a close condensation loop (WLHP). • Roof-top units are equipped with constant air volume fans. 106 . fan-coil four pipes (FC4P) and two pipes (FC2P). despite being non HVAC uses. Some basic remarks have to be kept in mind for central systems: • They have been designed assuming the same zoning and air distribution. This supposes that each central system is always handling the same amount of outdoor air.88) as heat source. The results of those simulations is given in the next chapter. Basic differences between climates are the following: for exterior walls insulation thickness varies from 4 cm in Seville to 8 cm in Milan. • Air transport efficiencies. at AHU level. For hydronic systems. They regard to chilled water temperature control and water transport. Water delta T for cooling and heating are 5 and 10 ºC respectively. The simulation covers seven different system types: Constant air volume (CAV). however VAV handle a variable supply air flow rate. fans. cooling and heating. • Air side economizer and exhaust air heat recovery are not installed when describing the stock CAC market.

77 15. The FC4P system has been kept to represent the typical fan coil system.40 58. the equivalent number of hours of operation (load in kWh divided by sizing in W.67 The cooling loads are different from one place to another but. since the sizing is not the same.77 1. by extending the post processing. Adjustment for chiller quality and options not covered in DOE software We post-processed the results given by DOE2 to make them flexible in terms of selection of a chiller. • Hydronic system except FC2P supplies chilled and hot water using independent circulation loops (four pipes facility).3 Consumption per physical square meter.34 London-CAV 20.17 kWh/m2 E/kWh) E/kWh) E/kWh) Seville-CAV 115.82 10.86 31.97 8.73 18.49 2. If the chiller behaves the same as in Seville.08 5.86 13. The two objectives don’t give the same results : nominal conditioned area is 4800 m2 while standardised areas for cost calculation and stock modelling are respectively 6200. LO: 80 W/m2. Finally we could also consider less frequent solutions that are presented as very efficient as VRV.91 1.87 32.09 London-RAC 15.00 1.60 Milan-CAV 73. RAC with primary air.02 18.20 Seville-RAC 104.04 848 28.10 (0. the one consuming less – with the same comfort level TAC) are given in Table 5. MI : 130 W/m2 –more detailed figures have been used by system types).42 13.82 6. the one consuming more.92 322 11.39 28. The sizing of the system has been adapted to each simulated climate (Seville. London.94 1. The post- processing consists in keeping all the auxiliaries given by DOE.86 202 6. for one square meter) is less variant for this same building in the three climates (SE: around 700 hours. and sometimes the standardised m2 when it’s related with consumption (standardised sizing of 120 W/m2). meter SCL compressor ty total Cost In Euros In Euros In Euros (kWh/m2) kWh/m2 SEC kWh/m2 SSEC kEuros (0. its influence is covered by the role of auxiliaries in London in physical terms (SSEER) but the cost of the system –and so the cost of the service.96 1.16 1008 34.53 36.• Each primary and secondary water loop is equipped with a constant flow circulation pump. The building conditioned area is about 4800 m2 under the form which has been simulated. as well as the level of definition have been largely improved.59 7. etc.07 1.53 2. This also allowed us to consider the case of wet cooling towers which represent a limited but non negligible share of the market.25 99.63 26. More detailed figures could be defined by system types.31 26.92 0.55 Milan-RAC 54. We see the difference in system initial price and the climatic conditions play a more important role than EER for the final cost of ownership.60 10. cooling demand. 3200 and 5200 m2 for this same building in the three climates( SE: 160 W/m2. chiller on natural water. Note that depending on our objective we have used the nominal square meter of the building (the one known in national statistics).52 1. Electricity Needs Electricity to Electrici Initial ALCC ALCC ALCC SSEER SEER Per sq. to cover the cooling load with any alternative chiller. Table 5.06 (0. Milano) and so the three locations do not display the same installed capacity for the same building shape.09 32. efficiencies and total cost of cooling one square meter of the office building.02 54. The figures given here are still consistent with the ones used for those places in EERAC. Efficiency figures may be found in the technical detailed report of Task 5.73 40.26 1. the main results of the simulation for the two extreme cases (CAV without humidity control. MI : around 500 hours). LO: around 200 hours.78 382 14.becomes very low in London. while the meteorological data and the software. We ran a specific program in which the load and climatic conditions remain the same but the quality of the chiller can be adjusted according to the findings of the techno-economic analysis (next chapter).73 70.05 59.3. Preliminary results for the cooling consumption of the office building Dealing with consumption. given the outside conditions extracted from the simulation.39 1. The EERAC figure did include a penalty for degradation of 107 .64 528 16.

The energy demand is 18% higher due to difference in occupation scenarios. MI : 119 W/m2 against SE: 160 W/m2. For instance Education buildings have a load curve similar to offices but not in July and August. We shall display separately the two aspects. A few key figures about the trade sector.8a). as formulated by the type of distribution and control) and from the efficiency of the equipment. All systems have been brought to the same AC quality since ventilation air is prepared centrally and then dealt to each zone. LO: 75 W/m2.(cooling) Houses Germany TAC Water Cooled +air +hum. bars and restaurants do not differ much on total demand but the peak is delayed by about three hours compared with office buildings and they open on Saturdays as well.(cooling) Trade France TAC Water Cooled with air dist.performance over time. When the two aspects are combined. Hotels.8a Total cooling consumption for CAC systems per square meter 108 .4 Simulated universe after extension of DOE results 15 countries Equal comfort level 18 systems 6 sectors Austria TAC Air Cooled with water distribution Hospitals Belgium TAC Air Cooled with air distribution Hotels & bar & restaurant Denmark TAC Air Cooled with air +humidity control Offices Finland TAC Water Cooled + water dist. LO: 80 W/m2. Some systems that propose moreover the humidity control and represent a very small part of the market have been treated.4 as reliably as possible by extrapolating the results of the two simulated buildings. Trade buildings display similar trends. Houses are very distinct (later use of AC in the day) but not very different on total. tuned on the Mall simulated in details : the sizing of chillers is not very different from the office buildings (SE: 133 W/m2. Fig 5. but we assumed here that CAC were perfectly maintained at their initial performance due to the larger building size.(cooling) Education Greece TAC Outside water + water dist Ireland TAC Outside water + air dist Italy TAC Outside water +air +hum Luxembourg TAC TWO LOOPS + CHILLER Netherlands TAC VRF Portugal TAC Packages & Splits large Spain TAC Rooftops Sweden TAC RACs on one loop UK TAC MS TAC Splits TAC Packages small TAC Single Ducts For all the extrapolations a few checks were made. system types and EU climates We covered the universe described by Table 5. the resulting scatter of electricity consumption is of the order of a factor 2. Hospitals have an office section but they work 7 days a week. Preliminary results for the cooling consumption of the office building The electricity consumption results from the load (demand. Extension to all economic sectors. The specific features of the load curves other than the ones simulated have been applied to the actual DOE load curves (offices). Table 5. The assumptions about the economic sectors (sizing and demand) have been tuned thanks to the simulation of the Trade building. For systems the description has been given previously. MI : 130 W/m2). for the same comfort level (figure 5.

8b enables to separate the part of the consumption differences between CAC coming from efficiency.( ng PA di on st um rd id er is + + Si C di er m er rd at PS ai s PA +h r hu at AC at er e w ai h at O w w at ir r+ t de i R w ith LO +a w w de ith si ai de w d + si w ut d O e ith ed si le ut d ol O TW d le ut oo w O le o l oo oo rC O oo d C le C C rC Ai er oo er er at Ai rC at at W W W Ai CAC Systems The main consumption factor is clearly the climatic area. pumps and cooling that is presented hereafter Fig 5.0 kWh/m2 80.0 20. It is shown on Figure 5.0 120. the total cooling load will differ because of the weight of supplementary fan energy released in the air to be treated and supplementary pumping energy released on the water loop when cooling.0 100. The second important parameter comes from the system type. thus. the load is treated either completely centrally or only partially .0 60.( c + ity ul +a is t.0 20.0 0. Total cooling load for each system type for the 3 climates LONDON MILAN SEVILLE Total Cooling Load 140.0 London Milan 60. The first part of these difference comes from the repartition of energy between fan.8a and 5.0 Seville 40.8b.0 0.(c + st ity +a C K& w is t.0 120.9.(c . 109 .0 100. Figure 5.0 40.( s w R C K& lp i M st on m d id rd er is + + al di C hu m er er rd at ai PS s Sm er PA u at at er AC w + ai +h h at w w O at ir t de i ith w R +a w LO w de ith r ai si de w d + w si ut d le O ith d le si d ut O d oo le TW e ut w oo le O l oo oo C O oo d C le C r C rC Ai er oo er er at Ai rC at at W W W Ai The difference between consumption and load representation Fig 5.(c . Reference office building unitary cooling consumptions 140. Depending on each CAC system type.8b.0 kWh/m2 80.0 S F e ps s l R ts op st t um n n l ) g) ) al tro is g ng lit rg VR io tio M LE uc di in in m rd to lo Sp ut +h la n i bu l ol l Ks D er R IL oo co oo e SP ai rib co at on tri H ir le C .0 n l t n st um g) g) op ps s em g) F s e s R ro is tio tio lit lit ge rg VR di LE lin lin lin rd nt o lo Sp Sp +h st la bu u ft ka r co oo o o IL te b ai e Sy SP co oo o tr i tri ac on ti ir H a .

and even larger in the case of London.26 1.44 1.4 3. London and Milan.4 3.85 0.5b SEER and SSEER. specific consumption in kWh/m2 for cooling in Seville.63 75.(cooling) 39.(cooling) 53.15 0.77 18.26 58.33 2.85 0.26 12.45 55.14 1. ranked by order of merit 110 . fans.47 1.43 18.51 0.9.44 0 94.74 73.6 2.78 15 TAC Multi Splits 54.64 1. Table 5.( + M di m TW e w er + is + is al C LO er m er ir rd si wat rd s Sm PA a at hu at er with ith O a w ir + O id e R w o o ith +a w w r C Co ith oo air d s w d + w d O e si le ut ed ol O d ut le l O oo W led C o C C Ai Air C er oo er r Ai at at W W CAC systems It is very impressive to see that the auxiliaries can reach the same order of magnitude than the real chiller consumption for central systems.89 12 TAC PACK&SPlarge 54.61 118. pumps) in Seville.(cooling) 41.89 100.Figure 5.26 2.4 3.52 1.61 99.(c c ul + at oled ate ity st C w ng .78 16 TAC Splits 54.01 8.3 116.92 1.78 14 TAC RACs on one loop 46.74 73.72 18 TAC Single Ducts 63.43 5.98 2.31 2.26 58.52 1.51 18. Repartition of total cooling consumptions in % As % of total kWh/m2 100% 80% Pump % 60% 40% Fan % 20% 0% Compressor % SP F AC Ro ge on ps um s es ts op R st s st n oo l g) n g) ) ro l p plit g lit K& VR di utio id utio LE uc di di lin o n f to lin Si kag r in lo nt Sp +h la S D er de +hu ool ir IL ut ter coo co e rib b o a ac at ti PS air H le tri ai t.56 1 TAC Air Cooled with water distribution 46.69 18.14 1.99 48.89 117.46 1. per physical square meter : SEER and SSEER.33 3.34 24.78 17 TAC Small packages 54.06 6.09 2. specific consumption in kWh/m2 in London.( s t.14 9 TAC Outside water +air +hum 49.52 1.54 7 TAC Outside water + water dist 36.91 1. The very high figures for SEER in London are partly due to climate and partly to the assumption that the square meter considered is a “standardised” square meter.73 76.91 1.51 10.85 0.52 1.88 5.97 Table 5.8 1.54 2.26 58.85 0.88 18.16 5 TAC Water Cooled with air dist.12 50.4 3.99 6 TAC Water Cooled +air +hum.26 58.8 0.91 1. The improvement in Air based systems should come from the improvement of secondary equipment and control.33 2. ranked by order of merit Comfort System Compressor Fan Pump T Cool SEER SSEER 11 TAC VRF 36.34 1.26 54.3 100.21 18.87 1.19 2.33 0.98 3 TAC Air Cooled with air +humidity control 63.65 18.43 48. The preceding analysis is based on the bare figures summarized in the 3 following tables respectively for Seville.5a Results for all systems.52 4 TAC Water Cooled + water dist.91 2.75 3.9 3.2 8.03 1.51 13 TAC Roof tops 49.51 18. Contribution of each piece of equipment in % of total consumption (per standardised square meter for cooling.09 2 TAC Air Cooled with air distribution 49.91 1.89 44.26 67.94 3.18 8.24 60.02 44.56 10 TAC TWO LOOPS + CHILLER 36.22 44.15 8 TAC Outside water + air dist 38.85 0.

09 0.3 0.35 2.74 69.39 32.92 16 TAC Splits 26.65 4 TAC Water Cooled + water dist.12 40.96 1.23 2.1 2.28 18.66 8 TAC Outside water + air dist 5.98 3.67 1. specific consumption in kWh/m2 in Milano.06 4.3 4.86 1.86 4.05 4.81 8.39 2.12 14.53 2.68 2.53 2.37 1.92 15 TAC Multi Splits 26.39 2.98 45.48 2. VRF are a good system.83 22.97 0.8 32.07 2.42 0 8. A regulation should be simply based on electricity consumption if we want to avoid such misunderstandings.79 17.71 0.09 2.07 1.09 1. 111 .61 1. Even Single Ducts are better than any collective system.39 1.35 3.66 3.87 45.94 2. Our feeling is that the designers should keep the right to use whatever system they need to cope with the project specificities but that they should be obliged to improve the system chosen to reach a certain level of consumption. in the middle of decentralised systems.39 1.64 2 TAC Air Cooled with air distribution 9.(cooling) 7.81 1.6 3.94 18. ranked by order of merit Comfort System Compressor Fan Pump T Cool SEER SSEER 12 TAC PACK&SPlarge 26.35 19.28 0.18 3.86 4.68 10.81 14 TAC RACs on one loop 22.92 6 TAC Water Cooled +air +hum.09 0.39 1.38 1.53 2.57 9 TAC Outside water +air +hum 7.03 1.14 0.93 36.61 36.36 14.39 1. despite the fact that we have added to them a primary air system to bring them up to the same comfort level than the other ones.62 7 TAC Outside water + water dist 19.86 17 TAC Small packages 7.86 36.88 Interestingly.57 13 TAC Roof tops 29.94 2.57 14 TAC RACs on one loop 7.73 4.5c SEER and SSEER.65 70.53 2.27 0 30.85 1.07 1.97 0.63 1 TAC Air Cooled with water distribution 27.98 0 66.64 1.(cooling) 21.23 4 TAC Water Cooled + water dist.92 11 TAC VRF 15.74 81.86 15 TAC Multi Splits 7.02 7.94 16.55 Table 5.73 3.67 22.75 79.(cooling) 32. but this is due to problems in the definition of “load” in DOE2 software with distinct systems.96 1.91 3 TAC Air Cooled with air +humidity control 40.98 10.93 5.(cooling) 9.91 22.64 6 TAC Water Cooled +air +hum.76 18 TAC Single Ducts 30.42 0 8.42 0 8.59 1. except for Seville where they show a benefit.64 2.1 1 TAC Air Cooled with water distribution 10.08 1.77 1.6 2.96 1.(cooling) 7.39 36.85 14.58 0.08 1.04 30.99 5 TAC Water Cooled with air dist.04 1.66 0.92 7.49 2.39 14. Wet cooling towers and systems with two water loops display the same performance as the best centralised system.15 0.8 37.42 0 8.53 4.74 24.98 46.08 13 TAC Roof tops 7.02 2.75 69.06 8 TAC Outside water + air dist 21.61 10 TAC TWO LOOPS + CHILLER 20.08 1.57 24.04 28.57 3 TAC Air Cooled with air +humidity control 12.39 2.86 0 11.78 0.9 1.13 7.09 0.49 22.25 3.35 1.35 11.97 45. but not overpass them.86 18 TAC Single Ducts 9.96 1.18 7.42 0 9.98 2.45 1.06 2 TAC Air Cooled with air distribution 30.33 10 TAC TWO LOOPS + CHILLER 4.06 10.71 1.16 2. the results are similar from one location to another.51 0. The decentralised systems have a large benefit.78 2.15 7 TAC Outside water + water dist 6.73 36.(cooling) 24.53 0.69 3.87 7.09 0.88 31.04 33.83 0.77 2.51 40.04 9 TAC Outside water +air +hum 28.8 1.85 5 TAC Water Cooled with air dist.45 1.42 5.37 1.99 1.09 0.65 83.29 2.07 1.55 1.53 0.97 0.56 0.21 0.68 24. SEER and SSEER do not follow the ranking based on consumption.07 1.04 28.29 44.96 1.97 0.35 0.35 11.93 1.38 1.54 11 TAC VRF 3.92 17 TAC Small packages 26.04 28.86 16 TAC Splits 7.04 28. Comfort System Compressor Fan Pump T Cool SEER SSEER 12 TAC PACK&SPlarge 7.73 31.84 40.39 2.

1998.8 32.0 4.9 2.12 shows the first two values (cooling and associated heating consumption by technical type) for the future.9 17.7 96.5d Annual average energy consumption per m2 by EU country (kWh/m2/year).1 0.0 UK 9.9 152.3 97.7 84.2 5.6 155.0 9.5.2 50.4 9.1 Neth 7.7 Spa 56.1 207.com produced cooling and heating degree day data for a number of EU locations (CDD threshold was 15. 112 .0 7. Figure 5.1 94.9 192.0 1.4 93.2 23.3 80.9 20.0 145.0 12.8 152.0 153.7 81.8 105.8 2.8 109.1 207.6 3.3 12.6 175.3 35.2 12.3 Energy consumption in 1990.3 9. An analysis had to be done based on cooling and heating degree day data (CDD & HDD).5 23.6 2.4 1.9 1.6 0.3 Gre 35.1 173.3 9.9 4.3 Ger 12.5 22.6 147.8 26.6 0.4 22.3 3.4 184. Table 5.2 11.5 144.9 1.3 3.1 170.6 10.6 0. the winters and summers in London are milder than those in a large part of central and northern continental Europe hence it is not possible to cover many climates using this station mixed with the more southerly locations.9 28.4 11.1 2.4 139.0 Ita 35.2 19.5 110.5 14.The simulations have then been extended to each of the 15 Member States by use of outside climatic information from an extended data base.5 139.8 152.9 0.0 83.7 94.3 3.7 2. For instance.1 135.2 145.7 119.9 155.8 3.7 Lux 9.0 192.8 49.2 8.7 94.9 19. HDD threshold temperature was 18.9 126. weighted for systems and sectors Cooling mode Heating mode Total Compress fans pumps Boiler fans pumps Cooling Heating Cooling & Heating or Aus 12. 2010 and 2020 Overall values The three main sections of our predictions relate to : • The actual cooling demand • The winter demand of the cooled areas if no reversible use took place • The winter demand of the cooled areas with the reversible use presently estimated.2 8.4 Ire 9. 5.i-wex.4 10. Using the office simulation results a linear relationship was established to predict energy consumption by cooling or heating equipment as a function of HDD and CDD and used to generate results fitting exactly with the borders of each country Using these combinations with the linear equations applied to the CDD and HDD data gives the following annual average unit energy consumption for the same office building by EU country.2 19.7 Por 36.7 14.0 Den 6. Searching on www.1 147.6 1.2 114.6 114.2 This detailed national treatment is translated finally in a set of weighting coefficients giving for each type of electricity consumption its expression as a weighed combination of the three simulations and allowing to compute the country specific impact of any variation made in the three original simulations as a result of the potential policy measures.9 0.2 26.3 0.8 108.4 34.9 0.0 3.5 17.0 Swe 5.6 0.2 Bel 9.4 23.6 140.1 27.1 Fra 19.9 19. while the most extreme cooling loads are covered through Seville.4 0. Table 5.7 15.5 °C).1 48.5 °C.5 28.6 9.4 22.5 Fin 5.8 3.5 119.3 9.5 135.7 11.7 136.8 22.1 2.

Figure 5.12 Energy for cooling and conventional heating associated with the cooled area -consumption by technical type for cooling Total cooling consumption by subtype 300 000 250 000 200 000 RAC PACK GWh 150 000 FCU CAV total conventionnal heating 100 000 50 000 0 1990 1995 2000 2005 2010 2015 2020 Figure 5. Total energy consumption by AC type in Europe in 2000 and 2020 for the three main quantities : cooling. Figure 5. heating (if no reversibility).13. heating (with present reversibility rate) 113 . country by country.13 gives details.

517 203.683 51.333 45.598 74.913 35.579 (Electricity only) Heating function 51. (CO : Cooling. (El.636 78.374 11.495 18. Note that gas is accounted for as a secondary energy.875 42. Total consumption by country .631 114.) Table 5.765 250. RH : Reverse Heating) Table 5.6 and 5.330 236. Heating function 7.442 111.879 33.727 109.103 94.BAU 60 000 50 000 40 000 CO 2000 CO 2020 GWh CH 2000 30 000 CH 2020 RH 2000 20 000 RH 2020 10 000 0 AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK Tables 5.6 Total energy demand generated by AC Electricity demand (TWh) 1990 1995 2000 2005 2010 2015 2020 Electricity and Gas Cooling function 22.844 Without REV.894 28. CH : Conventional heating.084 164.040 With present REV.7 give the main values (EUR15) for the three functions.7 Cooling only results by country and year (for comparison with national statistics) Total Cooling (GWh/ year) Year Pays 2000 2005 2010 2015 2020 AU 469 549 633 689 707 BE 274 422 559 681 708 DE 71 122 180 232 260 FI 206 210 229 242 246 FR 5 010 8 213 10 954 13 240 14 071 GE 2 286 4 012 5 542 6 785 7 415 GR 2 909 5 365 7 269 9 399 9 734 IR 127 180 222 252 264 IT 16 209 24 336 27 445 29 795 30 890 LU 11 18 23 27 29 NE 605 690 797 869 892 114 .

2020 BAU 45 000 40 000 35 000 30 000 TRA OFF G 25 000 HOU W h 20 000 HOS EDU 15 000 CAH 10 000 5 000 0 AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK 5. green house gases emission therefore we shall concentrate here on atmospheric emissions. However we should first explain and justify this reduction of perspective by considering one by one the effects not taken into account or simplified. R407C. PO 1 020 2 049 3 072 4 039 4 621 SP 19 689 28 333 33 573 38 719 39 915 SW 391 378 403 421 425 UK 2 359 3 227 3 826 4 241 4 401 Total 51 636 78 103 94 727 109 631 114 579 Energy by economic sector The contribution of each economic sector is described by Figure 5. R-134a and R-410A (these refrigerants are more or less compatible with the operating parameters of a traditional R22 unit). mostly Refrigerants and CO2. What are the interactions between refrigerant change and energy efficiency? For an optimist. Electricity consumption by economic sector in each country in 2020 Total cooling by country / sector . as this fluid has an ozone-depletion and a global warming potential its production is prohibited in developed countries. A/Cs can use alternative refrigerants such as R290. a higher energy efficiency and a more environmentally benign refrigerant will both result in a lower contribution to global 115 . R134a and R410A are the only refrigerant largely used in CAC directories for substituting R22.14 Figure 5. R22 is the most commonly used A/C refrigerant. TEWI (total equivalent warming impact) is the integrated index used to measure the global-warming impact of all gaseous emissions. R407C. however. including those from direct and indirect sources. Let's consider first the issue of refrigerants.14. ozone depletion.4 Global warming and other environmental impacts Atmospheric pollution reduced to CO2 Most environmental impacts of A/C take place in the atmosphere: acid pollution.

which all have a regional impact. its full implementation in Europe is not yet achieved and the market has not yet taken it into account. the average for OECD (440) and the exact figure for CO2 content per kWh are available for each country. set here at 350 gCO2/kWh. a trend which may have other causes or never happen. there is a trade-off between choosing a better fluid for direct emissions (leaks to the atmosphere) and a better fluid for indirect emissions (lower electricity use). furthermore there is a European bubble and the trends or measures considered here can gave directly positive or negative consequences on the achievement of the European objectives. For a pessimist. except nuclear plants. accidents in the case of nuclear plants. These evaluations strongly depend on the quality of the data. GWP : Global Warming Potential (kg CO2/kg fluid) m : annual mass of fluid emitted into the atmosphere (kg/year) E : yearly electricity consumption (kWh) b : CO2 emission per kWh of electric energy produced (kg CO2/kWh) n : duration of installation life (year). 116 . Two formulas can be used to evaluate the contributions. due to other policies on which we do not interfere. n (1) TEWI : kg of CO2 produced during the equipment lifetime. each kWh of electricity consumed implies a CO2 emission depending on the specific utility plant in that country. NOx and SO2. and consume about 3-4 kg of fresh water per kWh rejected (not only the part evaporated but also the poorly controlled salts (de-concentration). The TEWI index was introduced to compare the direct additional contributions of refrigerant system emissions and the indirect contributions due to the energy consumption of these systems. for which we add the uncertainty in the evaluation of the emissions and energy consumption. Atmospheric pollution from power plants is composed of dust. It's mostly the case for water cooled chillers using cooling towers. In fact. one can assume one average CO2 content of the European kWh. and CO2 with a world wide impact. simply estimate the total quantity and take it into the cost (1-3 euros/m3) in the optimisation. The direct contribution of A/C to global warming will drop independently from our action. the "ExternE" study gave recently values of the external costs of power plant pollution that we could use in case of necessity. We will not devote much time to the issue. b )] . Here we forget about other environmental effects : radio-elements. They are used in the about 12% of cooled area. Water use is another environmental impact of Air Conditioning to be taken into account. The uncertainty in the evaluation of TEWI are the same as that for the GWP (+/-35%). On this subject. Fortunately. The most simple is written: TEWI = [(GWP . TEWI (Total Equivalent Warming Impact) and leak rates of CAC systems The greenhouse effect contributions by the installation of refrigerant fluids should be evaluated on the total of their life cycle. the marginal rate with Combined Cycles which are likely to be installed in Summer peaking countries to cope with the new demand. The consideration for CO2 is different. etc because we have a marginal approach and nuclear is not the marginal energy. our study group is not in charge of the issue of change of refrigerants. Since there is a European bubble. the Kyoto protocol has been made recently. this can be taken into account in details if needed. The principal contribution to the greenhouse effect of the air conditioners comes from their energy consumption.warming. So we have decided to adapt our environmental considerations to our designated range of actions: energy consumption changes resulting in a lower indirect CO2 release. m) + (E . Why aren't we considering other atmospheric emissions than global warming by CO2. Actually. Its order of magnitude being 10-20 % the external cost effects can be estimated by computing the potential impact of electricity costs rising by 20%. Note that the external cost of the CC plants are among the lowest. We have assumed here that checking European regulation on acid pollution was not part of our objectives.

but those emissions are in some way unexpected (related with an unexpected demand for comfort) and concentrated on a few countrie (typically the five Mediterranean countries). 2005. So they should not be forgotten. 2005. and the simplest equation is enough. Practically. leaks during recovery are close to leaks during one year of operation. So our figures of impact are based on indirect CO2 emissions only. 2015 and 2020.6 : National cooling CO2 emissions of AC by country for 2000.6 kg/kW The order of magnitude of the effect of system type computed for one square meter cooled is : .08 TEWI units (kg CO2) due to leaks to be compared with about 100 indirect TEWI due to electricity consumption over 15 years . Kt CO2 2000 2005 2010 2015 2020 AU 164 192 221 241 248 BE 96 148 196 238 248 DE 25 43 63 81 91 FI 72 73 80 85 86 FR 1 754 2 874 3 834 4 634 4 925 GE 800 1 404 1 940 2 375 2 595 GR 1 018 1 878 2 544 3 289 3 407 IR 44 63 78 88 93 IT 5 673 8 518 9 606 10 428 10 812 LU 4 6 8 9 10 NE 212 242 279 304 312 PO 357 717 1 075 1 414 1 618 SP 6 891 9 916 11 751 13 552 13 970 SW 137 132 141 148 149 UK 826 1 129 1 339 1 484 1 540 Total 18 073 27 336 33 154 38 371 40 103 Figure 5. Packages etc.for other centralised systems 0. Numerical results about CO2 emissions for cooling in Europe Table 5. We propose the following values of yearly leak rates : .6 TEWI units due to leaks to be compared with 100 for indirect TEWI. 2010. Table 5. some leak before during or after repair) . Packages etc.for VRF 10% of 2 kg/kW (good reasons to leak despite a careful maintenance) .Some people make a difference between the annual emission rate and the fluid recuperation at the end of the appliance's life cycle. 33 Mt CO2 in 2010 may seem a small figure compared with the projected total around 3800 for EUR-15 in 2010.for VRF 5 TEWI units due to leaks to be compared with 100 for indirect TEWI .for RAC. 1% of 0.for RAC.15 show the results on the full stock.for other centralised systems 4% of 0.6 and Figure 5.15 : National cooling CO2 emissions of AC by country for 2000. 2015 and 2020. 117 . 2010. The proposed conclusion is that the TEWI penalty of VRF could be taken into account in the economic analysis. but not any other aspect . 0.3 kg/kW (some don't leak. taking into account the fact that the recuperation has become mandatory in numerous countries.

