Eeccac

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.
CONTENTS
ABSTRACT ................................................................................................... 11
SUMMARY OF RESULTS ............................................................................. 11

Definitions of all CAC systems found on the EU market have been given. ....................................... 12
All CAC equipment test standards have been reviewed and studied to assess their suitability to
represent energy efficiency under real operating conditions. ............................................................. 13
CAC market and stock data have been assembled for the first time................................................... 13
The present Energy Efficiency efforts have been reviewed ............................................................... 16
All the elements of a possible grading of Cooling market have been assembled ............................... 19
Splits and Packages are grouped in one single category ( .................................................................. 19
The impacts of BAU have been assessed ........................................................................................... 19
Optimisation of a chiller to improve its EER on the basis of capacity cost only ................................ 23
Optimisation of a chiller for its least LCC.......................................................................................... 23
Packaged units can also be improved a lot ......................................................................................... 24
System optimisation : all air systems.................................................................................................. 25
Part load performance has been quantified for the first time and the methods have been tested ........ 26
Impact of load reduction on the efficiency – a reporting format proposed to Eurovent ..................... 27
Magnitude of gain/losses due to part load .......................................................................................... 27
The simulations leading to the reference values of SEER (HSEER) .................................................. 28
EECCAC final figures for a European SEER method (ESEER) ........................................................ 29
EER alone is a poor selection tool ...................................................................................................... 29
IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for
comparison of chillers ........................................................................................................................ 30
The newly proposed ESEER method allows grading and ranking of chillers by order of merit ........ 31
Energy efficiency options have been defined for each system configuration ..................................... 31
Scenarios for energy efficiency have been established and quantified ............................................... 32
All the elements for an action plan on Air Conditioning are available in the full report .................... 33
1. INTRODUCTION ....................................................................................... 36
Selection of technical experts ...................................................................................................... 36
Participation of energy agencies, utilities, manufacturers and national experts ............... 37
2. CENTRAL AIR-CONDITIONERS IN EUROPE: DEFINITIONS

AND BASIC DATA .................................................................................... 38
2.1. Importance of AC for human health and productivity performance, link with
ventilation ........................................................................................................................................... 38
What is "comfort"? ............................................................................................................................. 38
Comfort level, Ventilation: our assumptions for the study ................................................................. 39
2.2. Basic definitions ........................................................................................................................ 39

RAC and CAC in competition ............................................................................................................ 39
Basic Thermodynamics at one instant ................................................................................................ 39
Main technologies for cold production ............................................................................................... 40
CAC systems types based on distribution .......................................................................................... 43
Classification of the systems .............................................................................................................. 43
2.3. Description of other aspects of systems ....................................................................................... 44

Terminal units and other peripheral equipment used.......................................................................... 44
General classification of systems based on chillers ............................................................................ 47
Number of water loops connected with the chiller ............................................................................. 49
2.4 Description of systems not using chillers ..................................................................................... 49

VRF (Variable Refrigerant Flow) CAC systems ................................................................................ 49
Water Loop Heat Pump CAC systems based on local packaged AC systems.................................... 49
Local package CAC systems: roof-tops and close control cabinets ................................................... 49
Inclusion of RAC in the present study................................................................................................ 49
Summary of choices in terms of local versus central systems ............................................................ 50
Sizing issues ....................................................................................................................................... 50
Free cooling ........................................................................................................................................ 51
2.5. Testing standards and performance standards .............................................................. 51

Chillers: the CEN and ARI approaches (at full load and IPLV) ......................................................... 51
Peripheral equipment of chiller based systems: testing and classification ......................................... 53
A proposal for a better characterisation of AHU ................................................................................ 54
Ventilation efficiency and Air Conditionning .................................................................................... 54
Testing and performance setting for packaged systems ..................................................................... 55
2.6. Overall view of energy performance ........................................................................................... 58

Year round thermodynamic balance ................................................................................................... 58
Definitions .......................................................................................................................................... 58
Full system efficiency......................................................................................................................... 60
2.7. Statistical databases used and information gathered ................................................................. 60

National surveys ................................................................................................................................. 60
Data from manufacturers associations ................................................................................................ 60
Correction and treatment of data ........................................................................................................ 61
3. MAIN FIGURES OF AIR-CONDITIONING IN EUROPE ................ 62

3.1. The demand for AC in Europe ........................................................................................... 62
A general growth ................................................................................................................................ 62
National differences in demand .......................................................................................................... 62
3.2. Technical response to the demand...................................................................................... 64

Market share of each technology ........................................................................................................ 64
Evolution of market shares of techniques ........................................................................................... 64
Comparisons with US market ............................................................................................................. 66
3.3. A few technical trends on the market ................................................................................ 68

The share between distribution systems in chiller based CAC ........................................................... 68
Reversible use of Air Conditioning .................................................................................................... 69
The choice between chiller-based systems and packages ................................................................... 69
The value and nature of the European CAC market ........................................................................... 71
Other stakeholders .............................................................................................................................. 72
3.4. Statistics on present Energy Efficiency on the EU market ......................................... 73

EER as a function of capacity and cooling medium for a chiller under 750 kW ............................... 73
Potential for efficiency gains of the selection of higher efficiency equipment................................... 75
EER for chillers over 750 kW ............................................................................................................ 76
4. FACTORS GOVERNING THE DESIGN, SELECTION, INSTALLATION

AND OPERATION OF CAC SYSTEMS ........................................................ 77
4.1 Actors involved with CAC systems ............................................................................................... 77
The main barriers to efficiency ........................................................................................................... 77
4.2 Practices and procedures adopted in CAC system design .......................................................... 77

Guidelines for the design of CAC systems ......................................................................................... 77
4.3Previous ............................................................................................................................................ 78
market-transformation efforts within the EU (equipment) .............................................................. 78

The Eurovent Certification programme .............................................................................................. 78
An example of a utility-led energy-efficient AC promotional campaign ........................................... 79
The UK Market Transformation Program .......................................................................................... 80
4.4 ........................................................................................................................................................... 81
Existing national regulations within the EU (which apply at the system level)............................... 81
Portugal: An example of a national scheme to promote energy-efficient AC through building thermal
regulations .......................................................................................................................................... 81
Summary of UK building regulations for space cooling and ventilation............................................ 82
The status of regulations in other EU Member States ........................................................................ 86
The Energy Performance of Buildings Directive (to be transposed nationally) ................................. 86
The draft Framework Directive for “Eco-design of End-Use Equipment” (to be adopted)................ 87
The draft Directive on Energy Demand Management (to be defined)................................................ 87
Practices and procedures adopted in CAC system operation.............................................................. 88
4.5 Regulatory structure and market transformation at the wider international level .................. 88
Minimum efficiency standards and energy labelling in the USA ....................................................... 88
ASHRAE 90.1: a comprehensive approach to raise CAC energy efficiency ..................................... 88
Mandatory HVAC Provisions in ASHRAE 90.1 ............................................................................ 90
Additional prescriptive HVAC requirements ..................................................................................... 91
Continuous maintenance of the ASHRAE standard ........................................................................... 92
Links between an ASHRAE standard and the US Energy Codes....................................................... 92
Australia, Japan, Korea and Taiwan ................................................................................................... 93
4.6 Choices and measures which could increase the efficiency of CAC systems ............................. 94
Measures which could increase globally the efficiency of CAC ........................................................ 94
Technical measures which could increase the efficiency of CAC systems ........................................ 95
Synthesis of policy measures to raise the efficiency of CAC systems ............................................... 98
First type: selection of more efficient components by whoever decides ............................................ 98
Second type: choice of the best general structure of the system......................................................... 98
Third type: improvement of the detailed structure of the system and control options........................ 98
Fourth type: reversible use of the system ........................................................................................... 98
Fifth type: maintenance and operation improved ............................................................................... 98
Sixth type: energy and power control ................................................................................................. 99
Seventh type: envelope and ventilation, other measures .................................................................... 99
5. PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO) .......... 100
5.1 AC Stock and market in 1990, 1998, 2010 and 2020.................................................... 100
Evolution of the market .............................................................................................................. 100
Some global results ...................................................................................................................... 101
Some national results................................................................................................................... 102
Sectoral market ............................................................................................................................. 103
The share between technical systems ...................................................................................... 104
5.2 Computation of energy consumption in European conditions ................................. 104

Real buildings for the simulation of CAC systems with DOE ......................................................... 105
Coverage of situations with the DOE software ................................................................................ 105
Adjustment for chiller quality and options not covered in DOE software........................................ 107
Preliminary results for the cooling consumption of the office building ........................................... 107
Extension to all economic sectors, system types and EU climates ................................................... 108
Preliminary results for the cooling consumption of the office building ........................................... 108
5.3 Energy consumption in 1990, 1998, 2010 and 2020..................................................... 112

Overall values ............................................................................................................................... 112
Energy by economic sector ........................................................................................................ 115
5.4 Global warming and other environmental impacts .................................................... 115

Atmospheric pollution reduced to CO2................................................................................... 115
TEWI (Total Equivalent Warming Impact) and leak rates of CAC systems .................................... 116
Numerical results about CO2 emissions for cooling in Europe ........................................................ 117
Use of water................................................................................................................................... 118
5.5 Heating, reversible or not .................................................................................................... 118
6. TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY

EQUIPMENT USED IN CAC ....................................................................... 122
6.1 Energy-engineering analysis of chillers ...................................................................................... 122
Chiller prices as a function of the refrigerating fluid and EER ........................................................ 122
Role of condensing medium ............................................................................................................. 122
Additional costs for reversibility ...................................................................................................... 122
Defining chiller part-load efficiency ................................................................................................ 123
Available data and simulation tools.................................................................................................. 124
Incremental costs as a function of efficiency.................................................................................... 124
Optimisation of the chiller used as baseline without any system consideration ............................... 124
Optimisation of a chiller in a system ................................................................................................ 125
Water cooled chillers ........................................................................................................................ 126
6.2 Engineering approach of the performance of Packaged units .................................................. 127

The US energy engineering analysis ................................................................................................ 128
Life cycle cost analysis ..................................................................................................................... 129
6.3 Energy Efficiency of Air Handling Units seen as tradable goods .......................................... 131
Fans integrated in AHU .................................................................................................................... 131
Heat recovery section of AHU ......................................................................................................... 132
7. TECHNICAL & ECONOMIC EVALUATION OF CAC SYSTEM PERFORMANCE AS A

FUNCTION OF THE DESIGN OF THE AC SYSTEM................................................................. 134
7.1 Comparison of different CAC systems ....................................................................................... 134

Energy consumption for a given comfort level ................................................................................ 134
Comparison of costs and sensitivities ............................................................................................... 135
7.2 The improvement of the efficiency of air handling systems in CAC ........................................ 135
Primary Air and ventilation .............................................................................................................. 135
Heat recovery on primary air ............................................................................................................ 136
Motors and fans efficiency ............................................................................................................... 136
Variable air flow and lower head losses ........................................................................................... 136
Terminal reheat issues ...................................................................................................................... 137
Air Side Free Cooling (Economiser) ................................................................................................ 137
Quality of Air Diffusion ................................................................................................................... 138
AHU improvement ........................................................................................................................... 138
7.3 Other cost & efficiency trade-offs ............................................................................................... 138

Water-side efficiency by sizing and control ..................................................................................... 138
Design of flow in water circulation .................................................................................................. 139
Influence of terminal equipment....................................................................................................... 139
Simultaneous demand of heating and cooling .................................................................................. 140
Heat rejection ................................................................................................................................... 141
7.4 The possible strength of regulatory efforts and the minimum LCC solutions ........................ 141
Concentration of efforts on Air based systems ................................................................................. 141
The result of optimisation ................................................................................................................. 141
8. EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE ........ 143

8.1 The importance and nature of part-load management measures ............................................ 143
Importance of establishing a EU method about part load ................................................................. 143
How to reduce the capacity of a chiller? .......................................................................................... 143
Staging of Part capacity (control issues) .......................................................................................... 145
High pressure control at part load..................................................................................................... 146
8.2 Is the IPLV approach directly applicable to European conditions? ........................................ 148
Buildings used in deriving the US-IPLV .......................................................................................... 148
Climate used in IPLV derivation ...................................................................................................... 148
Building cooling load calculation in US-IPLV................................................................................. 149
Calculating US weighing coefficients .............................................................................................. 149
Interpolation scheme needed to reduce testing time ......................................................................... 149
EMPE: an answer to a need for a European weighting with IPLV-like testing ................................ 150
Reduction of EMPE or IPLV to 2 points with extrapolation ............................................................ 151
8.3. Construction of a data base of EU chillers at part load –understanding part load ............... 152
Testing conditions and available testing results ............................................................................... 152
Impact of load reduction on the efficiency – a reporting format proposed to Eurovent ................... 153
Water cooled chillers –experimental results ..................................................................................... 153
Air cooled chillers –experimental results ......................................................................................... 155
8.4 Derivation of a new SEER method (ESEER) ............................................................................. 156

The simulations leading to the reference values of SEER (HSEER) ................................................ 156
Sizing issues for chillers rating as shown by the simulation of the buildings................................... 156
Reduction of European hourly load curves to a set of four conditions (based on the example of
Milano) ............................................................................................................................................. 158
Results for more extreme weather conditions (London, Seville, different distribution systems) ..... 160
Extrapolating to the European stock of chillers in use ..................................................................... 162
8.5 Is there a method good enough for classification of products by order of merit? .................. 164
EECCAC final figures -Simplification of the figures and uncertainty estimate ............................... 164
Classification : who is right? ............................................................................................................ 165
EER is a poor selection tool ............................................................................................................. 165
IPLV and EMPE are more accurate than EER for classification but do not give enough accuracy for
comparison of chillers ...................................................................................................................... 166
The proposed ESEER method allows grading and ranking of chillers by order of merit ................. 167
First way to realise the testing needed for the ESEER proposed certification method ..................... 167
Second way to realise the testing needed for the ESEER proposed certification method ................ 170
Final choice of the ESEER testing methodology.............................................................................. 171
Perspective of the proposed ESEER ................................................................................................. 171
9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY

CAC SCENARIOS ....................................................................................... 173
9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY

CAC SCENARIOS ....................................................................................... 173
9 ............................................................................................................................................................ 173
Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE ............... 173
Scenario 2 THE BEST CHOICE AMONG EXISTING COOL GENERATORS BASED ON FULL
LOAD INFO..................................................................................................................................... 173
Scenario 3 BAT- THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO ........ 173
Scenario 4 FREE COOLING ........................................................................................................... 174
Scenario 5 VAV ............................................................................................................................... 174
Scenario 6 British regulation on AC – heating, cooling and air movement- adapted for each EU
climate .............................................................................................................................................. 174
9.2 Results of scenarios....................................................................................................................... 174

General Evolution ............................................................................................................................. 174
Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE ............... 175
Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON
FULL LOAD INFO .......................................................................................................................... 175
Scenario 3 BAT- THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO ........ 176
Scenario 4 FREE COOLING ........................................................................................................... 176
Scenario 5 Variable Air Flow ........................................................................................................... 176
Scenario 6 British regulation on AC – heating, cooling and air movement- adapted for each EU
climate .............................................................................................................................................. 177
10. POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC

ENERGY PERFORMANCE ......................................................................... 178
10.1 Some fundamental considerations regarding policy measures .............................. 178
10.2 Policies and measures to encourage the selection of more efficient equipment .................... 178
Measures to provide information to end-users and equipment procurers ......................................... 178
A to G efficiency grading of central air conditioner components..................................................... 179
Market mixed statistics based on the scheme (splits and packages mixed) ...................................... 187
Removing less efficient equipment from the market (MEPS and voluntary agreements) ................ 188
Encouraging the selective acquisition of more efficient equipment by other means ........................ 189
10.3 Policies and measures to encourage the adoption of more efficient system structures ........ 190
Policy aims and potential measures targeting the adoption of more efficient system structures ...... 190
Legal basis for policy measures targeting more efficient system structures ..................................... 191
Specific recommendations................................................................................................................ 192
10.4 Policies and measures to improve system maintenance and operation .................................. 194
Policy aims and potential measures targeting improved O&M ........................................................ 194
Legal basis for policy measures targeting O&M .............................................................................. 194
Broadening the application of existing policy measures addressing O&M ...................................... 195
Specific recommendations................................................................................................................ 195
Definitions and general terms used in the study .............................................................................. 197
List of abbreviations ........................................................................................................................... 197
REFERENCES ................................................................................................................................... 199

Energy Efficiency and Certification of Central Air
Conditioners
ABSTRACT
Air-conditioning constitutes a rapidly growing electrical end-use in the European Union (EU), yet the
possibilities for improving its energy efficiency have not been fully investigated. Within the EECAC study
twelve participants from eight countries including the EU manufacturers' association, Eurovent, engaged in
identifying the most suitable measures to improve the energy efficiency of commercial chillers and AC
systems. Definitions of all CAC systems found on the EU market have been given. All CAC equipment test
standards have been reviewed and studied to assess their suitability to represent energy efficiency under real
operating conditions. European CAC market and stock data have been assembled for the first time. We can
keep a few figures in mind : 1200 Mm2 cooled in year 2000 (3m2/inhabitant), 2200 Mm2 in 2010
(5m2/inhabitant), with a share of reversibility around 25%.
The present Energy Efficiency efforts have been reviewed. They play a negligible role, 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). One thing can be done rapidly : all the elements of a possible
grading of chillers on the market, based on full load behaviour, have been assembled. Is there a margin for
further improvement ?
Optimisation of a chiller for its least LCC shows a large possibility, namely thanks to part load control. 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 3.00 and 3.50 and an initial overcost of +12% paying for itself
rapidly. For manufacturers, there are certainly other ways to reach 3.25 SEER than the ones investigated, less
expensive, but our objective was to find out if there is a margin for improvement. 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.
Packaged units can also be improved a lot. We show that the life cycle cost minimum occurs for large
packaged units with an EER of 3.22 W/W.
In terms of market transformation, EER is a poor selection tool ; the US IPLV and the Italian EMPE are
more accurate than EER for classification but do not give enough accuracy for comparison of chillers. The
proposed ESEER method allows perfect grading and ranking of chillers by order of merit . Energy efficiency
options have been defined for each system configuration and for the components outside the chiller.
Scenarios for energy efficiency have been established and quantified. All the elements for an action plan on
Air Conditioning are available in the full report.
SUMMARY OF RESULTS
Air-conditioning constitutes a rapidly growing electrical end-use in the European Union (EU), yet the
possibilities for improving its energy efficiency have not been fully investigated. As opposed to room air-
conditioners (RAC) central air conditioning (CAC) systems, which are defined as air conditioning systems
with more than 12kW of cooling capacity in the EU, are not bought or selected in a shop. They may be
selected by an installer of packaged units. They are usually designed by an AC engineer and the components
selected following the engineers’ recommendations.
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, which are relatively infrequent in Europe and
sized often under 12kW, a piece of equipment that we would call a RAC in Europe. European CAC systems
are commercial AC systems usually specified by engineers or technicians, who choose the system
technology without any direct influence from the customer, except for the specification of the desired
11
environment and other conditions such as maximum overall price, etc.. 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, like space heating, but also in the nature of comfort)
obtained.
Within the EECAC study twelve participants from eight countries including the EU manufacturers'
association, Eurovent, engaged in identifying the most suitable measures to improve the energy efficiency of
commercial chillers and AC systems. This study benefited from the co-operation between laboratories,
consultants and Eurovent, which was established during the conduct of the SAVE sponsored EERAC study
concerned with room air conditioners (EERAC 1999).
It was made easier by the existing information scheme by a subsidiary of the manufacturers’ association
called Eurovent-Certification. 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). To be really effective,
energy efficiency options have to be defined not on the basis of nominal operating conditions but at a variety
of part load conditions, which better reflects the CAC operating modes that occur in real use. 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.
Definitions of all CAC systems found on the EU market have been given.
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. The first choice determining a system is the type of the fluid
being centrally refrigerated and circulated. The most frequent (and really dominant option) is the use of a
chiller, 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. To transfer the
“cold” to air, Air Handling Units (AHU) are used. In the majority of situations however, chilled water is
circulated up to the rooms and the air of the room comes directly in contact with it. Even in this predominant
CAC system a choice must be made on how to transfer “cold” to the air of the room. 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.
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. In many cases large Unitary Air Conditioners
(or Packaged), which are self-contained direct-expansion (i.e. without using water as an intermediate heat
transfer vector) apparatus can be applied as can Multi-Split systems, which are a particular assembly of small
“split systems” and were originally investigated under the EERAC study. In addition a new variant of the
“split system” concept most commonly known as the VRF (Variable Refrigerant Flow) system, but more
generally as a modulated capacity system, is capable of significant energy savings and has occupied a market
segment. 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.
12
Figure 1. 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.H.U.s and INDUCTION FAN-COIL

DUCTWOR UNITS UNITS
2 pipes 3 pipes 4 pipes
MULTI- 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.
Eurovent-Certification, a branch of Eurovent has defined test conditions at which equipment energy
performance is to be reported by European industry, based upon performance testing at full load, in
accordance with CEN standards. The American Refrigeration Institute (ARI) and American Society of
Heating, Refrigerating and Air Conditioning Engineers (ASHRAE) have defined US-national test standards
in a similar way; 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. This has enabled US
legislative bodies to readily establish minimum performance criteria which are based upon comparative
performance under representative operating conditions. Thus in some way, the US standards have progressed
futher than the European ones, although it has been established in this study that they are not suitable for use
in European conditions without modification. There are ISO efforts for the testing aspect but the respective
standards are not all available.
CAC market and stock data have been assembled for the first time.
National surveys of the CAC market, usage and regulatory environment were conducted by the EECCAC
study participants for their country. 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. It received a very significant help from the manufacturers
13
associations. The resulting set of country reports for: Austria, France, Germany, Greece, Italy, Portugal,
Spain, United Kingdom (with special thanks to the BRE) provides a unique set of data at the national level.
The CAC market is expanding rapidly in Europe, as can be seen from the additional cooled building floor-
area installed from 1980 to 2000 for the EU-15, Figure 2 (including new systems and refurbishment).
Figure 2. Annual addition of building cooled-floor area by CAC in the EU (either really added or replaced)
AREA COOLED EU-15 added (or replaced.) m2

(YEARLY MARKET)
120,00
100,00
80,00
M
m2 60,00
40,00
20,00
0,00
1975 1980 1985 1990 1995 2000 2005
Because of a strongly differential growth rate across EU Member States, the relative share of the total EU
cooled floor-area of countries such as France or Germany, which was large in the 1980’s has become small
in the 1990’s. 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.
Figure 3. 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, have been determined in all usage sectors and for all years
between 1990 and 2020, see Figure 4 for example based on year 1998.
14
Figure 4. 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, the USA, from the CBECS
programme of the US Department of Energy’s Energy Information Administration. The figures cover the
same years (1999-2000) and the same type of building stock (non residential buildings in use); however, the
choice of AC equipment is very different. Packaged AC accounts for the majority of cooled floor area in the
USA while chiller based CAC systems dominate in Europe, Figure 5.
Figure 5a. 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. The share of cooled-floor area by AC type in non-residential buildings in the EU for 1999-2000
15
E UR (E E CCA C)
c hillers
pack ages
all RA C
Despite this difference in technical preferences, 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.
The present Energy Efficiency efforts have been reviewed

In the EU, the energy efficiency of the AC system is not presently a criterion that plays any major role in
the AC design and installation process; rather the efficiency improvements that do occur tend to
happen haphazardly. Through an analysis of data on chiller energy performance at full load test conditions,
which is taken from the Eurovent directory, the distribution of chiller EER1 as a function of cooling capacity
and condensation cooling fluid has been determined, Figure 6. From this data it is clear that there is no
statistical significant dependence of chiller efficiency on the chiller cooling capacity, however, units which
use water as the condensing medium are significantly more efficient than those that use air. In fact this
apparent difference is not internal to the chiller, but rather represents the temperature regime of cooling
towers, for which an arbitrary estimate is made in the standard.
Figure 6. Chiller energy efficiency (EER) at full load as a function of cooling capacity for chillers available on
the EU market in 1999. There are two groups of chillers, with distinct testing conditions (water cooled and air
cooled, that cannot be compared)
4.5
4.0
2
R = 0.0073
3.5
3.0
2.5
2
R = 0.0003
EER
2.0
1.5
1.0
air cooled
water cooled
0.5
Regression (air cooled)
Regression (water cooled)
0.0
0 100 200 300 400 500 600 700 800 900 1000
Capacity kW
The average EER is indeed 3.57 with water whereas it is 2.52 for the systems with air as a rejection
medium under the conditions of the testing Standard, see Table 1. One could conclude in favour of
a clear superiority of Water cooled systems over Air cooled systems. Nevertheless, water-cooled
systems are expensive (for both when a cooling tower or natural ground water is used), and will
1
Energy efficiency ratio, which is measured under full load conditions
16
therefore only be common among large capacity systems. The operating conditions on the field may
be very different from the testing conditions and reverse premature conclusions.
17
Table 1. 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 1.9 2.53 3.29
reversible air conditioning 1.9 2.48 2.96
Floor 3.31 3.34 3.39
Cooling only water conditioning 2.9 3.73 4.09
reversible water conditioning 2.9 3.57 4.09
Condenserless Cooling only water conditioning 2.76 3.21 3.69
Interestingly, the reversible systems, which have an average EER of 2.48 W/W, have an almost identical
energy efficiency to the cooling-only systems which have an average EER of 2.53 W/W.
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, including: the USA, Canada, Mexico, Korea,
Japan, Australia and New Zealand. Most policy measures have been aimed at packaged units, which are not
so important on the European market; however, the US IPLV approach, which has been found to not be
directly applicable in the EU, was -before this study and the AICARR’s EMPE proposal, the only attempt to
address the specific issues of chiller energy performance.
18
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, allowing a fair comparison of equipment despite of
testing conditions and technical differences. Statistics show for each segment (like figure 7 for the largest
segment) which part of the market falls in each grade.
Figure 7 Air cooled chillers, Cooling, below 750kW , statistics with proposed grading and for each refrigerant
40,0%
35,0%
30,0%
25,0%
R407C
R134a
20,0%
R22
HFC
15,0%
10,0%
5,0%
0,0%
A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)
Splits and Packages are grouped in one single category (Air-cooled air conditioners) for presentation of
results in table 2.
Table 2 Example of proposed grading in Cooling function -mixed statistics (splits and packages mixed) ;
present market average = 2.46
Class definition % on market Grade % with equal class

width
3.20 < EER 2% A 2%
3.20 > EER > 3.00 5% B 5%
3.00 > EER > 2.80 7% C 7%
2.80 > EER > 2.60 15% D 15%
2.60 > EER > 2.40 22% E 22%
2.40 > EER > 2.20 26% F 26%
2.20 > EER 23% G 11%
The impacts of BAU have been assessed
Projections of energy consumption have been made. The penetration of AC can be expressed in a variety of
standardised ways such as the unit cooled-area per inhabitant (in m2/hab.), Figure 8. The BAU definition is
the absence of large regulatory actions and of significant changes in consumers choice. We have estimated
the areas cooled in a way compatible both with manufacturers statistics (capacities, numbers of pieces) and
with national statistics (square meters cooled), table 3, while taking into account typical over sizing. We can
keep a few figures in mind : 1200 Mm2 cooled in year 2000 (3m2/inhabitant), 2200 Mm2 in 2010
(5m2/inhabitant), with a share of reversibility around 25%.
19
Figure 8. Average cooled-floor area per inhabitant for EU countries and the EU as a whole in 2000.
Total A/C in 2000
m2
/in
4
ha
bit
ant
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. Only trade and
offices really grow in relative terms and they may reach 70% of stock by 2020.
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 12,01 15,68 20,06 26,29 30,29 33,01 33,95
Mm² reverse 1,45 2,06 2,74 4,83 5,57 6,08 6,27
BE Mm2 cooling 4,03 8,98 20,36 32,41 42,77 52,09 54,29
Mm² reverse 0,84 1,84 4,03 6,46 8,43 10,24 10,73
DE Mm2 cooling 3,78 6,62 11,30 19,92 29,24 37,57 42,30
Mm² reverse 0,70 1,35 2,50 4,12 6,01 7,72 8,69
FI Mm2 cooling 15,88 24,06 36,43 43,28 47,28 50,19 50,99
Mm² reverse 1,35 2,28 3,71 7,49 8,21 8,74 8,89
FR Mm2 cooling 93,40 129,39 180,37 293,24 390,57 472,24 502,39
Mm² reverse 32,79 45,84 64,98 106,59 141,52 171,24 182,61
GE Mm2 cooling 34,07 66,29 127,64 216,74 298,51 365,63 400,13
Mm² reverse 4,88 9,54 18,81 30,61 41,65 51,09 56,23
GR Mm2 cooling 11,04 23,06 48,23 80,47 108,97 140,88 145,99
Mm² reverse 5,29 11,17 23,65 40,07 54,24 70,12 72,68
IR Mm2 cooling 5,03 6,81 9,37 13,84 17,07 19,39 20,37
Mm² reverse 0,75 1,08 1,78 2,30 2,83 3,22 3,41
IT Mm2 cooling 130,85 175,63 258,76 368,74 414,88 450,33 467,85
Mm² reverse 29,22 43,81 73,26 106,86 120,93 132,38 138,18
2
LU Mm cooling 0,25 0,43 0,87 1,34 1,76 2,07 2,20
Mm² reverse 0,07 0,10 0,17 0,26 0,35 0,40 0,43
NE Mm2 cooling 22,25 39,02 66,88 87,71 101,28 110,49 113,62
Mm² reverse 1,84 3,55 6,50 12,17 14,03 15,38 15,89
PO Mm2 cooling 8,46 12,51 18,73 34,84 52,08 68,41 78,27
Mm² reverse 4,67 7,27 11,25 18,47 27,53 36,11 41,31
SP Mm2 cooling 64,24 102,68 172,69 248,07 295,71 342,20 352,20
Mm² reverse 34,61 56,66 97,11 136,02 161,33 186,01 191,57
SW Mm2 cooling 38,41 53,26 69,38 78,17 83,23 87,28 88,21
Mm² reverse 4,08 6,14 8,74 14,90 15,88 16,68 16,92
UK Mm2 cooling 94,29 127,63 173,15 248,36 294,19 326,80 340,28
Mm² reverse 14,17 20,41 31,06 43,81 51,73 57,87 61,07
Total Mm² cooling 538,01 792,07 1214,23 1793,42 2207,83 2558,59 2693,04
Total Mm² reverse 136,71 213,10 350,28 534,96 660,23 773,29 814,88
20
Then we had to move from area statistics to energy use statistics. We computed the electricity consumption
of a square meter for AC depending on its location, its economic sector (typical use) and on the AC system.
In other words, 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.
Figure 9 Consumption of the 18 systems in three climates as simulated with DOE software
Reference office building unitary cooling consumptions
140,0
120,0
100,0
kWh/m2
80,0 London
Milan
60,0 Seville
40,0
20,0
0,0
S
F
ps
l
st
um
ts
op
t
n
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
ol
Ks
D
er
R
IL
co
oo
oo
e
ai
SP
rib
co
on
at
tri
H
ir
le
C
.(c
.(c
+
st
ity
+a
K&
w
t.(
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
w
ai
h
at
O
w
w
at
ir
r+
t
de
i
R
w
ith
LO
+a
w
de
ith
si
ai
de
w
d
si
w
ut
d
O
le
ith
ed
si
le
ut
d
TW
d
oo
le
ut
oo
w
O
le
l
oo
oo
rC
O
oo
C
le
C
C
rC
Ai
er
oo
er
er
at
Ai
at
at
W
W
r
W
Ai
CAC Systems
The three main sections of our BAU scenario predictions relate with : 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. Figure 10 shows the first two values (cooling and associated heating
consumption by technical type) for the BAU.
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
The tables 4, 5 and 6 give the main values (EUR15) for the three functions. Note that gas is accounted for as
a secondary energy, with the same value as electricity.
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 22,879 33,683 51,636 78,103 94,727 109,631 114,579

(Electricity only)
Heating function 51,598 74,442 111,084 164,517 203,330 236,765 250,844
Without REV.
With present REV. (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
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). Such impacts are not small, but limited if we compare them with other uses in buildings
(heating, home electronics, better lighting, etc.). Remember the figures given correspond to BAU, and that
there is no significant EE measure on that market. 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, packages), the cold generating plant (depending on its number of hours of operation, and climate),
and the full system, including distribution.
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. We introduce one by one the possible
improvements (better compressor, better evaporator, etc.) and we see how the price of the service rendered
(the kW of cooling capacity) varies. 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 11).
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 2,1 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 (assumed here 100 Euros/kW) as the present “worst
performer” 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 for its least LCC

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 SEER2 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 energy consumption. So it is interesting to define the “optimum” taking into account
consumption. The search for the optimum has been done in the same way as previously, through successive
additions, including part load options (Figure 12), with a 6% discount rate, electricity prices ranging from 6
2
Seasonal energy efficiency ratio, the energy efficiency ratio which reflects the real usage conditions of the equipnent
over the year
23
to 17 cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe), and equivalent
usage durations (at full load) taken as 400 or 800 hours/year .
Figure. 12 The annual cost of the service rendered by a chiller in terms of SEER
10
8
Total cost (Euros/m2)
ALCC17-800h
7
ALCC10-800h
ALCC17-400h
ALCC6-800h
ALCC10-400h
6
ALCC6-400h
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 the screw chiller considered is about 40% more efficient than the
present « bottom » of the market : it has an SEER between 3.00 and 3.50. It may correspond to a chiller with
a correct EER around 2.46 (enhanced evaporator and condenser, improved compressor) and a capacity split
between 3 or 4 compressors. For manufacturers, there are other ways to reach 3.25 SEER, less expensive, but
our objective was to find out if there is a margin for improvement.
Packaged units can also be improved a lot

For 26kWc packaged units, 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. The results of the
analysis taking these factors into account is shown in Figure 13 for the 26kWc unit, which is most
representative of the EU market. They show that the life cycle cost minimum occurs for large packaged units
with an EER of 3.22 W/W when a 6% real discount rate is applied. Although the overall life cycle cost per
kW are lower for the 52 kW unit the minimum still occurs for an EER of 3.22 W/W.
Figure 13 LCC curve of a 26kW package in Europe presented as an annual cost
€1,400
€1,200
€1,000
Life cycle cost (€/kWc
€800
€600
€400
€200
€0
2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
24
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. The designers need the whole range of AC solutions to cover the domain
of geometries and air quality requirements. 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, through successive additions,
including part load options (Figure 12), with a 6% discount rate, electricity prices ranging from 6 to 17
cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe), and equivalent usage
durations (at full load) taken as 400 or 800 hours/year. After sorting options and combinations, the optimal
trajectory of improvement of the annualised cost of ownership (ALCC) is given in figure 15.
Figure 15 Optimising with 6, 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
0,00% 20,00% 40,00% 60,00% 80,00% 100,00% 120,00%

% of reference
The optimum is very flat, 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.
25
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, the part capacity performance can be better than full load
at the same temperature. This results from the reduction in refrigerant flow –and consequent improvement in
heat transfer efficiency at part load. The compression ratio is decreased so that compressor isentropic
efficiency increases. There is much progress being made in the control of these part load phenomena. The
issue is : how to represent, certify and translate in a single figure those improvements. There was on the table
the original US-IPLV method and a European version called EMPE.
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. Further to this, 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, for instance- as can be shown by drawing the
loads in terms of outside temperatures).
The first remark in the Italian proposal is that the operating conditions are rather different from Southern
Europe conditions. And even, if Northern Europe countries may need air conditioning in summer, it cannot
be said that Italy would need air conditioning at 12.8°C as normal operating conditions. Therefore, AICARR
proposed a new energy index, named EMPE (Average Weighed Efficiency in Summer regime in Italian)
directly deriving from IPLV, with different energy weights and, in particular, with different temperatures at
the condenser inlet, more suitable for the European climate and requirements in the air conditioning field.
The AICARR proposal, EMPE was not based on a sufficiently large climatic and technical investigation. Its
strength (being very close to the existing US method, which aggregated many factors) was also its weakness.
We had the opportunity to go further by constructing a data base of EU chillers at part load, understanding
better part load, and proposing two separate methods, one for part load reporting and certification, the other
one for the computation of SEER.
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, due to the Us experience). The constraint
was to minimize the testing time while ensuring maximum precision, 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 new ESEER
method has been compared with the US-IPLV and EMPE proposal under both respects : time spent and
accuracy.
Original knowledge has been generated during the “Joint project” of EDF R&D facility and manufacturers
from Eurovent wanting to promote part load performance. The main tool used was actual testing of EU
equipment but a number of group meetings allowed to build a common thinking frame. The technical
description of the chillers tested follows on tables 7 and 8, split by condensation type.
Table 7. 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. 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
For all the tested chillers, some common testing points were made according to either the US-IPLV
or the EMPE conditions depending on the manufacturer will. For all chillers, 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, 3, 4 and 8,
only IPLV or EMPE points plus the CEN one were available. For the others as many testing points as
desirable have been obtained. In all circumstances a simple model has been used to draw the performance
maps from existing testing points.
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. It is so when there is one single compressor per chiller, or various identical circuits. A significant
market share of chillers have various compressors and a complex circuiting, leading to improved part load
performance. But a given part load regime has to be defined by the actual status of each piece of equipment.
For discrete stages chillers, it would be easier to describe performance at a given stage not at a given
percentage. For the very few continuously controlled chillers, fours stages can be defined in terms of input.
Since temperature and load can be tested independently and recombined, there is no need for combined
testing & weighting (like IPLV).
About certifying Part Load : what the manufacturers give to their customers is a « map » of performance,
not only values at the four arbitrary percentages and temperatures, plus the final Eurovent grading when it is
available, based on a SEER. The customer can rely on the Eurovent SEER computed from this map … or
compute its specific SEER for its specific demand. 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, as usual, and the
associated independent testing.
We arrived also at applicable conclusions on the way to report the SEER in the Eurovent directory. We
started from HSEER, the DOE reference that we generated. It is proven that each set of outside conditions
(for each sector, climate, type of chiller, type of secondary system) can be reduced to four or five external
conditions without loss of accuracy. The ESEER index proposed here is a set of 4 conditions given for
E.U. as a whole, but there can be as many similar indices as specific demands: sector, country, etc.
We have introduced a format for the description of the stages of a chiller, like in table 9 and following,
suitable for Eurovent specification. For each stage, the manufacturer has only to declare which piece of its
equipment is operating and to indicate CC , the cooling capacity and EP, the electric power absorbed. The
certifying body has only to check a few of the values, selected in the same conditions as usual. Note that this
procedure is in fact already used for some chillers with various speeds, namely “low noise” chillers with the
possibility of reduced fan speed.
Table 9 : Part load performance of water cooled scroll chiller N°4, 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
Circuit 2
EP (kW) 8,80 17,60 27,17 38,27
CC (kW) 37,50 78,00 112,50 150,00
EER 4,27 4,47 4,12 3,92
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.
27
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
1.2
1.1
1
N° 4
0.9
N° 1
N° 3
0.8
N° 6
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
given on figure 17 for the five water cooled tested units.
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
1.3
1.2
1.1
N° 5
1
N° 7
0.9 N° 8
N° 9
0.8
N° 2
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
The simulations leading to the reference values of SEER (HSEER)

Two buildings were simulated on computer, but buildings that do exist : an office and a commercial mall.
For each one, three climates have been simulated, adopting different envelope characteristics when moving
the building around Europe. The different systems identified in the stock and market study have been
simulated. CAC air and water distribution equipments have been simulated using the European average
efficiency values.
Hour after hour, the simulation uses then the characteristics of the real chillers modelled to compute the exact
yearly performance index : the HSEER (Hourly SEER), used then as a reference for other methods. At each
hour the outside enable to calculate all known stage capacities and respective electric powers, including the
high pressure control impact on each stage. Then the load is compared to each stage capacity. If the load is
lower than the smallest available capacity step, the cycling formula enables to calculate the electric power.
Otherwise, the weighting of electric power of each stage is found by the expression of the weighted average.
28
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, for all chillers, included single compressor units. However, it also shows that the
experimental uncertainty is quite high. It mainly comes from the uncertainty measurement on the
temperature difference at the evaporator. In order to simplify the application of the index, some rounding can
be done without modifying noticeably the ESEER figures obtained, largely under the experimental
uncertainty. A comparison of the conditions of the 3 available indexes is proposed Table 10 for air cooled
chillers.
Table 10. 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% 26.7 42 % 31.3 30 %
50 25 41% 18.3 45 % 27.5 40 %
25 19 23% 12.8 12 % 23.8 20 %
Temperatures of the ESEER are comprised between EMPE temperatures above and ARI temperature under.
ESEER weighting coefficients give more weight to the 25% point load than both index. For 50 and 75%,
coefficients are nearer to the EMPE index. The 100% coefficient is 3%, nearer from the IPLV one. A
comparison of the conditions of the 3 available indexes is proposed Table 11 for water cooled chillers.
Table 11. Comparison of the ESEER conditions with the EMPE and IPLV indexes for water cooled chillers
ESEER ARI EMPE
Part load ratio Temperatures (°C) Temperatures Temperatures
100 30 3% 29,4 1% 29.4 10%
75 26 33% 23,9 42% 26.9 30%
50 22 41% 18,3 45% 23.5 40%
25 18 23% 18,3 12% 21.9 20%
Temperatures of the ESEER are embedded by the EMPE ones above and ARI temperature beneath except
for the 25% point. The ESEER weighting coefficients give more weight to the 25% point load than both
index. For 50 and 75%, coefficient are nearer to the EMPE index. The ESEER 100% weighting coefficient is
nearer from the IPLV one.
EER alone is a poor selection tool

We shall compare now four classifications : according to EER, US-IPLV, EMPE, ESEER, using as a
reference the actual EU values obtained by simulation in the three locations and properly weighted. We take
the point of view of a user of the Eurovent certification system : by selecting a “better” chiller, am I really
selecting a better chiller? This is completely false if we base ourselves only on EER (figure 17).
29
Figure 17. Comparison of HSEER with EER on the tested chillers
HSEER versus EER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 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).
Figure 18 comparison of US-IPLV with HSEER for the tested chillers
HSEER versus IPLV
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
IPLV
30
Figure 19 comparison of EMPE with HSEER for the tested chillers
HSEER versus EMPE
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
EMPE
Based on similar assumptions, the two methods, 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; a piece of equipment graded better than an other one is
better or equivalent.
Figure 20 . comparison of ESEER with HSEER for the tested chillers