23 2.87 93.37 5.91 13. final (commercial) energies are added without conversion Comfort System Gas Fans Pum HP Heat RH C+H C+R 1 TAC Air Cooled with water distribution 25.97 0 10.37 5. This is influenced by the fact that we mix electricity and gas in the tables.15 33.84 9.(cooling) 7.33 18.(cooling) 6. Nonetheless the evaporation of this water produces a small improvement in energy performance. Use of water A wet cooling tower (which displays better energy performance) is more at risk of cultivating the legionella bacillus and consumes water (through evaporation.99 1. of the order of a few kWh/year per square metre cooled.43 12. reversible or not Here we are only interested in the heating of the cooled areas.23 2.42 96.15 33.99 1.06 5.33 18.97 H 10. What we discover is that the choice of a system for the cooling season decides on the energy use in the heating season with a high impact.(cooling) 25.23 2.58 3 78.5 Heating. CO2 emissions for cooling function in Europe 45 000 40 000 UK SW 35 000 SP PO 30 000 NE LU 25 000 IT kt CO2 IR 20 000 GR GE 15 000 FR FI 10 000 DE BE 5 000 AU 0 2000 2005 2010 2015 2020 Such impacts are not small. In the base simulations.0 110. namely for air based systems.43 12.48 93. home electronics.34 0. But also that coming forth to full year totals.48 111.37 5.06 5. Table 5.31 2 TAC Air Cooled with air distribution 6. but limited if we compare them with other uses in buildings (heating.5 107. etc.7c give the detailed results.97 5 TAC Water Cooled with air dist.39 3 TAC Air Cooled with air +humidity control 7.73 118 .05 8.84 9. not in heating in general.03 94. To heat reversibly or to heat with a boiler is not the most striking issue in the tables.76 0. heating has been provided by a gas boiler. Tables 5. The water consumption should be taken into account to perform cycle cost analysis of CAC systems cooled by water condensing chillers.76 0.19 7 80.53 83.7a to 5.48 115. formation of droplets and desalination) at the rate of 3-4 kg per kWh of heat rejected.97 4 10.7a specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in Seville.).28 6.28 6. As an option. there is a compensation and the performance of the systems becomes closer (the fan energy is fully recovered in winter for space heating).91 13. 5.42 98. which is equivalent to requiring a few hundred of litres of water per year for each cooled square metre in a typical southern European office building.6 4 TAC Water Cooled + water dist.1 7 TAC Outside water + water dist 25.11 96.23 2.34 0. better lighting.23 6 TAC Water Cooled +air +hum.15 33.99 1.33 18.11 98.05 8. reversible use of the cooling equipment has been considered.

64 6.6 127.30 18 TAC Single Ducts 26.2 75.07 166.91 84.23 11.93 1.86 143.84 8 TAC Outside water + air dist 54.4 91.5 115.77 6.24 78.66 1 82.09 126.93 5.90 8.16 117.85 110.84 8.48 13 TAC Roof tops 54.91 5.3 103.64 26.5 18 TAC Single Ducts 99.03 11.26 1.7 114.50 0.66 15.99 0.39 12 TAC PACK&SPlarge 99.23 11.23 2.13 1.93 1.9 73.79 107.(cooling) 65.93 4.38 1.05 136.77 14.78 106.76 3 TAC Air Cooled with air +humidity control 65.3 80.95 144.6 73.1 71.5 6 TAC Water Cooled +air +hum.81 63.48 35.53 1.64 26.94 15 TAC Multi Splits 99.16 12.4 63.17 74.50 0.53 1.9 89.(cooling) 88.50 59.71 15 TAC Multi Splits 108.05 136.30 1.30 17 TAC Small packages 26.50 1 71.1 101.91 9 TAC Outside water +air +hum 65.9 89.0 4 TAC Water Cooled + water dist.77 28.81 68.77 6.3 108.23 2.3 80.77 6.93 4.83 14.3 7.72 51.93 1.71 64.64 66.35 130.41 143.43 43.4 14 TAC RACs on one loop 106.93 4.1 84.2 103.10 83.71 21.38 126.83 14.17 110.71 38.94 4.66 15.48 35.36 98.6 71.(cooling) 54.95 144.1 101.6 76.7b specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in London.49 121.75 0.0 7 TAC Outside water + water dist 96.60 45.23 5 TAC Water Cooled with air dist.84 85.71 26.81 95.50 59.43 43.94 4.72 33.53 39.38 4.5 119 .71 64.34 0.53 1.1 14 TAC RACs on one loop 26.2 58.16 137.71 26. final (commercial) energies are added without conversion (HP : heat pump consumption) Comfort System Gas Fans Pump HP Heat RH C+H C+RH 1 TAC Air Cooled with water distribution 88.94 4.2 103.17 122.80 2 TAC Air Cooled with air distribution 54.71 136.69 12 TAC PACK&SPlarge 26.17 110.77 6.94 1.71 64.3 111.64 66.09 124.71 21.67 15 TAC Multi Splits 26.12 98.1 95.24 1 68.86 4.35 22.6 73.09 126.29 98.3 80.62 Table 5.48 18 TAC Single Ducts 108.29 31.3 143.41 142.20 116.0 91.64 26.15 12.3 7.50 62.77 6.48 17 TAC Small packages 108.3 106.1 101.38 4.44 144.17 74.71 135.50 0.28 6.6 71.3 7.(cooling) 54.23 20.66 15.14 146.30 1.73 142.71 71.69 37.8 TAC Outside water + air dist 6.38 4.91 13.57 1.64 21.30 13 TAC Roof tops 5.59 9 TAC Outside water +air +hum 7.11 14.51 20.53 108.6 81.30 16 TAC Splits 26.7 14 TAC RACs on one loop 117.16 42.53 108.67 78.77 28.84 38.2 109.39 8 TAC Outside water + air dist 54.64 66.54 145.05 8.16 117.73 31.84 9.22 11 TAC VRF 99.6 91.07 168.66 16.79 90.62 11 TAC VRF 108.23 20.17 74.73 31.60 1.66 15.09 124.60 132.26 1.49 121.0 3 TAC Air Cooled with air +humidity control 65.99 0.93 133.97 7.1 130.23 20.3 9 TAC Outside water +air +hum 65.83 14.75 77.30 1.71 26.16 117.73 31.78 96.23 11.5 129.29 31.44 7.2 7 TAC Outside water + water dist 88.86 93.5 78.17 2 TAC Air Cooled with air distribution 54.3 10 TAC TWO LOOPS + CHILLER 5.13 1.46 12 TAC PACK&SPlarge 108.84 38.10 59.71 39.2 54.64 21.09 126.1 101. final (commercial) energies are added without conversion Comfort System Gas Fans Pump HP Heat RH C+H C+RH 1 TAC Air Cooled with water distribution 96.71 64.9 72.14 67.4 59.09 126.23 11.2 54.23 6.85 144.6 87.64 66.49 117.26 1.8 112.17 99.68 31.29 98.5 16 TAC Splits 99.3 80.9 58.23 107.06 5.79 90.53 39.04 9.49 121.84 38.71 21.70 11 TAC VRF 26.50 0.93 1.3 57.31 5 TAC Water Cooled with air dist.3 7.5 13 TAC Roof tops 55.30 1.48 140.9 72.64 75.5 17 TAC Small packages 99.65 166.79 107.64 48.6 73.48 35.93 1.71 14.79 96.8 63.91 84.91 84.35 Table 5.4 91.9 79.4 63.79 107.32 6 TAC Water Cooled +air +hum.68 116.9 72.4 10 TAC TWO LOOPS + CHILLER 61.26 1.43 12.93 4.79 90.85 144.77 28.(cooling) 65.53 108.13 1.29 98.3 20.17 110.76 0.29 31.87 38.26 1.64 14.42 127.2 10 TAC TWO LOOPS + CHILLER 62.14 152.81 95.45 83.53 39.38 1.64 21.29 31.94 1.57 1.72 51.79 90.92 164.57 1.23 1.72 55.9 72.00 47.17 74.81 28.9 4 TAC Water Cooled + water dist.93 4.05 136.3 7.43 43.81 63.23 11.6 73.23 3.82 5.48 16 TAC Splits 108.79 107.93 39.(cooling) 96.9 78.05 136.54 125.1 97.85 2 10.64 42.71 88.4 69.7c specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in Milano.16 117.4 9.81 95.50 0.

6 VRF reversible 1.6 7.sober in Winter? Reversibility is not an easy task. If we summarise the comparison to SCOP and CO2 emissions.85 5.9 25.A policy interpretation of the figures is only possible if something is assumed about the competition between gas and electricity.4 tops.9 26. 6.3 28.8 SCOP and CO2 emissions of the heating function with comfort level TAC Seville London Milan SSCOP range CO2 emissions kgCO2/m2/year Boiler with Primary Air TAC Independent heating 6.7 WLHP (reversible) 1-1.9 22.5 22.7 21.5 8.6 22. or their CO2 content or their “primary” energy value.5 6.0 FCU non rev.8 25.. It bears the weight of the auxiliaries.5 28.7 20.3 21.7 3.7 CAV non rev 3. The best heating system is almost always a classic independent one.2 20. It’s an invitation to research : how to make air conditioning systems – a growing social demand. This is not a set of values against reversibility. 7. Table 5..5 Split systems reversible 1. and becomes a less efficient realisation of electric heating.7 20.7 20.8 27.. The coefficient 2.6-1.7 3. 120 .5-1.6 22.8 29. table 5. either their cost.Central Air Conditioners TAC Large packages (Roof 3.) reversible Large splits with primary 6. it’s one of the challenges of the next chapter.9 Very often more CO2 is emitted with a reversible system than with a non reversible one.6 22.8 30.5 system RAC with Primary Air TAC Split systems non rev.3-1.4 air Non reversible CAV reversible 1.1 23. and if we consider only the most frequent systems inherited from the past in the present stock.3-2..8 tops.2 25.2-1. a system which is structured and sized to face the very demanding conditions of Summer will consume more in Winter than the simplest systems used for heating only.2 CAC ..4 23.2 is sufficient to represent about all aspects and gives the exact CO2 value (the ratio 400:180). Simply.8 gives us interesting indications.4 25.) non rev Large packages (Roof 1.8 FCU reversible 1.4 23.

121 .

1 Energy-engineering analysis of chillers Chiller prices as a function of the refrigerating fluid and EER Chillers using the refrigerant R407C.7823 0 0 20 40 60 80 100 120 140 160 180 25000 Cooling capacity kW Prixce euro 20000 15000 10000 R22 R407C 5000 Regression(R22) Regression (R407C) 0 0 20 40 60 80 100 120 140 160 Cooling capacity kW Role of condensing medium By contrast there appears to be a much stronger relationship between chiller price and the choice of condensing medium. From this it appears that there is no additional cost for chiller equipment that uses R407C compared with those which use R22. Chiller cost versus cooling capacity. as a function of the refrigerant 40000 y = 230. The use of cooling with water can only be economically justified in large capacity systems. TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY EQUIPMENT USED IN CAC 6. Figure 6. condensing medium and price is shown in Figure 6.1.9228 40000 25000 Price euro 20000 35000 15000 R 2 = 0.3 2 R = 0. which has been developed as a zero ODP substitute to R22. The relationship between chiller capacity. the difference of cost between less expensive water condensation chillers (needing an outside tower) and more expensive (but complete) air condensation becomes small and does not pay for the additional equipment necessary for the system with condensation on water.2. judging from an analysis of their publicly quoted prices.98x + 5130. cooling capacity as a function of the type of condensing medium Additional costs for reversibility 122 . Figure 6.2 2 R = 0. Chiller price vs.46x + 4944.7719 10000 Air cooled 30000 5000 water cooled Regression(Air l d) Regression l d) (water R 2 = 0. For small capacities. 45000 Figure 6.7835 35000 45000 30000 y = 122.1 shows the price of the equipment as a function of its refrigerating power and refrigerant. on average have an identical energy performance and do not appear to be any more expensive to purchase.2.6.

The cost of a reversible chiller is on average 10% higher than the cost of a traditional cooling-only chiller.
The data in Table 6.1 show a comparison of prices and EER for a sample of 89 cooling-only models with 44
reversible models.

Table 6.1. Comparison of prices between cooling-only and reversible chillers

reversible cooling only
Cooling Price EER Price EER Price difference
Cap. kW (euro) (euro) in %
28 kW 8590 3.5 (water) 7629 4 (water) 12.6%
33.5 kW 12498 2.33 (air) 11629 2.33 (air) 7.4%
64 kW 18900 2.57 (air) 17025 2.34 (air) 11%
Although the cost of reversible equipment is on average 10% higher than the cost of a cooling-only system, it
may well be offset by the avoided installation cost of a stand-alone heating system.

Defining chiller part-load efficiency
An exercise done within AICARR (the HVAC engineers association of Italy) showed that the ARI
coefficients were completely unsatisfactory for use in Europe and this has lead to a proposal known as
EMPE (European Method for Part Load Efficiency). However the ARI standard is very important because it
is one step further than the ISO TC 86 / SC6 / WG9 part load testing standard which is being elaborated
internationally. Many chillers have been tested under the US IPLV approach and its introduction has
produced a significant market transformation impact in the US. Thus an integrated part-load testing approach
is no longer a hypothetical proposal, but a practical tool.
As the result of a CEC mandate, TC 113 of CEN is developing a part load test (part capacity in fact)
applicable to any AC equipment (CEN02). The difference between part load and part capacity lies in the
extent of the testing. In part-load testing the manufacturer manually adjusts the chiller to attain the required
testing load which is varied by conditions in the test chamber. In part capacity testing the entire chiller is
tested such that the chiller control unit, is used to adjust the cold generated by the chiller to a given
percentage of the full output, i.e. the real environmental test conditions are not applied and hence the
feedback between the outside and the chiller control is not tested. The draft standard (No. Part5) which it is
proposed would be added to the CEN 814 and 255 standards includes a part capacity test with 50% input
power and the same temperatures as the full load. This proposal received many negative remarks from the
European national standards bodies who vote on the adoption of CEN standards and thus will first be used as
an ENV text (to be used on a voluntary basis). A positive interaction took place between CEN TC 113 and
EECCAC, allowing a better representation of European interest in those subjects.
Italy has started its own part load testing (Italian Standard UNI 10963 " Air conditioners , chillers and
heat pumps- part load tests."). This Italian standard provides more ambitious a starting point than the CEN
draft standard resulting from the CEC mandate. Besides full load conditions, one temperature regime is
tested first at full load, then with part load capacity. In any case the laboratory shall run at least one test with
a capacity from 20% to 30% of the nominal full load capacity. The technician helps the equipment reach this
point by reducing the swept volume but if the unit has only on-off control, the test has to be run cyclically.
The duration of the cyclic run test is one hour and shall include at least two cycles.This gives an idea of the
chiller EER in the most common operating conditions and allows additional test runs to be performed for any
load. The manufacturer can run supplementary tests at different part loads, supplying more results in order to
get more accurate calculations and data. The connection between this testing standard and the EMPE (IPLV
like) method seems still to be established.
As part of the EECCAC study, EDF (Electricité de France), Eurovent-Certification and Armines launched an
experimental exercise in order to prepare the ground for a wider use of part load testing in Europe. The
exercise has two strands. First, some part capacity points will be tested in the EUROVENT certification
programme in order to gain some testing experience. EUROVENT could require their current full-load
measurements to be supplemented with testing at additional points in order to better characterise the annual

123

behaviour of the equipment. This would occur at a relatively low cost as the chiller is already being installed
on the test bench. Second: some chillers (one of which will be a reversible air to water heat pump) will be
fully tested (i.e. with full performance mapping) using EDF’s facilities.

Available data and simulation tools
Eurovent – Certification runs a directory of products on the EU market which gives good information of
product performance. The Directory has been used extensively. We have had access to a number of
simulation tools, two from manufacturers & THERMOPTIM, a thermodynamic software from Armines. We
have also used EUROVENT testing points (Joint project) and EDF’s experimental testing programme.

Incremental costs as a function of efficiency
Our first approach has been to disaggregate the total cost of the chillers into their main components and then
extrapolate the cost of each part assuming increasing efficiencies. This method overestimates the costs of
efficiency : we know that Energy Efficiency costs less than expected based on such a method when it's taken
on board by the companies because then R&D can intervene. But to start the process we need to find the
margin for self paying improvements on the market. We have based our analysis on a base line chiller with a
screw compressor operating with R134a.
Starting from such values one can seek the level of performance ensuring the minimum of LCC but one can
also estimate the overcost associated with some levels of standardised performance (moving from G, to F, to
E...) that the market cannot reveal (there is only one market price if we exclude the brand name effects). An
extrapolation could also determine the total overcost of the industry and the price increase to be expected
from EE improvements. The simulations have been performed with Thermoptim®. This extension of a
commercial software enables to perform non nominal performance calculations according to the following
description of the components of the modular chiller:

• compressor: isentropic efficiency and volumetric efficiency as a function of the compression ratio,

• evaporator: two zones, biphasic and vapour ; one correlation by zone gives the heat exchange
coefficient ; the parameters are only physical, exchange area, free flow area, hydraulic diameter and
an intensification factor to take into account specific surface enhancement or increase,

• condenser: 3 zones ; one correlation by zone gives the heat exchange coefficient ; the parameters are
only physical, exchange area, free flow area, hydraulic diameter and an intensification factor to take
into account specific surface enhancement or increase,

• expansion valve: the expansion process is supposed to take place at constant enthalpy.
The base case correspond to an air to water chiller with a screw compressor, working with the R134a
refrigerant, with a Cu-Al air coil and a shell and tube evaporator. The equipment has been designed to
represent similar behaviour, in terms variation of the EER with outside air temperature and water
temperature, to chillers’ manufacturer whose data were available.
The nominal full load efficiency has been decreased by decreasing the compressor isentropic efficiency and
the exchange coefficients at both heat exchangers, in order to represent the “bottom” of the market in terms
of performance and so to represent what it would cost to request a minimum performance to all chillers. To
complete the market reality, a similar work should be performed with an air to water scroll chiller with
R407C as the working fluid. A similar study should be made also on a water cooled chiller.

Optimisation of the chiller used as baseline without any system consideration
Now we can perform some economic calculation and compare the improvements proposed with the diversity
on the market. For a given electrical power the capacity varies proportionally to EER; for a given capacity,
the compressor can be reduced when EER increases. So the cost per kW decreases with the first steps of
performance and only increases later (see figure 6.3).

124

Figure 6.3 The cost of a chiller at nominal capacity according to its EER

Optimisation of Cost/kW final

110

100
Euro/kW

90

80
2 2,1 2,2 2,3 2,4 2,5 2,6 2,7 2,8 2,9
EER

Conclusion : the best chiller having the same cost (100 Euros/kW) as the present “worst” has an EER around
2.80. The range from 2.00 to 2.80 shows reasonable prices for a chiller judged only on capacity. It
corresponds exactly to the present market. The minimum cost chiller according to our analysis has the same
EER as the average market EER 2.50), which may be considered as a validation of our cost reconstruction.

Optimisation of a chiller in a system
The energy consumption of equipment will be more and more considered in the equipment design process.
One day, a definition of chillers performance based on SEER and SCOP will be substituted to the ones given
as EER and COP. The part load benefits will then be optimised and the optimisation can then be made on the
basis of SEER (computed here with the EMPE method). So it is interesting to define the “optimum” taking
into account consumption. The search for the optimum has been done in the same way, through successive
additions, including part load options (Figure 6.4).

Figure 6.4. The cost of the service rendered by a chiller in terms of SEER
10

9

8
Total cost (Euros/m2)

ALCC17-800h
7
ALCC10-800h
ALCC17-400h
ALCC6-800h
ALCC10-400h
6
ALCC6-400h

5

4

3
2,00 2,20 2,40 2,60 2,80 3,00 3,20 3,40 3,60 3,80 4,00
SEER

The optimal level of performance for a screw chiller is about 40% more efficient than the present « bottom »
of the market. We have seen that the chiller chosen in chapter 5 to represent the stock has a SEER around
2.00. The optimal chiller is already present on the market. It has a SEER between 3.00 and 3.50. One way to
reach this performance is an EER around 2.46 (enhanced evaporator and condenser, improved compressor)
and a splitting in 3 or 4 scroll units of the capacity of the compressor. We have estimated the associated

125

80 99.82 100. Then we have the use of natural water.04 to 1.2).22 1. The water cooled chillers are a screw unit with nominal 3. However there are strong electric auxiliaries for pumping and circulating the water.60 29.87 70. Part load optimisation would bring 10 to 20% decrease of the total bill of the office building simulated (SSEER from moving from 1.22 The energy benefits of water compared with air appear small.04 1.77 28.66 59. all conditions remaining the same (figure 6. from 0.36 1.66 0. not to be forgotten.73 Milan 1.1 performance of improved air cooled chillers kWh/m2 for cooling “poorest” WC “best” WC “poorest” WC “best” WC “poorest” AC “best” with the CAV system chiller with CT chiller with CT chiller with chiller with chiller AC TAC natural water natural water chiller Seville 100.73.1 hereunder. There is also a water consumption.37 32.16 to 1.26 85. with a higher initial cost.3%).23 32.78 28. from 1.5 EER with a poor part load control (uncontrolled screw).36 1. their water consumption and their cost.49 60.75 84.64 59. There is a large margin for improvement and central solutions are not condemned in comparison with packaged units if they improve their performance.25 SEER.66 69. However there are pressures against CT due to the legionella problem. or will they disappear slowly? The promoters of this solution should compare the LCC with and without CT.34 London 0. For the best range of products on the market we consider the same nominal EER but the best part load behaviour we found experimentally on a 4 scrolls chiller.19 84. What is important is the part load behaviour of the chiller not its type! 126 .97 London 31. Cooling tower use is desirable in principle because the EER seems better. Once again.31 EER for the Table 6.2 performance of improved water cooled chillers SSEER for cooling “poorest” WC “best” WC “poorest” WC “best” WC “poorest” AC “best” with the CAV system chiller with CT chiller with CT chiller with chiller with chiller AC TAC natural water natural water chiller Seville 1.74 0. manufacturers engineers may have other ways to reach 3.overcost at 12.15 1.64 0.23 1.64 0. once you include all the auxiliaries needed to reach natural water or to run a cooling tower and the climatic differences don’t change that comparison. An example can be find in table 6. The relative change at that bill level is half of what it is at chiller level due to the weight of auxiliaries remaining unchanged.05 1. The solution expands so slowly that we don’t have figures to prove any growth.71 0. For the bottom of the market we have used the same chiller as for the stock : 2. but our objective was to find out if there is a margin for improvement.64 to 0. namely in confined aquifers (the case represented here). less expensive.14 1. We have simulated for the previously defined office building both water cooled system.34.16 1.15 Table 6.22). Not only EER is better but it is kept most of the year due to the constant natural resource.05 1.66 Milan 69. The question is : will Cooling Towers remain used in the present 12% of cooled area. Water cooled chillers Another type of improvement could be to use a water cooled chiller (with a better nominal EER).3 Euros/kW (+12.