HSEER versus ESEER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
ESEER
Conclusion : the differences are relatively large between existing methods and reality, and not always in the
same direction. 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.
Energy efficiency options have been defined for each system configuration
Th possibilities are so numerous, and so system dependent that the process of “filtering” the most promising
was difficult. More than 20 basic systems and 100 variations were considered on a qualitative basis and then
"filtered". 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; their use is are not expanding; the literature
31
is already sufficient to enable the savings potentials and costs to be defined; or simulation is impossible. For
the others a detailed simulation has been organised.
High energy savings are possible, as well as significant CO2 emissions reductions, at a net negative cost,
such that all parties (manufacturers, consumers and utilities) would find a benefit in the deployment of
efficient CAC. 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; the
problem of certifying something built on site and only once; the lack of incentives for operators to optimise
efficiency; the absence or inadequacy of building codes addressing this new and rather complex equipment
segment, which is very difficult to model; different incentives caused by the separation between the plant
owner and the building renter, between the building renter and the CAC operator, etc. As a result the policy
measures to be proposed in the EECCAC study should not only address the offer of more efficient CAC
equipment, which in fact seems to be the smaller difficulty, 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.
The French measure is basically a minimum efficiency threshold above which AC systems can benefit
from EDF’s marketing and financial support. The Portuguese measures cover a number of sizing
obligations, and require the use of a central AC system with a number of energy saving features such as ’heat
recovery’, ’part load staging’, and ’free cooling’ over a certain cooling capacity threshold (usually 25kW).
Monitoring and maintenance are also included in the Portuguese RSECE the local regulation applicable to
CAC). The UK Market Transformation Programme is a comprehensive set of possible policy measures
including: obligations written into building codes (i.e. the imposition of limits on cooling demand);
information (through a national release of the Eurovent directory database and the later elimination of the
least efficient equipment) and voluntary "best practice " initiatives. The US ASHRAE 90.1 standard is a fully
integrated set of policy measures which include minimum energy performance thresholds, design guidelines
and specific system requirements. It is fully described in the report.
At the European level a draft Directive on Energy Efficiency of Buildings is currently under transposition.
This requires the calculation of building energy performance, which itself demands knowledge of Air
Conditioning (and other) system efficiencies. This will help the adoption of the best technical solution in new
buildings. Article 8 requires that central air-conditioning systems of greater than 12 kW cooling capacity
shall be regularly inspected. On the basis of this inspection, which shall include an assessment of the air-
conditioning efficiency and the sizing compared to the cooling requirements of the building, the competent
authorities shall provide advice to the users on the possible improvement or replacement of the air-
conditioning system and on alternative solutions. So there will be a movement towards improvement also in
the existing buildings.
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
part load is not taken into account. The policy measure associated is to ban some classes of equipment either
directly (Directive ) or by voluntary agreement. We can also expect that a certain number of years of
labelling and communication by energy agencies reaches the same point, nobody wanting to buy a « poor »
image equipment.
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); however part load is not taken into account in Eurovent
grading and so the corresponding improvement is not obtained. 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.
32
Scenario 3 BAT- 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). Part load IS taken into account in Eurovent grading and so the
corresponding improvement is obtained. 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). There is a reduction in cooling demand which is climate dependant and has been
expressed here by country and system.
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.
Scenario 6 BRITISH REGULATION ON AC – HEATING, 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; to know the
cost we have to evaluate the less costly options, which may be on either side, primary or secondary; national
values are different and have been derived from UK with corrections for DD and fitted to each country. The
policy instrument would be a strong and harmonised implementation of EPB directive. The less expensive
way of attaining the objective is the improvement of chillers. Starting from their present averages of EER
and SEER, this policy induces almost no extra cost for any stakeholder, and absolutely no cost provided it’s
applied to all manufacturers (and so that they all pass on the costs to the customer). To obtain this “free”
market transformation a prescriptive minimum should be applied to local manufacturers and importers at the
same time.
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. The
measures which can be considered to encourage the higher energy efficiency levels for new CAC equipment
are:
• Provision of information (labelling, grading, efficiency ratings)
• Removing less efficient models from the market (MEPS or voluntary agreements)
• Encouraging higher sales-weighed average efficiency levels through negotiated agreements (e.g.
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. (an EU equivalent to ASHRAE 90.1 and which like ASHRAE 90.1 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, b) can enable to compare of the relative benefits
of different system designs, c) can be used in building thermal regulations to demonstrate
compliance with requirements
33
• The development of EU benchmarks for CAC system efficiency expressed in terms of: building
function and size; occupancy and purpose; quality of comfort provision and climate (e.g. cooling and
heating degree days)
Further to this, 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.
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.
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).
34
35
1. INTRODUCTION
SAVE II is an EU programme to promote the rational use of energy within the European Community.
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 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). Since the EERAC study covered all types of AC of under 12 kW cooling capacity, 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.
The objectives of the study are:
• to estimate the electric power consumption of CAC,
• to estimate potential energy savings deriving from the use of more efficient CAC,
• to investigate ways in which these savings can be realised,
• to make appropriate recommendations, on the basis of a cost–benefit analysis.
The working party has been gathered and co-ordinated by Jérôme Adnot from Armines. The work has
been organised in tasks for which the best experts have been chosen as task leaders. A broad coverage of
energy agencies, utilities, 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.
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, with activities ranging from technological development to socio-
economic investigations.
• PW Consulting is a UK-based consultancy specialising in equipment energy-efficiency initiatives

and programmes around the world.
• INESTENE and –later- 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.
• The University of Athens, in particular the Group of Building Environmental Studies, is very active
in the field of solar cooling and energy conservation in buildings; the group carries out research,
specialised studies, application projects, education, and dissemination of information.
• Politecnico di Milano, and namely the Department for Energy studies is the main supporting
laboratory for the HVAC engineers gathered in AICARR,
• AICIA supports the research by ETSIS, the famous HVAC engineering school in Sevilla,
• UTCB is the Romanian Technical University for building sciences,
• ITF is an HVAC consultancy in the region of Chambery, France
36
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
and air quality. From an energy perspective, 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). In fact the desired comfort conditions may include thermal comfort (which is expressed in
terms of a mix of convective and radiative temperatures), humidity control and indoor air quality (IAQ),
which is usually obtained through ventilation, i.e. by the change of indoor air, and filtration components.
Figure 2.1 Essential quantities in the process of air-conditioning in summer, 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, or air quality, or humidity control) extract heat in
summer from inside the room (Pc), approximately equivalent to the value of the “load”, through the use
of electricity (Pe). Usually the heat rejected outside (Pr) has an energetic value equivalent to Pe + Pc.
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’. A similar index to the EER, the
coefficient of performance (COP), is applied to indicate the performance of reversible AC in the heating
mode. It is the ratio of the heat input into the conditioned space and the electric power consumed to
transfer it.
Main technologies for cold production

Evaporation of the liquid "refrigerant" (R22, R407C, R134a, etc.) creates the "cold" in the evaporator,
which subsequently absorbs heat from the refrigerated space. We shall describe the steps of the technology
and give the specific names for the chillers, the largest single equipment (figure 2.2). 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, or direct expansion evaporators consisting of a pack of finned tubes through
which the air is forced;
• liquid-cooled evaporators, 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.
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.
40
The centrifugal compressors will not be studied in details in this report. 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”.
A wide range of technologies are used to couple the compressor to the electric motor:
• open type or accessible compressors, presenting detacheable parts to access the compressor’s main
components and coupled to separate electric or thermal engines. They can be used with any
refrigerant but are generally employed in systems with medium to high cooling capacity.
• "Semi-hermetic" compressors that are similar to the open type compressors but have a common
casing with the electric motor; they are generally used for systems with medium cooling capacity.
Hermetic compressors, which have their body directly coupled to an electric motor cooled by the
refrigerant and enclosed in a totally sealed shell; these are generally used for systems with a small to
medium cooling capacity.
Figure 2.2 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. The
condenser technology depends primarily on the required application and the heat source. Condensers
used in CAC systems are divided into three categories:
• air-cooled condensers consisting of a finned tube heat exchanger (figure 2.3). The primary factor
which influences the performance of the condenser, is the outside air temperature.
• water-cooled condensers consisting of finned tubes with internal grooves to increase the heat
transfer surface area and the overall heat transfer coefficient. The temperature and flow rate of water
have the greatest influence on the condensing temperature. 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
They enable a high pressure ratio because of the absence of alternative compression. The compression ratios can vary
between 2 and 30. The turbine is called the impeller. If the fluid enters the impeller with a tangential component or
swirl, that would occur only at non-nominal or bad designed points, the speed of the refrigerant would be consequently
reduced as related to the speed of the impeller. The ratio of pressure producing work to kinetic energy output is known
as the impeller reaction and ranges from 0.4 to 0.7. That’s why, after the impeller, one can find a diffuser that ends
converting kinetic energy into pressure lift. The observed performance is around 6.00 in terms of EER and 5.7 in terms
of SEER, between 50 and 100 % better than the chillers of smaller size.
41
• Evaporative condensers, which are used in industrial applications and combine a condenser and a
cooling tower in a single apparatus.
Figure 2.3 Inside an air cooled chiller (Courtesy Airedale)
A water cooled chiller (figure 2.4) is generally used with a cooling tower. 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, which contain a conventional wet type tower in combination with an air-cooled heat
exchanger. They are especially used to reduce water vapour plumes and hence water consumption.
A wet cooling tower (which displays better energy performance) is more at risk of cultivating the
legionella bacillus and consumes water.
Figure 2.4 A water cooled chiller (Courtesy Carrier)
After condensation the refrigerant is expanded by an expansion valve, which is used to throttle the
refrigerant fluid back to the evaporator pressure and to control the refrigerant flow. Three systems are
used:
42
• expansion devices with a constant pressure difference.
• thermostatic expansion valves that are controlled via the superheating.
• electronic expansion valves that are also controlled via the superheating.
The consequences in terms of energy consumption of all these technical choices will be investigated in
the rest of the study.
CAC systems types based on distribution

The system is integrated by a number of interrelated elements controlling the total comfort, such as air
filtering, humidity treatment, central or local re-heat , etc.. In many cases the same system has to also take
care of the heating mode. Essential components of a CAC system are:
• Water chiller with an electrically driven vapour compressor,
• Air treatment central unit if we want to provide refrigerated air,
• Distribution structure including networks of fans, ducts and pumps for refrigerated air and water
circulation,
• Terminal room units for local air treatment (most frequent),
• Assembly of automatic controls to keep the requested conditions and general safety.
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, so as to create a common vocabulary and terminology in further EU
regulatory work on CAC.
Classification of the systems

Most large plants have to combine a number of systems, each of them addressing different parts of the space
having different loads, occupation scenarios, load levels. 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. Among
the many systems to consider (50 or so), some are obviously too complicated, 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.
The structure of a CAC system (and consequently its name) results from the accumulation of a number of
decisions on the essential components. The first choice determining a system is the choice of the fluid being
refrigerated centrally and circulated. The most frequent (and really dominant option) is the use of a chiller
that generates cold water (typically at 7°C), 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.
Even in this predominant CAC system a choice must be made regarding how to transfer “cold” to the air.
There are, for instance, “induction systems” or “fan-coil systems” and these can be used with a water
distribution network including two or three or four pipe assemblies.
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. In many cases large Unitary Air Conditioners (or Packages), which are
self-contained, direct-expansion (without water) apparatus can be applied as well as Multi Split systems, a
particular assembly of residential “split systems”, originally covered in the previous EERAC study. 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. These descriptions lead to the idea of a descriptive systems tree, of which
some significant branches are presented in Figure 2.5.
43
Figure 2.5. 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.H.U.s and INDUCTION FAN-COIL

DUCTWORK UNITS UNITS
2 pipes 3 pipes 4 pipes
MULTI- VRF
SPLIT
2.3. 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), fan coil units, air handling units, induction units, humidifiers
and de-humidifiers, balancing valves and dampers, controllers, etc.
The main devices which provide comfort are Fan Coil Units, which transfer heat from the air in the locally
cooled zone to a cold water circuit (Figure 2.6) and Air Handling Units which cool air more centrally before
its distribution and diffusion into the rooms (Figure 2.7).
Figure 2.6. Plan of a fan coil unit
44
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; the heat exchanger comprises one supply and one return pipe only
for refrigerated water;
• 2 pipe fan coil with change-over (2PR). 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. A
reversible chiller supplies either cooling or heating and changeover from one mode to the other is
centrally regulated according to season. This system cannot heat and cool simultaneously in two
different rooms and hence is used when the summer-winter transition is easily distinguishable;
• 2 pipe fan coil with electric heating (2PE). This system may be reversible or not, according to needs. 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. This system can heat or cool
simultaneously in two different rooms;
• 4 pipe fan coil (4P) with two coils frequently assembled together The same system can be used in all
zones, 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. A 3 pipes system as well as a 4 pipes system with
only one coil existed and disappeared.
There are also a wide-variety of air handling units (AHU –see figures 2.7 and 2.8) used for the remote
preparation of cold air.
Figure 2.7 Air Handling Unit with heating, cooling and variable air flow distribution
45
Return air
Outdoor air
variable flow boxes
Figure 2.8 The « coil » providing heating and cooling inside an AHU (Courtesy Trane)
Induction units (IU), less frequent nowadays, 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, Figure
2.9.
46
Figure 2.9 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, which may be: single duct or dual duct; use
fresh air only, or employ a heat recovery system; have terminal reheat or not; etc. Terminal reheating carried
out with an electric resistance heater or a gas furnace heater but never with a water coil. AHU systems can
also be distinguished between those that have a variable flow and those which have constant flow. The
EECCAC study has developed a specific terminology to describe this variety of systems.....
As mentioned before, systems using water chillers are the most frequent type of central air conditioning
system used in the EU.
A water chiller, comprising a group of equipment to refrigerate water, cools water, 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, it is useful to present all the possible systems which could
work with chillers in a synthetic table (2.1).
Table 2.1 Classification of central systems based on chillers
SYSTEM CLASSIFICATION SUMMARY
1. ALL AIR SYSTEMS

refrigeration: chiller (air or water cooled with/out cooling tower)
air treatment: central station type, air handling units
CONDITIONED AIR AIR AIR A.H.U.
ZONES DISTRIBUTION VOLUME TEMPERATURE CONFIGURATION
SINGLE-ZONE SINGLE DUCT FIXED VARIABLE CAV with terminal re-heat
with by-pass on re-
47
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 - cold air
second duct – hot air
DUAL CONDUIT High pressure systems
PRIMARY AIR FIXED VARIABLE
SECONDARY AIR VARIABLE CONSTANT
2. 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, induction units, radiant panels
48
Number of water loops connected with the chiller
The simplest solution is to have one single water loop in the conditioned space. It can be used for cooling
when connected to a water chiller and for heating, when connected to a boiler, the transition being called
“change over”. The same loop can be served by a reversible chiller provided refrigerated waterin cooling
mode or hot water in the heat pump mode. For more comfort, two water loops are installed (hot and cold)
with distinct generators, or alternatively to each side of the chiller (reversible heating and cooling). 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.
2.4 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. Although they are
similar to multi-split systems (a residential split system serving several rooms), they have not been accounted
for in the previous EERAC study due to their higher cooling capacity.
VRF systems are classified as built-on-site systems; 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.
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, VRF
systems are capable of being controlled to simultaneously satisfy different building zones requiring different
thermal conditions. They are able, in fact, to transfer “heat” and “cold” according to the local need with a
very low energy consumption. VRF units are available in three versions: cooling only, heat pump and heat
recovery.
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. 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. Although the use of a
central chiller and a central boiler is often also necessary, their sizing can be minimised. A more efficient
version of this system makes use of thermal storage and in some cases of ground water sources.
As for VRF systems, the installation needs to be adapted to the site. This system is particularly viable when
there are simultaneous cooling and heating needs in the building.
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, they don’t use water
pipes to distribute cold and have no, or very limited use of, air ducts. The two commonly used systems, Roof
Top Units and Close Control Cabinets, 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. They are
frequently used in super markets and in telephone central technical rooms although in other economies (i.e.
the USA) they are commonly used in much wider applications. Roof top units are always air cooled whereas
close control cabinets may be water or air cooled; in all cases they are self-contained units which are
completely assembled in a factory. “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.
Inclusion of RAC in the present study
49
Should we neglect RAC or include them here? To answer this question, figures about the market and level of
cooling provided by RACs drawn from the EERAC study have been utilised in the present study. RAC are
used for 78% in the economic sectors and for 22% in households. It is better to compare the shares of all
competing AC types in a single scenario for a given economic sector and country. This also helps to
understand the real basis of competition between RAC and CAC systems; and should thereby avoid
potentially distorting impacts from isolated policy actions.
Four types of air-cooled RAC are widely used:
• Split-packaged units; 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. The indoor unit includes the
evaporator and a fan, while the outdoor unit has a compressor and a condenser. There is a range of “large
split systems” over 12 kW of cooling capacity and therefore are usually classified separately from the
residential splits.
• Multi-split-packaged units; consisting of several indoor units (usually four or more) connected to one
outdoor unit. This family of AC equipment is partly under 12 kW of cooling capacity and partly over.
VRF (Variable Refrigerant Flow) systems can be considered as a version of multi-split systems but are
always over 12 kW in cooling capacity.
• Single-packaged units; are commonly known as ‘window’ or ‘wall’ RACs wherein one side of them is in
contact with the outdoor air for condensation, while the other provides direct cooling to indoor space by
means of an air circulation fan.
• Single-duct units; which are packaged AC appliances that are kept inside the room while cooling the
space, and reject hot air from the condenser to the exterior space through a duct.
Water-cooled units of any type under 12 kW cooling capacity were part of the previous EERAC study. The
water used in RAC could in principle be drawn from a natural water source, but this is seldom available; 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.
Summary of choices in terms of local versus central systems

A growing distance between the “centre” and the rooms increases losses and auxiliaries, leads to the choice
of a carrier but generates positive scale effects. Heating, ventilation and air conditioning systems can been
split in secondary (air-side) and primary (water side) systems. There is always an air side, but it can be
generated far from the room or close to it. To sum up, most system types can be classified according to the
air handling situation (central versus zonal) and cold source (hydronic versus package) criteria. This
classification based on the EE view is shown in table 2.2 .
Table 2.2 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. In particular, the potential problem of equipment over-sizing had to be treated explicitly. Since a
CAC system is designed and sized by a professional engineer who is usually concerned to minimise the
50
initial investment cost, we can expect that on average there is no over-sizing in this case. There are however
many examples of oversizing, which should be related to some specific factor (type of contractor, level of
expertise, time allocated for sizing, etc). We made the assumption of a fair and economic sizing. In the case
of RAC, 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, it is assumed that the
average European RAC is over-sized by a factor of 2 (i.e. 100% over sizing). 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). This value, being used twice
in opposed ways, has no influence on our statistics.
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, building, system type to allow national comparison. Note that the notion of conditioned area is
uncertain in national statistics : not the gross built area, not the strict area of activity rooms; conventions may
vary from one country to another, a fact that gives interest to our repeatable “standardised” area.
Free cooling
At some time during the year, outside air can be used directly to cool the space without any special thermal
treatment. There are control issues associated (flow rate, movement of dampers, nature of control : based on
temperature only or on enthalpy), etc.
2.5. Testing standards and performance standards

A branch of the trade association Eurovent has defined the set points for which performance is reported by
industry, thus allowing performance comparison at full load, in accordance with CEN standards. In the same
way the American ARI (American Refrigeration Institute) and ASHRAE (American Society of Heating,
Refrigerating and Air Conditioning Engineers) set US standards. Efforts have also been made by ISO to
define internationally applicable testing standards, but ISO standards are not currently available for all the
required equipment types.
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. The US test method is defined in
ARI 550/590-1998 ‘Water-Chilling Packages Using the Vapor Compression Cycle’. However the ARI
documents often mix pure testing standards, 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.
Thus in some ways, the US standards have progressed more than the European ones, although they are not
directly applicable in European conditions. 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, which is being developed by ISO TC 86, is not yet ready and is not
expected to be for some time.
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:
-- Cooling operation
Eurovent Certification (at ARI Standard 550/590 –

full load) 98
Leaving chiller water temperature 7°C 6.7°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) 29.4°C (85°F) (water
51
cooled)
Entering condenser air temperature 35°C (air cooled) 35°C (95°F) (air cooled)
-- 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. For this reason energy efficiency measurements at full load are directly comparable between
the two systems. The only difference is that the pumping power is “forgotten” in ARI values, which
overestimates by 1% air cooled chillers EER, and by 3% water cooled chillers EER. ISO harmonisation on
full load is easy. However the situation is not the same for part-load operation, which is the normal operating
condition for AC systems and the one where very significant energy efficiency improvements appear to be
possible.
ARI Standard 550/590 – 98 testing at part load conditions
The current European test standards do not include part-load ratings, whereas the chiller certification
programme operated by ARI in the USA does include part-load performance ratings. 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.
The weighing of part load points in ARI 550/590 are given in Table 2.3.
Table 2.3 Parameters used in the US IPLV
% load Air entering condenser Water entering Operating hours %

(DB) condenser
100 35 29.4 1
75 26.7 23.9 42
50 18.3 18.3 45
25 12.8 18.3 12
This load profile is the basis used to calculate the integrated part load value (IPLV), the seasonal average
efficiency of a chiller.
The Integrated Part Load Value is thus calculated using the following equation:
IPLV = 0.01A + 0.42B + 0.45C + 0.12D
Where A = EER at 100% of full load
52
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. Before our study a
first proposal by the Italian AICARR was made, called EMPE.
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 - Air handling units - Ratings and performance for units, components and sections’
prEN 1886: 1997 ‘Ventilation for buildings - Air handling units – Mechanical performance’
In the USA the ARI 430-1999 ‘Central station air-handling units’ test standard is used.
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, which must always be less than 3.96 l.s-1.m-2 at 400 Pa negative test
pressure for units that always operate under negative test pressure and should be less than 5.70 l.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. An additional test is conducted on the filter bypass leakage and the filter leakage
performance is classified according to the results.
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. Units
with a transmittance of less than 0.5 W.m-2.K-1 are classified as T1 while those with a transmittance of
greater than 2.0 W.m-2.K-1 are classified as T5. The T2 to T4 classes occur at intermediate values. 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.
Performance testing requirements are stipulated in prEN 13053 for fans, coils, heat recovery sections,
damper sections, mixing sections, humidifiers, filter sections and sound attenuation sections. 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’. In addition individual coils must be sealed within
the AHU casing such that the resulting air gaps do not exceed maximum permissible levels.
Three types of heat recovery section are recognised: recuperators, heat recovery sections with intermediate
heat transfer medium, and regenerators (heat recovery sections containing thermal accumulating mass). Heat
recovery sections are tested according to EN 305 ‘Heat exchangers. 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. Mixing sections are tested for air mixing efficiency.
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. An ISO test standard for air terminal units, known as ISO NP 17524, is under
development by ISO TC 86/SC 6.
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’.
An ISO Committee Draft on testing of ‘Air Conditioning Condensing Units’ has been produced as CD
13258. As of August 2001 a DIS was being prepared for ballot which would include refrigeration condensing
53
units. In the USA the following test standard is used: ARI 365 – 1987 ‘Commericial and industrial unitary air
conditioning condensing units’ .
ISO TC 86/SC 6/WG 8 has developed a proposed working draft PWD 16345 for ‘Cooling Towers – Testing
and Rating for Performance’. The Cooling Tower Institute in the USA has also issued a test standard to rate
the performance of Cooling Towers, CTI 201-1986 ‘Certification Standard for Commercial Water Cooling
Towers’.
A proposal for a better characterisation of AHU

An AHU is characterised by the following information:
1. Nominal flow, intended as the working flow recommended by the manufacturer expressed in m3/h.
2. Available static head, intended as the difference between the static pressure measured upstream and
downstream the AHU working at nominal flow without air recirculation.
3. Expected treatment typology: Heating; Cooling; Humidification Dehumidification
4. Filtration class, expressed according to the EN 779 standard.
5. Possibility of partial or/and total air recirculation
6. Presence of sensible or total heat recovery
7. Location of the AHU (indoor/outdoor)
8. Partial load efficiency, subdivided in:
a. Partial load exchange efficiency
b. Partial load air efficiency
9. Nominal power consumption, expressed in kW.
The listed information is necessary and sufficient to exhaustively define the functions and performances of
an AHU. A proposal of a comparative method has been made. 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. 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. The conventional nominal values would be reported in tables and
settled on the basis of criterion that allow for economics and technical factors. The approach will be tested as
a possible UNI (Italian) standard.
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.
The nominal efficiency value of an air diffusion system depends on the following factors:
1. Diffuser typology,
2. Diffuser arrangement,
3. Supply temperature.
These three factors are combined together and they can not be considered independently from the
performances standpoint. Yet, there are some directions, published by each manufacturer, which recommend
the right installation conditions and the maximum operating temperatures at nominal load.
54
In particular, two different categories of air diffusion system can be identified:
- Mixing system with the air introduced above the occupied zone.
- Displacement systems or mixing systems with the air introduced below, or inside, the occupied
conventional space.
The following tables 2.4 and 2.5 indicate the ventilation efficiency values for each system category, in the
hypothesis that every system is projected, realized and installed in conformity with the manufacturer
directions and that we remain at nominal conditions.
Table 2.4 – ventilation efficiency for mixing system with the air introduced above the occupied zone.
Diffuser typology ev ev
Dt < 0°C Dt ≥ °C
Helicoidal effect diffuser 1,00 1,00
Cones diffuser 0,90 0,75
Linear slot diffuser 0,75 0,60
Outlet with single or double fin rows 0,70 0,60
Table 2.5: ventilation efficiency for Displacer systems or mixing systems with the air introduced below, or
inside, the occupied conventional space.
Diffuser typology ev ev
Dt < 0°C Dt ≥ 0 °C
Floor helicoidal diffuser 1.2 1.1
Underseat diffuser or similar 1.3 1.3
Displacement diffuser 1.3 0.8
The efficiency value characterizes the share of the flow rate introduced that actually reaches the occupied
zone, contributing to cut the heat loads and the air pollutants down. There is much to gain by this approach,
embedded into a UNI standard but not yet quantified here.
Testing and performance setting for packaged systems

ISO 5151-1994(E) – 'Non-ducted air conditioners and heat pumps -- Testing and rating for performance' is
applicable to all packaged air conditioners. An equivalent standard exists for ducted equipment: ISO 13253-
1995 – Ducted air conditioners. 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. A new version is being prepared.
The ISO test procedure applies to packaged air conditioners of any capacity and type provided they are non-
ducted including cooling-only and reversible, single-phase and three-phase, and air-cooled or water-cooled
units. Testing for these systems was discussed in the previous EERAC study. 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. The standard
test conditions for the cooling capacity test are shown in Table 2.6. The test conditions are always at full-
55
load and with a single set of stable environmental conditions, thus the part-load performance of variable or
multiple speed drive units is not reflected.
56
Table 2.6: Test conditions for the determination of cooling capacity, 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.
2) Representative of equipment working with cooling towers. For equipment designed for other uses, 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, 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, Canada and Mexico have adopted the US approach. 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.
Central air conditioners in the USA include both ducted systems and ductless split systems (i.e. split-
packaged units) although ducted systems are predominant on the market. Small ducted systems are either
split systems or single package systems, but mostly the latter. The US efficiency regulations classify a
central air conditioner depending on its cooling (or heating capacity). 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. All ‘commercial’ units with capacities below 135 000 Btu/h (40
kW) are classed as ‘small’. Small commercial units with capacities between 65000 Btu/h (19.05 kW) and
135000 Btu/h (40 kW) are tested in the same way as large systems but have different efficiency
requirements. 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. All
central air conditioners, with capacities up to 65000 Btu/h, including split-packaged systems, are rated using
a seasonal energy efficiency ratio, SEER, that is based on the amalgamated results of testing cooling capacity
at four different test conditions. 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.
57
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, HSPF. The commercial (small and large) air conditioners and
heat pumps can be air, water or evaporatively cooled.
For small-commercial air conditioners and heat pumps with capacities above 65000 Btu/h but below 135
000Btu/h, the cooling performance is regulated for the EER (static test conditions) and Integrated Part-Load
Value (IPLV), which is a measure designed to reflect performance under part-load conditions. The heating
performance is measured and regulated using a static COP test.
For large commercial air conditioners and heat pumps, defined as those with capacities above 135 000 Btu/h
(40 kW), 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, an air cooled heat pump or a water/evaporatively cooled
air conditioner. Efficiency is measured and regulated in terms of the EER, COP and IPLV.
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. Both the
fixed and variable-speed air conditioners (either room or unitary) must satisfy MEPS (Minimum Efficiency
Performance Standards), energy labelling and are also subject to aspirational efficiency targets. The variable
speed units are tested and rated using a seasonal energy efficiency ratio (SEER). Reversible units are not
subject to COP requirements but are required to satisfy the cooling-mode performance requirements.
In Japan a central air conditioner would be classified as a unitary air conditioner. Larger room air
conditioners, of a packaged type, are classified as unitary air conditioners and are subject to energy
efficiency targets, not MEPS and labelling. The existing targets differentiate depending on whether an
appliance is integral (windows) or split-type and whether it is cooling-only or reversible. The targets for
reversible units are a combination of EER and COP targets. Multi-split systems are currently excluded but
are about to be included in new energy efficiency target regulations, due to come into effect between 2004
and 2007. 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.
2.6. 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, 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. The performance of the equipment at part load is significantly different from what it is at full
load. In many situations there is a possibility to take advantage of “free cooling”, which is available
whenever the outside air is cooler than inside air, furthermore there are plenty of ways of recovering
cold from some parts of the system in order to cool other parts. Free cooling increases the time during
which the building is only ventilated, but it’s not related with the ventilation function but with cooling.
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; 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.
Definitions
To facilitate understanding AC system energy performance, a common vocabulary has been defined based
upon the specification of seasonal or annual quantities. In establishing this balance, you separate the
auxiliaries according to the function being performed, that we consider being only two : heating with
ventilation, cooling with ventilation.
To apply this strict definition, we should come back to each time step and make some sophisticated
computation on the effects of the auxiliaries, mostly the ventilation auxiliaries at that time : do they act in the
58
direction of heating or in the direction of cooling? We have done something far simpler : when there is
heating only or cooling only, ventilation is integrated in the demand ; 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.
The quantities of interest are :
1. SCL: the Summer Cooling Loads or "cooling energy"; it is assumed that the total summer cooling load,
including the energy for cooling and humidity treatment, is completely satisfied by the AC system. The
SCL takes into account zone loads, outdoor air load, the heating of air that is passed through fans and the
real system operating schedule and thermostatic control. It is also sometimes called the ”coil cooling
energy” (kWh) or ”coil cooling load” (kW).
2. SEC: the Summer Electricity Consumption for cooling; SEC = the electricity consumption of the
compressor of the cooling equipment (e.g. the chiller, package, etc).
3. SSEC: the System Summer Electric Consumption for cooling; SSEC = electricity consumption of the
whole system i.e. that of the: Fans + Pumps + Primary equipment.
4. WHL: “Winter Heat Loads" represent the heat demand of the building , the equivalent of SCL in winter
season.
5. 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); reversible heating has some
limitations and not all the load can be satisfied by the cooling equipment.
6. WED: the Winter Electricity Demand (Consumption) for reversible use of air-conditioning, including all
auxiliaries.
7. WEC: the Summer Electricity Consumption for reversible cooling; WEC = the electricity consumption
of the compressor of the cooling equipment used to heat (e.g. the chiller, package, etc).
8. SEER: Seasonal energy efficiency ratio during the summer: the ratio of SCL to SEC. This parameter
could be miscalculated due to the large preponderance of electricity consumption by auxiliary
applications in most systems; the same may be said about the SCOP (the seasonal coefficient of
performance during the winter for reversible systems;
9. SSEER: System Seasonal EER for summer: the ratio of SCL to SSEC, the real index of performance;
10. SCOP the Seasonal COP for winter: the ratio of WHG to WEC.
11. SSCOP the System Seasonal COP for winter: the ratio of WHL to WED, the equivalent of SSEER for
winter.
This study is concerned with identifying beneficial means to increase the SEER, SCOP, SSEER, SSCOP and
partly WHL. 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.e. addressing the family of “free
cooling” options). 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; however, this is not the subject of
the present study.
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. 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 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.
59
Some countries, such as Switzerland, 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; however, this is made possible mainly because of the milder
climate. In some other countries, such as France and Portugal, 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. The Mediterranean countries, because of their geographical position, have to rely in
many cases on air conditioning to attain desired comfort levels in the existing building stock.
Full system efficiency

The total efficiency value results from a combination of all the system components efficiencies at different
load conditions, with an evaluation of the weight that each part load condition takes in comparison to the
whole yearly functioning cycle. We need a computer program to do this accurately. The simplified IPLV
approach, effective for the chiller, could also be extended to the whole plant, by evaluating with similar
procedure air water system and terminals, and the hot/cold water generators.
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. Simulation results, expressed by
airflow, temperature and humidity trend graphs, can be analysed to directly get the weights associated to the
different part load conditions. Thanks to the simulation the weights related to the partial load working
conditions for the space of a year are obtained.
2.7. Statistical databases used and information gathered

The conclusions of this study are based upon information gathered from four technical and market
databases, 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.
National surveys
The national surveys of the CAC market, usage and regulatory environment were conducted by the
EECCAC study participants for their country. 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. The resulting set of country reports for: Austria,
France, Germany, Greece, Italy, Portugal, Spain, United Kingdom (with special thanks to the BRE) provides
a unique set of data at the national level.
Data from manufacturers associations

Here are some details on the four databases used in the EECCAC study.
60
Manufacturers EUROVENT
directories Directories
National EUROVENT
confidential confidential
sales data sales data
Two database are permanently maintained by Eurovent at EU level. The first one is public: the directory of
certified products. The second one ("sales"), which is a data base on total units sold without reference to
performance, is confidential although some segments have been made available for use in this study, under
the condition that any data pertaining to a specific manufacturer is strictly anonymous. 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. 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, with the additional benefit of giving an idea of public
prices, at least in relative terms hence allow an indication of the relationship between cost and efficiency to
be established. 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.
Correction and treatment of data

This being said, the Eurovent market data provide an extraordinary means to reconstitute CAC stocks, both
in terms of numbers and cooling power. The data is generally based on the year 1998, except the particular
case of the AHUs which were for 1999. The Eurovent market figures need to be corrected for the following
three reasons:
• Eurovent does not incorporate all the market, 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?).
We used some estimated ratios to correct for these issues when no better data were available. Obviously real
national data from country reports have replaced these ratios whenever they became available from the
countries with a national participant. Checks on the consistency of various sources of information have found
them to be reasonably high (i.e. with discrepancies of a few tens of percent only). 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. Projections on the size of the stock going back in
time were made using rates based on measured data from 1996 to 2000, and estimated data from prior to
1996.
It should be stressed that all percentages are based on weighed statistics, with no figure being an arithmetic
average. The statistics are intended to give the right weight to the country, climate, type of AC system, etc.
61
3. MAIN FIGURES OF AIR-CONDITIONING IN EUROPE
3.1. The demand for AC in Europe

A general growth
The CAC market is expanding rapidly in Europe, as shown in Figure 3.1.
Figure 3.1. Apparent annual additional building floor area conditioned by CAC from 1980 to 2000, for the EU
(apparent means inclusive of additions and replacements)
EU-15 added (or repl.) m2
120,00
100,00
80,00
Mm2
60,00
40,00
20,00
0,00
1975 1980 1985 1990 1995 2000 2005
Source: EECCAC; 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. Economic growth in the South is resulting in AC levels rising in regions
where tertiary sector is important. In fact a number of central European countries (Belgium, Germany, etc. )
have experienced larger rates of growth in AC than some more Southern countries such as Portugal or France
(Figure 3.2). The figures are given here for the total market, including RACs, which are also mostly used in
the workplace), and also by country
Figure 3.2. Apparent annual additional building floor area conditioned by CAC from 1980 to 2000, by EU
Member State (apparent means additions and replacements)
30,00
25,00 Italy
M
m
Spain
2
ad 20,00
de
d
or
re
pl
ac 15,00
ed Germany
France
10,00
Others
UK
5,00 Greece
Portugal
0,00
1975 1980 1985 1990 1995 2000 2005
Source: EECCAC from Eurovent Experts and country reports
62
As a result of different growth, the relative weight of some countries like France or Germany as a proportion
of total installed AC within the EU, which was large in the 1980s has become small in the 1990s. Today just
two countries, Spain and Italy, account for more than 50% of the entire EU market Figure 3.3.
Figure 3.3. Apparent additional building floor area conditioned by CAC in 1998, 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. Figure 3.4 shows the share of
conditioned floor area by type of tertiary activity and country for CAC systems alone.
Figure 3.4. 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, (Figure 3.5).
63
Figure 3.5. Share of conditioned floor space by building type for each AC system type across the EU
100,00%
90,00%
80,00%
70,00%
60,00% education
50,00% houses
40,00%
30,00% trade
20,00% offices
10,00%
0,00% hotels&bar
Rooftops
RAC
VRF
chillers
packages&splits
hospitals
This market is centred on offices and trade. 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. The
only exception are VRF type systems, maybe due to their flexibility of use and installation, corresponding to
hotels, bar and existing medium office buildings.
3.2. 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 3.6 (these
fugures exclude room air conditioners).
Figure 3.6. 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 3.7 indicates which systems and segments are experiencing the largest growth (from 1996 to 2000).
64
Figure 3.7. Average annual rate of growth in conditioned floor area by type of CAC for the period 1996-2000
0,16
0,14
0,12
0,1
0,08
0,06
0,04
0,02
0
Large Chillers Packages Rooftops VRV Small
Splits +8,5% +2,5% +9% +13% A/C
+14% +10,5%
The average growth rate for large splits of 14%, for VRF of 13% and small AC for 10.5% are very different
from the overall average growth rate of 9%. The competition is focused on the "new" segment of smaller
buildings (trade, small offices, etc), which have correspondingly smaller average loads. Figure 3.8 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. 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". We see a growing competition of RAC against chiller based solutions
and an adaptation of solutions for the treatment of smaller sites.
Figures 3.8. The percentage of AC supplied by each AC type by user sector for the EU in 1998
120,00%
100,00%
80,00%
VRF
60,00%
Rooftops
40,00% packages&splits
chillers
20,00%
RAC
0,00%
offices
trade
education
hotels&bar
houses
hospitals
Figure 3.9 shows the growth in conditioned floor area by each type of CAC&RAC system from 1980 to 2000
across the EU.
65
Figure 3.9. 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, 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, Figure 3.10.
Figure 3.10.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, 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, which is the world’s largest. The US and EU figures cover the same years
(1999-2000) and the same building stock (non-residential buildings in use); however, the preferred technical
solutions are very different, with packages dominating in the US, and central chillers dominating in Europe,
as shown by comparison of the data in Figures 3.11 and 3.12.
66
Figures 3.11. 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 3.13.
67
Figure 3.13. 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)
3.3. 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, it is necessary to be able to determine the
proportion of conditioned floor area which: is treated by reversible AC systems; uses water distribution
systems; uses air distribution systems, and to have data on the growth rates of each. It is necessary to have an
image of the global industry and the main stakeholders.
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, i.e. the installed area of
installations with water distribution divided by the total area of installed AC. In fact this value is equivalent
to the installed area of installations with FCU divided by the total area of installed AC. Figure 3.14 shows the
large variation in the share of chillers using water distribution systems by EU country.
Figure 3.14. 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
m
ga
an
nd
ai
nc
ec
Ita
do
...
st
Sp
rtu
m
a
re
la
m
Au
ng
Fr
er
er
Po
G
iu
ki
h
G
lg
U
et
Be
68
Reversible use of Air Conditioning
One important aspect of this study is the reversible use of the cooling equipment for heating, Figure 3.15. On
a packaged or split product, 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. The statistics on chiller reversibility are derived from
data on a number of system sub-types:
water cooled chillers including water-to-water heat pumps
air-cooled chillers including condenserless water-cooled systems
air-to-water heat pumps with reverse cycle
water-to-air heat pumps with reverse cycle on a water loop
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. It is further assumed that the pure air cooled and the centrifugal chillers are none
reversible.
Figure 3.15. 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 3.16.
69
Figure 3.16. Market shares of AC technical solutions in four European countries (based on installed
conditioned floor area in non-residential buildings) in 1998.
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. Packages
have a comparatively large market share in Spain as do VRV systems in the UK. The average size (cooling
capacity) of chillers is smaller in Italy and the other Mediterranean countries, 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 3.17.
Figure 3.17. Average cooling capacity of chillers in four EU countries, 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). As a result, 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. Competition between “local” systems, VRV and mini-chillers for the medium-size building
market is the dominant issue for the future.
70
Figure 3.18. Growth rates are high in countries with both small and large systems and smaller in countries
with only large hydronic systems (UK)
180,00
160,00
Index relative to 1996
140,00 France
120,00
100,00 UK
80,00
60,00 Italy
40,00
20,00
0,00
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, it appears that many manufacturers operate on an EU-
wide level. The largest are usually foreign owned companies, resulting from the fact that a number of the
countries where they originate have a large and mature internal market (e.g. Japan and the USA) which
results in a transfer of technology and experience to their European branches. This does not mean that these
local companies of foreign corporations have no technical autonomy, but it partly explains the operation of
the market.
The European CAC equipment industry is self-sufficient within Europe and is fairly concentrated although
less than the car industry. In terms of market share the manufacturers can be categorised into three broad
groups, Table 3.1.
Table 3.1. 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, their centres of
development in Europe constitute a rather autonomous technical base of the European AC industry, which is
71
the third largest in the world after the USA and Asia. This base, joined to the existence of average-sized
manufacturers which are 100% European, ensures a great autonomy of supply for Europe. It results in the
independence and equality of the European manufacturers’ association, Eurovent, compared with its
American equivalent, ARI.
Table 3.2 shows the estimated value of the European CAC market without taking into account the imported
contents of the equipment (i.e. without considering the value of components imported from outside Europe
compared with those produced in Europe).
Table 3.2. 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% 6.8 0.25 1700
&CT
splits 90% 5% 85% 1.1 0.2 200
> 12kW
Packages 90% 10% 80% 0.9 0.2 200
rooftops 95% 5% 90% 0.7 0.15 100
VRF 50% 10% 40% 1.25 1 250
FCU/AHU 100% 0% 100% 6.8 0.1 650
RAC 75% 0% 75% 5.0 0.2 1000
Total 87.7% - 84.2% 21.6 - 4100
weighted
in income
These figures do not include the value of installation but do include the profit margin of the equipment
suppliers. The US and Japanese markets are worth about 10 000 MEuros per annum on the same basis.
Other stakeholders
Installers, designers and operators all have to adapt to the customer demands. They have to display a
competitive cost, or be able to guarantee a high reliability (better servicing, better contracts) in order to
compete. 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.
Utilities are important stakeholders. Summer peaking may be a problem for some Southern European utilities
but is often seen as a market opportunity for Northern European utilities.
Governmental agencies and ministries are responsible for the development of building codes. Thermal
insulation, which is often introduced into building codes to limit heating requirements, very often also helps
lower cooling needs; however, in some cases increased insulation can aggravate summer discomfort and
increase the need for AC.
Building thermal regulations usually aim to minimise AC energy demand but often "don't know how” it
should be done. There is a hesitation between a pure prescription on some elements (an obligation of means)
and a global limitation of demand, leaving the designer free to choose the elements and to assemble them to
reach the target (the obligation of results). The problem arises from the lack of energy consumption
calculation methods that are applicable to a wide range of systems. European countries cope with this
problem in different ways, but nobody appears to be happy with their current regulations.
Extension to EU accession states of the CAC market is already a reality. 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
Romania. This has given some insights into how the findings may be applicable in the rest of the CEEC. 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.
3.4. 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. The Directory is meant as an instrument to direct the buyers by giving certified
performance information. In a first moment information was limited to electric power and cooling capacity.
Now EER and COP are highlighted to promote Energy Efficiency. The intention is to go even further by
making use of part load information for a more appropriate selection of products. Note that in terms of
chillers the directory is limited to 750 kW capacity which practically leaves uncovered the centrifugal
chillers only, however this type is sold always on specific request. 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, either for chillers or packaged air conditioners. However, 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.
Using data drawn from the Eurovent directory as well as a few individual manufacturers product directories,
a complete data-set of CAC capacity and nominal energy performance (at full load) has been assembled and
analysed for chillers under 750 kW. Over this capacity the companies are very few and we have used directly
data provided by some of them.
EER as a function of capacity and cooling medium for a chiller under 750 kW
Figure 3.19 shows the EER as a function of capacity for chillers on the EU market according to their mode of
condensation.
Figure 3.19. Chiller EER as a function of cooling capacity for 1998. There are two groups of chillers, with
distinct testing conditions (water cooled and air cooled, that cannot be compared)
4.5
4.0
R2 = 0.0073
3.5
3.0
2.5
R2 = 0.0003
EER
2.0
1.5
1.0
air cooled
water cooled
0.5
Linéaire (air cooled)
Linéaire (water cooled)
0.0
0 100 200 300 400 500 600 700 800 900 1000
Puissance frigorifique kW
Source: Eurovent directory
73
Tables 3.3 and 3.4 indicate the average and range of EER values for CAC systems found on the European
market.
Table 3.3. Average and extreme EER values for chillers on the EU market, split according to Eurovent
internal categories, for year 1998
Categories Cooling Capacity CC Number of EER EER EER