The US Department of Energy imposed minimum energy performance requirements for large packaged air conditioners (known as ‘unitary air conditioners’ in the USA) through the EPCA in 1992.5 tons (26 kW). such as ‘roof tops’. Share of large packaged air conditioners as a function of cooling capacity in the EU and US markets (source: Eurovent and ARI databases) 127 .78 W/W for both the 26 kW and 52 kW units.46 W/W in 1998. found in the European market are either identical to. Bearing in mind this technological similarity and the resource constraints applying to the current study a decision was made to adapt the results of existing techno-economic energy engineering analyses conducted for this type of equipment in the USA for use in Europe rather than conduct a fresh European analysis. or share the same technology as. Figure 6.2 Engineering approach of the performance of Packaged units For the most part the large packaged air conditioners.9 W/W.5 and 70. models available for sale in wider international markets. As a result the smaller 26 kW base case unit is much more representative of the type of models found on the EU market than the 52 kW unit. Figure 6.5 kW is 2. the minimum EER was 1. The lowest EER considered in the US energy engineering analysis is 2.1 standard proposed minimum energy performance requirements for the same appliances and these have since been made mandatory requirements at the state level by a large majority of US states. As a result of the existing US regulations the minimum permissible EER for packaged AC units with a cooling capacity between 19 and 39.6.2 kW. and 2) a roof-top unitary air conditioner having a cooling capacity of 15 tons (52 kW). An analysis of the US market for large packaged central air conditioners established that the market could be adequately represented by a techno-economic energy engineering analysis of two fundamental models: 1) a roof-top unitary air conditioner having a cooling capacity of 7. By contrast the average efficiency of packaged units in the EU market and within the Eurovent database was 2.5 shows the distribution of models by cooling capacity in the two markets. As recently as 1999 the non-binding ASHRAE 90.61 W/W and for those with a cooling capacity between 39.5 W/W and the average efficiency was 2.5. such as the USA.9 kW while that in the USA is 36.78 W/W and the maximum EER was 3. In 2000 the US DOE launched a revision process for the existing EPCA MEPS which aims to set more stringent MEPS from 200X. The maximum EER level found on the US market in 2003 was 4 W/W.3 kW is 2. The large difference in the lower and average efficiency levels can be ascribed to the impact of the US policy measures and the absence of equivalent measures in the EU. Following the US MEPS development process a full techno- economic energy engineering analysis has been conducted for large packaged air conditioners. In 2003 the lowest efficiency unit which was active on the US market had an EER of 2.41 W/W. The only other significant difference in the products found on the two markets concerns the average energy efficiency levels. which forms the basis for the results reported in this section.58 W/W. A parallel analysis of the European market shows that the average cooling capacity of large packaged AC units in the EU is 28.

warranty costs. The cost efficiency relationships established in this way were found to follow an exponential growth curve. indirect labour. The inputs to the cost model were reviewed by individual manufacturers and the values adjusted if appropriate. cost of capital and overhead labour and ‘variable’ factory expenses such as: manufactured materials. Each of these curves was in turn regressed to a single market-average exponential curve to give the results shown in Figures 6. general & administration costs. The selected units were broken down (physically or using catalog/design data) to create a bill of materials that was fed into a cost model. purchased materials. representing several manufacturers and a wide range of efficiency levels were examined and four units chosen. The goal of the US energy engineering analysis was to develop cost versus efficiency curves of large packaged AC units to guide policy development. 40% 35% 30% EU 1998 Share of models 25% US 2003 20% 15% 10% 5% 0% 0 50 100 150 200 250 Cooling capacity (kW) The US energy engineering analysis This section draws heavily from (TIAX 2002). The incremental cost of 26 kWc packaged air conditioners as a function of their efficiency on the US market in 2001 (US$) (source: TIAX 2002) 128 .7 below. net profits. thus the data for each manufacturer was regressed to a exponential curve. These figures also show the upper and lower 95% confidence intervals as broken dotted lines about the average line (solid). It also itemises corporate expenses such as: research and development.6. fabrication labour. taxes and sales and marketing costs. equipment maintenance. In particular the intention was to determine the life cycle cost of packaged AC units as a function of their EER. Figure 6. tooling amortisation.6 and 6. shipping and indirect materials. utilities. The cost model itemises ‘fixed’ factory expenses such as: equipment and plant depreciation. assembly labour. The methodology used was as follows: A total of eighteen large packaged AC units.

5 EER (kW/ton) In addition to using this market-based reverse engineering approach a classic design option analysis was also conducted to explore the cost efficiency relationships of products with potentially higher efficiency levels than those found on the existing US market .0 12. distributors and mechanical contractors were determined (the latter separately for small and large contractors operating on new.7. Life cycle cost analysis This section draws heavily on LBNL 2002. $1. Data has been gathered in the EU on: the average efficiency of packaged AC systems. in order to 129 .5 12.0 10.0 9.5 10. However. distributor and contractor mark-ups to the incremental ex-factory equipment costs estimated in the energy-engineering analysis.5 Btu- hr/W (=3.5 9.0 10.000 Eq $1.or replacement-construction markets).5 10.5 12. typical building load factors and marginal electricity prices.5 11.5 EER (kW/ton) Figure 6.500 ui p m $1. The results indicate that on average the life cycle cost minimum in the USA occurs for an EER of 11. Incremental changes in the total installation costs as a function of the energy efficiency of the large packaged AC units were estimated by applying the incremental wholesaler.0 9. The incremental cost of 52 kWc packaged air conditioners as a function of their efficiency on the US market in 2001 (US$) (source: TIAX 2002) $2.0 11. The data on manufacturing cost as a function of efficiency derived from the energy-engineering analysis were converted into life cycle cost vs. efficiency curves in the following manner: Mark-ups for wholesalers. that are all necessary inputs if the life cycle costs of European packaged systems are to be determined.5 9.37 W/W) for both the 26 and 52 kW units.0 12. the typical price and installation costs of packaged systems as a function of their cooling capacity.000 en t Co $500 st De lta ($) $0 -$500 8.0 11.000 Eq $750 ui p m $500 en t Co st $250 De lta ($) $0 -$250 8.5 11.

22 W/W when a 6% real discount rate is applied. installation and maintenance costs are the same. the results were found to agree to within 0.4 3.6 2.25W/W derived from the values quoted in the tables of section 2. However. They show that the life cycle cost minimum occurs for large packaged units with an EER of 3.2.13: Estimated average life cycle cost per m2 of cooled space per year vs. For 26kWc units the US analysis implied an average equivalent of 2097 hours of full load operation per year while 800 hours per year is deemed more likely for the EU.8 3.0 3. EER for large packaged air conditioners on the EU market (based on a 52kWc unit) 130 . which is most representative of the EU market.establish the relationship between these life cycle costs and the nominal efficiency of the packaged system it is necessary to adapt the US cost-efficiency data to reflect the circumstances in the EU in the absence of equivalent European data. when the life cycle cost results produced in this manner were compared with those produced for large packaged systems with an EER of 2. The assumptions used in this calculation imply that not only the trend in relative manufacturing cost vs.14: Estimated average life cycle cost per m2 of cooled space per year vs. The results of the analysis taking these factors into account is shown in Figure 6. EER for large packaged air conditioners on the EU market (based on a 26kWc unit) €17 €16 €15 ALLC (€/m2/yr €14 €13 €12 €11 €10 2.2 3.22 W/W.4 2.3%! This implies that the adapted US equipment cost versus efficiency relationships are reliable for use in the EU. efficiency for packaged AC units is the same in the EU as in the USA.2 2.6 EER (W/W) Figure 6. The comparable results for the 52kWc unit are shown in Figure 6. but also that the relative trends in distribution.14.13 for the 26kWc unit. Figure 6. Although the overall life cycle cost per kW are lower for the 52 kW unit the minimum still occurs for an EER of 3.

Decreasing the Life Cycle Cost means in fact in the same time an important energy saving. There are two effects in the same direction of the electricity consumption in fans : direct electricity consumption and increase of cooling loads leading to an indirect increase. In many cases it may be better to select a larger unit in order to reduce the operating costs. essentially because large differences exist between European countries concerning climate conditions and also energy prices (electricity. the pay back will be better.8 3.6 2. in London the figures become smaller but the difference larger (25% in cooling demand). Investment and maintenance together make only about 20% of the Life Cycle Costs.4 3. It was emphasised that the most important parameters influencing the Life Cycle Cost are: • size of internal area • pressure loss in the duct system • heat recovery device • control systems for regulation of the actual demands of the ventilation. The internal pressure drop will decrease and efficiency of other functioning parts will be better. €17 €16 €15 ALLC (€/m2/yr €14 €13 €12 €11 2.4 2. The Eurovent Working Group prepared a Recommendation for selection and design of an Air Handling Units in order to reduce or minimise Life Cycle Cost. Such a Recommendation was not simple to make. It is not a small effect : in Seville the system with less fans demands 95 kWh/m2 in cooling as opposed to 105 for the system with more fans. The fact is that the energy consumption (operating) costs represent more than 80% of the total cost during the lifetime of the unit. For instance a reduction of pressure loss from 400 to 250 Pa ( which may be relatively easily obtained ) will give much better results than increasing the efficiency of the fan by 5% ( which may be difficult to achieve ). Table 6. members of the Eurovent WG 6C “Air Handling Units “ took a very active part in the SAVE project concerning Life Cycle Cost of equipment used in ventilation. Fans integrated in AHU A well designed duct system with low pressure losses will greatly reduce the electrical consumption of fans. 131 .2 3. Even if the investment cost increases with a larger unit.0 3. Higher efficiency values are strongly recommended for longer operation periods.6 EER (W/W) 6.3 Energy Efficiency of Air Handling Units seen as tradable goods Major European manufacturers.2 2.3 indicates the minimum efficiency levels for fans recommended by Eurovent ( Efficiency figures valid for most common operation hours (around 3000 hours/year). air conditioning and refrigeration. energy for heating and cooling).

50 45000 44% 45% 47% 48% 49% 51% 52% 53% 54% 54% 55% 16. A certain level of heat recovery is recommended by Eurovent. by taking advantage of the results of the Eurovent LCC study. Minimum efficiency levels of fans recommended by Eurovent MINIMUM TOTAL EFFICIENCY* FAN/MOTOR COMBINATION [%] Air flow rate Available static fan pressure [Pa] ** 3 3 [m /s] [m /h] 250 315 400 500 630 800 1000 1250 1600 2000 2500 1. cold or hot climates. Table 6.00 7200 33% 34% 35% 36% 38% 39% 40% 41% 42% 43% 43% 2.00 57600 45% 46% 48% 49% 50% 52% 53% 54% 55% 55% 56% 20.Table 6.30 22680 40% 41% 42% 44% 45% 46% 48% 49% 50% 50% 51% 8.00 28800 42% 43% 44% 45% 47% 48% 49% 50% 51% 52% 52% 10.00 90000 47% 48% 49% 50% 51% 53% 54% 55% 56% 56% 57% Heat recovery section of AHU By using Heat Recovery Systems it is possible to reduce the energy consumption and consequently the Life Cycle Cost tremendously .especially with extreme climate conditions in both.00 36000 43% 44% 45% 47% 48% 49% 51% 52% 53% 53% 54% 12.50 9000 34% 35% 36% 37% 39% 40% 42% 43% 43% 44% 45% 3.4.3.00 14400 37% 38% 39% 40% 42% 43% 45% 46% 46% 47% 48% 5.15 11340 35% 36% 38% 39% 40% 42% 43% 44% 45% 45% 46% 4.60 5760 32% 33% 34% 35% 36% 38% 39% 40% 41% 42% 42% 2.25 4500 31% 32% 33% 34% 35% 37% 38% 39% 40% 40% 41% 1.00 3600 30% 31% 32% 33% 35% 36% 37% 39% 39% 40% 40% 1.00 18000 39% 40% 41% 42% 43% 45% 46% 47% 48% 48% 49% 6.00 72000 46% 47% 48% 49% 51% 52% 53% 54% 55% 56% 56% 25. 132 .

Fans using the inverter.min.4.max.6. depending on the number of hours of operation per year (h/a) Annual hours of operation (h/a) h/a ≤ 3000 3000 < h/a ≤ 6000 6000 < h/a ≤ 8760 Heat recovery wheel .min. dry efficiency* 45% 50% 55% .max. give possibility to maintain the optimum speed for different air flow rate demands during the day.Table 6. Fan consumption remains constant. The combination between heat recovery and free cooling has been proven by simulation to be a simple addition of savings.max. dry efficiency* 45% 50% 55% .min. dry efficiency* 40% 45% 50% . pressure drop 300 Pa 250 Pa 200 Pa Run around coil loop . 133 .min. pressure drop 200 Pa 150 Pa 125 Pa Plate heat exchanger without bypass .max. pressure drop 300 Pa 250 Pa 200 Pa Heat pipe . pressure drop 250 Pa 200 Pa 150 Pa Plate heat exchanger with bypass .max. dry efficiency* 65% 70% 75% . In the Eurovent Recommendation there are many examples for various European conditions. pressure drop 300 Pa 250 Pa 200 Pa * Dry efficiency based on a mass ratio of 1 Running the Air Handling Unit at a speed which is needed for the actual demand will also save energy. Heating saving potential is very important in every climate and it is over 40%.min. dry efficiency* 40% 45% 50% . Levels of heat recovery as recommended by Eurovent. Cooling savings are less important and only significant in hot summer locations (3% in Seville). It is also possible to see how different parameters influence the Life Cycle Cost and the consultant or purchaser may look at the special conditions that are valid just for his particular system for the case of a cross flow sensible heat recover device with an efficiency of 0.

7.40 200000 120000 17.Central Air Conditioners TAC Large packages (Roof 2. noise.) Large splits with primary 2.30 336000 221538 34. Table 7.25 130000 128000 12.76 VAV 1. To give an idea of the cost of comfort. geometry of building) and the comparisons could in principle only take place between 2 or 3 systems.33 2P FCU 2..91 tops. we computed here the ALCC of a few systems installed in an hypothetical 2000 m2 building (800 hours equivalent. b) evaluate the relative energy performance of different systems to enable fair comparisons to be made between systems. TECHNICAL & ECONOMIC EVALUATION OF CAC SYSTEM PERFORMANCE AS A FUNCTION OF THE DESIGN OF THE AC SYSTEM We need to simulate different CAC systems in order to a) understand the relative importance of each aspect of the CAC system in influencing overall CAC energy consumption.00 326000 144000 28. optimised SSEER).80 348000 102857 31. once the building is designed.53 4P FCU 2.1 Hypothetical ALCC of a few air conditioning systems for a 2000 m2 building with comfort level TAC Components of cost ALCC Investment Energy Euros/15 y Euros/15 y Euros/m2/y RAC with Primary Air Hypothetical TAC SSEER Multi Split systems 2. We can however study the imaginary situation where all systems are feasible and where decision would be made on the basis of energy and cost only.77 (under windows) CAC ..1 Comparison of different CAC systems Energy consumption for a given comfort level The selection of a system takes into account a number of specific factors (customer’s demand.25 188000 128000 16.25 274000 128000 22.10 WLHP 2.75 Packaged systems 2.25 248000 128000 20.70 352000 169412 34.48 CAV 1.78 Note: Sizing = 120W/m² for CAC. 240 for RAC 134 .00 318000 144000 27. 7.30 VRF with primary air 2.

01 Note: Sizing = 120W/m² for CAC.30 (under windows) CAC . Moreover the importance of full and part-load operation was stressed and led us to detailed computer simulation (DOE software). All investment has been taken as amortised over 15 years which is a heavy assumption.Table 7. 240 for RAC Comparison of costs and sensitivities Note that the solutions not providing total air conditioning but just cooling.00 220000 144000 17.2 Hypothetical ALCC of a few air conditioning systems for a 2000 m2 building with comfort level TC Components of cost ALCC Investment Energy Euros/15 y Euros/15 y Euros/m2/y RAC without Primary Hypothetical Air SSEER TC Multi Split systems 2.65 VRF without primary air 2.25 160000 128000 13. They have two functions and we could as well decide to make economic calculations on the cooling function alone. In the second one (apparently Southern States).80 260000 102857 22. filter selection). but also on pressure drop in the distribution system. The overall efficiency of the air-handling system relates in principle to the efficiency of each of its consituent components.25 220000 128000 16. The sensitivity to electricity price is significant : –2/+ 3. 7.23 4P FCU 2.5 Euros per year and square meter if we consider the extremes of the prices on the EU market. the efficiency of air diffusers.40 100000 120000 9. the decision of A/C 135 . ventilation comes first as an hygienic necessity and then a further decision leads to cool the space or not.00 228000 144000 17. This doesn’t get the prominence that it deserves: because fan energy is proportional to the square of the velocity and velocity is (inversely) proportional to the square of duct diameter.77 2P FCU 2.68 WLHP 2.2 The improvement of the efficiency of air handling systems in CAC The influence of the main elements of the air handling system have been quantified. Primary Air and ventilation Two philosophies of ventilation seem to exist in Europe also responding to local natural conditions : in the first one (adopted by Northern countries).69 Packaged systems 2. The TAC solutions are in the range of 17 to 35 Euros (except the rooftop which is less expensive but has specific geometric constraints and the water loop heat pumps) and that solutions providing TC only are in the range of 15 to 20 Euros (if you are not ready to show packaged RAC in your facade). the air tightness rating and the influence of thermal insulation.25 186000 128000 14. duct sizing can be very important (and of course. The values obtained here seem very high compared with what we can read in some places. Fan specific consumption depends not only on fan efficiency.Central Air Conditioners TC Large splits 2. the operating mode. system configuration and operating conditions. Energy represents 30 to 50% of total expenditures. specially if we think of the systems providing ventilation. cost 5-6 Euros less than the others. in addition the efficiency depends on ducting: specific pressure drop rating.

we have to consider them. Heat recovery on primary air All what has been said about the improvement of all-air systems and of the heat recovery section of AHU remains true for the primary air of mixed systems. A similar rule exists in the USA for the ventilation aspect of rooftops. This is not the case for VAV systems where differences are due to flow regulation.1 Fans energy over months in a CAV and in a VAV system 136 .57 W/(m3/h) for VAV (SPF = 2. temperature.57 W(/m3/h).be applied to the whole air supply (and extract) system in a proper building thermal regulation. The only difference is due to the number of working days per month. and to controlled ventilation.05 W/(l/s)).1. In fact.4 W/m³/h) may be achieved by the use of high efficiency fans with an overcost of 2 Euros/fan kW.47 W/(m3/h) and VAV=0. parameters like the power/flow ratio or the combined efficiencies of motors/fans used by Eurovent in AHU should . As previously commented. etc. we used an average specific consumption of 0. fan consumption suffers a 15% reduction but also there is a decrease in cooling due to fan heat that ranges from 4 to 12. Obviously V allows a better air quality (dust. Figure 7. The direct comparison with DOE simulation of the effects in Europe of variable flow (option called VAV. The energy impact of the two philosophies is large but not really part of our study.in principle . in American English) has been possible. Variable Air Volume. Variable air flow and lower head losses In our study. the classic constant flow all-air system serving as a basis and has been assumed to represent the full EU STOCK of Air only systems (34% of CAC). Note that W/(m3/h) is a measure of Delta P/efficiency. the exhaust taking place in the “very polluted” rooms. The reason why the consumption figure is higher for VAV than for CAV is due to higher head losses in the air distribution network mainly caused by higher air velocities and VAV terminal boxes. Central ventilation (V) has been the base of our study. As a result Local extraction (LE) is the dominant feature in some countries. This is the normative parameter in the USA.comes first and leads to more air changes with the outside. In this section we study the sole active unit of HVAC air-side. For existing CAV system type. There are also in some countries obligations of ventilation in cascade. it includes the fan as a component and the design effort on the air circuit. Efficiency index is specific consumption and energy use will be expressed in kWh/m². Conclusions of DOE simulations: logically.47 W/m³/h. as in the US regulations (CAV=0. As an example we represent monthly fan consumption for CAV and VAV system types. You cannot install air distribution systems if SPF is not under a certain value : 0. while V (central ventilation) dominates others.25 W/(m3/h) for improved systems against 0. every system type but VAV is equipped with constant air volume fans. More speculatively.47 W/(m3/h) for CAV (SPF = 1. We have not investigated further this option.) but is more costly.5% and that we have taken as 8% on average. Motors and fans efficiency Overall Fans efficiency can be treated with a power/flow ratio like 0.75 W/(m3/h) for the worst. see figure 7.7 W/(l/s)) and 0. The flow rate is lower and you have to create the exhaust air circulation to the heat recovery station at a cost. but the capacity of heat recovery being very different in the two situations. etc. we assume that a 15% reduction of SC (0. However fans are different in the two systems. This end-use consumption is almost constant along the year for CAV fans. even if its market share is now declining.

59 3.19 4. Terminal reheat issues On air-side it’s also possible to provide evidence (simulation or literature) about the order of magnitude of terminal reheat . It is related also with the choice of an optimal blowing temperature : the higher . Conservatively. Dual Duct final distribution systems (rather inefficient by principle) are only allowed with variable flow in some countries and this measure could be extended to all countries. The cost and performance of heat recovery from condenser being known. During the summer there is also a large saving potential that is always higher than 60 %. Portugal). VAV boxes operate at minimum flow.39 4.49 0. the better for EER and for FC.99 VAV 0.98 1.79 4. 137 . There are discussions about the reality of those results on the field due to O&M issues : balancing.39 4. controllers tuning.78 0. The potential given by simulation is far higher but we have taken this conservative assumption due to potential O&M uncertainties. The present use of this feature in EU systems is far from 100%.39 3. large potential. etc.53 0. which corresponds to outdoor ventilation air ratio.39 3.19 4.5 VAV consumption reduction percentage is higher during winter because of lower loads.63 1. Free cooling simulations in DOE have been complemented with cost functions. control) for a reduction of chiller electricity by 20% (no auxiliaries increase). free cooling is very climate and control dependent. Conclusions: VAV saving percentage varies from 75 to 80 % despite specific consumption for fans (0. The main reason for this large saving potential is the adaptation of zone flow rate to real load conditions which are normally under design values due to load calculation oversizing. It seems not to present any specific potential for our study because they are very uncommon. new connections. We are using by default the classic temperature control.79 4. CAV and VAV fan consumption for SEVILLE 5 4 kWh/m² 3 2 1 0 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC CAV 4. There is a very limited overcost of 2% of overall system costs (ducting. Of course. Air Side Free Cooling (Economiser) Air side Free Cooling (FC) is the key option in a cost benefit analysis : small cost.57 0.66 0.93 1. Many regulations make it more or less compulsory (USA. In fact it's a promising option to have that perimetral heating or terminal reheat in Air systems (or part of the reheat in an AHU) made from heat recovery on condenser and not from a boiler. this option could enter the C/B-analysis. some think it should be completely banned in regulations (except when provided from some renewable source of energy like condenser heat). Europe should take such a measure when applying its “Energy Performance of Buildings” new directive.99 3.5 0.47 W(/m³/h)).54 0.19 4. we made calculations with a 50% saving value. The Ashrae approach requests variable speed in all air ducts not only blowing through a target of consumption to be reached at full load and another one at a specified part load value.57 W/(m³/h)) is higher than that for CAV (0. On average we could say that VAV save 80% of fan energy in Europe.11 0.

Pumping consumption is also reduced due to cooling load decrease and pumps "on demand" control. There is also the possibility to operate them directly from a cooling tower where cold is generated by evaporation. Displacement is now reported to be less costly by 15% in installation costs but. 3.3 Other cost & efficiency trade-offs Water-side efficiency by sizing and control Oversized FCU of classic type will allow higher operating temperatures at chiller’s level hence an improved EER. • Mandatory air leakage tests (air leakages of more than 20% are common).) since the larger part of the fan pressure is dissipated in the AHU. and even for very hot climates like Seville.0 m/s and 3. we have a very energy efficient treatment of demand. The quality of the duct system should be controlled in some way. AHU improvement We should promote the investigation on the development of low pressure drop AHU components (filters. etc. Quality of Air Diffusion We have in the Italian UNI standard a set of interesting figures. 7. It will at the same time reduce temperature difference between room and FCU from 12. 138 .5 K to 11. For this reason by introducing cool air at low speed at the bottom of the room and leaving it move upwards when there is a heat source. the fan energy consumption and the unit cost of the options for the air speed in the unit (2. We have investigated only the first phenomenon (temperature regime) and only in one case.5 °C to 10. The option of “displacement” allows to lower significantly the loads. This process can be extrapolated by a few more degrees but two phenomena appear : the exponential nature of heat exchange and an additional demand for blowing power inside the FCU. entering in a different manner the equation of human comfort. • Use of low pressure drop connection pieces. Every AHU should be equipped with a small variable speed pump that delivered the correct water flow.5 °C.5 m/s. the system is unable to answer to the total heating needs in winter and an overcost appears somewhere else. at each moment (the control signal that normally goes to the control valve would go to the variable speed pump). How to consider the option of radiant panels? since the benefit claimed comes from the change of temperature regime in the water distribution equipment. Research has not yet produced perfectly consistent values for this approach which already works in practice (a few percent of new installations). but no cost data. • Air balance criteria and equipment. for example. There is a second order effect coming from the fact that part of the cooling is radiant. For instance moving from the classic 7/12 °C regime to 8/13 °C will increase average from 9. coils. heat exchangers. costs and benefits have been evaluated by extrapolating the temperature effect to the new data both in summer and winter. In any case.5 m/s). The piping systems should be designed so that all AHUs should not have balance valves and flow control valves.5 K (with 22 °C inside) and increase requested area. In reality a large portion of the loads doesn’t impact comfort because they are too high in the room. as in the case of chilled ceilings. sound traps.0 m/s. They allow an increase in distribution temperature by using large areas for heat exchange.The real figures given by DOE show that the cooling saving potential is strongly dependent on climate. This measure eliminates the energy pressure dissipated in the balance valves and the flow control valves. • Velocity or pressure drops limits. Oceanic weathers like London offer a more than 80 % reduction of cooling consumption. 2. Radiant panels or beams are one step further in the same direction. The size of the AHU should be determined after a LCC analysis considering the annual operating hours. saving potential is over 20%.

High pumping savings range from 60 to 72% (relative to pumps consumption). In a few DOE simulations.6%) and we abandoned this solution in the study. Another philosophy (advocated by one company) explains that for the same average fluid temperature in the final FCU (consequently its cooling capacity) one should look for the largest possible DeltaT between inlet and outlet. Fans situation and performance is quite unclear and we decided to wait for the next step of development of the system to include it in the analysis. In principle it could be simulated by generating extreme behaviours. they are mounted under a plenum with injected air. The electricity saving is very low (0. Design of flow in water circulation Efficiency of pumps for water circulation is the first obvious issue. and present average values are 0.6 to 5. Every equipment should be equipped with a small variable speed pump that delivered the correct water flow. High performance values are 0. The assumption was to determine performance improvement with a small ΔT difference in a FCU4P system (8/13°C instead of 7/12). ventilation is started. a local unit (having the appearance and the cost of a local radiator or FCU of good quality) combines ventilation (controlled room by room) and radiant cooling.1 to 1. The inversion from cooling function to heating function and vice versa is a very general problem of systems and should be manually operated or automated carefully.1% (less demand since the heat is not dissipated by the pumps). difficult to regulate because very dependent on experience gained with a specific building and climate.62. The modification of the temperature regimes on the chilled water loop (set point and point of application of the set point : departure or return) do change consumption. The piping systems could also be designed so that all equipment should not have balance valves and flow control valves. This supposes a decrease in Chiller consumption and an increase in Pump consumption.85 and 0. it’s a good practice to “reset” it off design. we valued the saving potential of using two-way valves and variable flow pumps. We have simulated with DOE software this option.7 to 1. we will save pumping energy. at each moment (the control signal that normally goes to the control valve would go to the variable speed pump). Pumping saving potential of this measure is 13%. recently patented by the firm Van Holsteijn en Kemna (VHK). since the relative returns will diminish if we move further. On the examples taken from company documents. that is. the cost seems the same as the one with FCU. respectively.1 to 0. there is a significant benefit at system level from doing this. If the cooling capacity of the primary air and the capacity of the chilled ceiling are enough together to ‘treat’ the load. The electricity saving is very low (0. but we decided this was too uncertain and we abandoned the idea of simulating or regulating this type of intervention. This measure eliminates the energy pressure dissipated in the balance valves and the flow control valves. If we improve pumps efficiencies.In a special configuration of the ceiling panels. Entering water temperatures under room temperatures by only 10K would be enough to deliver 200 W/m2 of radiator. It may have a low energy penalty if done properly : this is a behavioural more than technical issue. Our base case assumption is that every system has three-way valves and for this reason circulation loop and pump flow are constant. Two types of information are available : load and outdoors temperature.3%) and we abandoned this solution in the study. pump is controlled by a variable frequency drive. In fact the difference appears in winter because the ceiling looses part of its heating capacity compared with a regular FCU and another system is needed in addition.8 and 0.67 for motor and mechanical efficiencies. Total saving potential is low (from 0. Influence of terminal equipment 139 . Decrease in cooling consumption due to pump heat ranges from 3.7%) due to the limited importance of pumping energy. When radiant cooling is not enough. likely for increasing the relatively small cooling capacity. Pump efficiency is the result of motor and mechanical effects. respectively. Even if temperature cannot be increased at design conditions. In another realisation of the radiant system.