in kW models min ave max
Packaged, cooling only, air ≤50kW 174 1.9 2.55 3.3
cooled, conditioning 50kW<CC≤100kW 102 1.93 2.49 3.04
100kW<CC≤300kW 99 2.12 2.53 2.85
300kW<CC≤500kW 4 2.52 2.56 2.59
>500kW 6 2.41 2.57 2.66
Packaged, reverse cycle, air ≤50kW 105 1.9 2.48 2.96
cooled, conditioning 50kW<CC≤100kW 35 1.99 2.43 2.84
100kW<CC≤300kW 14 2.1 2.49 2.73
300kW<CC≤500kW 5 2.56 2.73 2.86
Packaged, reverse cycle, air ≤50kW 6 3.31 3.34 3.39
cooled, floor cooling/heating
Packaged, cooling only, ≤50kW 8 3.31 3.75 4.06
water cooled, conditioning 50kW<CC≤100kW 10 3.55 3.77 3.96
100kW<CC≤300kW 31 2.9 3.72 4.05
300kW<CC≤500kW 20 3.16 3.79 4.04
>500kW 15 2.9 3.62 4.09
Packaged, reverse cycle, ≤50kW 8 2.99 3.28 3.5
water cooled, conditioning 50kW<CC≤100kW 5 2.9 3.18 3.5
100kW<CC≤300kW 3 2.9 3.45 3.8
300kW<CC≤500kW 5 3.85 3.94 3.98
>500kW 7 3.84 3.98 4.09
Remote condenser, cooling ≤50kW 6 3.13 3.32 3.53
only, water cooled, 50kW<CC≤100kW 3 3.16 3.2 3.25
conditioning 100kW<CC≤300kW 14 2.96 3.27 3.7
300kW<CC≤500kW 6 2.87 3.18 3.46
>500kW 7 2.76 3.03 3.29
Source: Eurovent statistics
Table 3.4. 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 1.9 2.53 3.29
reversible air conditioning 1.9 2.48 2.96
Floor 3.31 3.34 3.39
cooling water conditioning 2.9 3.73 4.09
reversible water conditioning 2.9 3.57 4.09
Condenserless cooling water conditioning 2.76 3.21 3.69
Statistically there is no relationship between chiller EER and its cooling capacity; however, on average there
is significantly higher EER for chillers which are cooled with water compared with those that use air. In fact
this improvement is not inherent to the chiller, but rather represents the temperature regime found in cooling
towers. The values used for testing the two types of system are somehow arbitrary and it may be that the
74
apparent benefit from water cooling is not fully realised in practice. Based on the standard test data the
average EER for water-cooled chillers is 3.57 W/W whereas for air-cooled chillers it is 2.52 W/W.
Nevertheless, 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. Interestingly, the average EER of the reversible systems is almost the same as for the cooling-only
systems .
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; however, when a piece of equipment is compared with its
direct peers (as expressed through the product categories described in Tables 3.3 and 3.4) there is still a wide
spread in product efficiency, as shown in Figure 3.5.
Figure 3.5. Distribution of EER/EERave (where the average EER is average for the same product category)
for chillers on the EU market in 1998
25
21.06
20.34
20
15
% 13.47
13.04
of models 11.6
11.17
10
5.3
5
2.44
1.58
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. From a total of 698 chillers in the Eurovent database the type of compressor was known for
304 of them. The following compressor types can be distinguished:
- "scroll" (orbital)
- "screw"
- "reciprocating" (i.e. with pistons)
Table 3.6 shows the proportion of chillers by compressor type within this subset of models as a function of
the condensing medium. Table 3.7 gives additional data that enable a comparison of chiller efficiency as a
function of compressor type and condensing medium.
The average performance of chillers with air-cooled condensers is almost independent of the compressor
type at ~2.5 W/W. 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 ~3.9 W/W compared
to ~3.5 W/W for those using scroll or reciprocating compressors. We note also that the best air cooled
chillers (the top runner, not the best on average) are the ones with scroll, then reciprocating, then screw.
75
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
Figure 4.1 Label stamped on Eurovent certified appliances.
Eurovent Certification is based on a few principles:

- all products in the defined scope must be certified (“certify-all principle”)
- regular testing by independent bodies must be carried out on randomly selected units.
- Test results must be very close to claimed performance characteristics – otherwise the catalogues data must
be re-rated.
Certify-all principle is not yet fully applied in all certification programmes but the goal will be achieved in a
few years. When all products presented by a manufacturer in the given scope are certified, the image of
certification is clear and there is no possibilities for misunderstanding. However for some equipment, a
progressive implementation of this principle has been necessary.
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. In order to be able to compare products from
different manufacturers, testing must be performed according to precisely defined procedures. 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.
Test standards contain specifications for test installation, incorporation of unit to be tested, test conditions
(temperatures, flow, humidity, etc.) and method of calculation of performance characteristics. Acceptable
deviations from test conditions and required accuracy of measuring devices are also specified.
Everything being well defined, any laboratory applying the same test method should in principle obtain the
same results when testing the same units. In practice, it is not possible to achieve such an ideal work. There
are always subtle differences between laboratories and human factor plays an important role.
Therefore in order to avoid possible discrepancies, the principle of single laboratory for a given product was
introduced in Eurovent Certification. This is applied at ARI which served as a model for Eurovent
procedures.
However, this principle could not be applied for all programmes for various reasons. Sometimes, historically
several laboratories had been used by participants and it was difficult to select a single test house. For other
programmes, the testing capacity (expressed in number of tests to be performed during a year) of available
laboratories was not sufficient.
Finally Eurovent Certification now uses nine independent laboratories from five countries for its thirteen
programmes.
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.
79
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). EdF and Eurovent have an agreement
wherein Euroevent will do additional certification testing for the RACs that are in the EdF programme.
Supplying the data to Eurovent means that it can be independently verified, through their certification
process, before being used by EdF. EdF also obliges products promoted within the directory to attain a
minimum energy efficiency value under each test condition. An example of these requirements are given in
Table 4.1. but there are many sets of requirements for the various equipment types. All values used are
Eurovent certified values, subject to independent testing.
Table 4.1. 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>2,2
mode
Heating 7 6 40 45 COP>2,5
mode
-7 -8 * 45 COP>1,5
* 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.
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. A
key feature of the programme is that the policy development is open to public comment, and the involvement
of key stakeholders is actively sought. For air-conditioning, the programme has thus far modelled the carbon
emissions consequences of three scenarios, and identified a range of market transformation options that can
be expected to constrain the expected increase. Some of the policy options are measures that have already
been agreed upon, for example, revisions to building regulations. Others are highly desirable but require
legislation, perhaps at European level –e.g. mandatory minimum efficiency standards for AC equipment. In
addition, the analysis and debate has identified measures that can be implemented without legislation. One
specific activity is to place the Eurovent performance data on an (existing) interactive product performance
database. In addition to providing system-specifiers with easier access to the data, this provides a means of
introducing indicative performance levels ("good practice", "best practice") that can form the basis of
voluntary procurement programmes. Information on this programme is accessible at www.mtprog.com
80
4.4Existing 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”. There is a hesitation between a pure prescription on some elements
(obligation of means) and a global limitation of demand, leaving the designer free to choose the elements and
to assemble them to reach the target (obligation of results). The problem arises from the lack of consumption
calculation methods applicable to a wide range of systems. Different countries cope with this problem in
different ways.
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. These are described in detail below.
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. 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).
The RSECE defines regulations that need to be complied with when HVAC systems are sized and
installed, so as to ensure (i) a minimum energy efficiency level of systems and equipment used to meet
indoor thermal comfort and air quality requirements, (ii) quality and safety of the facilities, and (iii) the
respect for environment. 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. In these
circumstances it obliges the use of central AC systems rather than distributed RACs.
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, the RCCTE. 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. However, the improved thermal characteristics and solar factors set out
in the RSECE are a design option and are not mandatory. Rather they are one of the aspects taken into
account in the calculation of the nominal power of the system to be installed. Thus, the designer is free
to choose other options (e.g. better lighting systems) to compensate for not attaining the 20%
improvement in envelope performance.
The RSECE sets out a number of limits to be applied to the equipment energy use:
Limits on the Joule effect. The electric heating power provided by the Joule effect cannot exceed 5%
of the total heating power installed, nor 25kW by independent zone.
Limits on terminal re-heating. Terminal re-heating is allowed for cooling-only systems but cannot
exceed 10% of the installed cooling power.
Limits to individual AC appliances. Individual air conditioning appliances for heating or cooling are
only allowed in spaces with special internal conditions, i.e. conditions which are different from the
rest of the independent zone.
Energy recovery is promoted. Energy recovery from the exhaust air is mandatory during the heating
season, when the rejection air has a thermal power greater than 80 kW. Free cooling is mandatory in
"all-air" systems with a ventilation air flow greater than 10000 m³/h.
Power stages should allow adaptation to the demand. To minimise the heating and cooling
generators from working at partial load, these equipment are required to have a number of stages
depending on the power, according to the Table 4.2.
81
Table 4.2. 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
The attainment of minimum efficiency levels for AC equipment is compulsory. Compression

machines must have a COP (EER) greater than 2.0. Energy Management Systems (EMS) are
required for systems with a thermal power greater than 250 kW. For systems with a thermal power
greater than 500 kW, the EMS should allow the centralised optimisation of the parameters.
The regulation imposes the adoption of a maintenance plan and a monitoring system. The energy
consumption of all equipment with an electric power greater than 12.5 kW should be independently
metered. Commissioning tests are required for boilers, chillers (power and efficiency), cooling towers,
pumps, hydraulic tests, heat exchangers, controllers, noise levels and overall functionality.
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.
The UK employs three alternative methods for demonstration of compliance with the national building
thermal regulations for tertiary buildings, offering increasing design flexibility in return for greater demands
in terms of calculations. These are:
The Elemental Method, which considers the performance of each aspect of the building individually (e.g. by
imposing minimum u-values). Some flexibility is provided for trading off, for example, insulation levels and
heating system efficiency.
The Whole-Building Method. This mainly applies to offices, and requires that the heating, ventilation, 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. There are less detailed whole-building methods for
schools and hospitals.
The Carbon Emissions Calculation Method. This also considers the performance of the building as a whole,
but is applicable to any building type. 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.
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. This may be satisfied by limiting glazing area,
providing adequate shading, or designing for night cooling operation.
For spaces with glazing facing only one orientation, 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
Horizontal 12%
Alternatively, it is acceptable to show either that:
- 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 2.5% of
occasions for July (between 1976 and 1995)
- showing by (acceptable) calculation that, in the absence of mechanical cooling or mechanical
ventilation, 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 0.068 0.065
Other fuels 0.091 0.088
(all figures in kgC/kWh)
However, these figures may be exceeded if building insulation levels are increased beyond the minima.
Carbon emission factors for different fuels are tabulated: significant figures are
natural gas 0.053
oil 0.074
electricity 0.113
In effect, 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.
For offices, air conditioning systems should have a satisfactory "Carbon Performance Rating" (CPR).
Notionally, this is a limit on the carbon emissions per m2 of floor area from air conditioning or
mechanical ventilation systems under standard operating conditions. In practice, it operates as a limit on
the installed cooling power and fan power per m2 of floor area, 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.
The maximum allowable ratings (all in kgC/m2/year) for new installations are:
New building Existing building
Air conditioning 10.3 11.35
Mechanical ventilation 6.5 7.35
For substantial modifications to existing systems, the performance must be the least demanding of either
these values or a 10% improvement on the original value.
Key calculation parameters:
Equivalent hours of full load operation:
83
- mechanical ventilation or air conditioning fans 3700 hours per year
- 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. These factors depend on the level of plant monitoring provided. They are
multiplicative.
Column C figures apply if no plant monitoring is provided; column B when any of: energy metering,
run-hour metering, internal zone temperature monitoring are provided; column A when, in addition, the
monitoring system has the ability to draw attention to "out of range" values.
Air distribution systems:
A B C
Operation in mixed mode with natural ventilation 0.85 0.90 0.95
Controls which restrict the hours of operation of the system 0.90 0.93 0.95
Efficient means of controlling air flow rate 0.75 0.85 0.95
"Mixed mode" requires the provision of sufficient opening windows and an interlock to prevent air-
conditioning operating when windows are open. It is only permissible when the perimeter zone of the
space accounts for more than 80% of the floor area.
"Air flow rate control" requires variable speed drives or variable pitch fan blades: throttling or inlet
guide vanes do not qualify for the allowance.
Refrigeration plant:
A B C
Free cooling from cooling tower 0.90 0.93 0.95
Variation of fresh air using economy cycle or mixed mode 0.85 0.90 0.95
Controls which restrict the hours of operation of the system 0.85 0.90 0.95
Controls which prevent simultaneous heating and cooling in the same zone 0.90 0.93 0.95
Efficient control of plant capacity, including modular plant 0.90 0.93 0.95
Partial ice storage 1.80 1.86 1.90
Full ice thermal storage 0.90 0.93 0.95
"Efficient control of capacity" requires good part load efficiency (without defining this): hot gas bypass
does not qualify.
"Full ice storage" requires chillers to operate only at night: if day and night chiller operation is intended,
this is "partial ice storage".
For buildings other than offices, there are no explicit requirements for air conditioning system
efficiency. This causes some difficulties when applying the Carbon Emissions Calculation Method (see
later) and guidance on the calculation of SSEER is being prepared.
However, there is a general requirement that components such as fans, pumps and refrigeration
equipment are reasonably efficient and appropriately sized, so there is scope to introduce performance
requirements.
84
There are requirements for SPF, Specific Fan Power, the electricity demand (of motor rating) per
airflow unit (expressed as W/litre/second), and encouragement of variable flow control.
For ACMV systems in new buildings, the SPF should not exceed 2.0, and preferably 1.5, and for new
systems in existing buildings or substantial alterations to existing systems should not exceed 3.0. There
are exceptions for non-comfort applications
Whole-building method
For offices, as an alternative to the air-conditioning CPR, a similar calculation may be calculated for the
combined emissions from the heating, lighting and air-conditioning systems. This provides greater
building services design flexibility, but the building envelope requirements of the elemental method
must also be complied with.
The required values of whole-building CPR are:
Building type New office refurbished office
Naturally ventilated 7.1 7.8
Mechanically ventilated 10.0 11.0
Air-conditioned 18.5 20.4
Calculation procedures are explained in BRE Digest 457.
Carbon Emissions Calculation Method
This route permits the greatest design flexibility. 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, designed to comply with the Elemental Method. There are, however, still some
constraints on acceptable values for some parameters - for example, air leakage of the building.
There is no prescriptive list of acceptable calculation methods, but a completed copy of Appendix B of
CIBSE AM11 "Building Energy and Environmental Modelling" is an acceptable demonstration.
Although this route is expected to be used only for a few, probably high-profile, buildings, it has already
generated some criticism. These relate to its implementation, rather than the principle. The main
problems are:
- the designer is required to design two buildings and their systems in order to demonstrate compliance
- 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.
Guidance on both these points, in the form of a simplified calculation for the notional building, has been
prepared (CIBSE,2004). Of particular relevance to the EECCAC study is the need to make explicit
assumptions about - or calculations of - about the seasonal system energy efficiency ratio (and seasonal
system coefficient of performance).
Information Provision: metering and logbooks
The Regulations require that the owner and/or occupier of the building be provided with a logbook that
contains, amongst other things, the design assessment for CPR or other benchmarks, commissioning
details, operating instructions, and details of all meters provided. A recommended template for this has
been published by CIBSE.
There is also a requirement that sub-metering be provided. This includes separate metering for tenancies
of more than 500 m2 (though for tenancies below 2500 m2, proportioning of cooling may be
85
acceptable). Generally, 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.
The status of regulations in other EU Member States

France
France is in the process of developing building thermal regulations, 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. The approach under consideration is based on the standardisation of algorithms used in a
simulation tool known as Consoclim. In the absence of a full calculation method, no regulation is applied,
not even the one on comfort without air conditioning. 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, reduction of lighting and ventilation loads, reduction of solar gains, separation of air
renewal of different buildings, a number of stopping devices, and a number of local control obligations.
Germany
Germany has recently initiated a process to develop national regulations to limit the energy consumption of
building active cooling systems. The new Energy Saving regulation of 2001 requires, that the cooling load
should be as low as possible according the state of the art, but does not specify how this is to be determined.
Accordingly, it is intended, to generate within the DIN 4601 a new part (Part 11), 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.
Italy
In Italy, thermal regulations affecting the building envelope and internal energy-using systems, including the
heating and cooling systems, fall within the framework of a general law, Law N. 10 of 1991. This all
encompassing law, 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. Under Italian
legislation, laws are followed by Law-Decrees which establish ways and means to fulfil law dictates, and
finally official Standards (e.g. UNI standards and similar) which give the physical parameters of
applications.
A law decree (no. DPR 412/93) addressing the heating systems of buildings has been established but the
same has not yet been done for the cooling system. Studies, proposals and recommendations are still under
examination, yet none has been transformed into official documents so far.
The Italian Engineering Association AICARR and CTI, Italian Thermal Committee, are preparing guidelines
for the application of the European Energy Performance of Buildings Directive (see below) that shall
subsequently be transposed into Italian Law.
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. Spain also has a voluntary
building energy labelling scheme which includes specifications for cooling system energy performance.
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.
The objective of the Directive is to promote the improvement of the energy performance of buildings within
the European Community, taking into account outdoor climatic and local conditions, as well as indoor
climate requirements and cost effectiveness. 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. Transposition of the Directive into national law is required by January 2006; however, member
states have some flexibility concerning the exact manner in which it is to be implemented. The full
86
application of requirements for boiler inspection, air conditioning system inspection and building
certification (see below) may not be delayed beyond January 2009.
The Directive aims at reducing the consumption of energy in new and renovated buildings, excluding
industrial buildings, through the following:
A) Establishment of a general framework and common methodology for calculating the integrated
energy performance of buildings.
B) The development and application of minimum energy performance standards to new buildings and to
certain existing buildings when they are renovated.
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.
D) Specific inspection and assessment of boilers and heating/cooling installations.
All these measures are to be taken nationally, before January,4 2004. The field of application of (D) for
CAC i.e. > 12 kW is exactly that of this study. Buildings of more 1000 m2 in floor area are to
addressed in A, B and C. The method of calculation to be established in (A) is the responsibility of each
Member State and thus is not unified across Member States. It should include the energy consumption
of AC and ventilation. There could be obvious advantages in harmonising such national building codes,
namely in the very technical field of CAC. 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. Indeed, heating remains the essential
concern of EU building codes, even after the harmonisation.
Article 8 requires central air-conditioning systems over 12 kW to be regularly inspected. 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. 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. 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.
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, which amongst other
measures would give the Commission the right to establish mandatory minimum energy performance
standards (MEPS) for end-use equipment. 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. The determination of this is to
be based on the results of a technical-economical analysis. As yet there is no clear time line regarding
when this draft will be submitted to the council of ministers and parliament for approval.
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, especially as it is provided by utilities and service
companies in the form of energy services. It is envisaged that Member States will set targets to promote
and support energy efficiency services, (e.g., third party financing) and programmes, especially for
smaller energy consumers such as households and SMEs. This includes a supportive framework for
implementation and financing of energy services, adapted to each Member States’ liberalised market. 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. This proposal is in lieu of
additional public service obligations in the Amended Internal Market Directives and the Commission’s
87
Amended Rational Planning Techniques Directive proposal from March 1997.
Practices and procedures adopted in CAC system operation
The operation and maintenance of CAC systems is usually contracted out to a specialist company. Two
broad types of contract are used:
Contracts of “means” Within the framework of a contract with obligation of means, the building owner
entrusts the execution of specific tasks to a 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. It is a little bit out of fashion due to the existence of
other typical contracts.
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. Their importance
derives from the importance of the public markets.
Thus, the company gives its estimate on operational budgets, its guarantee on the quality of air
conditioning and well-being in the buildings, on the maintenance of the materials which are entrusted to
them and the compliance with the code of practice. It implements the means that it judges necessary, as
it is needed, until obtaining the contracted result.
Whereas a contract of means can be of low duration, the contract of results can be only a contract of
long duration. Indeed the guarantee of the results implies a perfect knowledge of the installations but
also, very often, significant investments in time for the knowledge, commissioning and adjustment of
the installations. A contract of results is incontestably the form which it is advisable to give to a
technical management contract when there are, by nature, expensive and complex air conditioning
installations.
4.5 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.
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 .
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.
However, 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, as
described below.
ASHRAE 90.1: a comprehensive approach to raise CAC energy efficiency

The objective of the ASHRAE 90.1 standard (ANSI/ASHRAE/IESNA 90.1-1999) 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. It does not apply to low-rise
residential buildings, which are covered under the ASHRAE 90.2 standard. The 90.1 standard provides:
88
(a) minimum energy-efficiency requirements for the design and construction of; 1. New buildings and their
systems, 2. New portions of buildings and their systems, and 3. New systems and equipment in existing
buildings.
(b) criteria for determining compliance with these requirements.
The provisions of the standard apply to:
(a) the envelope of buildings provided that the enclosed spaces are: 1. heated by a heating system whose
output capacity is greater than or equal to 3.4 Btu//h*ft2 (10W/m2), or 2. cooled by a cooling system whose
sensible output capacity is greater than or equal to 5 Btu/h * ft-2 (15 W/m2);
(b) the following systems and equipment used in conjunction with buildings: 1. heating, ventilating, and air
conditioning, 2. service water heating, 3. electric power distribution and metering provisions, 4. electric
motors and belt drives, and 5. lighting.
Moreover, Standard 90.1 focuses on comfort conditioning rather than industrial, manufacturing, or
commercial processes. Note, too, that the stated purpose of the standard is to provide minimum
requirements; a designer or owner can always exceed these basic conditions for compliance.
The latest version of Standard 90.1 which was issued in 1999 has several differences from the previous 1989
version. It has been reorganised for ease of use, such that the new standard clarifies requirements and
provides a simplified compliance path for small commercial buildings. More importantly, the 1999 edition
expands the standard's scope to include both new and existing buildings and building systems. For alterations
and additions, the 90.1 User's Manual notes that, “In general, the Standard only applies to building systems
and equipment…that are being replaced.” A life-cycle-cost analysis was used to define the criteria in the
1999 edition and thereby balance energy efficiency with economic reality.
Standard 90.1—1999 addresses building components and systems that affect energy usage. The technical
sections of the standard, Sections 5 through 10, specifically address components of the building envelope,
HVAC systems and equipment, service water heating, power, lighting, and motors. Each technical section
contains general requirements and mandatory provisions; some sections also include prescriptive and
performance requirements.
To comply with Standard 90.1—1999, the prospective design must first satisfy the general requirements and
mandatory provisions of each technical section. But that's not all. 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.
The ECB method permits tradeoffs between building systems (lighting and fenestration, 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. Using the ECB approach requires simulation software that can
analyse building energy consumption and model the energy features of the proposed design. Standard 90.1
sets minimum requirements for the simulation software. Suitable programs include BLAST, DOE-2, and
TRACE™.
Energy-conscious comfort in ASHRAE 90.1

The present HVAC section of Standard 90.1, Section 6, has been substantially reorganised compared with
the 1989 edition. HVAC-related requirements are presented in order of complexity, beginning with the
simplest and most common design obligations. Because the HVAC section is 21 pages only the key
requirements are summarised here. Section 6 of ASHRAE/IESNA 90.1—1999 describes mandatory and
prescriptive requirements for commercial heating, ventilating, and air-conditioning systems. It also defines
three methods for compliance:
1. A Prescriptive Path, which comprises mandatory provisions and prescriptive requirements
2. An Energy Cost Budget method, which combines mandatory provisions and a computerised methodology
that permits tradeoffs between various building systems and components
89
3. A Simplified Approach option, which consists of a subset of all mandatory provisions and prescriptive
requirements
For small buildings, the “simplified approach” consolidates the provisions on roughly two pages so that
design professionals can quickly locate all applicable requirements. The difference lies in ease of use and the
degree of flexibility allowed. Eligibility for this approach requires that the building occupy less than 25000
sq ft of gross floor area and not more than two stories. Another prerequisite (there are others) is that each air-
cooled or evaporatively-cooled HVAC system serves only one zone.
Mandatory HVAC Provisions in ASHRAE 90.1

Mandatory requirements for HVAC systems include mechanical equipment efficiencies, controls,
construction, insulation, and completion. These requirements are an integral part of every compliance path.
Mechanical equipment efficiency. The 1999 standard upgrades the minimum efficiency requirements for
many types of HVAC equipment and adds efficiency requirements for heat-rejection equipment, ground-
source heat pumps, and absorption chillers. Standard 90.1—1999 also provides tables for centrifugal chillers
selected at non-standard conditions (leaving chilled water temperatures, entering condenser water
temperatures, or condenser water flow rates). For equipment covered under the previous edition, the 1999
standard allowed the 1989 efficiencies to apply until October 29, 2001, Table 4.3.
Table 4.3. Summary of the revised of ASHRAE 90.1 energy performance requirements
Equipment Type Per90.1—1989 After29-Oct-2001 Test Procedure

Rooftop air conditioner, 8.5EER 9.7EER ARI340/360
15tons
Water-source heat pump, 9.3EER (85°FEWT) 12.0 EER (86°FEWT) ARI320d(ARI/ISO-13256-
4tons (cooling mode) 1after29-Oct-2001)
Water-cooled screw chiller, 3.80 COP 3.90 IPLV 4.45 COP 4.50 IPLV ARI590
125tons
Water-cooled centrifugal 5.20 COP 5.30 IPLV 6.10 COP 6.10 IPLV ARI550
chiller, 300tons
In the case of centrifugal chillers, both the full-load COP and IPLV must be 6.10 SI or better, that is 0.576
kW/ton or less [kW/ton figure of merit = 3.516/COP, with COP in W/W].
Controls. The 1999 standard also contains extensive HVAC control requirements regarding deadbands,
restrictions for set-point overlap, and off-hour controls. Stipulations for off-hour controls include all of the
following:
1. 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.)
2. Zone isolation capabilities that permit areas of the building to continue operating while others are shut
down
3. Automatic shutdown
4. Setback controls
5. Optimum start controls after the system airflow exceeds 10000 cfm
Construction, insulation, and completion. Mandatory HVAC requirements also address construction (duct
sealing, leakage tests) and insulation of ducts and piping. Climate and placement dictate insulation
requirements for ducts. For piping, the requirements depend on pipe location and the operating temperature
range of the fluid.
Drawings, manuals, and a narrative of the system operation must be supplied to the building owner, which
makes a lot of sense. Even if an engineer designs a great system, it's unlikely that energy savings will accrue
if the operator doesn't know how the system should work.
90
The standard also addresses balancing for air systems larger than 1 hp and for hydronic systems larger than
10 hp. It further requires control elements to be calibrated, adjusted and in proper working condition for
buildings that exceed 50000 sq ft.
Additional prescriptive HVAC requirements

Under the Prescriptive Path, a prospective HVAC design must satisfy specific prescriptive requirements in
addition to the mandatory provisions reviewed above.
Economisers (automated free cooling). Climate and equipment size dictate the prescriptive requirements for
airside and waterside economisers. The economiser must also be integrated, that is, capable of operating in
conjunction with mechanical cooling. In addition, 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.
An economiser can be omitted from unitary equipment if its performance is efficient enough. For example,
the requirement for a 20-ton rooftop air conditioner in Tucson, which has 6921 Cooling Design Days—base
50 (CDD50), is an EER of 9.7. If the EER of the selected rooftop air conditioner is 11.1 in US units, i.e. 2.8
SI or better, an economiser is unnecessary.
Simultaneous heating and cooling. Although the 1999 standard limits this practice, it does not ban
simultaneous heating and cooling. Exceptions provide sufficient flexibility to maintain either temperature or
humidity control. For example, unlimited reheat is permitted if at least 75 percent of the reheat energy
originates from a site-recovered or on-site solar energy source. 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.
Air system design and control. Fan power limitations, now expressed in terms of nameplate power, must be
met when the total fan power for the system exceeds 5 hp (about 3.6 kW). 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.
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. This requirement will almost certainly prompt increased
use of variable-speed drives or vaneaxial fans in systems of this size.
Another notable addition to this set of prescriptive requirements is the following:
Set Point Reset. For systems with direct digital control of individual zoned boxes reporting to the central
control panel, static pressure set point shall be reset based on the zone requiring the most pressure; i.e. the set
point is reset lower until one zone damper is nearly wide open.
Also known as fan-pressure optimisation, the basic premise of set point reset is that the static-pressure set
point can be reduced dynamically, which lets energy savings accrue rapidly.
Hydronic system design and control. Like the fan on the air side of the system, the 1999 standard requires
that the pump in a variable-flow system draws substantially less power at part load. Unless there are three or
fewer control valves in the system, 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. This requirement will undoubtedly prompt greater use of variable-speed drives.
Supply-temperature reset is required, too—but not for variable-flow systems nor where it “…cannot be
implemented without causing improper operation of heating, cooling, humidifying, or dehumidifying
systems.”
Heat-rejection equipment. For heat-rejection equipment such as cooling towers, the fan must be able to
reduce its speed to two-thirds or less if its motor is 7.5 hp or larger. Beyond this power limit, a cooling tower
with less than two cells must be equipped to reduce fan speed on all cells — perhaps with pony motors, two-
91
speed motors, or variable-speed drives. If the cooling tower has three cells, at least two of them must be
equipped with speed control.
Energy recovery. Systems larger than 5000 cfm that bring in lots of outdoor air (at least 70 percent of design
airflow) must include energy recovery; the means of recovery must be at least 50-percent effective. This
proviso will probably lead to the increased use of energy recovery in air handlers dedicated to ventilation,
particularly in retrofit applications in which ventilation airflow is brought into compliance with
ANSI/ASHRAE 62.1.
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.
Heat recovery for service water heating is required in facilities that operate 24 hours a day, where the heat
rejection capacity exceeds 6 million Btu/h, and where the service-water heating load exceeds 1 million Btu/h.
Continuous maintenance of the ASHRAE standard

As a continuous-maintenance standard, ASHRAE/IESNA 90.1 remains a dynamic document. Rather than
periodic updates (every five years, for example), committee members can request changes to the standard at
any time. Public proposals submitted by February 20 are considered at the ASHRAE annual meeting (usually
held in June). If the committee sees merit in a proposed change, it issues an addendum for public review and
comment. When consensus is reached, the addendum is incorporated in the standard.
Links between an ASHRAE standard and the US Energy Codes

The US Energy Policy Act or EPAct (P.L. 102-486) requires states to certify that their energy codes meet or
exceed the requirements of ASHRAE Standard 90.1—1989. EPAct also requires that the U.S. Department of
Energy (DOE) evaluate subsequent revisions of Standard 90.1 to determine whether they improve energy
efficiency in commercial buildings. The U.S. DOE posted the results of its quantitative analysis on its Web
site, www.eren.doe.gov, in a report entitled ‘Commercial Buildings Determinations — Explanation of the
Analysis and Spreadsheet’. The report observes that “Overall, considering those differences that can be
reasonably quantified, the 1999 edition [of ASHRAE/IESNA Standard 90.1] will increase the energy
efficiency of commercial buildings.”
In fact, both the report and SSPC 90.1, the ASHRAE committee responsible for maintaining the standard,
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. In some instances, the 1989
standard was either unjustifiably stringent (in the case of metal roofs, for example) from a cost standpoint or
did not adequately reflect real-world conditions (in the case of warehouse lighting). 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.
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.
Aggregation of the energy use intensities produced by the simulations was done as follows: 1) extract zone-
based energy use intensities from simulations; 2) aggregate results by required economiser usage in each
climate; 3) map simulation results by climate 4) scale simulation results to existing building stock floor area
by building type and region; 5) weight results for frame and mass wall construction ; 6) weight results for
heating fuel 7) convert energy use intensities by fuel type to site energy, source energy, and energy cost
intensities; 8) weight energy use intensity results by building construction floor area estimates. Table 4.4
shows the estimated energy savings from application of the revised standards ASHRAE/IESNA 90.1—1999
Table 4.4. Percentage Change in Whole-Building Energy Use Intensity (EUI) and Dollars Use Intensity ($UI) through application of
ASHRAE/IESNA 90.1—1999
Building Type Floor Area Electricity Gas Site EUI Source EUI $UI (USD)
Weighting
Assembly 0.068 9.5% -5.3% 4.4% 7.2% 7.5
92
Education 0.218 11.4% -6.3% 5.2% 8.6% 9.0
Food 0.027 -1.2% 1.7% -0.4% -0.8% -0.9
Lodging 0.079 0.2% -6.5% -1.7% -0.6% -0.5
Office 0.190 10.6% -12.7% 8.2% 9.7% 9.8
Retail 0.246 15.7% -30.7% 12.7% 14.7% 14.9
Warehouse 0.173 -71.6% -11.3% -45.1% -58.7% -59.7
National 1.000 7.3% -8.6% 3.9% 5.9% 6.2
Australia, Japan, 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.
They have introduced MEPS for packaged air conditioners with cooling capacity between 7.5 kW and 65
kW. At the same time the Australian government assessed the requirements. The most stringent MEPS being
applied internationally were found to be the US ASHRAE 90.1-1989 requirements and the current Australian
requirement are loosely based on these, see Table 4.5. Australia currently has no MEPS for chillers and has
no energy labelling requirements for packaged air conditioners. Australian test standards for packaged AC
units are compatible with ISO and European test standards.
Table 4.5. Minimum energy performance requirements for packaged air conditioners with a cooling capacity
between 7.5 and 65 kW in Australia
Cooling Capacity Minimum cooling

(kW) COP (W/W)
7.6-10.0 2.25
10.1-12.5 2.30
12.6-15.5 2.35
15.6-18.0 2.40
18.1-25.0 2.45
25.1-30.0 2.50
30.1-37.5 2.55
37.6-45.0 2.60
45.1-65.0 2.65
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. Thus far Japan has developed the following requirements for central AC
systems:
Table 4.6. 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) 2.88 2004
Unitary AC (heating and cooling) 3.06 = (EER+COP)/2 2004
93
These requirements apply to all unitary (i.e. packaged) AC equipment within the specified cooling capacity
range and hence applies to large room AC units, multi-splits, VRF units and classical packaged systems
(rooftops and cabinets). Japanese test conditions for packaged AC units are mostly compatible with ISO and
European test standards. As yet there are no measures for chillers and there are no labelling requirements for
this kind of AC equipment.
Korea
Korea only has MEPS in place for unitary split-packaged AC units of between 10 and 17.5 kW in cooling
capacity. These are required to attain a mandatory minimum EER of 2.25 W/W, but in addition the
government expects manufacturers to attain a minimum sales-weighted efficiency level of 2.93 W/W. 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 4.7). Taiwanese test
conditions for chiller units are compatible with ISO and European test standards.
Table 4.7. Minimum energy performance requirements for chillers in Taiwan
Type of chiller EER COP Cooling

capacity range
(kW)
Water-cooled chillers (volumetric compressors) 3.01 3.50 <528
Water-cooled chillers (volumetric compressors) 3.10 3.60 528 to <1760
Water-cooled chillers (volumetric compressors) 3.44 4.00 >1760
Water-cooled chillers (centrifugal type) 3.70 4.30 <528
Water-cooled chillers (centrifugal type) 4.10 4.73 528 to <1760
Water-cooled chillers (centrifugal type) 4.51 5.25 >1760
Water-cooled chillers (volumetric compressors) 2.06 2.40 all
4.6 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, some are known for their better Energy Efficiency. One way of
improving EE is to promote some specific system types. We have discarded such an option, 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.(cooling)
5 CAC Cooling towers CAV/AHU TC Water Cooled with air dist.(cooling)
6 CAC Cooling towers CAV/AHU TAC Water Cooled +air +hum.(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
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, for equality of comfort. In fact it is only
installed in some situations.
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.
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, 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. 2
V2 Motor Eff. 1
V3 Selection of fans characteristic curves and the pressure control
V4 Improved local tangential ventilator
V5 Improved fans on condensers, AHU,
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, some under-sizing acceptable
D&S3 Quality of the service of "sizing" (for example search for alternatives, sizing of fans, 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, giving to electricity a weight close to its price
95
D&S7 Regulation is expressed in terms of carbon intensity, 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, 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, rooftops, RAC, etc. used for homogeneous zones
PACK1 Local Free cooling
PACK2 Changing control set-points (T,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, "decoupling", variable speed in distribution
WS7 Improved pumps
WS8 Pipes insulation reinforced
WS9 Less head losses, 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,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. 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, like window opening
C5 Quality level
Alternative Equipment strategies
97
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, literature and country reports have highlighted a number of different "philosophies"
to approach CAC energy efficiency. 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, but big cabinets, rooftops, etc.) reveals that the more efficient products on the
market are always +20-50% better than the average. 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. Presently the selection
of equipment is made by professionals based on the initial cost per kWc or, in a few cases, based on rated
EER although ideally it should be based on a SEER (or IPLV) rating.
Second type: choice of the best general structure of the system

Within RACs on one hand and CAC on the other, there are families or types that have different average
efficiency. 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); for CAC the
potential seems larger. 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. It is part of
EECCAC terms of reference to define what can be reasonably included in building codes. 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.
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.g. making use of "free cooling"). These gains are not completely attainable
through building codes, but mostly through good engineering work. They are also partly related to the type
of equipment, the existence of dampers, controllers, etc. (see ASHRAE 90.1). There are also savings related
to efficiency of fans selected, variable speed options and the quality of control. The EECCAC study has to
review these options and present them in a structured way. 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).
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. However such reversibility options are poorly
known, infrequently realised and not very well controlled. Condenser heat recovery is better known and
documented, although not prescribed.
Fifth type: maintenance and operation improved
98
The gains reported in the previous types will only be achieved over the long term therefore the maintenance
or improvement of performance, by technical measures or contractual means (such as Energy Performance
Contracting) or by periodic audit, is an interesting family of options.
Sixth type: energy and power control

Gains related to the reduction of peak power, as opposed to energy consumption, should not be overlooked
especially, as some data show, there is a significant pressure on utilities investment to cope with peak
demand up to 2010. The energy consumption allowing peak power management may be a little higher, but
shaped completely differently and thus can make the whole energy system more efficient. This could not be
investigated in the EECCAC study because it was outside the scope of the study.
Seventh type: envelope and ventilation, 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. However, building envelope and ventilation
choices made to minimise winter demand (insulation and air tightness) and, even more, the choices made to
minimise summer demand (solar control, night time ventilation) have a large influence on AC energy
demand. Measures taken to lower the specific electricity consumption in lighting or office equipment have a
very large "double dividend" in avoided air conditioning. The regulatory approach used in Switzerland is to
state that, generally speaking, efforts made to improve the energy performance of the building envelope and
ventilation system can avoid the need for most artificial air conditioning.
99
5. PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO)
5.1 AC Stock and market in 1990, 1998, 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. So the BAU scenario is based on an absence of any policy measure. The year 2010 has been
chosen as a reporting reference point due to the Kyoto deadline. Results are also reported up to the year
2020. 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.
Evolution of the market

As previously described, the extrapolations were based on 1998 figures and national evolution trends.
Validation of the results projected into the past were limited to the most reliable comparisons i.e. 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. The model was primed
using: the market data from Eurovent for 1998, and the historical data on national market and stock sizes for
the years 1975 to 1990.
The air conditioned areas installed in each country, namely the AC stock, is not an easy figure to give by
country : one of the problem being the definition itself of an area effectively cooled, another one the
obtention of data. 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), Table 5.1. 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.
Table 5.1. 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, E, EL,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.
In fact the AC saturation levels are very different from one sector to another: for example, in Austria the
hospitals are air-conditioned but the houses are not, etc. However there is not enough data to model each
sector. The residential sector, which was already treated in the EERAC study, has not been remodelled here
and the other economic sectors (hospitals, hotels & bars, offices, trade, houses, education) are modelled as if
they followed the same "learning curve", even if the starting point and the final saturation level (at time
infinity) are different from one sector to another. 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.
For the past the model extrapolates back to 1970 to generate stocks for the periods 1970-75, 1975-80, etc...
which will be renewed 15 years later when the AC equipment is removed. The overall growth in the size of
the stock is the difference between the apparent market and level of renewal.
100
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.
Some global results