In the same way.41% 4. Such advanced multizone systems can be justified by its actual benefits.24% 0.92% -0. Terminal units demand what they need and the evaporator ‘sees’ a variable flow.…).Improvement of flow efficiency in FCU and AHU water circuits is possible in various ways : Variable flow pump. 4P is around 15%).78% PACK 32. Evaporators can accept flow rates decreased by 40% typically. one of the reversibility approaches. where various facades receive differently the sun.03% FC2P -6. The DOE simulation allowed us to understand the real order of magnitude of simultaneous heating and cooling. Simultaneous demand of heating and cooling Some systems have a capacity to transfer heat from one zone to another. VRF. More precisely.24% -5. We have gathered some elements on advanced multizone strategies (WLHP. Four pipes fan-coil. But it is also an interesting system for transfer between zones demanding heat and cold (but this not always realised).22% 7.08% WLHP 31.91% 1. etc. Finally VRF is one step further in the same direction.88% 7. It is one way of operating at variable speed (see part on packaged systems). We have considered the basic Air and Water System types (Two pipes fan-coil.74% 4. better Pump efficiencies.06% 7. It seems a relatively frequent solution in commercial malls because it allows individual metering of consumption by each user.26% -0. the effect corresponds to only a few percent of the demand.16% -23. In this relatively complex office building where internal and external zones are treated separately. we have computed for each hour with simultaneity the lowest of the two quantities : cooling demand.3.59% 0. Table 7.5 % of total cooled area) and presents specific energy conservation features : transfer from one zone to another. Water loop heat pump).3 Importance of simultaneous heating and cooling In % of heating CAC system LO MI SE CAV -0.62% PACK -5. TWL. either cooling or heating.67% -18. A remarkable option at system level is to go for “local pumping systems”. Note that we can estimate that on the market those two solutions are still frequent (2PE was 25% a decade ago and seems now at 10% only. The system using RAC on a water loop (WLHP) is relatively frequent (1. the uncommon TWL (a promising two water loops system experimented in France and in the UK allowing simultaneous heating and cooling) can provide simultaneous heating and cooling.so it’s the main enemy of reversibility.33% -4.23% -4.19% -5.92% -3. Using four pipes FCU or using the two pipes system based on “change over” will allow to feed the four pipes FCU from heat recovered from the condenser. adequate Pump performance curve.90% VAV -0.45% -6.62% FC4P 29.53% 140 .61% FC4P -7. The importance of additional consumption generated by the 2 pipes FCU with electrical heating versus the 4 pipes –if we want to insure the same temperature all the time. heating demand and expressed it in percent of demand.89% -17. table 7.50% FC2P 26.47% -21.49% In % of cooling CAC system LO MI SE CAV 0.can be estimated.17% WLHP -6. high EER and COP year round.73% 4.15% 0.21% VAV 5. The gains in pumping cost become significant.66% 7.. Also we suspect that the electrical resistance is used for main space heating not just for this adjustment….42% 3.

Control is easy since underground temperatures are constant. in systems with more than one cooling tower. etc. Using natural water –river. Costs and administrative problems are reported as enormous in Italy and Spain while France maintains such a policy (Aquapac). ground water. The designers need the whole range of solution to cover the domain of geometries and air quality requirements. 7. There is also the possibility to generate DHW (Domestic Hot Water) at a small cost from condensing heat.The economics of the transfer are as favourable as expected.1 Bottleneck of air systems ALCC Euros/m2/year air packages water rac SPECIFIC CONSUMPTION kWh/m2/YEAR The result of optimisation Table 7. So the bottleneck to the expression of a global reduction in consumption will be the point (shown hereunder by an array) where the improved air based solutions start not to pay for themselves : the designers will find it is too heavy a constraint. In some circumstances the chiller becomes useless and the natural water can cool directly the building (see system TWL as an example). all cooling towers should work simultaneously at all times (this strategy reduces drastically the cooling tower fan energy consumption) Temperature control should be modified. The homogeneity of figures between air systems on one hand and all other systems on the other hand is interesting..2 OPTIMISATION OF AIR BASED IN SEVILLE For For Total Initial Cost ALCC ALCC ALCC cooling secondary electricity of of of and fans system system system 141 .4 The possible strength of regulatory efforts and the minimum LCC solutions Concentration of efforts on Air based systems We have concentrated our efforts on the air system which show presently (under the CAV form) the most consumption and the highest cost. Heat rejection Cooling tower fans should have variable speed drives and.as a heat rejection medium is very beneficial in energy terms because the high heat exchange coefficients and low temperature at the condenser improve EER. Figure7.

00% 120.76 1 062 390 30.76 135.17 (MWh) E/kWh) E/kWh) E/kWh) 0 – CAV 121.55 1 090 090 31. -3% +6 E/m2 5 –Lower HL 121.79 1 126 540 32.67 35.65 29.84 32.12 156.34 37.76 159.00% % of reference The optimum if very flat.78 33.00% 100.88 140. specially if we get interested with the highest cost electricity.16 32.67 18. 118.61 in AHU – 7%/+12E/m2 6 –Optimised 73.80 31.70 32.3 Optimising with 6.81 30. The regulatory measure could be taken anywhere between a 0% and a 60% reduction without generating overcosts (in the LCC definition) in Seville.06 (0.67 37. 142 . Figure 7.59 31.56 30.63 33.00 37. the optimal trajectory of improvement is given in figure 7.10 1 074 710 31.60 chiller – see chapter 6 After sorting and combinations.73 1 – VAV-50%/ 121.77 34.00% 60. +2 E/kW 4 – HR 60% .94 32.02 37.00% 80.03 1 054 140 30.10 (0.00% 20.76 110.32 29.–15 %. 10 and 17 cEuro/kWh a full all air system 39 37 35 33 ALCC17 Euros/m2 ALCC10 31 ALCC6 29 27 25 0.00% 40. pumps (MWh) (0.60 200Pa.89 29.76 155.05 30.43 1 053 640 30.17 +6 E/m2 2 – FC –20% 97.78 1 090 090 31.3.10 144.31 .61 /+2 Euros/m2 3 – Fans –8% 111. The optimal regulation would request a 50% cut in electricity consumption compared with a standard CAV design.

Such a figure is largely available on the US market and is called there an IPLV (Integrated Part Load Value). The EMPE methodology is not different from the IPLV one.1 optimisation exercise). In Italy. 2002]. 1997] for heating and [CEN.8. it is to say that the error coming from the reduction of the data to single points should be inferior to the testing uncertainty. How to reduce the capacity of a chiller? The performance of each of the capacity steps will differ per se even if the operating conditions (entering air or water at the condenser and leaving water temperature) are identical. or grading chillers from A to G in an undiscussable manner. The new ESEER method is compared with the US-IPLV and EMPE proposal under both respects : time spent and accuracy. 1998] for cooling. as described in chapter 2. hence the proposed name. a provisional standard that could become a full standard (on which certification could be based) in case experience is gained about performing part load tests. it is an EnV. The compressors treated hereafter are of three types. For a large set of modelled chillers. - 30 % in electricity consumption in our chapter 6. Given that the US-IPLV climatic conditions are not relevant for Europe. We shall propose such an ESEER in the coming pages. defined as the ratio of the cooling capacity of one stage to the full load capacity stage. a comparison is drawn between EMPE and IPLV figures.5. The constraint is to minimize the testing time while ensuring maximum precision. About part load behaviour of centrifugal chillers we give only qualitative indications. i. This couple of standards has to be replaced soon. by its own revision [CEN. Only a ESEER largely agreed can be accepted as a basis for comparing chillers.e. screw or scroll compressors.1 The importance and nature of part-load management measures Importance of establishing a EU method about part load Until now. Consumption is not governed by full load EER given in such a standard but by the average part load EER. Part load performance is tested at different part load ratios. the gains and losses obtained by part load management can be mixed in some manufacturers documentation with more ‘artificial’ or ‘conventional’ gains and losses due to temperature conditions in testing. the ESEER (European SEER). however. the chillers and other vapour compression cycles have been tested using the full load standard [CEN. An extension to the new full load EU standard contains some information about part load testing. 2002]. The resulting index is called EMPE. leading to show the direct application of the IPLV to Europe would give overestimated values for the chillers’ seasonal efficiencies. The goal of this chapter is to define an ESEER method that enables to calculate the seasonal efficiency for all European chillers (centrifugal units are not treated explicitly in this document by lack of specific information but seem likely to be covered by the proposed method). a part load performance standard has already been defined and accepted [UNI. As long as a good method is not agreed. reciprocating. The potential gain associated with part load management is high (for instance +30 % in EER. called often a Seasonal EER (SEER). Both depend on the kind of compression circuit. Italian manufacturers have made a proposal for using the same methodology as for the IPLV but using different conditions for air and water condensing temperatures. A good SEER definition is the essential tool for achieving actual and comparable gains not artefacts. The one we are looking for should represent the universe of EU buildings and climates. The evolution of the efficiency with the part load ratio is still a subject for research and we have proposed here original results. We need to describe the means of reducing the capacity to understand the reduced temperature and the part capacity behaviours. Centrifugal compressors at part load 143 . EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE 8.

The valve at leaving opens only when the pressure in the chamber reaches the pressure of condensation. a slide valve or an equivalent steps system (corresponding to discrete steps for the bypass) is used to control the capacity. There are different modes for unloading the centrifugal chillers : prerotation inlet vane guides. The part load performances seem always less than full load ones in what we have investigated. this compressor adapts its discharge and evaporative pressure to external conditions. Thus. The Vi is in fact directly related to the compression ratio since it is the ratio of inlet to outlet gas volumes of the compression chamber. the inlet valve remains open as long as the pressure in the chamber is lower than the suction pressure. Speed of rotation : centrifugal chillers cannot be operated for small flow rates since the rotation speed needed would be too high. from [PILL85]. the following stage is called variable Vi. and thus the cooling capacity of the chiller. the second valve enables to adapt the compression ratio to the condensing pressure for each capacity step while the first one enables to adapt the swept volume. the valves of both the others being closed. This is the temperature aspect of part load. the two stages compressor enable to unload at a lower step. Thus. forcing the manufacturers to increase the full load speed to be able to reduce it at part capacity operation. At suction. the compressor has no means of adapting the pressure of exit of compression to the pressure of condensation. Part load : the surging phenomena occurs at low part loads. It makes it possible to shunt part of gases off the compression chamber to adapt the swept volume to the one needed to match the load. The last option should be to use a variable speed drive for the motor. a hot gas bypass strategy is adopted leading to still poorer performances at very low loads. Screw compressors at part load In its basic configuration. However. and Vi variable (b). For part capacity behavior. Reciprocating compressors at part load The reciprocating compressor owns a spring valve at inlet and outlet. By associating in series two compressors the surge limit goes under the single compressor one (10% instead of 20 or 25% load). any difference between the discharge pressure and the condensation pressure is synonymous of energy losses. this induces a heating of the engine. In a centrifugal compressor. The other treatment of part load is through capacity reduction. contrary to the reciprocating compressor. when the load becomes inferior to the surging load. This induces pumping losses. This slide valve (Figure 8.1) is generally controlled by discrete steps. In practice. But the option is scarcely expanded because of its cost. 144 . the valves of both the others remaining open. given that reducing the speed of this compressor decreases the tightness of the lobes of the screw compressors. Hence. Mechanical realization of part capacity for a compressor with four pistons and two stages of compression is : • only two pistons compress the refrigerant. the flow comes back through the impeller leading to a cyclic phenomena badly known so that manufacturers forbid the chiller to work in these conditions. the end of the piston allows gas into the high pressure side. However. before the impeller.The efficiency of this kind of chillers at design point depends on : the size of the compressor (less stages = better efficiency because of intermediary losses). the inlet vane guide enables to create more or less swirl to reduce capacity to match the load. Some solutions with variable drive speed exist on the market. to enable cutting off the compressor. the fluid which passes in the pistons in open position is pumped. variable speed. • or only two pistons compress the refrigerant. when external conditions vary. This is a mechanical type of unloading.1 : slide valve position experiencing low Vi and high Vi (a). These two are the more common ones. Figure 8. At that time.

Thus. The corresponding operating times for each capacity step 145 . but differed by more or less 1°C and sometimes even more from user selected values. The set point is always a control parameter to be entered by the user whereas dead-band can be fixed by the manufacturer or not. dynamic testing installations are not available and would need long debates to be specified and then adapted by certifying laboratories. Figure 8. one uses several compressors in parallel on the same circuit and make them cycle. Generally. variable Vi unloading (for screw chillers only). one can use variable speed drive (for screw chillers only in our scope). capacity staged chillers are controlled using a water inlet or water outlet set point control with a dead band. the water temperature varies more or less between 6 and 8°C for all stages. water temperature measurements do not correspond to real temperatures and the control behaviour can be far different from Figure 8. Typical control of a 5 capacity stages chiller in the cooling mode. Staging of Part capacity (control issues) Generally.2. the one represented by the scheme Figure 8. To perform this latter control of the refrigerant flow in the cycle. depending on the water loop inertia and on the condensing temperature that modifies the cooling capacity of stages. neither adaptation to varying condensing pressure.2. even if some experimental testing of dynamic capabilities of chillers have shown that chillers did not always behaved this way [AFCE. different technologies are used to control part capacity stages. two capacity steps plus the full load are available if nominal capacities of each of the two compressors differ.2 shows the control scheme for a five step capacity chiller on the inlet water temperature. one can find unloading by varying the number of available circuits or by varying the flow rate in one circuit. available on screw and reciprocating as a supplementary mean). On installed units such equipments are not installed. or shutting down a compressor over two or more (sole option for scroll. unloading of multistage compressor (screw or reciprocating). either very long straight pipes are needed so that a homogeneous flow may be reached before measurements or. set point control temperatures were not respected. Depending on the kind of compressor circuit. the cooling load is the weighted average of the two steps cooling capacities CC for the same inlet condensing and outlet temperature.Scroll compressors at part load The rationale is about the same one as for the basic screw : nor unloading is available for chiller applications. It has also been observed that most of the time. pieces of equipment have to be installed at the inlet and outlet of the evaporator to enhance the turbulence. Cycling between stages at part load The following representation of the chiller performance when load is higher than the smallest capacity step is adopted : if the load lies between two capacity steps. For correct measurements of inlet and outlet water temperature (like the ones used in standardised testing).2. For two scroll compressors on the same circuit. Dead-band 1°C Part load (%) 100 80 60 Dead-band 1°C 40 20 Inlet water temperature °C 0 6 7 8 9 10 11 12 Given that chillers generally operate at full load and nominal inlet condensing temperature with a 5°C between inlet and outlet. However. 2002]. Stages are successively triggered when water temperature increases and moves apart from the set point. the chiller will operate on each one of the two neighbouring steps . We will assume here perfect control. Figure 8.

Supplementary work could be performed to set the experimental testing conditions enabling the calculation of this coefficient from the manufacturers specifications of the minimum water loop volumes and of the smallest capacity step available. the compressor has to establish the pressure difference between low and high pressure sides.9 full load efficiency 0. On the contrary.1 0 0 0.3. the chiller operates only part of the time. we selected (figure 8.3 0. establishing the full electric power is quite instantaneous. 2002] Equation named (2) hereunder.9 is a proposed default coefficient. Degradation of the reduced efficiency versus the part load ratio (same sources temperatures ) Ratio of part load efficiency to 1 0. This leads to an energy loss at the starting of the chiller.3) the Italian standard [UNI. the superheating of the refrigerant has to stabilize : only at that time the full capacity of the step is reached. thus fitting its cooling capacity to the load.2 0. The unit only begins to cool water when the average refrigerant evaporating temperature is lower than the average water temperature.2 0.4 0.9 1 Ratio of part load capacity to full load capacity (same source temperatures ) The corresponding curve is represented on the figure 8.C cyc + 1 − C cyc CC FL With Ccyc=0.enable to determine the electric power absorbed EP and thus the efficiency for each hour.5 0.3. (1) 8760 ∑ CC i SEER = i =1 8760 ∑ EP i =1 i where CC is the cooling capacity and EP the electric power absorbed in each operating condition actually met. High pressure control at part load The high pressure cannot go too low.3 0.6 0. Classical control consists of maintaining the high pressure above 15 bar by cycling the 146 . It is generally maintained high enough by controlling the air flow rate for air cooled chillers. Figure 8. Ratio of part load efficiency to full load efficiency for the same inlet condensing temperature and outlet water temperature as a function of part load capacity in the same conditions. At each starting. It is used here to compute the part load performances of single circuit units and the performances of multi- staged units when load is inferior to the smallest capacity step available. it corresponds to the following formula : CC EER = CC FL (2) EER FL CC .5 0.8 0. Yearly efficiency is calculated with Equation (1). When the cooling load is lower than the smallest capacity the equipment can deliver. In that case. Then. each starting is an energy loss.1 0. Ccyc=0.7 0.4 0.9.8 0. As a consequence of a review of all existing experimental evidence.7 0.6 0.

EP/Epnom.95 0. The impact of this phenomena has been measured while carrying over a test campaign on a scroll unit.75 0. due to the mechanical flow rate reduction. The following point was to determine how the efficiency varies with the outside air temperature when the chiller operates at reduced flow rate : does the reduced efficiency increase with the same slope than with the full air flow rate ? Given the NUT-epsilon curve of the heat exchanger. one compressor in operation on each circuit.05 1 EP/EPnom 0.2 1.5.9 1 % of nominal flow rate The decrease of the efficiency while varying the flow rate is reported Figure 8.8 0. Then. Evolution of the chiller performances when reducing the condenser air flow rate presented under a reduced form (EER/EERnom.85 0.fan or by switching one fan OFF (the first one being the better one since the whole area remains in use for the heat exchange) with a fixed dead band.7 0. 75%.15 1.7 0.9 CC/CCnom 0.1 EER/EERno m 1.5 0.6 0.4. Figure 8.8 0. CC/Ccnom in terms of reduced flow rate) Reduced performances while varying flow rate 1. but in a more complex manner. The 70% flow situation corresponds to 2 fans among 3 being ON for each circuit. the high pressures are higher than at 50% load. 50% and 25% . The 50% flow rate corresponds to one fan functioning on one circuit and two on the second. the increase of the efficiency with the outside air temperature decreases faster than at full flow rate. Decreasing the flow rate at the condenser increases the high pressure and thus decreases the performance. in fact. 50% and 25% capacity steps.4. Figure 8. This fact is observed in reality. The efficiency decreases with a square tendency when the air flow rate is reduced. At full load. Condensation pressure control effect on reduced EER (EER/EERnom) in terms of outside air temperature 147 . Each circuit has 3 fans on its condenser (line configuration). Each circuit has a tandem scroll compressor. the 75%. are slightly higher than this theoretical staging. which means two steps by circuit. the capacity steps available are 100%. The unit is divided into 2 distinct and symmetrical refrigerating circuits. at reduced flow rate. Precise measurement of the flow rate was not available. The chiller was operated at 50% load.

an analysis of the method shows that the ARI part-load temperature testing points are "sized" to be "representative" of US buildings (cooling even in negative Celsius temperatures. cooling above –17.5 we can understand that the high pressure control will impact differently single circuit and double separated circuits units. the 25% load point does not differ from the 50% since one compressor is in operation on one circuit. single 1. chapter 2.2 and 12. for that double circuit unit.2 °C.as can be shown by drawing the loads in terms of outside temperatures). Buildings used in deriving the US-IPLV This standard covers all the tertiary sector buildings for air conditioning application on the whole US territory.60 1.70 FL curve 1. • Group 4.40 PL curve. Thus. Climate used in IPLV derivation 3 [ARI98] states that these cities represent 80% of the installed chillers in the US.2 and 12. free-cooling between –17. Further to this. one circuit is operated at full load while the other is operated at 50% load. A single building has been “averaged” to be representative of buildings of 29 cities3 chosen to be representative of the places where chillers are installed in the US. occupation 24h/day : 7days/week.5 for 25% and for 75% load on a single circuit.2 Is the IPLV approach directly applicable to European conditions? The percentage of operating hours spent at each part load condition. • Group 2. 148 .50 PL curve. occupation 12h/day : 5days/week. For double circuit and 75% load operation.8 °C.30 circuit 75% 1.10 PL curve. cooling above –17. single circuit VSDF 1. (1% etc…) is intended to be representative of the US climate and buildings but not of the European ones. single circuit 50 % 1. cooling above 12.2 °C. • Group 3. occupation 24h/day : 7days/week. For single circuit units. free-cooling between –17. for instance. Four building groups have been identified depending on the occupation scenario and the possibility to use free-cooling or not : • Group 1. the efficiency variation with outside air temperature is shown Figure 8.20 PL curve.00 15 17 19 21 23 25 27 29 31 33 35 Without high pressure control OAT (°C) On figure 8. given in our description of US-IPLV. cooling above 12. 8. Thus not only has the load been varied as in the draft CEN part-load test standard which is currently under discussion but also the temperatures.8 °C. On the contrary.8 °C. Reduced EER versus OAT for different load ratios DOE2 curve 1. single circuit 25 % 1. occupation 12h/day : 5days/week.8°C. the 25% load would correspond to still higher ambient temperature for triggering fan cycling or reduction speed.5. the 50% being strictly symmetrical.

one obtains the energy needs curve (figure 8. the results are weighed by the representative coefficients of the groups amongst US building studied. The load curve of group 1 is given figure 8. the missing efficiency will be obtained by interpolation of the two closest embedding efficiencies available within testing point according to ARI testing load temperature curve.8°C. Internal loads represents 38% of the total load above 12. an occurrence curve of dry bulb and wet bulb air temperature is drawn by 5 °F (2. four integration intervals are then defined: The sizing interval (for temperature higher than 95 °F : 35 °C) and the three other intervals : [0.The mean climate used to perform the bin method is an average of the 29 cities climates. Each group is weighted in function of its relative weight in the US from a statistical study.6. Then.75].2°C) for groups 1 and 3 either is null under 12.8 °C) bins. The real unity is the hour but this name enables to remember that it characterizes the repartition of the energy to be delivered versus outdoor temperature. What is shown is actually the product of the hours by bin and of the normalized capacity in %. or 35°C. • On this curve. Coming from this average climate. Calculating US weighing coefficients The following calculations depend now on the group chosen. temperature higher than 95°F. two possibilities exist : either it can unload at a lower capacity than the missing step or it cannot.4 °C) approach. from [ARI98]. Then. The nominal full load sizing for the four groups correspond to the highest bin. The climatologic data were averaged without any weighting of energy or capacity installed in each city. 149 . between –17. one integrates the energy curve that will give the energy weight of the four testing points called respectively A. C and D values. The water temperature is deduced from the wet bulb temperature using an added 8 °F (4. [75. [55. the load curve is multiplied bin by bin by the number of hours experienced in each bin considering the average climate. • Taking group 1 as an example.6 : annual cooling needs as a function of outside air temperature. Building cooling load calculation in US-IPLV The details of the calculation are not explained. If it can. Interpolation scheme needed to reduce testing time If the unit cannot unload to one of the specificied steps. The ARI 550/590 unit for energy is the ton-hours. For each interval.2 °C and 35 °C.6). The building load experiments either a 20% of maximum load at 0°F (–17.8°C for groups 2 and 4.95]. The only indications are given in the text. Figure 8.55]. group 1. B.

the IPLV exact figures would be obtained. 2001] Italian proposal is that the operating conditions are rather different from Southern Europe conditions. The graph differs from the representation used in the IPLV standard for part load.Figure 8. first stage] and the [26.7 : Interpolation procedure illustration 35 Outside air temperature (°C) ARI air condensation 30 curve 75 25 50 25 20 15 10 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Part load ratio (%) One example of the interpolation can be described with the help of figure 8. 150 . The solid and very black line is the ARI assumed load/temperature curve. And even. the capacity step chillers being penalized for not supplying continuous unloading. the global seasonal figure is underestimated by the interpolation procedure whereas. 2 and 3). second stage] points. Thus. second stage]. it cannot be said that Italy would need air conditioning at 12. first stage] due to temperature decrease. first stage] and [18. the 50% point efficiency is underestimated in that case since efficiency of the [26. The less solid black line is determined by the capability of the four steps of the chiller to reach the four specified temperatures.7 illustrates the interpolation procedure for unit N°7. To avoid the multiplication of the number of testing points.8 °C as normal operating conditions. The corresponding point is located at the crossing point of the ARI curve and the horizontal plain arrow at 18°C and 50% load ratio. The method is interesting since it enables to reduce the testing points number and its effect will be discussed further hereunder. Thus the efficiency of the 50% point 18°C should be calculated as the weighted average of the two points: [18. Figure 8. EMPE: an answer to a need for a European weighting with IPLV-like testing The first remark that led to the EMPE [AICARR. if Northern Europe countries may need air conditioning in summer. Thus instead of weighting the two previously mentioned points. the ARI procedure uses only the testing points (full circles) to perform interpolation of the efficiencies for specified load points. the chiller will cycle between the 50% and 25% stages.7 : when the outside air temperature is 18°C and the load between about 30% and 62%. the standard proposes to weight the [18. second stage] point is lower than the [18. for a continuous control screw unit. The consequence is that for that unit number 7. Three straight lines representing the capacity stages relative to full load 35°C capacities are drawn for the stages 1. The load ratio in % is related to full load and 35°C inlet air temperature.

one for part load reporting and certification. which aggregated many factors) was also its weakness. understanding better part load. Two testing points only are performed.1: Table 8. Reduction of EMPE or IPLV to 2 points with extrapolation The following method has been proposed by one manufacturer. in particular. and proposing two separate methods. The EMPE formula is absolutely similar to IPLV. We had the opportunity to go further by constructing a data base of EU chillers at part load. Substantially it is assumed that the efficiency changes linearly with the load (the temperature also decreases as in the IPLV or EMPE or at other conditions). The EER value at 75% of the load is calculated as follows: 151 . named EMPE (Average Weighed Efficiency in Summer regime in Italian) directly deriving from IPLV. with different temperatures at the condenser inlet. with different energy weights and. Its strength (being very close to the existing US method. the other one for the computation of SEER. but the values of energy weights and inlet temperature to the evaporator and the condenser are those indicated in table 8.1: calculation conditions for EMPE The AICARR proposal.Therefore. The EER values at 75% and at 25% of the load. It is an amendment to IPLV or EMPE. fitter for the European climate and requirements in the air conditioning field. would be obtained by interpolations and extrapolations on the IPLV or EMPE curve. the first one at 100% load and design temperature and the second one at 50% of the load and the associated reduced temperature. necessary to define the index. AICARR proposed a new energy index. The manufacturer could then decide if making 4 tests or only 2. EMPE was not based on a sufficiently large climatic and technical investigation.