The total cooled area is given in Figure 5.1 and will rise from the present 1000 Million m2 to around 2000
Million m2 in 2010.
Figure 5.1. 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, for instance in terms of the cooled area (m2) per
inhabitant as in Figure 5.2.
Figure 5.2. Cooled area per European in 2000 by Member State and for the EU as a whole
m2/inhabitant
7
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). The estimated cooled area per
European in the future is shown in Figure 5.3.
101
Figure 5.3. Cooled area per European in 2020 per Member State and for the EU as a whole
m2/inhabitant
18
16
14
12
10
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.
Some national results

We can follow national evolutions on figures like figure 5.4. Saturation as well as differentiation
between countries appear with their real importance.
Figure 5.4. 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 5.2 with the most important national values, because this may be a useful
tool in some national frameworks. However we have to comment accounting methods. 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. All conditioned areas hereunder are « standardised » areas,
corresponding to the typical European sizing ratio (120 W/m2 for CAC, 240 W/m2 for RAC). No country
has exact statistics of conditioned areas, but some are close to it. The values hereunder cannot be compared
directly to such « national » statistics for two reasons:
1- the sizing ratios vary according to climate, national habits, sector, etc. (this has been partly corrected for)
102
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, adjacent rooms, etc.); so
both our “standardised” figures and the national figures have a margin of uncertainty and should be
compared with caution.
Table 5.2 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 15,68 20,06 26,29 30,29 33,01 33,95
Mm² reverse 1,45 2,06 2,74 4,83 5,57 6,08 6,27
BE Mm2 cooling 4,03 8,98 20,36 32,41 42,77 52,09 54,29
Mm² reverse 0,84 1,84 4,03 6,46 8,43 10,24 10,73
DE Mm2 cooling 3,78 6,62 11,30 19,92 29,24 37,57 42,30
Mm² reverse 0,70 1,35 2,50 4,12 6,01 7,72 8,69
FI Mm2 cooling 15,88 24,06 36,43 43,28 47,28 50,19 50,99
Mm² reverse 1,35 2,28 3,71 7,49 8,21 8,74 8,89
FR Mm2 cooling 93,40 129,39 180,37 293,24 390,57 472,24 502,39
Mm² reverse 32,79 45,84 64,98 106,59 141,52 171,24 182,61
GE Mm2 cooling 34,07 66,29 127,64 216,74 298,51 365,63 400,13
Mm² reverse 4,88 9,54 18,81 30,61 41,65 51,09 56,23
GR Mm2 cooling 11,04 23,06 48,23 80,47 108,97 140,88 145,99
Mm² reverse 5,29 11,17 23,65 40,07 54,24 70,12 72,68
IR Mm2 cooling 5,03 6,81 9,37 13,84 17,07 19,39 20,37
Mm² reverse 0,75 1,08 1,78 2,30 2,83 3,22 3,41
IT Mm2 cooling 130,85 175,63 258,76 368,74 414,88 450,33 467,85
Mm² reverse 29,22 43,81 73,26 106,86 120,93 132,38 138,18
LU Mm2 cooling 0,25 0,43 0,87 1,34 1,76 2,07 2,20
Mm² reverse 0,07 0,10 0,17 0,26 0,35 0,40 0,43
NE Mm2 cooling 22,25 39,02 66,88 87,71 101,28 110,49 113,62
Mm² reverse 1,84 3,55 6,50 12,17 14,03 15,38 15,89
PO Mm2 cooling 8,46 12,51 18,73 34,84 52,08 68,41 78,27
Mm² reverse 4,67 7,27 11,25 18,47 27,53 36,11 41,31
SP Mm2 cooling 64,24 102,68 172,69 248,07 295,71 342,20 352,20
Mm² reverse 34,61 56,66 97,11 136,02 161,33 186,01 191,57
SW Mm2 cooling 38,41 53,26 69,38 78,17 83,23 87,28 88,21
Mm² reverse 4,08 6,14 8,74 14,90 15,88 16,68 16,92
UK Mm2 cooling 94,29 127,63 173,15 248,36 294,19 326,80 340,28
Mm² reverse 14,17 20,41 31,06 43,81 51,73 57,87 61,07
Total Mm² cooling 538,01 792,07 1214,23 1793,42 2207,83 2558,59 2693,04
Total Mm² reverse 136,71 213,10 350,28 534,96 660,23 773,29 814,88
The differences are very small on total between the two types of figures (standardised or not) but not for a
specific country. Only the “standardised” values are used in the rest of the present report.
Sectoral market
The evolution of the various economic sectors and their demand for comfort vary a lot (figure 5.5). Only
trade and offices really grow in relative terms and they may reach 70% of stock by 2020.
Figure 5.5. The evolution of cooled floor-area by EU economic sector from 1985 to 2020
103
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 5.6) no factor will influence the evolution of the share of techniques on the
market and the effects of past trends disappear gradually. 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. This
is also neglected in our analysis because socio-economic research on the subject is scarce.
Figure 5.6. Evolution of cooled floor area provided by each AC type in the EU from 1985 to 2020
COOLED AREA IN Mm2
3000,
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
5.2 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. We have to rely on simulations that are made reliable by
the years of experience gained in the US, the care in using them and some validation on a few field results.
104
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.
Real buildings for the simulation of CAC systems with DOE

A Spanish office building has been analysed then simplified and simulated in details (Figure 5.7a) both for
envelope and equipment. It has an L floor shape and it is basically dedicated to office areas, but includes
other complement uses like cafeteria, medical room and toilets. The cooled area is 4800 m2.
Figure 5.7a The shape of the office building used as main reference.
Offices account for 50% of cooled surfaces in Europe. In order to cover the second most important sector,
Trade, we have simulated a second building. The shopping mall is a real building located in Seville in the old
railway station “Plaza de Armas”. It was rebuilt as a shopping mall after the 1992 Seville Universal
Exhibition. It is composed of shops, restaurants, cinemas, a supermarket, etc.. Figure 5.7b shows a 3D view
of this mall. The cooled area is 12 300 m2.
Figure 5.7b 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.
EU climate has been represented by three climates: Seville, London and Milan.
105
Envelope has been adapted to each country construction habits. Insulation cannot be the same in the various
countries. The final decision regarding this matter has been taking the building’s envelope from the
“TRIBU” study of Building Codes (TRIBU, 1994). Basic differences between climates are the following: for
exterior walls insulation thickness varies from 4 cm in Seville to 8 cm in Milan; we have used double glazing
in every climate, but clear in London, low emissive in Milan and low emissive and reflective in Seville.
Thermal comfort and indoor air quality have been guaranteed for every system to allow comparisons.
The simulation covers seven different system types: Constant air volume (CAV), variable air volume (VAV),
Roof-top units (RT), fan-coil four pipes (FC4P) and two pipes (FC2P), package terminal air conditioners
(PTAC) and water to air heat pumps connected to a close condensation loop (WLHP). Every HVAC end use
is covered, namely, fans, pumps, cooling and heating.
Lighting and plug equipment, despite being non HVAC uses, have been also considered because they
generate a large share of the cooling load.
After an exhaustive filtering process, six WS EEO have been studied and ranked. They regard to chilled
water temperature control and water transport. Basic AS EEO have been valued, including air transport
efficiency, air side economiser and exhaust air heat recovery. The results of those simulations is given in the
next chapter.
Some basic remarks have to be kept in mind for central systems:
• They have been designed assuming the same zoning and air distribution. So there is the same number of
AHU's for CAV, VAV and RT. This implies that zone supply air flows are also the same for each
climate.
• Roof-top units are equipped with constant air volume fans.
• Same ventilation level (that is same zone outdoor air flow rate) has been considered. For VAV systems,
the minimum supply setting of each VAV box equals the design outdoor ventilation rate, and, at AHU
level, outdoor flow rate is always maintained constant. This supposes that each central system is always
handling the same amount of outdoor air, however VAV handle a variable supply air flow rate.
• Air transport efficiencies, expressed in terms of specific consumption (W/m3/h) is equal to 0.47 for
constant volume fans (SPF = 1.7 W/(l/s)) and 0.57 for variable volume ones (SPF = 2.05 W/(l/s)). A
variable speed motor is used to control supply flow for VAV.
• Air side economizer and exhaust air heat recovery are not installed when describing the stock CAC
market.
With regards to zonal systems, the following issues should be pointed out:
• One (or some of equal size) terminal unit is installed for each thermal zone.
• Ventilation is guaranteed using a primary air AHU that provides neutral (22 ºC) outdoor air directly to
every building zone. Heat recovery is not used for this AHU.
• Air transport efficiencies, expressed in terms of specific consumption (W/m3/h) equals to 0.15 for FC
and WLHP terminal units. PTAC fan consumption is considered as cooling consumption since
manufacturers' data include this consumption in EER.
For hydronic systems, main remarks follow:
• Chilled water loops provide water at 7 ºC to cooling coils while hot water is supplied at 60 ºC. Water
delta T for cooling and heating are 5 and 10 ºC respectively.
• An air-condensed screw chiller (EER = 2.6) is used to provide chilled water to cooling coils, and a gas
standard hot water boiler (Eff = 0.88) as heat source.
106
• Each primary and secondary water loop is equipped with a constant flow circulation pump. Efficiency
figures may be found in the technical detailed report of Task 5.
• Hydronic system except FC2P supplies chilled and hot water using independent circulation loops (four
pipes facility). The FC4P system has been kept to represent the typical fan coil system.
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. 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). The post-
processing consists in keeping all the auxiliaries given by DOE, to cover the cooling load with any
alternative chiller, given the outside conditions extracted from the simulation. This also allowed us to
consider the case of wet cooling towers which represent a limited but non negligible share of the market.
Finally we could also consider less frequent solutions that are presented as very efficient as VRV, chiller on
natural water, etc. by extending the post processing.
The sizing of the system has been adapted to each simulated climate (Seville, London, Milano) and so the
three locations do not display the same installed capacity for the same building shape. The building
conditioned area is about 4800 m2 under the form which has been simulated. Note that depending on our
objective we have used the nominal square meter of the building (the one known in national statistics), and
sometimes the standardised m2 when it’s related with consumption (standardised sizing of 120 W/m2).
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, 3200 and 5200 m2 for this same
building in the three climates( SE: 160 W/m2, LO: 80 W/m2, MI : 130 W/m2 –more detailed figures have
been used by system types).
Preliminary results for the cooling consumption of the office building

Dealing with consumption, the main results of the simulation for the two extreme cases (CAV without
humidity control, the one consuming more, RAC with primary air, the one consuming less – with the same
comfort level TAC) are given in Table 5.3. 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. If the chiller behaves the
same as in Seville, 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- becomes very low in London.
Table 5.3 Consumption per physical square meter, cooling demand, efficiencies and total cost of cooling one
square meter of the office building.
Electricity Needs Electricity to Electrici Initial ALCC ALCC ALCC
SSEER
SEER
Per sq. meter SCL compressor ty total Cost In Euros In Euros In Euros
(kWh/m2) kWh/m2 SEC kWh/m2 SSEC kEuros (0,10 (0,06 (0,17
kWh/m2 E/kWh) E/kWh) E/kWh)
Seville-CAV 115,05 59,25 99,26 1,94 1,16 1008 34,86 31,73 40,34
London-CAV 20,82 10,87 32,77 1,92 0,64 528 16,77 15,73 18,60
Milan-CAV 73,53 36,73 70,49 2,00 1,04 848 28,31 26,09 32,20
Seville-RAC 104,02 54,40 58,52 1,91 1,78 382 14,86 13,02 18,09
London-RAC 15,59 7,97 8,39 1,96 1,86 202 6,08 5,82 6,55
Milan-RAC 54,63 26,39 28,53 2,07 1,92 322 11,60 10,42 13,67
The cooling loads are different from one place to another but, since the sizing is not the same, the equivalent
number of hours of operation (load in kWh divided by sizing in W, for one square meter) is less variant for
this same building in the three climates (SE: around 700 hours, LO: around 200 hours, MI : around 500
hours). More detailed figures could be defined by system types. The figures given here are still consistent
with the ones used for those places in EERAC, while the meteorological data and the software, as well as the
level of definition have been largely improved. The EERAC figure did include a penalty for degradation of
107
performance over time, but we assumed here that CAC were perfectly maintained at their initial performance
due to the larger building size.
Extension to all economic sectors, system types and EU climates

We covered the universe described by Table 5.4 as reliably as possible by extrapolating the results of the two
simulated buildings. For systems the description has been given previously. All systems have been brought
to the same AC quality since ventilation air is prepared centrally and then dealt to each zone. Some systems
that propose moreover the humidity control and represent a very small part of the market have been treated.
Table 5.4 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.(cooling) Trade
France TAC Water Cooled with air dist.(cooling) Houses
Germany TAC Water Cooled +air +hum.(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. The assumptions about the economic sectors (sizing and
demand) have been tuned thanks to the simulation of the Trade building. The specific features of the load
curves other than the ones simulated have been applied to the actual DOE load curves (offices). For instance
Education buildings have a load curve similar to offices but not in July and August. Hotels, 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. Trade buildings display similar trends. Hospitals have
an office section but they work 7 days a week. Houses are very distinct (later use of AC in the day) but not
very different on total. A few key figures about the trade sector, tuned on the Mall simulated in details : the
sizing of chillers is not very different from the office buildings (SE: 133 W/m2, LO: 75 W/m2, MI : 119
W/m2 against SE: 160 W/m2, LO: 80 W/m2, MI : 130 W/m2). The energy demand is 18% higher due to
difference in occupation scenarios.
Preliminary results for the cooling consumption of the office building

The electricity consumption results from the load (demand, as formulated by the type of distribution and
control) and from the efficiency of the equipment. We shall display separately the two aspects. When the two
aspects are combined, the resulting scatter of electricity consumption is of the order of a factor 2, for the
same comfort level (figure 5.8a).
Fig 5.8a Total cooling consumption for CAC systems per square meter
108
Reference office building unitary cooling consumptions
140,0
120,0
100,0
kWh/m2
80,0 London
Milan
60,0 Seville
40,0
20,0
0,0
S
F
ps
l
R
ts
op
st
um
n
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
ol
Ks
D
er
R
IL
oo
co
oo
e
SP
ai
rib
co
at
on
tri
H
ir
le
C
.(c
.(c
+
st
ity
+a
K&
w
is
t.(
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
w
ai
h
at
O
w
w
at
ir
r+
t
de
i
R
w
ith
LO
+a
w
de
ith
si
ai
de
w
d
si
w
ut
d
O
e
ith
ed
si
le
ut
d
ol
TW
d
le
ut
oo
w
O
le
l
oo
oo
rC
O
oo
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. The second important parameter comes from the
system type. Depending on each CAC system type, the load is treated either completely centrally or only
partially ; thus, 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. It is
shown on Figure 5.8b.
Figure 5.8b. Total cooling load for each system type for the 3 climates
LONDON MILAN SEVILLE

Total Cooling Load
140,0
120,0
100,0
kWh/m2
80,0
60,0
40,0
20,0
0,0
n
t
n
st
um
g)
g)
op
ps
s
em
g)
s
e
s
R
ro
is
tio
tio
lit
lit
ge
rg
VR
di
LE
lin
lin
lin
rd
nt
lo
Sp
Sp
+h
st
la
bu
ft
ka
r
co
oo
IL
te
b
ai
e
Sy
SP
co
oo
o
tr i
tri
ac
on
ti
ir
H
a
.(c
.( c
+
ity
ul
+a
is
t.(
s
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
Sm
er
PA
u
at
at
er
AC
w
+
ai
+h
h
at
w
w
O
at
ir
t
de
i
ith
R
+a
w
LO
w
de
ith
r
ai
si
de
w
d
+
w
si
ut
d
le
O
ith
le
si
d
ut
O
d
oo
le
TW
e
ut
w
oo
le
O
l
oo
oo
C
O
oo
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 5.8a and 5.8b enables to separate the part of
the consumption differences between CAC coming from efficiency.
The first part of these difference comes from the repartition of energy between fan, pumps and cooling that is
presented hereafter Fig 5.9.
109
Figure 5.9. Contribution of each piece of equipment in % of total consumption (per standardised square
meter for cooling, fans, pumps) in Seville.
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
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
D
er
de +hu ool
ir
IL
ut ter coo
co
e
rib
o
a
ac
at
ti
PS air
le
tri
ai t.(c
ul
+
at oled ate ity
st
C
w
ng
.(
s
t.(
M
di
TW e w er
+
is
+
is
al
C
LO er
m
er
ir
rd
si wat
rd
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
O
e
si
le
ut
ed
ol
O
d
ut
le
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, and even larger in the case of London. The improvement in Air based
systems should come from the improvement of secondary equipment and control.
The preceding analysis is based on the bare figures summarized in the 3 following tables respectively for
Seville, London and Milan. 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.
Table 5.5a Results for all systems, per physical square meter : SEER and SSEER, specific consumption in
kWh/m2 for cooling in Seville, ranked by order of merit
Comfort System Compressor Fan Pump T Cool SEER SSEER

11 TAC VRF 36,21 18,51 0,26 54,98 2,87 1,89
12 TAC PACK&SPlarge 54,4 3,85 0,26 58,52 1,91 1,78
15 TAC Multi Splits 54,4 3,85 0,26 58,52 1,91 1,78
16 TAC Splits 54,4 3,85 0,26 58,52 1,91 1,78
17 TAC Small packages 54,4 3,85 0,26 58,52 1,91 1,78
14 TAC RACs on one loop 46,01 8,06 6,24 60,31 2,26 1,72
18 TAC Single Ducts 63,33 3,85 0,26 67,44 1,64 1,54
7 TAC Outside water + water dist 36,65 18,51 18,74 73,9 3,14 1,56
10 TAC TWO LOOPS + CHILLER 36,69 18,51 18,74 73,94 3,14 1,56
1 TAC Air Cooled with water distribution 46,77 18,51 10,63 75,91 2,46 1,52
4 TAC Water Cooled + water dist.(cooling) 39,34 24,26 12,73 76,33 2,92 1,51
13 TAC Roof tops 49,89 44,44 0 94,33 2,09 2,09
2 TAC Air Cooled with air distribution 49,22 44,43 5,61 99,26 2,34 1,16
5 TAC Water Cooled with air dist.(cooling) 41,12 50,18 8,89 100,19 2,8 1,15
8 TAC Outside water + air dist 38,02 44,43 18,3 100,75 3,03 1,14
9 TAC Outside water +air +hum 49,43 48,88 18,3 116,6 2,33 0,99
6 TAC Water Cooled +air +hum.(cooling) 53,45 55,2 8,89 117,54 2,15 0,98
3 TAC Air Cooled with air +humidity control 63,99 48,88 5,61 118,47 1,8 0,97
Table 5.5b SEER and SSEER, specific consumption in kWh/m2 in London, ranked by order of merit
110
12 TAC PACK&SPlarge 7,97 0,42 0 8,39 1,96 1,86
15 TAC Multi Splits 7,97 0,42 0 8,39 1,96 1,86
16 TAC Splits 7,97 0,42 0 8,39 1,96 1,86
17 TAC Small packages 7,97 0,42 0 8,39 1,96 1,86
18 TAC Single Ducts 9,51 0,42 0 9,93 1,64 1,57
11 TAC VRF 3,87 7,86 0 11,73 4,03 1,33
13 TAC Roof tops 7,83 22,27 0 30,09 1,99 1,99
Table 5.5c SEER and SSEER, specific consumption in kWh/m2 in Milano, ranked by order of merit

12 TAC PACK&SPlarge 26,39 2,09 0,04 28,53 2,07 1,92
15 TAC Multi Splits 26,39 2,09 0,04 28,53 2,07 1,92
16 TAC Splits 26,39 2,09 0,04 28,53 2,07 1,92
17 TAC Small packages 26,39 2,09 0,04 28,53 2,07 1,92
11 TAC VRF 15,85 14,35 0,04 30,25 3,45 1,81
18 TAC Single Ducts 30,94 2,09 0,04 33,08 1,77 1,65
13 TAC Roof tops 29,61 36,98 0 66,59 1,85 1,85
Interestingly, the results are similar from one location to another. The decentralised systems have a large
benefit, 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. Even Single Ducts are better than any collective system. VRF are a good
system, in the middle of decentralised systems, except for Seville where they show a benefit. Wet cooling
towers and systems with two water loops display the same performance as the best centralised system, but
not overpass them. SEER and SSEER do not follow the ranking based on consumption, but this is due to
problems in the definition of “load” in DOE2 software with distinct systems. A regulation should be simply
based on electricity consumption if we want to avoid such misunderstandings. 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.
111
The simulations have then been extended to each of the 15 Member States by use of outside climatic
information from an extended data base. For instance, while the most extreme cooling loads are covered
through Seville, 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.
An analysis had to be done based on cooling and heating degree day data (CDD & HDD). Searching on
www.i-wex.com produced cooling and heating degree day data for a number of EU locations (CDD
threshold was 15.5 °C, HDD threshold temperature was 18.5 °C). 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, Table 5.5.
Table 5.5d Annual average energy consumption per m2 by EU country (kWh/m2/year), weighted for systems
and sectors
Cooling mode Heating mode Total