Moreover.2 : Comparison of the proposed 2 points methodology for scroll air condensing units with the EMPE Air cooled scroll chiller n° 5 6 11 13 14 15 2 3 EMPE 3. For all chillers. The technical description of the chillers tested follows on tables 8. some always benefit. For chillers n° 2. Tested water-cooled chillers Name Type Circuits Compressor Available Stages N° 1 Screw 2 3 8 N° 3 Screw 2 2 4 N° 4 Scroll 2 4 4 N° 6 Screw 1 1 Continuous For all the tested chillers.4.82 3.3. Table 8. 3. EER100% + EER50% EER75% = 2 The EER value at 25% of the load can be calculated as follows: EER25% = EER50% + ( EER50% − EER75% ) This system is representative when a control step is placed at 50%.3.26 3.92 3. Table 8. 4 152 .39 2.22 3. 2002].59 3.39 2 points method 3.2). We have tested this simplification extensively on a set of chillers (table 8. Construction of a data base of EU chillers at part load –understanding part load Testing conditions and available testing results Original knowledge has been generated during the “Joint project” of EDF R&D facility and manufacturers from Eurovent wanting to promote part load performance. Moreover a bias is introduced in the classification : some always loose. This methodology suffers the same bias that was introduced by the interpolation process.42 3. The main tool used was actual testing of EU equipment but a number of group meetings allowed to build a common thinking frame.85 3.3 and 8.32 3.50 3. some common testing points were made according to either the ARI or the EMPE conditions depending on the manufacturer will.16 2. even for the very commune air scroll range.48 Relative difference with EMPE 14% 6% 12% 7% -5% -14% 16% 3% The uncertainties that have been generated are too large compared with the accuracy expected for the seasonal index.35 3. again increased. practical limits as well as the impossibility to predict cycling keep us from recommending that method.4. a supplementary point was added to fulfil the CEN EnV requirement : nominal inlet condensing temperature (35°C for air and 30°C for water) and 50% load ratio referred at this nominal inlet condensing temperature [CEN. The evolution of part load efficiency at reduced temperature cannot be modelled simply by a linear regression : it depends on the unit. 8.80 3. split by condensation type. Tested air-cooled chillers Name Type Circuits Compressors Available Stages N° 5 Scroll 1 2 3 N° 7 Scroll 2 4 4 N° 8 Herm rec 2 2 2 N° 9 Scroll 2 4 4 N° 2 Screw 2 2 Partially continuous Table 8.91 3.

only IPLV or EMPE points plus the CEN one were available. as usual. based on a SEER. as a whole.50 150. We arrived also at applicable conclusions on the way to report the SEER in the Eurovent directory. the electric power absorbed. plus the final Eurovent grading when it is available. We started from HSEER. No need to test every condition reported in the “map”: the benefit of Eurovent is the fair and independent choice of a few points on the map. but there can be as many similar indices as specific demands: sector. etc.and 8. the air cooled chiller stage efficiencies can suffer different fan pattern and/or circuit separation that do not infer for water cooled chillers. selected in the same conditions as usual. not only values at the four arbitrary percentages and temperatures.U. The customer can rely on the Eurovent SEER computed from this map … or compute its specific SEER for its specific demand. country. It is proven that each set of outside conditions (for each sector. But a given part load regime has to be defined by the actual status of each piece of equipment.17 38.47 4. Indeed. Let’s note that it is far easier to analyse the part load behaviour of water cooled than it is for air cooled chillers. type of chiller. In all circumstances a simple model has been used to draw the performance maps from existing testing points. like in table 8. climate. fours stages can be defined in terms of input.5 : Part load performance of water cooled scroll chiller N°4. and the associated independent testing.27 4. It is so when there is one single compressor per chiller. We have introduced a format for the description of the stages of a chiller. the cooling capacity and EP.92 Now we shall present examples of the proposed procedure. Note that this procedure is in fact already used for some chillers with various speeds. For the others as many testing points as desirable have been obtained. Water cooled chillers –experimental results 153 . A significant market share of chillers have various compressors and a complex circuiting. The ESEER index proposed here is a set of 4 conditions given for E. it would be easier to describe performance at a given stage not at a given percentage. suitable for Eurovent specification. For each stage. or various identical circuits. About certifying Part Load : what the manufacturers give to their customers is a « map » of performance. the DOE reference that we generated. For the very few continuously controlled chillers. type of secondary system) can be reduced to four or five external conditions without loss of accuracy. For discrete stages chillers. as could be reported in Eurovent part load certification scheme N° 4 // WT : 30°C STAGES 1 2 3 4 Compressor 1 0 0 0 1 Circuit 1 Compressor 2 0 1 1 1 Compressor 3 0 0 1 1 Circuit 2 Compressor 4 1 1 1 1 EP (kW) 8.00 EER 4.5 and following.00 112. The chiller part load behaviour is described first for water cooled units. The certifying body has only to check a few of the values. there is no need for combined testing (like IPLV). the manufacturer has only to declare which piece of its equipment is operating and to indicate CC .27 CC (kW) 37. namely “low noise” chillers with the possibility of reduced fan speed. leading to improved part load performance. Table 8. then air cooled units. Impact of load reduction on the efficiency – a reporting format proposed to Eurovent The main finding is that a percentage (like 50%) is not enough to characterise a part load behaviour of a chiller.60 27.50 78. Since temperature and load can be tested independently and recombined.80 17.12 3.

23.8 1 1. for higher than 50 part load ratios and on a single circuit.7) : it is a double circuit water screw chiller with one compressor by circuit. At 75%. the efficiency decreases somehow faster. Table 8. reducing the refrigerant flow rate at the minimum for part load ratio smaller than 50%. This complex part load behaviour can be summed up under one form per temperature – see Table 8.The overall performance improvement (or degradation) at part load (temperature effects being substracted) is given on figure 8. the symmetry of the chiller would impose identical performances.6 0. N°4 and N°1.5 under the form proposed to Eurovent. Reduced efficiency while decreasing part load ratio (same source temperatures) for the testedwater cooled chillers Reduced efficiency of the part load stages for water cooled chillers 1. one compressor by circuit is activated. with the same symmetrical tandem on each circuit. At 25%. one being the specific configuration of the plate heat exchanger : the two distinct refrigerant circuits use the same brazed heat exchanger in order to cut the costs. One can see easily the difference between the two type of unloading.8 N° 6 0. or one single table with all temperatures –table 8.4 0.2 0. one circuit is at full load and the other at half load.2 Part load ratio Let’s give more explanation about two of the tested chillers.9 N° 1 N° 3 0. only one circuit works at half load. Logically. In that case. Chiller N°4 is a two circuit four scroll compressor chiller. Now let’s consider chiller N°1 (Table 8.8.6 0.6 : Part load performance of water cooled scroll chiller N°4 N° 4 // WT : 30°C STAGES 1 2 3 4 Compressor 1 0 0 0 1 Circuit 1 Compressor 2 0 1 1 1 Compressor 3 0 0 1 1 Circuit 2 Compressor 4 1 1 1 1 EP/EPFL 23% 46% 71% 100% CC/CCFL 25% 52% 75% 100% EER/EERFL 109% 114% 105% 100% Let’s note here that the presentation under this format ensures that the manufacturer has consciously chosen this staging as optimum and hopefully that it has been prioritised as factory default control parameters. Stages configurations and performances are gathered Table 8.2 1.5 0 0. as examples of real life issues. The bias can come from many causes. at 25% and at 50%.7 0.8 for the four tested units. symmetrical.6. Figure 8.7 : Part load performance of water cooled scroll chiller N°1 N°1 // WT : 30°C STAGES 1 2 3 4 5 6 Circuit 1 Compressor 1 0% 0% 0% 59% 101% 100% 154 . Table 8.1 1 N° 4 0. The efficiency increase show that at 50%.

Reduced efficiency while decreasing part load ratio (same source temperatures) for the water cooled chillers Reduced efficiency of the part load stages for air cooled chillers 1. Table 8. as examples of real life issues.9 : it is a screw double circuit unit. Each compressor can unload partly continuously from about 100% to 75% and then two supplementary stages at 66% and 33% are available for each one.8: Part load performance of air cooled scroll chiller N°5 N°3 // OAT : 35°C STAGES 1 2 3 Circuit 1 Compressor 1 1 0 1 Compressor 2 0 1 1 Fan 1 1 1 EP/EPFL 38% 52% 100% CC/CCFL 46% 64% 100% EER/EERFL 120% 124% 100% Now let’s consider N°2 in table 8. N°5 and N°2. Only the four tested stages are reported.2 0.8 N° 2 0.4 0.8 1 1. There is one screw compressor by circuit.2 1.3 1. Figure 8.6 0.9. This part load behaviour just confirms that unloading on a single circuit with a slide valve is very inefficient as compared to the full load efficiency of the screw. Table8.6 0. Only the 25% can.2 Part load ratio Let’s give more explanation about two of the tested chillers.7 0. which means three capacity steps. The increased efficiency with reducing the refrigerant flow rate from stage 2 to stage 1 is counterbalanced by the relative increasing weight of the fan consumption. The percentage for each compressor corresponds to the ratio of the cooling capacity of the compressor to its full load capacity (half the chiller capacity).1 N° 5 1 N° 7 0. Circuit 2 Compressor 2 51% 58% 100% 56% 56% 100% EP/EPFL 33% 36% 51% 70% 87% 100% CC/CCFL 25% 28% 49% 58% 79% 100% EER/EERFL 74% 78% 96% 82% 91% 100% It clearly appears that 3 stages are used for compressor 2 (100.8 gives testing results on chiller N°5 : it is a single circuit unit with an asymmetrical scroll compressor tandem.9 for the five tested units. Air cooled chillers –experimental results The overall performance improvement (or degradation) at part load (temperature effects being substracted) is given on figure 8.9 N° 8 N° 9 0. 155 .5 0 0. The 50% and 75% capacity stages cannot be allocated to each circuit. Refrigerant circuits are separated. 56 and 50) and 2 only for compressor 1 (100 and about 60).

administration …). three climates have been simulated. it led to an incommensurable number of simulation. the weighting of electric power of each stage is found by the expression of the weighted average. CAC air and water distribution equipments have been simulated using the European average efficiency values. hostels. the cycling formula enables to calculate the electric power. Available hourly load curves Office building Climate System type London Milan Seville CAV CAV-FC VAV FCU 4P Sizing issues for chillers rating as shown by the simulation of the buildings The Milan CAV hourly load curve is presented Figure 8. Milan office building CAV system hourly load curve 156 . adopting different envelope characteristics when moving the building around Europe. Otherwise. The different climatic conditions and central air conditioning systems available are described Table 8. At each hour the outside enable to calculate all known stage capacities and respective electric powers. In fact this computation has been performed in two steps : first.10. Table 8. However. without mentioning the number of buildings to be entered in the used building code.4 Derivation of a new SEER method (ESEER) Given the complexity of the subject. Table 8. Then those conditions are used a number of times with different chillers’ and equipment quality. hospitals. but buildings that do exist : an office and a commercial mall.10. including the high pressure control impact on each stage. The tendency that links load and temperature is very clear even if an important scatter is observed.10. the simulation uses then the characteristics of the real chillers modelled to compute the exact yearly performance index : the HSEER (Hourly SEER). Figure 16. Some studies are available in Europe giving the description of the commercial building stock and a very few countries have also developed the buildings within simulation tools. the EECCAC group adopted the building simulation tool DOE2 to simulate representative buildings of the European stock market. when multiplying the simulation cases including building types (offices. The simulations leading to the reference values of SEER (HSEER) Two buildings were simulated on computer. malls. The different systems identified in the stock and market study have been simulated. If the load is lower than the smallest available capacity step. Then the load is compared to each stage capacity.9 : Part load performance of air cooled screw chiller N°2 N°2 // OAT : 35°C STAGES 1 2 3 4 Circuit 1 Compressor 1 30% ? ? 100% Fans 3 3 3 3 Circuit 2 Compressor 2 0% ? ? 100% Fans ? 3 3 3 EP/EPFL 28% 55% 64% 100% CC/CCFL 15% 50% 71% 100% EER/EERFL 52% 89% 111% 100% 8. used then as a reference for other methods. chiller hourly load curves were extracted as well as the coincident hourly OAT and the specific humidity. We had to make some decisions. Hour after hour. the climatic conditions and the different systems. For each one.

Given the part load characteristics of the chillers. For real world installations.11. for the screw chiller N°2. Hourly load curve 550 500 450 400 350 300 kWh CAV-MILAN 250 200 150 100 50 0 10 15 20 25 30 35 OAT (°C) When trying to calculate air cooled chiller performances from hourly load curve. The seasonal efficiencies for the MILAN CAV load curve are presented Table 8.12 30% 3. because the reduced part load efficiency is still higher than 1 at 25% and hardly lower than 1 for the 50% load reduced efficiency. an iteration process on hourly load and chiller capacity has been adapted to make sure that the cooling capacity is enough all the year long. On the contrary. so that all the chillers may be compared on the same load and temperature repartition. In the ARI standard. So the real optimal design rule is : the maximal chiller capacity and the corresponding temperature corresponds to the (load. other load pattern intervene as the solar loads. These results confirm that the 157 . it could happen that the sizing could lead to non-satisfied needs .81 60% 3. This is the first limitation of ARI sizing assumptions. that point is solved by assumption since all the straight line load curves have the same maximal temperature of 35°C that also corresponds to the maximum load.83 2. It clearly appears Figure 8.76 2. Of course.11 for 30% and 60% oversizing for the air cooled chillers N°7 and N°2 .10 that the ARI hypothesis is not verified. security coefficients are generally applied to the simplified sizing method leading to huge oversizing up to 100%. two problems appear that will show up two significant limitations of the ARI method. the thermal inertia and the dehumidification loads that are the sources of the non explained variance by the load and temperature correlation. The load curves must enable to calculate the consumptions for all chillers. 30°C) chiller with this simple law has been drawn Figure 8. Table 8. Impact of oversizing on seasonal performances for Milan CAV hourly load curve SEER values Oversizing N°7 N°2 0% 3. it seems obvious that consequences for the seasonal efficiency will also be very important. if maximal load were at lower temperature. The capacity variation with OAT of a perfectly sized (500 kW.81 3.10 and shows sizing is correct.60 The differences are limited for the air scroll chiller N°7. OAT) couple that enables the chiller to cover all the cooling needs. The explanation is that even if temperature is correlated to the load. the load must be divided by the sizing load. a sharp efficiency degradation with the load had been noted. The sizing realized for the three climates led to 30% constant oversizing for the 3 climates for the office building. Thus.

But this methodology can –and will.12. Figure 8.10 with % load instead of kWh on the Y axis.11. as already explained. Then.12 gives for instance the discrete load curve for the hourly simulation of the office building in Milan. all the points between the full load stage and the step immediately inferior are shared between the two steps. 50%. In each rectangle.be applied to all conditions obtained by simulation. 75% and 100% of the design load is operating. as assumed in the ARI standard. where k varies between 1 and K. the load curve is put under a grid format. the number of temperature bins and i varies between 1 and I. the outside air temperature is binned. the number of % load bins.11). the load ratio is binned. with 4 capacity steps at 25%. It shows that it is not far from a straight line. the relative kWh are added.05. The ARI methodology of reduction would lead almost to a null energy share for the full load stage. leading to the possibility of a four points representation of any condition or of the EU average of operating conditions. The Milan-CAV load curve is used to illustrate the methodology. which is not true for staged chillers but approaches the truth for continuous control chillers. On the Y axis. Milan-CAV 158 . from the hourly load curve. Figure 8. However there would be no weight for the 100% stage applying the ARI methodology from now on. it is supposed that a virtual chiller. the average load ratio is kept and called here after binL(i). giving weight to the full load. In the ARI methodology. Reduction of European hourly load curves to a set of four conditions (based on the example of Milano) To reduce the load curve to 4 points. giving a repartition of the cooling energy needs on the grid (Figure 8. Reduced load curve for the office building. On the X axis. Figure 8. only the 100% and more than 35°C OAT couples are kept. STEP 2 of the load curves reduction process A statistical reduction method enabled us to get. Whereas in our calculation methodology. And it also shows that no optimum sizing can be done without the exact knowledge of the chiller part load performances. This is the second limitation of ARI sizing assumptions. for reducing the load curve and temperature occurrences to 4 points.sizing is a key factor for seasonal performances analysis. the representation of a discrete load curve and a discrete weighting curves. Grid representation of the Milan-CAV load curve (load ratio bin length : 0. For each line. STEP 1 of the load curves reduction process The sizing described above enables to transform Figure 8. temperature bin length : 2°C) The average temperature is kept for each column and called hereafter binT(k).

Figure 8. for the Milan-CAV office building hourly load curve 159 . At this level.14.13 gives the representation of the energy weight of each class as a function of temperature classes.7). Milan-CAV STEP 3 of the load curves reduction process The ARI methodology load curve reduction would not give any weight for the full load point. 50. the chiller load and the associated weighting energy coefficients are still not null at temperatures as low as 15°C. It can be seen that even for Milan. We know this is not true and we have used a simple 4 stages chiller to determine the weight of each stage and then the corresponding operating temperatures. for each capacity step.13.Figure 8. 75 and 100) at each temperature bin. Each stage weight and OAT repartition is represented Figure 8. method 3rd Step. Reduced weighting curve for the office building. Figure 8. It is equivalent to the ton hour curve given in the ARI standard for Group 1 (Figure 8.14. we have obtained weighting coefficients that represent the energy weights to be associated to the % load required (25. Reduced weighting by stage under the PT format.

Table 8.So now.2 3% 75 28.4272 0. the weighting coefficients are known : SEERP= [0.0611 28.12. without high pressure condensation control • Without high pressure condensation control. D (nominal full load) 100 31.2024 0.3369 0. B.12.0514] The results for the reduction for the Milan office building load curve are presented Table 8.0 34% 50 23. with cycling • Without cycling.9997 23.1 43% 25 18.10) : • CAV-FC system in Seville • CAV system in London However. C. the seasonal performances are calculated successively for the 4 following configurations : • Without cycling.1 20% Results for more extreme weather conditions (London. Reduction of the chiller hourly load curve for the office building in Milan using the CAV system for air cooled chillers Energy weighting Part load (%) Reference Inlet air temperature (°C) coefficients A.13. with high pressure condensation control • With both phenomena The reduction results of the two specified load curves are presented Table 8. The two extreme load curves reduction presented show that both weighting coefficients and temperature conditions greatly vary with climatic conditions and systems.0627 31.0335] STEP 4 of the load curves reduction process The average temperature for each stage is calculated and the following operating temperatures are found : SEERT=[17. Seville. different distribution systems) The hourly calculation methodology has been applied to two different load curves among the twelve available (Table 8. 160 . in order to separate the quality of the reduction by itself from the non linear models representing the chillers.

8% 1.15 4.3% 3.2% 0.9% deviation HSEER 2.4% -1.18 2.8 11 % 17.05 3.6% 4.90 3.11 3.1 48 % 24.22 3. • For both conditions. B.82 2. C.0% 1.3% 4.86 3.88 4.14 Cycling only ESEER 3.40 2.1% deviation HSEER ranking 4 1 5 2 3 3 1 4 2 5 ESEER ranking 4 1 5 2 3 3 1 4 2 5 The following remarks comment the key points of Table 8.4% 0.20 4.4% 3. it shows the important impact of load on the seasonal efficiency.12 2.47 2.18 2. + 5.58 3.31 Relative 1.10 3.7% 2.5% deviation HSEER 2. Table 8.71 3.30 4. We translate the information in terms of ranking : chiller ranked 1 is better than chiller ranked 2.4% -0.07 Relative 2. the third for the Seville sequence and the last for the London sequence. both impacts are logically very low for the CAV-FC load curve in Seville.9% 3.2% 1. Accuracy of the reduction for the tested air cooled chillers Conditions Seville CAV-FC load curve London CAV load curve Chillers N° 5 N° 7 N° 8 N° 9 N° 2 N° 5 N° 7 N° 8 N° 9 N° 2 EER 2.39 2.6 1% 75 32.59 3.43 2.07 Relative 1.7% 3.19 No cycling.73 3.31 2.35 2.31 Relative 2.0% 5.91 2.70 4.4% 1.5% deviation HSEER 2. and so on.72 4.08 4.4 37 % 20. However.46 4.22 3. The nominal full load efficiency (at 35°C) is reported for each chiller and so is the hourly seasonal efficiency ratio (noted HSEER for hourly). + 4%].4% 2.98 3.7% 0.56 4. However.72 3. The relative maximal bias created for Seville by the methodology reduction is [+ 2%.08 4.Table 8.1%.34 2.69 3. • The relative maximal bias created for Seville by the methodology reduction is [+ 2%.73 3.4% 2.51 2.9 42 % 25 22.51 2.9% 4.1%].96 Fan cycling only ESEER 3.93 3.59 2.92 Cycling and fan cycling ESEER 2. D Inlet air temperature (°C) coefficients A.1% 2.15 3.0% 0.14 3.7 4% 27.26 3.14.80 4. B.93 2. conformingly to table 9 reduction results.01 3.73 3.1 47 % The results of the reduction methodology are presented Table 8.59 2. the ESEER follows the HSEER classification of seasonal efficiencies.21 3.00 3.31 2.93 3. C.6% 0.60 3.6% 1.93 EER ranking 5 2 3 4 1 5 2 3 4 1 HSEER 3.11 3. the reduced index figure (noted ESEER) and the relative efficiency difference between the two seasonal figures.4% -1.10 3.03 2.13 4.7% 2.05 3.99 3.90 3.11 3.0% 2.44 2.38 2.31 3.3% 2.14.4% 1. No policy measure should be based on it only as far as capacity staged or continuous capacity control chillers are concerned.5% 2.8 10 % 50 27.90 3.0% 2.8% 0. • Cycling and fan cycling impacts cannot be neglected for the London curve conditions as it is shown by the HSEER evolution.01 3.47 2.03 3.7% 5. it appears that the chiller classification is largely modified .98 3.14 : • The full load nominal efficiency appears to be a poor indicator of the seasonal efficiency. D (nominal full load) 100 36.2% 3.06 3.13. The chiller N°2 is the first for the EER sequence.30 4.11 3.9% 1. 161 .43 3.05 3. the ESEER values are always higher than the HSEER values by 2% to 5.0% 0. using the CAV system Seville CAV-FC London CAV Energy weighting Energy weighting Part load (%) Reference Inlet air temperature (°C) coefficients A.00 3. Reduction of the chiller hourly load curve for the office building in Seville using the CAV-FC system for air cooled chillers and in London.09 3. no fan cycling ESEER 3.14 3.66 3. • For the two different load curves.08 4.71 3.

free cooling and VAV specific applications (WC : weighting coefficient) Climate London Milan Seville CAV 5.3 19. Figures of the two first lines of Table 8.6% 30.16.3% 9. For the values for CAV+Free Cooling and VAV applications.1% 19.1 0.3% VAV 27.6 24.7% 20.9% 3.8% WC Free Cool.4% 0.3% 16.2 28.3% VAV 31.4 22.5% 24.7% 1.6 28. Cooling energy needs hourly load curves weighting coefficients for the air cooled chillers.1 24.4% 2.6% 40.7% 48.8% CAV-FC 31.8 3.1% 0.8% Thus it can be concluded that the methodology proposed is a qualified tool to classify the air cooled chillers at the condition to respect seasonal efficiency classes wide at least of 5% of the market average ESEER absolute figure.7% 9.Table 8.1% 38.1% CAV-FC 27.2 28.9 17.2 2.3 26.8 22.4 17.2% 39.4% 31.7 32.9% 23.6% 2.1% 7. 6.3% 49.5% 8.6% 51.5% FC4P 27.1 23. Extrapolating to the European stock of chillers in use Results of this weighting (described in chapter 5) are given Table 8.0% 0. Final weighting coefficients for the available load curves types are presented Table 8.0 24.4 28.5% 41.8% FC4P 31.8% 0.6% 31.6 22.7% 39.1 27.7% FC4P 36.15 below enables to check that the load curves selected for the sensitivity analysis represent indeed extreme conditions for temperature and load.0% 0.1 26.1% 34.8 4.1 18.2% 22.0 5.3% 0. only the air distribution systems are concerned .0% 3. For the Seville CAV-FC system.1% 32.3% 10.17 for CAV and FCU are to be understood as a complete set of coefficients for 6 load curves.6% 33. For the London CAV system.8% 35.6 25. 162 .2% Seville CAV 36.1% 18.7 3.2% CAV-FC 36.17.5% WC FCU 0.8% 1.4% 0.1% 0.1 17.6 20.7% 29. Table 8.2% 9.5% 37.1% Milan CAV 31.5 19.15.9% 0.4 4. the temperatures are the lower (except for the FC4P system) and weighting coefficients maximum at low loads.1 1.7% 42.1 0.2% 21.16 for air cooled chillers.6 24.6 26.7 32. Cooling energy needs national weighting coefficients for air cooled chillers for CAV and FCU (WC : weighting coefficient) Country Aus Bel Den Fin Fra Ger Gre Ire Ita Lux Neth Por Spa Swe UK CAV 0.1 3.7% 43.5% 47.7% 37.8% Southern Europe country visibly represent most of the cooling energy needs in Europe.9 24.1% CAV (air distribution) and FCU (water distribution) systems enough to represent the present stock. The 3 weighting coefficients issued from CAV market shares are here to generate scenarios of improved efficiency.5% 27.0% 41. The sum of the 6 weighting coefficients equals 1.4% 0.2% 4.2% 26.7% 54.3 22.9 32.8 20.0% 40.2% 5. for CAV and FCU.8% 0.1% 37.9% VAV 10. We can reduce this information to a set of 6 coefficients that will be used to weight the SEER obtained from London.0 21.3% 40. Applying the reduction methodology to the set of available load curves Temperatures (°C) Weighting coefficients Load 100% 75% 50% 25% 100% 75% 50% 25% Climate System London CAV 27.6% 0. Table 8.7% VAV 37.2% 29.6% 30.4% FCU 8.5 19.5% 38.1% 8.6 1.0% 0.0 23.2 33.4 28.1% 62.5% 42.3% 0.1 2. respective reverse conclusions on temperatures and load can be made. Milan and Seville load curves.17. Table 8.9% 41.4% 47.1 16.

Here.At last. real towers have been sized using the cooling towers electronic catalogue of the leading manufacturer. considerations on control of equipment and recent correlations. these coefficients enable to weight the results of Table 8. the lower stage temperature increases while weightings move towards higher loads. despite of the Legionella disease that certainly greatly modified the sales. For both options. for water cooled units.6 23% 21. since they represent 80% of the European stock.18. Only the open type towers have been considered here. the average coefficients only slightly moves. o the tower performance curves.1 45% 31.6 2% 75 30. On Figure 8.1 28% The differences for the VAV application for weighting coefficients can be neglected while each stage temperature increases by about 1°C. o the specific humidity.9 4% 34.7 41% 25. confirming that the system driving efficiency factor is respectively the load avoided for the free cooling option and the fan consumption avoided for the VAV option. For free cooling. the water temperature at the condenser inlet depends not only on the OAT but also on : o the condensing water flow rate.15 and to present final ESEER average European conditions in Table 8.0 3% 33. The values of the two first arrays are our proposal for EU chillers rating.18.2 30% 50 24. The load curve is the Milan CAV one.1 33% 30.0 15% 20. the part load ratios versus the condenser inlet water hourly temperatures are drawn. 163 . completed with the NUT-epsilon heat exchanger theory. However. Table 8. which is false. Transformation of the ambient conditions for the modelled default cooling tower for a specific screw chiller extracted from a manufacturer electronic catalogue (the Milan CAV hourly load curve is used). The effects of the chosen control scheme clearly appear : the inlet condenser temperature does not fall under 15°C while the cooling tower control uses the 21°C set point.14.5 35% 26. Application of the method to derive a ESEER for air cooled chillers used with the free cooling or VAV options ESEER Free-cooling VAV Weighting Weighting Weighting Part load ratio Temperatures Temperatures Temperatures coefficients coefficients coefficients 100 34.14.3 41% 25 18. In a similar way the values have been defined and validated for water cooled chillers. Figure 8. it was supposed that the approach (in that case defined as the temperature difference between the inlet water temperature at the chiller and the air wet bulb temperature) was constant for all conditions of operation. for a specific screw chiller extracted from the manufacturer catalogue. Within the ARI standard. The other results show how free cooling and VAV application could change those recommended values.

97% 22. A comparison of the conditions of the 3 available indexes is proposed Table8.5 4. included single compressor units.6 18. It has to be recalled that the cooling towers have been sized for maximum wet bulb temperature with 1% of yearly occurrence.4 2% 28. largely under the experimental uncertainty.8 0.4 18.96% 35.5 Is there a method good enough for classification of products by order of merit? We are comparing now the numerical results and the way each of the existing methods would sort chillers by order of merit.0 26.3 24.15% 45. However.8 34% 26.26% 42.0 26.4 2.9 18.20.75% 29. In order to simplify the application of the index.3 0.87% 32.2 22. EECCAC final figures -Simplification of the figures and uncertainty estimate Our work clearly shows also that the methodology for air and water cooled chillers enabled to extract seasonal operating temperature conditions with errors on the seasonal efficiencies that are inferior to the experimental uncertainties.84% Seville CAV 29.7 2% 75 25.6 29% 8.8 21. However.8 26.9 21.20% 38.6 2% 29.28% 42. some rounding can be done without modifying noticeably the ESEER figures obtained.3 24.38% 46.6 2.5 23.66% 21.21 for air cooled chillers. because a constant share of water cooled and air cooled systems has been used.54% VAV 28.07% 9.89% FC4P 28.2 22.5 2.19.6 12% 18.7 26.9 26. the final weighting for the ESEER for water cooled chillers slightly differs from the air cooled chillers coefficients.4 26.54% 30.2 22.1 22.13% CAV-FC 27.34% 44.2 26.29% 22.2 44% 26.3 0.05% FC4P 27.2 30% 50 22.38% 10.28% Milan CAV 28.67% 25.9 41% 22.69% 38.It clearly appears Table 8.0 1. 164 .5 23. it also shows that the experimental uncertainty is quite high.5 0.5 18.8 19.5 25.90% 40.3 20.3 2.53% 47.02% The same market shares can be used for water cooled chillers than for air cooled chillers.5 21. Table 8.42% 27.20.53% 39. Applying the reduction methodology to the set of available load curves Condenser inlet water temperatures (°C) Weighting coefficients Load 100% 75% 50% 25% 100% 75% 50% 25% Climate System London CAV 28.03% 30.06% 9.23% 63.8 2.3 21. Table 8.6 3.30% 36. for all chillers.5 18.96% VAV 26.3 40% 22.2 19.85% CAV-FC 30.1 22.90% 32.6 18. Application of the method to derive a ESEER for air cooled chillers used with the free cooling or VAV options ESEER (Water) Free-cooling VAV Condenser inlet Weighting Condenser inlet Weighting Condenser inlet Weighting Part load ratio temperatures (°C) coefficients temperatures (°C) coefficients temperatures (°C) coefficients 100 28. Table 8.8 18.5 24% 19.2 18.89% FC4P 28. since Table 8.19 exhibits different load weighting coefficients.16% 19.39% 43. The main differences amongst climates and systems are found for weighting coefficients differences. the same for all countries. It mainly comes from the uncertainty measurement on the temperature difference at the evaporator.96% 41.23% 22.8 39% 25 18.35% 42.2 22.8 25.7 18.32% 10.45% VAV 28.85% CAV-FC 29.9 26.04% 43.31% 38.65% 14. The lower temperature results in Seville as compared to Milan show that on average wet bulb temperatures are higher in Milan.19 that the temperature dependency is not linked first to climatic conditions but to the sizing and to the cooling tower choice.51% 34.

22. Table 8. Table 8. We take the point of view of a user of the Eurovent certification system : by selecting a “better” chiller.9 30% 50 22 41% 18. Comparison of the ESEER conditions with the EMPE and IPLV indexes for water cooled chillers ESEER ARI EMPE Weighting Weighting Weighting Part load ratio Temperatures (°C) Temperatures Temperatures coefficients coefficients coefficients 100 30 3% 29. am I really selecting a better chiller? EER is a poor selection tool Figure8. US-IPLV.9 20% Temperatures of the ESEER are embedded by the EMPE ones above and ARI temperature beneath except for the 25% point.8 12 % 23.15.3 45 % 27. ESEER. using as a reference the actual EU values obtained by simulation in the three locations and properly weighted. The ESEER weighting coefficients give more weight to the 25% point load than both index. For 50 and 75%.3 12% 21. For 50 and 75%. Classification : who is right? We shall compare now four classifications : according to EER. Comparison of HSEER with EER on the tested chillers 165 .22 for water cooled chillers.5 40% 25 18 23% 18.4 10% 75 26 33% 23. coefficient are nearer to the EMPE index. The ESEER 100% weighting coefficient is nearer from the IPLV one.3 45% 23.21.5 40 % 25 19 23% 12. A comparison of the conditions of the 3 available indexes is proposed Table 8.4 1% 29.7 42 % 31. EMPE. nearer from the IPLV one.8 20 % Temperatures of the ESEER are embedded by EMPE temperatures above and ARI temperature beneath.9 42% 26. coefficients are nearer to the EMPE index.3 30 % 50 25 41% 18. ESEER weighting coefficients give more weight to the 25% point load than both index. The 100% coefficient is 3%. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers ESEER ARI EMPE Weighting Weighting Weighting Part load ratio Temperatures Temperatures Temperatures coefficients coefficients coefficients 100 35 3% 35 1% 35 10 % 75 30 33% 26.

5 1 1.16.5 0 0 0.17. HSEER versus EER 4 3.5 HSEER 2 1.5 1 0.5 IPLV Figure 8.5 1 0.5 4 4.5 3 2. comparison of EMPE with HSEER for the tested chillers 166 .5 3 3.5 3 3. comparison of US-IPLV with HSEER for the tested chillers HSEER versus IPLV 4 3.5 0 0 0.5 2 2.5 EER IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for comparison of chillers Figure 8.5 1 1.5 3 2.5 2 2.5 HSEER 2 1.

1998]. and not always in the same direction. 167 . DMT and manufacturer laboratories visited. HSEER versus EMPE 4 3.5 1 0.5 3 3.5 HSEER 2 1.5 3 2. that contains some remarks about testing.5 4 ESEER Conclusion : the differences are relatively large between existing methods and reality.5 1 1. Since there is no European specific standard to perform part load testing. First way to realise the testing needed for the ESEER proposed certification method At this point there is still a choice to be made between an experimental approach based on ARI-IPLV (knowing that it will be completely changed in a few years due to the arrival of an ISO standard) or based on the draft CEN standard close to publication and more consistent with the upcoming ISO standard.5 3 2.5 3 3. The proposed ESEER method allows grading and ranking of chillers by order of merit Figure 8.5 0 0 0. • the experience gained during the “Joint project” at EDF R&D facility. 1998].5 0 0 0.18 . the analysis is based on : • the full load testing definition [CEN.5 2 2.5 1 1. comparison of ESEER with HSEER for the tested chillers HSEER versus ESEER 4 3.5 HSEER 2 1.5 2 2. IPLV and EMPE have the same advantages and disadvantages. the two methods.5 4 EMPE Based on similar assumptions.5 1 0. The newly proposed ESEER method is more accurate in a noticeable manner and satisfies the needs of Eurovent certification process as well as the expectations of the DGTREN in a market transformation effort. • the IPLV standard [ARI.

if we follow the present ARI approach the manufacturers have to give to Eurovent and to the testing laboratory the expected results close to the ESEER conditions.3 25 CC 176. three points are taken “randomly” and checked. the electric powers and the efficiencies of each one of the point that will be tested and the inlet fluid temperature at the condenser according to the ESEER temperature load “curve”. 2ND TESTING POINT Then. the manufacturers must give to Eurovent the cooling capacities.7 3.8 P3a 22.0 93.4 57% 88.9 P2a 39.0 2. Point 2. and efficiencies.3 Fan cycling 3. the table allows to answer directly to the question : what happens to performance when the load (resp. Decreasing capacity Toe = 7 (°C) Stage 4 Stage 3 Stage 2 Stage 1 Decreasing % of FL % of FL % of FL % of FL OAT (°C) at 35°C at 35C at 35°C at 35°C 35 CC 100% 153.T1) noted after P1.5 Fan cycling 4.9 14.0 61% 93. it is enough to allow the use of the full table and the calculation of the ESEER.2 EER 2. more time is needed to guess out the proper conditions for testing in such an ARI approach.9 EP 43. (75%. In a certification approach.1 41.8 32.6 3. A % tolerance on the cooling capacity is calculated and consequently. that will be explained just after.0 P3’ (P3b) 11. Table 28 gives an example. the time needed to perform all tests.4 EP P1 60. part load ratio (% of full load.0 137.8 81.6 3.7 EER 3.2 12.1 3. For continuous control chillers : the chiller can supply the adapted capacity within the allowed tolerance. stage capacities. The testing must fulfil the following associated constraints : • it must enable to start the part load certification next year. the tolerance on the inlet water temperature at the evaporator is known. for chiller number 7.2 Fan cycling 4. First.5 38.7 116. (100%.0 1ST TESTING POINT The testing begins with the full load nominal temperature. For chiller N°7.6 19 CC 185.9 EER 3.4 28% 42.3 82% 126.6 2.2 3. the real cost factor.9 35. the inlet evaporator water temperature is 12°C and the outlet is 7°C.9 EP 48. • it must minimize the number of testing points. both the outside condensing fluid and the inlet evaporator temperatures are decreased to the 75% load point. so as to minimise interpolation and iterations.0 43. We shall present afterwards another way of doing. the computation of ESEER will be only a weighting/interpolation of testing results.9 EER Fan cycling 4.4 P2’ / P2b 24. In case we want a “blind” checking.75°C.23. • it must respect the chiller ESEER sequence.0 29% 43.2 Fan cycling 4.3 134. electric power.7 4. 168 .T2) noted after P2. which seems to us more in the “spirit” of Eurovent certification : once the manufacturer gives the full table of performance.8 20. OAT = 35°C).7 P4 10. Under this format. the temperature) decreases. • it must not require the manufacturer presence for the testing. it is to say.7 30 CC 166.6 EP 53. It means the inlet water temperature at the evaporator is decreased until 10.8 26. Table 8.

For capacity staged chillers, we guess in general, the chiller cannot supply the required capacity within the
allowed tolerance ; thus, either the step is above, either the step is beneath. If a step has been triggered above,
it is not counted, and only the step just beneath will be. To reach the step, the inlet evaporator water
temperature is decreased until a capacity stage under 10.75°C has been reached. Correlated inlet at the
condenser according to the ESEER load versus temperature curve is then imposed, for instance, (70%, T2’)
P2’.
The procedure enables not to be forced to check the performances of the step above 75%, thus economizing
testing time, but at the cost of a further interpolation procedure.

Then, for both chillers, with continuous or discontinuous capacities, the following acquisition is made
respectively at P2 and P2’ :
[ARI, 1998]
“C3.1.2 To confirm that steady-state conditions have been established at the specific set of conditions and
within the tolerances set forth in C6.2.1, three sets of data shall be taken, at a minimum of five-minute
intervals. To minimize the effects of transient conditions, test readings should be taken as nearly
simultaneously as possible.”

The procedure is then repeated for the 3rd and 4th testing points.

For the chiller N°7, the points needed to calculate the ESEER are noted P1, P2’, P3’ and P4. From these
points, the interpolation scheme and cycling correction are then applied.

The testing time and precision are gathered Table 8.24. Evaluation of the time needed for the complete
testing procedure is given with and without the interpolation procedure. In this latter case, two points are
needed for the 75% and 50% points. Depending whether two stages embed the 25% load point or not, it will
respectively lead to 7 and 6 testing points. The higher option is kept. The points needed at each % load are
noted P2a, P2b. It would lead in that case to 6 points.

Table 8.24. Evaluation of the first ESEER testing methodology (the +1 testing point corresponds to the nowadays
non nominal testing point defined in the Eurovent testing procedure)
4 points with interpolation 4 points without interpolation
Testing Time Precision Testing Time Precision
CTS (-) (-)
DS (-) (-)
ST [CEN, 1998] 1 hour (+) 1 hour (+)
2 hours 2 hours
ST, PID
(P2’,P3’,P4’) (P2ab,P3ab,P4ab)

1 hour 1 hour
FC, PID
(P3’,P4’) (P3ab,P4ab)
IP WITH : not satisfying WITH : satisfying
Testing points 4 (+1) 7 (3 at 1 hour) (+1)
Set up 1 hour 1 hour
Disassembling 1 hour 1 hour
TOTAL 2 days 3 days

Consequently, the interpolation scheme leads to a non precise enough index for classifying the chillers. But
the results can be obtained in 2 days only.
When applying the method without interpolation, the ESEER sequence is exact but the testing time increases
to 3 days.

For both methods, the PID, sensor temperatures and fan cycling problems are likely to create insolvable
testing problems.
For fan cycling there is no guarantee that the behaviour of the chiller be the same for the tested unit and for a
sold unit, since the parameter could be modified to get favourable behaviour for testing.
The same rationale applies to chillers that have several possible staging to output the same capacity.

169

These two formats of the same scenario are not satisfying ; as a consequence, the second scenario is
proposed.

Second way to realise the testing needed for the ESEER proposed certification method

The manufacturer must give the Table 8.25 to Eurovent for the chiller that will be tested. All the stages
programmed in the soft of the chiller must be supplied. For continuous control chillers, 6 capacity steps at
least must be supplied.
Table 32 is the Eurovent tested chiller N°7 : a 2 circuits, 4 scroll compressors (1 tandem by circuit), 3 fans on
each circuit.
• C1 and C2 are the 2 distinct circuit.
• The percentages refer to the full load point at 35°C OAT for the cooling capacity, the electric power
and the EER.
• The percentages for fans and compressors refers to the circuit full load and not to the chiller full
load. Here they are electric power ratios.
For chiller specific configurations, supplementary information should be gathered for testing :
• For fans, their position should be given to the experimenter if they supply the air for different part of
the air coil and that consequently stopping 1 fan is not equivalent to stopping another one. For
variable speed chillers, the manufacturer should also explain to the experimenter how to reach the
published points.
• For reciprocating chillers, supplementary information should be supplied to the experimenter to
enable to make the difference between compressor unloading or compressor ON-OFF.
• For screw chillers with slide valve, the manufacturer should explain to the experimenter how to
activate the slide valve (access and postion).

Table 8.25. Scenario 2, part load and reduced temperature performance table
Toe =
7 (°C)
Stage 1 Stage 2 Stage 3 Stage 4
% (full % (full
OAT load load
% (full load % (full
(°C) 35°C) load 35°C)
35°C) 35°C)
19 C1 C2 CC 29% 43.9 C1 C2 CC 61% 93.0 C1 C2 CC 89% 137.0 C1 C2 CC 120% 185.0
Fan 66% 0% EP 18% 10.9 66% 66% EP 35% 20.7 66% 66% EP 55% 32.8 66% 66% EP 73% 43.8
Comp 50% 0% EER 157% 4.0 50% 25% EER 175% 4.5 50% 100% EER 163% 4.2 100% 100% EER 165% 4.2
25 C1 C2 CC 28% 42.9 C1 C2 CC 61% 93.4 C1 C2 CC 87% 134.0 C1 C2 CC 115% 176.3
Fan 66% 0% EP 20% 11.9 100% 100% EP 37% 22.0 100% 100% EP 60% 35.8 100% 100% EP 82% 48.9
Comp 50% 0% EER 141% 3.6 50% 25% EER 166% 4.3 50% 100% EER 146% 3.7 100% 100% EER 141% 3.6
30 C1 C2 CC 27% 41.6 C1 C2 CC 57% 88.1 C1 C2 CC 82% 126.4 C1 C2 CC 108% 166.3
Fan 100% 0% EP 21% 12.7 100% 100% EP 40% 24.2 100% 100% EP 66% 39.4 100% 100% EP 90% 53.9
Comp 50% 0% EER 128% 3.3 50% 25% EER 142% 3.6 50% 100% EER 125% 3.2 100% 100% EER 120% 3.1
35 C1 C2 CC 25% 38.4 C1 C2 CC 53% 81.5 C1 C2 CC 76% 116.8 C1 C2 CC 100% 153.7
Fan 50% 0% EP 24% 14.2 100% 100% EP 45% 26.9 100% 100% EP 73% 43.8 100% 100% EP 100% 60.0
Comp 25% 0% EER 106% 2.7 50% 25% EER 118% 3.0 50% 100% EER 104% 2.7 100% 100% EER 100% 2.6

Three points are tested on the whole map. Indeed, since the complete performance map is known, it is not
needed any longer to test the ESEER specific points to be able to calculate the index value. As a
consequence, 3 points (+1 for the non nominal temperatures full load point) only can be chosen randomly by
the experimenter.

Then the ESEER can be calculated following the scheme:
• 25%, 50% and 75% load point efficiencies are calculated according to equation (7a) and (8a),
• If needed, cycling correction is done according to Equation (2) with Ccyc= 0.9.

170

Using this scenario, most of the testing problems are solved .It still remains to manufacturers that build
chillers with variable speed fan control to make sure a precise fan speed reduction can be set manually.
As for the precedent scenario, it is not guaranteed the sold chillers will have exactly the same characteristics.
But this problem is part of the Eurovent certification scheme.

Table 8.26. Evaluation of the second ESEER testing methodology
4 points with interpolation
Testing Time Precision
CTS NO
DS NO
ST [CEN, 1998] 1 hour (+)
ST, PID NO
FC, PID NO
IP (+)
Testing points 3 (+1)
Set up 1 hour
Disassembling 1 hour
TOTAL 1 day

Final choice of the ESEER testing methodology

This testing procedure highly reduces the testing time as compared to the scenario 1. Moreover, it will enable
to easily set up other seasonal performance indexes that the average ESEER.

If the performance map (Table 8.25) is published, it will be a huge and needed progress :
• it will allow the buyer to compare the chillers on specific site conditions ; at the moment only the
EER information is available, and we have seen it contained in fact little average efficiency
information,
• it will also enable the buyer to optimise the chiller size as a function of the specified site load curve.

At the knowledge of the authors, 4 manufacturers on the European market have already achieved similar to
Table 8.25 performance maps for each chiller. Only the fan and compressor information have to be added to
enable the testing.

Waiting that performance maps be ready for all manufacturers, the certification procedure can begin with
only the information that enable to characterize the ESEER points, it is to say the highlighted testing points
Table 8.25.

Perspective of the proposed ESEER

The tool is not gifted of any prediction power of the yearly consumption for any real installation. It is just an
indicator of the seasonal performance, whose only aim is to classify the chillers a fairer way the simple EER
does.

It has been shown that the reduction methodology enabled to successfully extract four weighted temperature
conditions : the bias introduced was lower than the experimental uncertainty. As a consequence, it could be
applied to any other stock of load curves to build other indexes for other domains. This method also enables
to increase the number of points ; however, for some of the load curves considered, more than 5 points would
give a useless 100% load point.

Given the differences within the weighting coefficients for the 3 selected climates, different values should be
published at least as a function of the country. To do so, a simple method based on cooling degree days or
more simulations could be developed, if manufacturers require it, to adapt the coefficients and temperatures
by country. Similar spreading could also be done by type of building. Moreover, the present study enabled to

171

the integration of dynamics into the part load testing index should also be considered. As given. • Then. also a developing market in France and Germany. The same philosophy could also be applied to ground water condensing chillers and heat pump. the more developing end-use in Europe nowadays. The default cycling law could then be revised for all chillers.show that the free cooling and VAV options should be differentiated from typical CAV installations. It has been shown however that the methodology could be applied to these units for the load curves treated. supplementary work has to be performed to reach a such far away goal from actual chillers characterization. a method must be adapted to determine the cycling degradation versus the load. the seasonal performance index is thought to be a huge and necessary progress to strengthen energy efficiency of chillers. Certifying seasonal performances for chillers means indicating an average efficiency generally higher than nominal efficiency as has been largely figured. as nominal full load EER is. Future versions should consider the extension to single stage units that generally operate at different conditions. 172 . Similar work has to be performed in the heating mode since air to water reversible chillers is an increasing end use in Europe. The reduction methodology for a dedicated load curve associated to the presentation of individual testing results or more generally of part load performances and reduced temperature efficiencies for chillers would be a highly efficient simple selecting tool for buyers when following the choice method steps hereafter : • Hourly simulation of the project gives the building or chiller hourly load curve for the specific project. mainly because of the differences in the weighting coefficients. Spain and Greece represent 85% of the installed chiller based systems according to the EECCAC stock and market study). The load weighting coefficients are the main seasonal efficiency drivers. the reduction methodology enables to characterize 4 capacity step points and operating temperatures. But it also means to avoid the efficiency competition may be based on non-representative indicators. Temperatures can be shifted easily of 1°C if needed (as practised to round the operating temperatures). In order to approach field reality. This index could be used for single circuit units. Thus. This work could also serve as a basis for developing a seasonal index for room air conditioners. Nevertheless. However. The applicability to each country and building type should also be studied in order to give a full range of testing conditions and weighting coefficients nearer from specific and climatic applications. the coefficients are nearer from Southern Europe operating temperatures and weightings (Italy. • The presentation of part load results by the manufacturers enables to select the chiller on 4 efficiency points.

Chillers AC : the average being 2.05 on full market average and the factor is 98.46.23 on full market average and the factor is 96.50 so the factor is around 96% Scenario 2 THE BEST CHOICE AMONG EXISTING COOL GENERATORS BASED ON FULL LOAD INFO On average packages and chillers reach in 2005 the EER level corresponding to the minimum LCC (BAT with present information) but part load is not taken into account in Eurovent grading and so the corresponding improvement is not obtained. that will be provisionally taken as the same factor RAC either the average moves from 2. that will be provisionally taken as the same factor.4% There will be a consequence for reversible winter heating.1% There will be a consequence for reversible winter heating.22 by a voluntary agreement or the classes B to G are banned.1-Scenarios Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE All packages RAC and chillers presently under average reach by 2005 the EER level corresponding to the average of present market but part load is not taken into account in Eurovent grading and so the corresponding improvement is not obtained.0% to be applied to compressor consumption.10 on an EER around 2.0% to be applied to compressor consumption.Packages and large splits either the average moves from 2.2% to be applied to compressor consumption. Chillers WC : the average being 3.00 (present market) to 3.9.50 (present market) to 3. Packages and large splits the average being 2.78 on full market average so the factor is around 76. the classes E F and G should be banned. the gain is 0. Scenario 3 BAT.3%.80 or the classes D E F and G are banned. RAC the gain on SEER is the same that the gain on EER – we use the same benefit as the one gained with scenario 2 (inverters excluded) 173 . 9. the weighted gain is 0. the factor is 89.50. The scenarios corresponding to some policies and measures are defined by impact time and can be translated at that time in new specific consumptions for the market after that time. the weighted gain is 0.4% Packages and large splits the gain on SEER is the same that the gain on EER –we use the same benefit as the one gained with scenario 2. Chillers either the average moves from 2. peak power demand and CO2 emissions will be produced and the results compared with the base case for period of 1990 to 2020. nobody wanting to buy a « poor » image equipment.46 (present market) to 3.THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO All packages and chillers reach in 2005 the SEER level with the minimum LCC (BAT with upcoming information given by part load testing).64 and the gain is 0. We can also expect that a certain number of years of labelling and communication by energy agencies reaches the same point.7 on full market average so the factor is around 78. the classes E F and G should be banned.24 on full market average (+18% of which 12% may be obtained as well with scenario2) .175 on full market average and the factor is 93.85. The policy measure associated is banning some classes of equipment either directly (Directive ) or by voluntary agreement. The policy measure associated is banning many classes of equipment or a negociated agreement on average part load performance like ACEA agreement for cars Chillers the average SEER moves from 3.20 by a voluntary agreement or the classes B to G are banned. the classes E F and G should be banned. RAC classes E F and G should be banned and the average gain corresponding is 0. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY CAC SCENARIOS In this section. the weighted gain is 0. the factor is 82. the gain is 0.50 (present market) to 2. The policy measure associated is banning many classes of equipment or a negotiated agreement on average full load performance like ACEA agreement for cars. Part load is taken into account in Eurovent grading and so the corresponding improvement is obtained. efficiency policy and technology scenarios which project energy consumption.

we apply then -12% to each item of AC (fans. at comfort level TC) There is a reduction in cooling demand which is climate dependant but has been expressed here in relative terms on table 8.2. national values are different and have been derived from UK with corrections for DD.2 Results of scenarios General Evolution The scenarios are ranked in the following order : 1. The less expensive way of attaining the objective is the improvement of chillers.adapted for each EU climate Introduction of a MEPS on total electricity used for Heating ventilating and AC in kWh/ m2. primary or secondary.5.3.5 Consumption of each system. To obtain this “free” market transformation a prescriptive minimum should be applied to local manufacturers and importers at the same time. pumps etc). The overall reduction being a 12% reduction. relative to present. cooling and air movement. after introduction of scenario 4 FACTOR to be applied depending on Compressor demand system reduction water distribution (50/50) 95% air distribution 90% VRF 100% PACK&Splarge (50/50) 95% Rtops 90% RACs on one loop 95% MS 95% Splits 95% PACKsmall 95% Single Ducts 95% Scenario 5 VAV Obligation of variable air flow in air distribution systems There is a reduction in cooling demand which is climate dependant but has been taken here as 30% of auxiliaries consumption in Air based systems Scenario 6 British regulation on AC – heating.Scenario 4 FREE COOLING Obligation of introducing free cooling on air side of air based distribution systems at a certain value of air flow (Portuguese regulation and Ashrae) even for primary air (which is the case of our simulations. Starting from their present averages of EER and SEER.4.5 Table 8. The impact has been calculated with the assumption of a weighted mix of both situations : new buildings and new installations.6. The policy instrument would be a clear and harmonised implementation of EPB directive. this policy induces almost no extra cost for any stakeholder. The maximum flexibilty in demand is around 20TWh and can be obtained by introducing policy measures related with part load efficiency of chillers. which may be on either side. However a strong commitment on full load achieves about the same gains. 9. 174 . and absolutely no cost provided it’s applied to all manufacturers (and so that they all pass on the costs to the customer). to know the cost we have to evaluate the less costly options.

2 Figure 9. We can also expect that a certain number of years of labelling and communication by energy agencies reaches the same point.00 20.2 Savings in 2020 (CAHORE stands for Cafes.00 100.1 Demand evolution.3 shows that this potential is more on RAC and packages and so that the savings induced benefit more to households and trade than with the previous one.3 savings in 2020 in scenario 2 2020 175 .00 Scenario 6 Scenario 3 40.Figure 9.00 Scenario 1 TWh for AC Scenario 4 Scenario 5 Scenario 2 60.00 0.00 2000 2005 2010 2015 2020 Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE The policy measure is banning some classes of equipment either directly (Directive ) or by voluntary agreement. Restaurants) Scenario 1 4000 3500 3000 2500 GWh saved in AC 2000 1500 1000 500 0 CAHORE Education Hospitals Households Offices Trade Total Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON FULL LOAD INFO The policy measure associated is banning many classes of equipment or a negotiated agreement on average full load performance like ACEA agreement for cars. depending on scenarios 140.00 BAU 80. The effects are strong on trade and offices as shown on figure 9. Fig 9. Hotels. nobody wanting to buy a « poor » image equipment. Figure 9.00 120.

75 5.93 9.2 show a real potential for this improvement.3 176 .15 Cooling Scenario 5 Variable Air Flow Obligation of variable air flow in air distribution systems translated in national values for savings in table 9.68 6. The policy measure associated is banning many classes of equipment or a negotiated agreement on average part load performance like ACEA agreement for cars.46 24.86 10.1 Savings (TWh) 2005 2010 2015 2020 Chiller based 4.53 19.32 6.70 Total savings 14. We see on table 9. Table 9. The results of table 9.2 Savings (TWh) 2005 2010 2015 2020 Savings due to Free 2. Table 9.53 14. Scenario 2 in 2020 16000 14000 12000 10000 GWh 8000 Série1 6000 4000 2000 0 CAHORE Education Hospitals Households Offices Trade Total Scenario 3 BAT.00 3.77 15.64 26.21 12.1 a large influence on chiller based systems due to part load and also on the other systems even if they operate at full load.THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO Part load is taken into account in Eurovent grading and so the corresponding improvement is obtained.15 in SC3 Scenario 4 FREE COOLING There is a reduction in cooling demand which is climate dependant and has been expressed here by country and system.45 systems in SC3 Others in SC3 10.

9 148.2 Germany 225. cooling and air movement.3 134.4 Portugal 201.7 207.45 VAV Scenario 6 British regulation on AC – heating.3 Finland 253.6 164.45 scenario 6 Such a regulation provides a gain not so high as a very strong prescription.4 180.4 148.74 9.4 148.1 Netherlands 196.5 202.1 161.4) for a consistent effort of all countries in the same way as the UK is doing.3 Savings (TWh) 2005 2010 2015 2020 Savings due to 3.7 Sweden 253.5 181.1 Denmark 224.6 164. The gains are represented by the figures in table xx Table 9.4 Belgium 200.62 15.0 183.4 180.6 164.7 207.2 France 198.2 UK 182.2 163. 177 .Table 9.9 176.4 Greece 201.2 163. but is safer because the designer can put the effort on any segment of the project.4 Spain 165.8 164.5 181.6 202.28 13.5 Savings (TWh) 2005 2010 2015 2020 Savings due to 4.07 16.89 7.4 Luxembourg 200.85 12.0 183.5 181.1 Table 9.9 183.6 202.0 228.8 164.6 178.adapted for each EU climate The gains are expressed nationally (table 9.0 228.0 159.4 kWh/m² Present average New buildings New installations Austria 225.1 Italy 201.3 Ireland 182.

The measures which can be considered to encourage a higher energy efficiency levels for new CAC equipment are: • Provision of information (labelling. Proposals for each of these are made in section 10. efficiency ratings) • Removing less efficient models from the market (MEPS or voluntary agreements) • Encouraging higher sales-weighed average efficiency levels through negotiated agreements (e. Manufacturers generally test the efficiency of their equipment and report the results in their product literature and catalogues. and minimum energy performance requirements (either mandatory or voluntary in nature). energy labelling. The most advanced national building thermal regulations addressing the energy consumption of central air conditioning systems are in the UK and Portugal. As CAC systems are diverse in nature. installation and operation of specific CAC systems. 10. there are many more areas for action than are specified within it. the application of energy labelling for central air conditioning equipment would either require an amendment of the existing energy labelling Directive to include energy-using equipment destined for usage sectors other than just households. Many Member States are in the process of revising their building regulations to take account of the requirements of the Energy Performance in Buildings Directive and thus this offers an ideal opportunity for them to co-operate at least regarding measures applying to the energy performance of building cooling systems.1-1999 standard. In the case of minimum energy performance requirements. These type of measures include certification. fleet-average efficiency targets) • Financial and/or fiscal incentives for higher efficiency equipment • Public procurement and general market transformation programmes Measures to provide information to end-users and equipment procurers At the current time there are no mandatory requirements to provide end-users with information on the energy efficiency of central air conditioning equipment. these could be developed within the mandate of the proposed Directive on Ecodesign of End-Use Equipment. Without independent verification 178 . they only address the efficiency of the end-use equipment used in CAC systems as determined under standard test conditions and will not realise many of the potential energy savings which are related to the design. grading.1 Some fundamental considerations regarding policy measures The results of the energy scenario analyses of the preceding chapter have illustrated that there are significant potential energy savings to be attained by the optimisation of CAC systems.3. yet even these are not as mature or as encompassing as the US ASHRAE 90. viable policy measures will need to take account of the diverse circumstances which apply to them in order to realise the potential energy savings. or it would require the issue of a new primary Directive giving authority to the Commission to develop energy labels for commercial and tertiary equipment. 4 and 5. like other tradable goods are subject to the terms of the European single market and therefore it is appropriate for the European Commission to develop policy measures which will raise the average efficiency of new equipment sold within the European Union. are often designed on-site rather than simply being factory made packaged-systems and are installed and operated in diverse circumstances. Important as these measures are. CAC equipment. Specific proposals regarding this are made in sections 10.10.2.2 Policies and measures to encourage the selection of more efficient equipment The analysis presented in this study has shown that there is a significant variation in energy efficiency for all types of CAC equipment that have been investigated when tested under standard test conditions. however.g. POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC ENERGY PERFORMANCE 10. Policy measures which can realise these savings at the design and installation stage are typically provided through building thermal regulations. however. The new Energy Performance in Buildings Directive places some obligations on Member States to develop policy measures which will address some part only of these savings.

For the time being there is no such scheme in place and it may be some time before one is formally developed. An effective way of indicating the relative energy efficiency of products in a simplified manner is to use a categorical energy label. CAC equipment are generally large and are not generally ‘on display’ in a shop window at the moment of purchase. It is therefore strongly recommended that mechanisms be put in place which will allow such information to be passed through the equipment procurement and usage chain. By so doing it would allow manufacturers to express the relative energy performance of their products and. which could be applied by them on a voluntary basis (e. a key question is how that information should be presented to the public? In the case of products having split-incentives such as CAC components do. as has been seen for household appliances. i. In the spirit of aiding the rapid adoption of an efficiency grading scheme for central air conditioning equipment. ensuring that the purchaser and subsequent end-user be able to see the relative efficiency of the equipment is likely to be an effective market transformation tool. to apply a label which rates the energy performance of the equipment into one of a limited set of efficiency ranges which are graded using a simple scale. and that the label information should be indicated on the fixed metal rating plate that is applied to the equipment before it leaves the factory. 179 . which grades appliance energy efficiency using an A to G rating scale. however. While Eurovent Certification reports the energy efficiency levels of most CAC equipment as tested under standard test conditions it does not yet compare the efficiency of equipment in a simplified form. The grading follows the A to G approach used in the EU energy label for household appliances and also applied in some national regulations to rate the energy performance of buildings and cars. the remainder of this section contains explicit recommendations regarding the thresholds which could be applied to denote the A to G efficiency grades for each equipment type examined in this study. it could be made mandatory within Eurovent for all manufacturers who wish to place their products in the Eurovent catalogue to report the A to G grading of their equipment). It is therefore debatable whether there is much advantage in applying a removable energy label to the product itself. Even though central AC equipment are subject to the classic split incentives situation where the purchaser or procurer is unlikely to be the entity responsible for paying the energy bills the provision of relative efficiency information is still a fundamental component supporting more complicated policy measures that may aim to bridge the split incentives problem. The current EU energy label applied to household appliances. A to G efficiency grading of central air conditioner components This section gives proposals for the energy efficiency grading of the principal types of CAC cooling equipment (the equipment which generates the primary cooled or heated fluid) separated from the rest of the CAC system (the cold or heat transfer and conditioned air distribution system) under standardised load conditions. however. including on-line sales. would allow greater differentiation of products within the market place. have expressed an interest in adopting such a comparative grading approach. industry associations. It certainly makes sense to exploit the high brand recognition and public understanding of the current A to G energy-label efficiency scale for the comparative rating of CAC components. it is desirable that not only the procurer should be aware of the comparative energy efficiency of the equipment they are purchasing but also that subsequent potential users of the piece of equipment should be able to see this easily.g. The current EU energy labelling Directive is restricted to household appliances hence would require amendment to address this issue. Alternatively it seems essential that the information on the comparative energy rating of the equipment (A to G) should be presented in all product catalogues and literature. is an effective and widely known example of this. such as Eurovent.. The European Commission could take advantage of this informal adoption of a grading system to prepare the ground for a formal efficiency rating scheme in the years ahead.these values may lack credibility therefore Eurovent currently operates a certification scheme of their members products in order to ensure the reliability of claimed performance levels. The information provided in a comparative energy label provides the basic language of energy efficiency that enables many other market transformation efforts to be realised. As mentioned a removable paper label as currently applied to household appliances is not likely to satisfy these requirements.e. However. thus the user of the information is required to have a high degree of expertise to be able to interpret the reported efficiency level.

were the energy consumed by the cooling tower to be included) it would be possible to use the same scale for both. The analysis of Chapter 8 has led to a proposal for an EU IPLV for chillers based upon specific rating conditions. It is clear from the results reported in Chapter 8 that a part-load efficiency scale is more representative of the true energy efficiency of CAC components when in real usage conditions and therefore the ultimate goal should be to develop ratings based on part-load performance. In the future. Heating and cooling-modes. Accordingly this report presents proposals for A to G efficiency grades based on the analysis of full-load performance data. The efficiency ratings are not directly comparable between these 7 distinct sets of test conditions. Admittedly the additional consumption of the necessary fan for the condenser is not included according to the test standard but this omission is arbitrary (efficiency of 0. Two separate efficiency scales letters (one for the cooling function and one for the heating function) are already used for RAC in the EU RAC energy labelling Directive. the uncertainties regarding the tower control and the origin of the natural water being used are such that it is impossible to make a meaningful comparison. which degrade the EER. The results of Chapter 8 have indicated that commonly. The separation for units currently in the catalogue is not needed since the screw units (mostly) follow the same design as units with less than 750 kW of cooling capacity and the product ranges overlap across both capacity ranges. It is proposed to introduce a specific classification for them. Furthermore. A key question is whether it is appropriate to mix all the chillers into a single product group for efficiency grading or to adjust the efficiency scales depending on the product sub-category? Were the systems using water and air completely comparable (i. even if this is adopted without controversy and in the shortest imaginable time scale it is still likely to be years before there are a large number of EU IPLV ratings available from which to develop an IPLV rating scale. • floor-feed systems (in cooling and/or heating-mode for reversible units). however. Product categories. the part-load ratings are usually equivalent to or slightly higher than the full load ratings. those with a capacity less than 750kW and those with a capacity greater than 750kW. it would be appropriate to review the appropriateness of these grades for translation into part-load grades.3) and in fact heat 180 . the relative efficiency of equipment determined at full load is indicative of its relatively efficiency at part-load. • air cooled (in cooling and/or heating-mode for reversible units).e. Some markets require ducted condensers. which inherently generate incomparable figures as follows: • water cooled (in cooling and/or heating-mode for reversible units). Therefore the same approach is followed here. however. at such a time when EU part-load ratings are widely available. which suggests that the same efficiency range may be applicable to both the full and part-load efficiency grades. therefore there is little risk of misleading the public by adopting an initial grading system based on full load performance. The basic chiller types to be separated correspond at least to the different testing conditions applied in the test standard.Structural issues There are certain technical structural issues that need be addressed prior to the formulation of proposals for an A to G efficiency grading. Thus a more pragmatic solution would be to adopt an efficiency grading scale based on an analysis of the currently available full-load efficiency ratings and to use this in the near term. however. Some centrifugal units are already integrated in the catalogue. The chillers could also be separated into two categories as a function of cooling capacity say. but not always. Full or part-load? The first issue to consider is whether the scale should be based on full or partial load ratings. which are intended for use in the near term. at the present time there is no accepted part-load test and only full load efficiency ratings are generally available. • condenser-less units. although it seems they can be graded on the same scale.

for air-cooled units. • use of limits of classes ending by 0. Therefore it is recommended that manufacturers should supply a common static pressure for testing ducted units. However.1. is strictly superior to 3.00 D/E 2. The average EER and COP values are of course based on data which include units using HFCs.75 3. Floor Floor A/B 3.65 3.25 4. Table 10. A. Remote condenser Floor heating and Floor heating and cooling cooling A/B 3.25 D/E 2. is strictly superior to 3.10 E/F 2.25 3.30 3. Water Cooled Water Cooled.90 3. it is proposed to make the average efficiency level of the products on the market correspond to the threshold between the D and E grades for each chiller type whenever possible.85 3.40 C/D 2.05 4.1 Proposed efficiency grades for chillers in the cooling-mode EER boarders Air Cooled Air Cooled.45 4.50 F/G 2.25. classified by order of importance: • use of the same classes width (for simplicity). Impact of the proposed grading on the chiller cooling market 181 .25 C/D 2. • centering on the average of the catalogue (equal treatment between types).50 B/C 3. Table 10. intends to respect the following rules.25 Note: for borders.15 2.90 3.65 5. From these basic rules come the following proposals (Tables 10. for air-cooled units.05 3.85 4.05 4. and the correction of the standards cannot translate the reality.75 E/F 2. For this reason. A.45 4.60 3.55 B/C 2.1 and 10.45 3.45 3.50 4.30 3.80 Note: for borders.55 3.75 3.65 4. The following classes and values regarding R22 are given as an information only since the European market will no longer have R-22 chillers.90 3.00 2.25 3.10.15 4.70 3. the available static pressure is set by the design of the manufacturer.exchange drops in this case. Establishing a classification based on specific statistics for ducted units will eliminate the two problems.45 2. The construction of the scale for the different categories.05 4. Water Cooled Water Cooled.10 3.05 or 0.85 4.35 4.60 3.2). the testing does not give comparable figures.10 2.05 4. • adjustment of the extremes (G of about 10 %. A of about 1 %).95 F/G 2. Proposed grading scale In order to require the same effort for each type of chiller.2 Proposed efficiency grades for chillers in the heating-mode COP boarders Air Cooled Air Cooled.20 4.25 4.20 3.95 3. when testing ducted units.50 3.

0% R22 HFC 15.41 refrigerant R134a 2.1. cooling-mode.5) E (>2.0% A (>3.0% 35.9) C (>2.3% 19.59 A (>3.0% 30.Figure 10.1) Total SUM R407C 4 30 104 296 411 310 121 1276 % R407C 0% 2% 8% 23% 32% 24% 9% 100% SUM R134a 0 11 39 27 30 27 4 138 % R134a 0.0% 25.0% 8.1) B (>2. Air cooled chillers.9% 100.3) F (>2.55 refrigerant HFC 2.0% 0.6% 21.7% 19.7) D (>2.0% 28.3) F (>2.0% 5.0% SUM R22 14 20 118 163 108 33 30 486 % R22 3% 4% 24% 34% 22% 7% 6% 100% %HFC 2% 5% 25% 30% 22% 10% 5% 100% 182 .1) G (<2.0% R407C R134a 20.0% 10. < 750kW 40.5) E (>2.42 refrigerant R22 2.1) refrigerant R407C 2.1) G (<2.6% 2.1) B (>2.9) C (>2.7) D (>2.

25) D (>3.85 refrigerant R22 3.0% 0.0% 20.Figure 10.85) E (>3.45) F (>3.3 and 10.85) E (>3.0% 100.05) G (<3.2% 29.05) G (<3.3% 25.79 refrigerant R134a 4.45) F (>3. cooling-mode.0% 14.34 refrigerant HFC 3.8% 14.2% 16.0% SUM R22 5 3 15 19 28 14 10 94 % R22 5. < 750kW 45% 40% 35% 30% R407C 25% R134a R22 20% HFC 15% 10% 5% 0% A (>5.05) Total SUM R407C 0 31 35 26 45 87 17 241 % R407C 0% 13% 15% 11% 19% 36% 7% 100% SUM R134a 0 6 11 7 4 0 0 28 % R134a 0.88 A (>5.25) D (>3.65) C (>4.6% 100. 183 .4.4% 39.0% 21.3% 3.65) C (>4.05) B (>4.05) refrigerant R407C 3.2.3% 0.0% %HFC 0% 14% 17% 12% 18% 32% 6% 100% When we compare the model-based statistics from the directory with a sample of 2001 confidential market- based figures we find the values given in Figures 10.05) B (>4. Water cooled chillers.9% 10.

9 4.9 3. Sales and Market statistics Sale statistics R407C Eurovent catalog values R407C 2.1 2.6 2.7 2.7 3.1 1.5 4.8 2.9 3.5 Sales and Eurovent catalogue statistics for air condensing units are convergent: sales statistics show a better efficiency.1 Sales and Eurovent catalogue statistics for water condensing units are convergent: sales statistics show a better efficiency.2 2.1 3.7 5.3 100 5 0 0 4.6 2. Figure 10.4 2.9 2.7 2.7 3.4.7 4.7 4.3 3.8 150 3. 184 .5 3.5 4.5 2.2 2.3 4.1 300 10 200 5.7 3.3.3 15 3.3 250 1.9 2.1 4.7 3.3 4.5 2.9 4.1 1000 1.9 400 5.8 2.5 3.7 2.1 2.4 2.8 3.9 20 5.3 5.2 1500 1.Figure 10.3 2.2 200 1.9 100 500 3 2 3 50 2 0 0 2.6 3.6 1.9 3.3 2000 1.5 500 2.3 2..1 3. Sales and Market statistics Sale statistics R407C Eurovent catalog values R407C 1.1 5.

00 > EER (2) Water-cooled air conditioners – cooling-mode Table 10.5 Proposed efficiency grades for split and multi-split packaged AC in the cooling-mode Energy efficiency class Energy efficiency ratio A 3.cooling mode Table 10.40 < EER B 4.80 D 3.50 E 3.10 D 3.50 F 2.6 Proposed efficiency grades for unitary packaged AC in the cooling-mode Energy efficiency class Energy efficiency ratio A 4.20 G 2.60 > EER > 2.40 > EER > 2.50 > EER > 3.20 < EER B 3.50 > EER > 2.30 > EER > 3.3 Proposed efficiency grades for large-split packaged AC in the cooling-mode Energy efficiency class Energy efficiency ratio A 3.80 > EER > 2.20 > EER > 2.10 > EER > 3.80 > EER > 3.40 E 2.20 > EER Table 10.20 G 2.00 > EER > 2.80 D 2.20 > EER > 2.00 < EER B 3.10 C 4.10 > EER > 2.90 G 2.Proposal of grading of packaged AC in Europe (extension of the RAC labelling scheme) (1) Air-cooled air conditioners .40 > EER > 2.60 < EER B 3.80 > EER > 2.20 F 2.80 > EER > 2.60 D 2.20 > EER > 3.00 C 3.40 > EER > 4.00 > EER > 2.20 > EER Table 10.60 > EER > 3.80 E 2.90 > EER 185 .60 E 2.80 C 2.20 F 3.60 > EER > 2.30 C 3.40 F 2.4 Proposed efficiency grades for large unitary packaged AC in the cooling-mode Energy efficiency class Energy efficiency ratio A 3.00 G 2.

50 F 3.40 > COP > 3.40 < COP B 3.40 > COP > 3.60 E 2.40 > COP > 3.20 > COP > 3.40 C 4.00 D 3.20 > COP (4) Water-cooled air conditioners – heating mode Table 10.40 >COP > 4.80 > COP > 2.80 E 2.20 C 3.8 Proposed efficiency grades for large unitary packaged AC in the heating-mode Energy efficiency class Coefficient of performance A 3.50 G 2.10 E 3.00 < COP B 4.70 C 3.20 > COP 186 .70 > COP > 3.20 > COP > 2.60 < COP B 3.40 F 2.00 > COP > 3.00 > COP > 2.40 > COP > 2.60 F 2.60 > COP > 3.50 > COP > 3.80 > COP > 2.10 > COP > 2.10 D 4.70 < COP B 4.40 > COP Table 10.80 E 3.7 Proposed efficiency grades for large-split packaged AC in the heating-mode Energy efficiency class Coefficient of performance A 3.10 > COP > 3.50 > COP Table 10.60 > COP > 2.40 D 3.40 C 3.60 > COP > 2.20 D 3.(3) Air-cooled air conditioners – heating mode Table 10.20 G 2.20 G 3.10 Proposed efficiency grades for unitary packaged AC in the heating-mode Energy efficiency class Energy efficiency ratio A 4.80 F 2.80 > COP > 3.40 G 2.9 Proposed efficiency grades for split and multi-split packaged AC in the heating-mode Energy efficiency class Coefficient of performance A 4.70 > COP > 4.

60 < COP 3% A 3.80 46% D 23% 2.40 5% B 5% 3.40 11% F 11% 2.20 6% C 6% 3.Market mixed statistics based on the scheme (splits and packages mixed) Table 10.11 Air-cooled air conditioners – cooling-mode -mixed statistics (splits and packages mixed).87 W/W.20 26% F 26% 2.5. Class definition % on market Grade % with equal class width 3. Market average efficiency = 2.80 > EER > 2. Class definition % on market Grade % with equal class width 3.Cooling function -mixed statistics % of market(density) 30% 26% 25% 22% 20% 15% 15% 11% 10% 7% 5% 5% 2% 0% A B C D E F G Table 10.60 > EER > 2.00 5% B 5% 3.Heating function -mixed statistics (splits and packages mixed).60 > COP > 2.40 > COP 8% G 5% 187 .20 > EER > 3.00 > EER > 2.46 W/W.40 22% E 22% 2.40 > EER > 2.80 > COP > 2.20 < EER 2% A 2% 3.60 > COP > 3. Air-cooled air conditioners .60 15% D 15% 2.40 > COP > 3.60 17% E 17% 2.12 Air-cooled air conditioners . Market average efficiency = 2.80 7% C 7% 2.20 > EER 23% G 11% Figure 10.20 > COP > 2.

13 lists the full-load efficiency levels for CAC equipment associated with the least life cycle cost for the final user as determined in this study. Table 10. Air-cooled air conditioners .22 RAC 3 to 12 3. Table 10. because most companies supplying the EU 188 . This implies the introduction of energy efficiency requirements for new equipment.15 0. which could be mandatory (MEPS) or voluntary in nature but attaining the same goal. Adopting similar measures based upon an EU part-load performance indicator (EU IPLV and/or EU SEER) would produce energy savings in 2020 at about 18% of the Business As Usual scenario total. Removing less efficient equipment from the market (MEPS and voluntary agreements) The draft Ecodesign of End-Use Equipment Directive proposes that policy measures should be enacted which bring the market average efficiency of equipment up to the least-life cycle cost for the final user. The adoption of policy measures which would move the average new equipment efficiency levels to those indicated in Table 10.20 An alternative approach to MEPS setting is to harmonise levels with those applied internationally.3 from 2005 onwards was simulated in Scenario 2 as reported in Chapter 9 and would lead to energy savings of about 11% by 2020 compared with the Business As Usual scenario. From a commercial perspective there is some logic to this approach.13 Full-load efficiency levels associated with the least-life cycle cost by CAC component (W/W). Equipment type Cooling capacity EER at least life range cycle cost (kW) (W/W) Water-cooled chillers standard 12 to 750 4. It's a pity that our statistics mix the various subtypes (packages and splits) because it seems that there are really two populations in the data.50 Water-cooled chillers centrifugals 750 upwards 5-6 Air-cooled chillers 12 to 750 3.25 0. Note also that we have here only full load COP and EER.6.1 0.00 Large packaged AC (cooling mode) 12 to 75 3. By contrast simply setting MEPS at the current average full-load efficiency levels for CAC equipment would only save energy equivalent to 3% of all CAC energy consumption by 2020: see Scenario 1 in Chapter 9.05 0 A B C D E F G The last column gives a frequency independent of class width (so divided by two in the intermediate class for heating.2 0.Figure 10.Heating function -mixed statistics % of market(density) 0.

g.3 Water-source heat pump. which could achieve similar objectives. 300tons Of course there are other options than applying a mandatory minimum energy performance threshold. only class A equipment should be procured for use in the public sector) • Develop corporate procurement guidance documentation. Some of potential measures are as follows: • Establish public sector procurement guidelines (e.10 COP (6.5 to 39.02 19.5 to 70.51 Values about 14. This would have the benefit that it would not have to wait that the Ecodesign of End- Use Equipment Directive be implemented.5 Packaged AC (reversible) 2. which could take advantage of any efficiency grading system that is introduced.84 39. for packaged AC units.ukepic. Were the EU to adopt the same MEPS requirements as currently apply in the US through ASHRAE 90.1- 1999 (as either a mandatory or voluntary measure) it would imply the following efficiency thresholds: Table 10.1 4tons (cooling-mode) Water-cooled screw chiller. which suggests that the US efficiency thresholds might be appropriate for adoption in the EU.45 COP (4.3 Packaged AC (reversible) 3.1 MEPS levels for CAC equipment (W/W) Equipment Type MEPS level (W/W) Capacity range Packaged AC (cooling-only) 3. the least life cycle cost efficiency level has been found to be the same.14 ASHRAE 90.02 19.g.market are multinationals with headquarters outside the EU and the CAC equipment they supply to the EU market is based on the same product platforms that are used in other international markets. 3. From a programme management perspective this has the advantage that the MEPS efficiency thresholds are tried and tested having been successfully applied elsewhere.5 to 70. as in the UK web-site www.org) • Develop low cost credit lines for more efficient equipment (e. analytical tools and training materials to explain and quantify the advantages of procuring more energy efficient CAC equipment • Develop and promote on-line directories of efficient equipment (e. Furthermore in the only case where the life cycle cost analyses have been directly compared.g. as in the UK Enhanced Capital Allowance scheme) • Arrange training programmes for associations of designers and installers to explain the cost-benefits of more efficient AC equipment and to ensure lines of access to efficient equipment • Provide rebates on efficient equipment (e.5 to 39.84 39.10 US IPLV) Values about 1056 chiller. as in the Dutch rebate scheme for class A household appliances) 189 .g.5 Packaged AC (cooling-only) 2. 4. Encouraging the selective acquisition of more efficient equipment by other means There are many actions that Member States can initiate to encourage the selective procurement of more efficient central air conditioning equipment. One approach would be to negotiate fleet-average efficiency targets with the European industry.50 US IPLV) Values about 440 125tons Water-cooled centrifugal 6.

These limited results imply that it might be feasible to develop simple benchmarks of overall CAC system performance as a function of the level of cooling and air quality required. Furthermore they do not address the activities of the installation engineer who is responsible for executing the system design and commissioning the system. method which requires the use of one of a number of designated detailed building thermal simulation tools. The comparison of the energy performance of eighteen different CAC systems. or corporate tax breaks for manufacturers and/or corporate procurers) 10. Once the fuel of the heating and cooling system has been fixed these two approaches are effectively transposable. these can be supported by the following measures: • The provision of analytical tools and technical guidance enabling the energy efficiency of CAC systems to be optimised • Training of system designers and installers on the options regarding energy efficient CAC systems • The provision of financial and fiscal incentives to help overcome split incentives such as the provision of cheap credit for efficient systems Building code requirements are articulated in quite different ways among EU Member States. As such building codes are the primary policy measure which promote the adoption of more efficient system types. the carbon approach provides an additional degree of freedom for designers to satisfy the requirements through optimisation of the choice of fuel used by the system. each designed to provide total cooling in a typical EU office. which are designed to save energy. environmental and air quality requirements of the client within acceptable cost boundaries. however. the building type and of the cooling and heating degree days. Such measures are necessarily focused on the individual components and not on the performance of the system as a whole. In Portugal.1-1999 standard. • Create favourable tax differentials for efficient equipment (e. In the UK the primary policy goal is carbon abatement and therefore the requirements are expressed in terms of maximum allowable emissions of carbon per m2 per year. has shown that there can be a difference of a factor of 2 in the total energy consumption per m2 of cooled space for typical configurations of CAC equipment using average efficiency components. Following the simple prescriptive measures is an easy way for a designer to demonstrate 190 .g. the absolute annual energy consumption per m2 shows a much smaller variation and follows a trend that can be related in a roughly proportional manner to the annual cooling and heating degree days. however. lower VAT levels. The same results have also indicated that while the proportion of energy required for heating and cooling may vary appreciably from one climate and Member State to another. In some Member States the building codes are articulated in terms of limits regarding the maximum allowable energy consumption per m2. Both approaches have their merits and indeed can even be integrated as is the case in the US ASHRAE 90. as outlined in section 10. Policy measures would aim to encourage the adoption of more efficient CAC systems while maintaining the freedom of the system designer to achieve a solution which meets the cooling. One difference regards how the requirements for energy using systems in building codes should be expressed. The approach of setting limits for either energy consumption or carbon emissions per m2 leaves the designer almost complete freedom to decide how they are going to satisfy the requirements.3 Policies and measures to encourage the adoption of more efficient system structures Policy aims and potential measures targeting the adoption of more efficient system structures Measures that aim to encourage the adoption of more efficient components.2. In this standard the designer is obliged to follow some general requirements and mandatory measures for each technical section but thereafter has a choice of two final compliance pathways: following a further set of simple prescriptive requirements or demonstrating compliance by satisfying the “energy cost budget”. will only realise some of the potential to save energy for CAC systems. there are no requirements on overall annual energy or carbon but rather a set of simple prescriptions to follow. however. therefore they do not encompass the freedoms and constraints applying to the system designer in trying to design an efficient yet effective CAC system.

it's by itself one of the most important energy saving features. existing methods in some countries could be provisionally approved as ways to reach an A or a B. here by the will of the households. Ultimately such a code could be the EU equivalent to the US ASHRAE 90. The ENPER-TEBUC study within SAVE deals with the issue of harmonisation in European Building Codes and has set up a platform for information exchange on Energy Performance (EP) standardisation and legislation among the prominent national players. but for most other Member States it will require completely new regulations to be developed addressing the cooling system. Higher grades such as E or D could be reached on the basis of a certain value of SEER.1 standard. This implies that public domain simulation tools would need to be developed to support the model building code. Legal basis for policy measures targeting more efficient system structures As mentioned. who may need more freedom to meet the same energy goals than would otherwise be allowed through application of a set of simple prescriptions.25 %) or B (-50 %) or even A (-75%) compared with the average European performance level. readily available EER values. simultaneously meets the needs of designers “in a hurry” dealing with standard design briefs and those who have specific and complicated design briefs. building codes are the primary policy measure available to encourage the adoption of more efficient system types. which is non-binding in itself but can be brought either wholly or in-part into national regulations as deemed appropriate by the authorities in each Member State. This and others can make it possible to go quickly for the third generation suggested here. As any other standard this one would be applicable only voluntarily. The need to develop minimum energy performance requirements within building codes for so many EU states raises the question of whether it would be appropriate for Member States to co-operate with each other. through the use of detailed simulation software. and reached a certain performance level like C (. however. companies. local authorities wishing to have buildings consuming ¾ or ½ of European average of the new buildings in 2000 and to have this environmental effort certified from outside. Sharing development and analytical effort makes considerable sense for such a major undertaking and is a key recommendation from this study.e. In the longer term the EU model code could be designed to enable a energy efficiency labelling or grading system for CAC systems of say an A to G type.their compliance with the standard. 191 . In absence of CEN standardisation on many subjects. the prescriptions are relatively rigorous compared with the requirements when a detailed simulation tool is used. Part load is not only a phenomenon to be computed. The recent Energy Performance in Buildings Directive obliges Member States to develop mandatory minimum energy performance requirements for buildings and specifies that these should encompass the energy used by mechanical cooling systems. Following these measures the importance of system design cannot be neglected. taking part load optimisation efforts into account. The satisfaction of the requirements would need to be determined at the design/installation stage through verification using simulation tools. The requirement that a simple calculation method should be constructed against which the compliance of the MEPS is to be judged may also require modification of the Portuguese regulations. just as is currently the case in the US Energy Star for buildings scheme. Therefore the standard creates an incentive for designers to use detailed building thermal simulation software. The intention has been to systematically collect and summarise the different approaches and to develop suggestions for a European 'model code'. In the case of the UK this Directive may simply require minor revisions of the existing regulations for building cooling energy performance. The combination of a prescriptive compliance pathway and a pathway based on meeting overall energy limits demonstrated. The difficulty in obtaining the best grades (closer to A) should be increasing and involve not only the manufacturer but also other elements of the chain. There are many such efforts under way in individual Member States and there would be considerable value in co-ordinating national efforts within an EU umbrella project. Moving from G to F or E might be based only on full-load ratings i. A designer could refer to a design procedure proving he has considered all cost effective options of the project.

g.0) • Adequate sealing and insulation of ducting • The usage of energy (heat) recovery systems For HVAC control systems introduce building code measures which encourage: • Restrictions on dead-bands • Avoidance of set-point overlaps (e.g.1 and which like ASHRAE 90.Specific recommendations The European Commission and/or a coalition of willing Member States should consider: • the development of an EU model building code that addresses air conditioning amongst other energy end-uses. the integration of airside and waterside economisers which are capable of operating in conjunction with mechanical cooling).g. simultaneous heating & cooling) • Stipulations for off-hour controls including: 192 . quality of comfort provision and climate (e.g.g. variable speed drives or variable pitch fan blades • Proper sizing of components such as fans (e.g.g.5) • Limits on the maximum SPF of mechanical ventilation systems in existing buildings (e. b) can enable the comparison of the relative benefits of different system designs. requirements for maximum installed fan power (expressed as W/litre/second)) • Variable flow control (an obligation to do this could produce energy savings worth 10% of all CAC energy consumption by 2020: see Scenario 5 in Chapter 9) • Limits on the maximum SPF of mechanical ventilation systems in new buildings (e. (an EU equivalent to ASHRAE 90. that the SPF should not exceed 1.g. c) can be used in building thermal regulations to demonstrate compliance with requirements • The development of EU benchmarks for CAC system efficiency expressed in terms of: building function and size.1 is subject to continuous maintenance) • The development of practical public domain CAC system design tools which: a) can aid system designers to develop energy efficient CAC designs. that the SPF should not exceed 3. occupancy and purpose. ensuring that if ‘passive’ free-cooling is enabled mechanical cooling does not operate in those zones using free-cooling) • The enablement of automatic free-cooling (e. Note : provisions must be included to ensure their proper functionality otherwise energy losses could occur through this measure (an obligation to do this could produce energy savings worth 5% of all CAC energy consumption by 2020: see Scenario 4 in Chapter 9) • Efficient means being able to control air flow rates e. cooling and heating degree days) Member States should undertake a revision of their building thermal regulations to address the following specific issues aimed at reducing CAC energy consumption: For air-distribution systems introduce building code measures which encourage: • Operation in mixed-mode with natural ventilation (e.

including modular plant (e. 1. controllers. Automatic shutdown 4. cooling towers. good part-load efficiency) (e. electric heating power provided by the Joule effect should not exceed 5% of the total heating power installed. noise levels and overall functionality In addition the following specific measures which impact of CAC system design and installation are required under the Energy Performance in Buildings Directive: 193 . • Limits on terminal re-heating.g.) 2. hydraulic tests. pumps.g.g. including modular plant (i.e.e.g. Optimum start controls after the system airflow exceeds a minimum level For refrigeration plant systems introduce building code measures which encourage: • Free cooling from cooling towers • Variation of fresh air using economy cycle or mixed mode • Controls which restrict the hours of operation of the system • Controls which prevent simultaneous heating and cooling in the same zone • Efficient control of plant capacity. nor 25kW by independent zone). For the installation and commissioning of CAC systems introduce building code measures which encourage: • commissioning tests to be conducted for boilers. Setback controls 5. Shutoff damper controls that automatically close when the systems or spaces served are not in use (these dampers should also satisfy a maximum allowable leakage rate. Zone isolation capabilities that permit areas of the building to continue operating while others are shut down 3. (e. the use of power stages to allow output to be adapted to the demand) • Efficient control of heat rejection equipment capacity. heat exchangers. chillers (power and efficiency). good part-load efficiency for cooling towers) • Full cold thermal storage (i. terminal re-heating is allowed for cooling-only systems but can not exceed 10% of the installed cooling power). chillers would only operate at night) • Proper sizing of components such as pumps and refrigeration equipment • Adequate insulation of piping • The use of energy (heat) recovery systems • The use of variable flow hydronic systems and wherein pumps draw substantially less power at part- load than full-load • Heat recovery for service water heating For CAC integrated heating systems introduce building code measures which encourage: • Limits on the Joule heating (e.

• Energy certification of new and existing buildings to verify their compliance with minimum energy performance requirements • Certificates are to be made available to the owner or prospective tenant when the building is constructed and when it is sold or rented out. Member States are obliged to ensure that this inspection is conducted in an independent manner by qualified and or accredited experts. In fact this is already common practice today although there can be a bewildering array of contractual arrangements on offer and little access to independent assessment of the results produced. It would therefore be appropriate that policy measures be developed to encourage good practice in this regard. 194 . Legal basis for policy measures targeting O&M Article 9 of the recent Energy Performance in Buildings Directive specifies that Member States must introduce mandatory regular inspection of AC systems above 12 kW in capacity. It shall be accompanied by recommendations for the cost-effective improvement of the energy performance . If implemented in a intelligent manner these requirements could go some way to ensure that existing maintenance and operation contracts are appropriate and are properly observed. flow rates and temperatures of installed systems and their components) • That individual items of equipment which are performing at lower than rated efficiency levels are clearly reported in the certification procedure. Most end-users are unlikely to have the required expertise in house to optimise the efficient use of the CAC system and therefore it would be beneficial wthat they be encouraged to undertake suitable service contracts with specialist system operators. Recommendations about these requirements are that Member States should consider: • Ensuring that a company or entity independent from the designer and installer should conduct the building certification • That certifiers are trained and clear certification procedures have been developed and adopted • That the certification would be automatically triggered for all new installations and would verify that both the installed systems and their individual components meet the energy performance requirements and attain their pre-declared performance ratings (this would require be able to measure the electrical energy. It has been estimated that without proper maintenance the efficiency of CAC systems deteriorates by 2% every 5 years. Certificates shall not be valid for more than 10 years. Optimal operation of the system can be encouraged through intelligent control regimes which in turn can be encouraged through appropriate energy performance contracting. The inspection shall include an assessment of the AC system efficiency and sizing relative to the cooling demand of the building and advice will be provided to the users regarding possible improvement or replacement of the AC system and on alternative solutions. As maintenance is relatively unexpensive and straightforward it is appropriate to implement policy measures which encourage regular competent maintenance to minimise this deterioration in performance.4 Policies and measures to improve system maintenance and operation Policy aims and potential measures targeting improved O&M The maintenance or improvement of performance. 10. • The certificate shall include reference values such as legal standards and benchmarks. by technical measures or contractual means (Energy Performance Contracting) or by periodic audits can result in significant energy savings for CAC systems. In particular the service and cleaning of heat exchangers should be properly encouraged.

The standard also addresses balancing for air systems larger than 1 hp and for hydronic systems larger than 10 hp. Broadening the application of existing policy measures addressing O&M The Portuguese regulations impose the adoption of a maintenance plan and a monitoring system for CAC systems such that the energy consumption of all equipment with an electric power greater than 12. especially for smaller energy consumers such as SMEs. it is important that Member States implement this measure in such a way that a review of operation and maintenance contracts are encompassed within the inspection. (e. Member States should consider enacting measures to ensure that: • Building owners and occupiers are provided with a logbook and an adequate operation guide for the CAC system deployed 195 .. but some would be common to all Member States. the design assessment for CPR or other benchmarks. It is envisaged that Member States will set targets to promote and support energy efficiency services. any chiller installation (which may include more than one chiller) of greater than 20 kW input power should be separately metered. proportioning of cooling may be acceptable). The objective of the proposed Directive on Energy Demand Management (also known as Energy services) is to complete the internal market for energy by developing and encouraging energy efficiency on the demand side. However. especially as it is provided by utilities and service companies in the form of energy services. This includes separate metering for tenancies of more than 500 m2 (though for tenancies below 2500 m2. so that the findings could be fed into the national provisions being drawn-up by Member States. The UK regulations also require that the owner and/or tenant of the building be provided with a logbook that contains. Some aspects of these guidelines would necessarily be specific to the situation applying in individual Member States. If the Commission is unable to initiate this process actors at the Member State level charged with implementing the Energy Performance of Buildings Directive could take the initiative to establish a working party of willing Member States. In parallel it would be very useful that efforts be made to define best practice in operation and maintenance so that best practice guidelines can be issued against which ever existing contracts would be compared. It also requires control elements to be calibrated. it's unlikely that energy savings will accrue if the operator doesn't know how the system should work. third party financing) and programmes. Therefore it would be very beneficial were the European Commission to take a lead in organising an EU working party charged with defining best practice in the operation and maintenance of AC systems. and any motor control centre providing power to fans and pumps of more than 10 kW input power. This is a sensible provision because even if an engineer designs a great system.000 sq ft. There is an additional requirement that sub- metering should be provided. This could certainly be used as an opportunity to improve the O&M of CAC if appropriate rules can be defined. The US ASHRAE 90. and in proper working condition for buildings that exceed 50. commissioning details.This requirement for “independent” inspectors presumably precludes that the inspection should be conducted by the entity holding the operation and maintenance contract and therefore should provide some measure of independent verification of the proper conduct of those contracts.g.5 kW should be independently metered.1 standard requires that drawings. adjusted. amongst other things. operating instructions. manuals. Generally. and a narrative of system operation must be supplied to the building owner. and details of all meters provided. Specific recommendations The European Commission and/or a coalition of willing Member States should consider: • Making efforts to define best practice in operation and maintenance • Making efforts to define best practice in operation and maintenance performance contracting With an aim of informing national building thermal regulations and the implementation of the Energy Performance in Buildings Directive.

196 . impose a requirement for regular maintenance and independent metering of CAC systems above a minimum size) • That a competent inspectorate be developed capable of carrying the provisions of the Energy Performance in Buildings Directive applying to AC systems • One of the roles of the inspectorate required under the terms of the Energy Performance in Building Directive should be the independent review and evaluation of CAC system operation and maintenance contracts Member States could also consider the development of low cost finance mechanisms to encourage the adoption of good practice for CAC operation and maintenance. • A regular maintenance and monitoring system be adopted for all CAC systems (e.g.

Categorical energy label An energy label that classifies product efficiency into one of several classes. Design temperature The temperature within a conditioned space that needs to be achieved during a test for the energy-consumption measurement. MEPS Minimum energy performance standards (sometimes known as ‘minimum energy efficiency standards’). List of abbreviations AC Air Conditioning ACEA EU association of car makers ACMV Air Conditioning & Mechanical Ventilation (UK regulations) ADENE Portuguese energy-conservation agency AHAM US Association of Home Appliance Manufacturers AHU Air Handling Unit AICIA Association conducting research under the auspices of ETSIIS in Seville AICARR Italian Association of Air Conditioning. Examples include the EU’s energy labels. Korea. or it can be the converse in which the future savings are discounted using a real discount rate. Brazil. Payback period (PBP) The period of time it takes for a consumer to recover the extra investment made in a higher-efficiency appliance through savings in operating costs. Control cycle The period between two successive starts or two successive stops of the compressor in a refrigerating system. Top Runner The term applied to the Japanese appliance energy-efficiency policy. Net present value (NPV) The monetised value of future costs expressed in terms of their discounted value at the present time. Refrigerating and Air Conditioning Engineers BAT Best Available Technology BAU Business As Usual BRE Building Research Establishment 197 . ALCC Annualised Life Cycle Cost ANSI American National Standards Institute ARI American Refrigeration Institute AS/NZS Joint test standards issued by Standards Australia and Standards New Zealand ASHRAE American Society of Heating. which rank efficiency from A to G. The payback period can be ‘simple’ in that no discounting of future savings is applied. Iran.Definitions and general terms used in the study Appliance category A group of appliances or equipment that have similar technical characteristics from the perspective of their user utility. wherein MEPS have been set at efficiency levels equivalent to those of the highest efficiency appliance on the market. Thermal bridge A high thermal conductivity pathway. Defrost cycle The period between two successive starts or two successive stops of a defrost heater in a appliance with an automatic defrost system. Mexico and India have all developed categorical energy labels. and Australia’s energy label. which ranks efficiency from 1 to 6 stars. Thailand. Heating and Refrigerating Engineers. Montreal protocol The internationally binding agreement to phase out ozone-depleting substances such as CFCs.

air-conditioning and ventilation equipment manufacturers FCU Fan Coil Unit GB Great Britain (excludes Northern Ireland) GEA Group for Efficient Appliances CPR Carbon Performance Rating (UK regulation) HFC Hydrofluorocarbons IDAE Spanish energy-conservation agency IEC International Electrotechnical Committee IPLV Integrated Part-Load Value ISO International Standards Organisation LBL-MIM Lawrence Berkeley Laboratory – Manufacturer Impacts Model LCC Life-Cycle Cost LLCC Least Life-Cycle Cost MEPS Minimum Energy Performance Standards (same as MEES. Minimum Energy Efficiency Standards) NPV Net Present Value ODP Ozone-Depletion Potential RAC Room Air Conditioners (in the wide sense) R&D Research and Development 198 . Restaurants CEC Central and Eastern European Countries or Commission of the European Communities CECED European major household appliance manufacturers’ association CECOMAF See Eurovent CEEC Central and Eastern European Countrie CEN Committee European de Normalisation (European Committee for Standardisation) CFC Cloroflurocarbons COP Coefficient of Performance CPR Carbon Performance Rating (UK regulation) CPR Carbon Performance Rating (UK regulation) DD Degree Days EDF Electricité de France EER Energy-Efficiency Ratio (W/W) EMPE Italian Method for part load rating EMS Energy Management Systems EPB Energy Performance in Buildings (EU regulation) CPR Carbon Performance Rating (UK regulation) ESCO Energy Service Companies EU European Union Eurovent European association of refrigeration. Hotels.CAC Central Air Conditioners CAHORE Cafes.

2001... Procedings of the AFCE conference. France. Air Handling. CEN. Rivière P.Bouteloup. standardCenelec EN 12055. C Vecchio. Seasonal efficiency of primary air conditioning systems. 1996. CEN. 199 . 1998. Permanent Technical Committee for "Refrigeration" in AICARR's Technical Activity Commission.F Kreider. J. Becirspahic S.. 1998. Heating and Cooling of Buildings. M. standardCenelec EN 255. E. Air-conditioning. Vizzotto. Bacigalupo. Mc Graw-Hill Book Company. “Les Editions Parisiennes”. Water Chilling Packages using the vapor compression cycle.. 1997.Le Guay. Design for Efficiency. 1998. 1997. SSEER system seasonal energy-efficiency ratio (W/W) TAC Total air conditioning (one level of comfort) TC Total cooling (one level of comfort) TEWI Total Equivalent Warming Impact UK United Kingdom (includes Northern Ireland) UoA Univeristy of Athens US DOE US Department of Energy US EPA US Environmental Protection Agency VA Voluntary Agreements VAV Variable Air Volume VRF Variable Refrigerant Volume VVT Variable Volume and Temperature REFERENCES AFCE. in Watts (of motor rating) per litre/second of airflow. J. « Air conditioners and heat pumps with electrically driven compressors – Cooling mode”.Liguen. 1997. Colomines F. AICARR. 1999. M. Sept 2002.2001 Average weighed efficiency of compression chillers: AICARR’s proposal for a calculation method AICARR. Standard 550/590. M. A. Adnot J. Marchio D. « Air conditioners and heat pumps with electrically driven compressors – Heating mode”. 2002. J. Ecole des Mines de Paris.. 1998. ARI. 4 vol. Average weighed efficiency of compression chillers: a proposal to AICARR for a calculation method..Rabl.RSECE The Portuguese building thermal regulations which include requirements for AC systems SEER Seasonal Energy-Efficiency Ratio (W/W) SPF Specific Fan Power. 2001. 1994. ASHRAE Handbooks Fundamentals and Systems. Vio.

TRIBU.S. 1994. December 2002. Preliminary European Guide on HVAC design. ADNOT. 1998. October 1999 200 View publication stats . RIVIERE. Eric Auzenet. LBNL (2002) ‘Commercial Unitary Air Conditioner & Heat Pump: Life-Cycle Cost Analysis - Inputs and Results’. 2004. Commercial unitary air-cooled air-conditioners and air-source heat pumps’. E&FN Spon. and Ventilation" Prepared by Detlef Westphalen and Scott Koszalinski. (1994). Paris: Ecole des Mines de Paris et al. (2002) UNI Standard 10963 : Air conditioners . REHVA. Eurovent. Control Guide for Ventilation and Air-conditioning. England. proposed revision of [CEN98] and [CEN97]. Annuaire des Produits Certifiés. TIAX LLC. ventilating and air-conditioning. Little. thermic school.part load tests. ORPHELIN Tiax (2002) ‘Engineering-analysis cost-efficiency curves. J. Auxiliary Equipment. France.com. Dominique Marchio. EERAC [1999] Energy efficiency of room air-conditioners. Inc. Arthur D. ECEEE 1997 summer study proceedings Rivière. M. "Guidance for the use of the carbon emissions calculation method: CIBSE TM32:2003" (ISBN 1 903287 41 3) Ph Davy de Virville. Department of Energy. www. 1997 . CEN. RHEVA. 2001 general presentation and use of a method of calculation of consumption of the RAC (Room Air Conditioners) by Ph. CIBSE. for DG- TREN.eurovent- certification. May. "Etude Comparative des Réglementations Thermiques Européennes des Bâtiments non résidentiels". The international dictionary of heating. For U.026. 1994 for Ademe UNI. 1998 Eurovent-Certification. 1993. 2002. prEN 14511 – 1 à 5. Cambridge Massachusetts for the US Department of Energy.103/D/97. Paris. SAVE contract DGXVII4. chillers and heat pumps. January 2002. 2002. 1999) "Energy Consumption Characteristics of Commercial Building HVAC Systems Volume II: Thermal Distribution. Lawrence Berkeley National Laboratory for the US Department of Energy. TRIBU. Westphalen. the Commission of the European Communities .