Compress fans pumps Boiler fans pumps Cooling Heating Cooling & Heating
or
Aus 12,3 12,6 1,2 114,9 28,3 3,8 26,1 147,1 173,2
Bel 9,3 9,4 0,8 108,5 23,8 3,2 19,5 135,6 155,0
Den 6,0 7,6 0,6 140,2 26,6 3,5 14,1 170,4 184,5
Fin 5,2 8,9 0,9 152,3 35,0 4,7 15,0 192,1 207,1
Fra 19,4 11,9 1,4 93,9 17,9 2,8 32,6 114,6 147,3
Ger 12,6 9,3 0,9 126,2 23,0 3,4 22,8 152,6 175,3
Gre 35,6 10,9 1,7 84,4 10,7 2,1 48,3 97,2 145,4
Ire 9,3 9,6 0,7 94,5 22,1 2,9 19,5 119,4 139,0
Ita 35,0 12,8 2,3 80,7 11,6 2,2 50,1 94,5 144,7
Lux 9,3 9,1 0,8 109,4 23,3 3,2 19,1 135,9 155,1
Neth 7,0 9,9 0,8 105,1 27,3 3,5 17,7 136,0 153,7
Por 36,2 12,4 1,0 83,2 11,0 1,8 49,7 96,0 145,7
Spa 56,9 20,8 3,8 22,2 5,6 0,7 81,5 28,5 110,0
Swe 5,2 8,9 0,8 152,4 34,9 4,7 14,9 192,1 207,0
UK 9,4 9,6 0,7 94,4 22,1 2,9 19,7 119,5 139,2
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.
5.3 Energy consumption in 1990, 1998, 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.
Figure 5.12 shows the first two values (cooling and associated heating consumption by technical type) for the
future.
112
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 5.13 gives details, country by country.
Figure 5.13. Total energy consumption by AC type in Europe in 2000 and 2020 for the three main quantities :
cooling, heating (if no reversibility); heating (with present reversibility rate)
113
Total consumption by country - 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 and 5.7 give the main values (EUR15) for the three functions. Note that gas is accounted for as a
secondary energy. (CO : Cooling, CH : Conventional heating, RH : Reverse Heating)
Table 5.6 Total energy demand generated by AC
Electricity demand (TWh) 1990 1995 2000 2005 2010 2015 2020
Electricity and Gas
Cooling function 22,879 33,683 51,636 78,103 94,727 109,631 114,579
(Electricity only)
Without REV.
With present REV. (El.)
Table 5.7 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
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 5.14
Figure 5.14. Electricity consumption by economic sector in each country in 2020
Total cooling by country / sector - 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
5.4 Global warming and other environmental impacts

Atmospheric pollution reduced to CO2
Most environmental impacts of A/C take place in the atmosphere: acid pollution, ozone depletion, green
house gases emission therefore we shall concentrate here on atmospheric emissions, mostly Refrigerants and
CO2. However we should first explain and justify this reduction of perspective by considering one by one
the effects not taken into account or simplified.
Let's consider first the issue of refrigerants. R22 is the most commonly used A/C refrigerant; however, 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, R407C, R-134a and R-410A (these refrigerants
are more or less compatible with the operating parameters of a traditional R22 unit). R407C, R134a and
R410A are the only refrigerant largely used in CAC directories for substituting R22. TEWI (total equivalent
warming impact) is the integrated index used to measure the global-warming impact of all gaseous
emissions, including those from direct and indirect sources.
What are the interactions between refrigerant change and energy efficiency? For an optimist, a higher energy
efficiency and a more environmentally benign refrigerant will both result in a lower contribution to global
115
warming. For a pessimist, 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). Fortunately, our study group
is not in charge of the issue of change of refrigerants. The direct contribution of A/C to global warming will
drop independently from our action, due to other policies on which we do not interfere.
Why aren't we considering other atmospheric emissions than global warming by CO2. Atmospheric pollution
from power plants is composed of dust, NOx and SO2, which all have a regional impact, and CO2 with a
world wide impact. We have assumed here that checking European regulation on acid pollution was not part
of our objectives. On this subject, the "ExternE" study gave recently values of the external costs of power
plant pollution that we could use in case of necessity. The consideration for CO2 is different; the Kyoto
protocol has been made recently, its full implementation in Europe is not yet achieved and the market has not
yet taken it into account; 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.
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. Since there is a European bubble, one can
assume one average CO2 content of the European kWh, set here at 350 gCO2/kWh, the marginal rate with
Combined Cycles which are likely to be installed in Summer peaking countries to cope with the new
demand. In fact, the average for OECD (440) and the exact figure for CO2 content per kWh are available for
each country, this can be taken into account in details if needed. Here we forget about other environmental
effects : radio-elements, accidents in the case of nuclear plants, etc because we have a marginal approach and
nuclear is not the marginal energy. Note that the external cost of the CC plants are among the lowest, except
nuclear plants. 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%, a trend which may have other causes or never happen.
Water use is another environmental impact of Air Conditioning to be taken into account. It's mostly the case
for water cooled chillers using cooling towers. They are used in the about 12% of cooled area, 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). We will not devote much time to the issue, simply estimate the total quantity and
take it into the cost (1-3 euros/m3) in the optimisation.
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 principal contribution to the greenhouse effect of the air conditioners comes from their
energy consumption. Actually, each kWh of electricity consumed implies a CO2 emission depending on the
specific utility plant in that country. 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.
The uncertainty in the evaluation of TEWI are the same as that for the GWP (+/-35%), for which we add the
uncertainty in the evaluation of the emissions and energy consumption. These evaluations strongly depend
on the quality of the data.
Two formulas can be used to evaluate the contributions. The most simple is written:
TEWI = [(GWP . m) + (E . b )] . n (1)
TEWI : kg of CO2 produced during the equipment lifetime.

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
Some people make a difference between the annual emission rate and the fluid recuperation at the end of the
appliance's life cycle, taking into account the fact that the recuperation has become mandatory in numerous
countries. Practically, leaks during recovery are close to leaks during one year of operation, and the simplest
equation is enough.
We propose the following values of yearly leak rates :

- for RAC, Packages etc. 1% of 0.3 kg/kW (some don't leak, some leak before during or after repair)
- for VRF 10% of 2 kg/kW (good reasons to leak despite a careful maintenance)
- for other centralised systems 4% of 0.6 kg/kW
The order of magnitude of the effect of system type computed for one square meter cooled is :
- for RAC, Packages etc. 0.08 TEWI units (kg CO2) due to leaks to be compared with about 100 indirect
TEWI due to electricity consumption over 15 years
- for VRF 5 TEWI units due to leaks to be compared with 100 for indirect TEWI
- for other centralised systems 0.6 TEWI units due to leaks to be compared with 100 for indirect TEWI.
The proposed conclusion is that the TEWI penalty of VRF could be taken into account in the economic
analysis, but not any other aspect . So our figures of impact are based on indirect CO2 emissions only.
Numerical results about CO2 emissions for cooling in Europe
Table 5.6 and Figure 5.15 show the results on the full stock. 33 Mt CO2 in 2010 may seem a small figure
compared with the projected total around 3800 for EUR-15 in 2010, 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). So they should not be forgotten.
Table 5.6 : National cooling CO2 emissions of AC by country for 2000, 2005, 2010, 2015 and 2020.
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 5.15 : National cooling CO2 emissions of AC by country for 2000, 2005, 2010, 2015 and 2020.
117
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, but limited if we compare them with other uses in buildings (heating, home
electronics, better lighting, etc.).
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, formation of droplets and desalination) at the rate of 3-4
kg per kWh of heat rejected, 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. Nonetheless the evaporation of this
water produces a small improvement in energy performance, of the order of a few kWh/year per square
metre cooled.
The water consumption should be taken into account to perform cycle cost analysis of CAC systems cooled
by water condensing chillers.
5.5 Heating, reversible or not

Here we are only interested in the heating of the cooled areas, not in heating in general. In the base
simulations, heating has been provided by a gas boiler. As an option, reversible use of the cooling equipment
has been considered. To heat reversibly or to heat with a boiler is not the most striking issue in the tables.
Tables 5.7a to 5.7c give the detailed results. 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, namely for air based systems. But
also that coming forth to full year totals, there is a compensation and the performance of the systems
becomes closer (the fan energy is fully recovered in winter for space heating). This is influenced by the fact
that we mix electricity and gas in the tables.
Table 5.7a specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in Seville; 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 25,37 5,99 1,97 H
10,15 33,33 18,11 98,53 83,31
2 TAC Air Cooled with air distribution 6,06 5,76 0,23 2,43 12,05 8,42 98,03 94,39
115,0 110,6
4 TAC Water Cooled + water dist.(cooling) 25,37 5,99 1,97 4
10,15 33,33 18,11 96,19 7
80,97
5 TAC Water Cooled with air dist.(cooling) 6,06 5,76 0,23 2,43 12,05 8,42 96,87 93,23
6 TAC Water Cooled +air +hum.(cooling) 7,28 6,34 0,23 2,91 13,84 9,48 111,5 107,1
7 TAC Outside water + water dist 25,37 5,99 1,97 0
10,15 33,33 18,48 93,58 3
78,73
118
8 TAC Outside water + air dist 6,06 5,76 0,23 2,43 12,05 8,79 96,86 93,59
9 TAC Outside water +air +hum 7,28 6,34 0,23 2,91 13,84 9,85 110,3 106,3
10 TAC TWO LOOPS + CHILLER 5,82 5,99 0,23 1,16 12,04 9,50 1
71,24 1
68,70
11 TAC VRF 26,93 5,99 0,23 6,72 33,15 12,71 88,14 67,69
12 TAC PACK&SPlarge 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
13 TAC Roof tops 5,91 5,75 0,23 3,03 11,90 8,78
106,2 103,1
14 TAC RACs on one loop 26,86 4,64 6,85 2
10,87 38,35 22,36 98,66 1
82,67
15 TAC Multi Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
16 TAC Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
17 TAC Small packages 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
18 TAC Single Ducts 26,93 4,50 0,23 11,68 31,66 16,17 99,10 83,62
Table 5.7b specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in London; 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 96,23 20,38 4,84 38,49 121,4 63,71 136,5 78,80
3 TAC Air Cooled with air +humidity control 65,43 43,53 1,71 26,17 110,6 71,41 143,2 103,9
4 TAC Water Cooled + water dist.(cooling) 96,23 20,38 4,84 38,49 121,4 63,71 135,9 78,23
7 TAC Outside water + water dist 96,23 20,38 4,84 38,49 121,4 69,16 137,1 84,84
10 TAC TWO LOOPS + CHILLER 61,51 20,38 1,71 14,45 83,60 45,12 98,10 59,62
11 TAC VRF 108,3 20,38 1,71 39,35 130,4 59,73 142,1 71,46
12 TAC PACK&SPlarge 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,48
13 TAC Roof tops 54,93 39,60 1,71 38,78 96,24 78,38 126,3 108,4
14 TAC RACs on one loop 106,4 9,44 7,00 47,17 122,8 63,60 132,9 73,71
15 TAC Multi Splits 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,48
16 TAC Splits 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,48
18 TAC Single Ducts 108,3 7,93 1,71 71,49 117,9 79,42 127,9 89,35
Table 5.7c specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in Milano; 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 88,83 14,94 4,48 35,53 108,2 54,95 144,4 91,17
3 TAC Air Cooled with air +humidity control 65,73 31,13 1,64 26,29 98,50 59,07 168,5 129,0
4 TAC Water Cooled + water dist.(cooling) 88,83 14,94 4,48 35,53 108,2 54,95 144,6 91,31
7 TAC Outside water + water dist 88,83 14,94 4,48 35,53 108,2 58,54 145,1 95,39
10 TAC TWO LOOPS + CHILLER 62,11 14,94 1,64 14,67 78,69 37,68 116,2 75,22
11 TAC VRF 99,77 14,94 1,64 42,20 116,3 57,14 146,6 87,39
12 TAC PACK&SPlarge 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,5
13 TAC Roof tops 55,81 28,30 1,64 48,84 85,75 77,14 152,3 143,7
14 TAC RACs on one loop 117,97 7,84 8,16 42,93 133,9 58,92 164,9 89,94
15 TAC Multi Splits 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,5
16 TAC Splits 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,5
18 TAC Single Ducts 99,77 6,26 1,64 75,23 107,6 81,48 140,7 114,5
119
A policy interpretation of the figures is only possible if something is assumed about the competition between
gas and electricity, either their cost, or their CO2 content or their “primary” energy value. The coefficient 2.2
is sufficient to represent about all aspects and gives the exact CO2 value (the ratio 400:180). If we
summarise the comparison to SCOP and CO2 emissions, and if we consider only the most frequent systems
inherited from the past in the present stock, table 5.8 gives us interesting indications.
Table 5.8 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
6.6 22.7 20.5
system
RAC with Primary Air
TAC
Split systems non rev. 6.6 22.7 20.5
Split systems reversible 1.2-1.5 6.3 28.8 29.2
CAC - Central Air Conditioners
TAC
Large packages (Roof
3.4 25.7 21.4
tops...) non rev
Large packages (Roof
1.3-2.7 3.5 28.8 30.8
tops...) reversible
Large splits with primary
6.6 22.7 20.4
air Non reversible
CAV reversible 1.6-1.7 3.8 25.2 20.7
CAV non rev 3.9 26.3 21.8
FCU reversible 1.3-1.6 7.2 25.5 22.0
FCU non rev. 7.8 27.4 23.7
WLHP (reversible) 1-1..5 8.9 25.4 23.6
VRF reversible 1.5-1.85 5.1 23.9 22.9
Very often more CO2 is emitted with a reversible system than with a non reversible one. The best heating
system is almost always a classic independent one. This is not a set of values against reversibility. Simply, 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. It bears the weight of the auxiliaries, and becomes
a less efficient realisation of electric heating. It’s an invitation to research : how to make air conditioning
systems – a growing social demand- sober in Winter? Reversibility is not an easy task, it’s one of the
challenges of the next chapter.
120
121
6. TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY EQUIPMENT
USED IN CAC
6.1 Energy-engineering analysis of chillers
Chiller prices as a function of the refrigerating fluid and EER
Chillers using the refrigerant R407C, which has been developed as a zero ODP substitute to R22, on average
have an identical energy performance and do not appear to be any more expensive to purchase, judging from
an analysis of their publicly quoted prices. Figure 6.1 shows the price of the equipment as a function of its
refrigerating power and refrigerant. From this it appears that there is no additional cost for chiller equipment
that uses R407C compared with those which use R22.
45000
Figure 6.1. Chiller cost versus cooling capacity, as a function of the refrigerant
40000
y = 230.98x + 5130.2
2
R = 0.7835
35000
45000
30000
y = 122.46x + 4944.3
2
R = 0.9228
40000
25000
Price
euro
20000
35000
15000
R 2 = 0.7719
10000
Air cooled
30000
5000
water cooled
Regression(Air
l d)
Regression
l d)
(water R 2 = 0.7823
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. The relationship between chiller capacity, condensing medium and price is shown in
Figure 6.2.
For small capacities, 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. The use of cooling with water
can only be economically justified in large capacity systems.
Figure 6.2. Chiller price vs. cooling capacity as a function of the type of condensing medium
Additional costs for reversibility
122
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
8
Total cost (Euros/m2)
ALCC17-800h
7
ALCC10-800h
ALCC17-400h
ALCC6-800h
ALCC10-400h
6
ALCC6-400h
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
overcost at 12.3 Euros/kW (+12.3%). Once again, manufacturers engineers may have other ways to reach
3.25 SEER, less expensive, but our objective was to find out if there is a margin for improvement. There is a
large margin for improvement and central solutions are not condemned in comparison with packaged units if
they improve their performance.
An example can be find in table 6.1 hereunder. For the bottom of the market we have used the same chiller
as for the stock : 2.5 EER with a poor part load control (uncontrolled screw). 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. Part load optimisation would bring 10 to 20% decrease of the total bill of the office
building simulated (SSEER from moving from 1,16 to 1,34, from 0,64 to 0,73, from 1,04 to 1,22). The
relative change at that bill level is half of what it is at chiller level due to the weight of auxiliaries remaining
unchanged.
Water cooled chillers

Another type of improvement could be to use a water cooled chiller (with a better nominal EER). Cooling
tower use is desirable in principle because the EER seems better. However there are pressures against CT
due to the legionella problem. There is also a water consumption, not to be forgotten. The question is : will
Cooling Towers remain used in the present 12% of cooled area, or will they disappear slowly? The
promoters of this solution should compare the LCC with and without CT, their water consumption and their
cost.
Then we have the use of natural water, with a higher initial cost. The solution expands so slowly that we
don’t have figures to prove any growth. Not only EER is better but it is kept most of the year due to the
constant natural resource. However there are strong electric auxiliaries for pumping and circulating the
water, namely in confined aquifers (the case represented here).
We have simulated for the previously defined office building both water cooled system, all conditions
remaining the same (figure 6.2). The water cooled chillers are a screw unit with nominal 3.31 EER for the
Table 6.1 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 100,19 84,82 100,75 84,80 99,26 85,97
London 31,78 28,23 32,60 29,37 32,77 28,66
Milan 69,64 59,66 69,66 59,87 70,49 60,15
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 1,15 1,36 1,14 1,36 1,16 1,34

London 0,66 0,74 0,64 0,71 0,64 0,73
Milan 1,05 1,23 1,05 1,22 1,04 1,22
The energy benefits of water compared with air appear small, 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.
What is important is the part load behaviour of the chiller not its type!
126
6.2 Engineering approach of the performance of Packaged units
For the most part the large packaged air conditioners, such as ‘roof tops’, found in the European market are
either identical to, or share the same technology as, models available for sale in wider international markets,
such as the USA. 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.
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. As recently as
1999 the non-binding ASHRAE 90.1 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. In 2000 the US DOE launched a revision process for the existing EPCA MEPS which
aims to set more stringent MEPS from 200X. Following the US MEPS development process a full techno-
economic energy engineering analysis has been conducted for large packaged air conditioners, which forms
the basis for the results reported in this section. 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 7.5 tons (26 kW), and 2) a roof-top unitary air conditioner having a cooling capacity of 15 tons
(52 kW).
A parallel analysis of the European market shows that the average cooling capacity of large packaged AC
units in the EU is 28.9 kW while that in the USA is 36.2 kW. Figure 6.5 shows the distribution of models by
cooling capacity in the two 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.
The only other significant difference in the products found on the two markets concerns the average energy
efficiency levels. As a result of the existing US regulations the minimum permissible EER for packaged AC
units with a cooling capacity between 19 and 39.5 kW is 2.61 W/W and for those with a cooling capacity
between 39.5 and 70.3 kW is 2.41 W/W. In 2003 the lowest efficiency unit which was active on the US
market had an EER of 2.5 W/W and the average efficiency was 2.9 W/W. The maximum EER level found on
the US market in 2003 was 4 W/W. The lowest EER considered in the US energy engineering analysis is
2.78 W/W for both the 26 kW and 52 kW units. By contrast the average efficiency of packaged units in the
EU market and within the Eurovent database was 2.46 W/W in 1998, the minimum EER was 1.78 W/W and
the maximum EER was 3.58 W/W. 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.
Figure 6.5. Share of large packaged air conditioners as a function of cooling capacity in the EU and US
markets (source: Eurovent and ARI databases)
127
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 goal of the US energy engineering analysis was to develop cost versus efficiency curves of large
packaged AC units to guide policy development. In particular the intention was to determine the life cycle
cost of packaged AC units as a function of their EER. The methodology used was as follows:
A total of eighteen large packaged AC units, representing several manufacturers and a wide range of
efficiency levels were examined and four units chosen. 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.
The cost model itemises ‘fixed’ factory expenses such as: equipment and plant depreciation, tooling
amortisation, equipment maintenance, utilities, indirect labour, cost of capital and overhead labour and
‘variable’ factory expenses such as: manufactured materials, purchased materials, fabrication labour,
assembly labour, shipping and indirect materials. It also itemises corporate expenses such as: research and
development, net profits, general & administration costs, warranty costs, taxes and sales and marketing costs.
The inputs to the cost model were reviewed by individual manufacturers and the values adjusted if
appropriate. The cost efficiency relationships established in this way were found to follow an exponential
growth curve, thus the data for each manufacturer was regressed to a exponential curve.
Each of these curves was in turn regressed to a single market-average exponential curve to give the results
shown in Figures 6.6 and 6.7 below. These figures also show the upper and lower 95% confidence intervals
as broken dotted lines about the average line (solid).
Figure 6.6. 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
$1,000
Eq $750
ui
p
m
$500
en
t
Co
st $250
De
lta
($) $0
-$250
8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5
EER (kW/ton)
Figure 6.7. The incremental cost of 52 kWc packaged air conditioners as a function of their efficiency on the
US market in 2001 (US$) (source: TIAX 2002)
$2,000
Eq $1,500
ui
p
m $1,000
en
t
Co
$500
st
De
lta
($) $0
-$500
8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5
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 .
Life cycle cost analysis

This section draws heavily on LBNL 2002.
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,
distributors and mechanical contractors were determined (the latter separately for small and large contractors
operating on new- or replacement-construction markets). 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, distributor and contractor mark-ups to the incremental ex-factory equipment costs
estimated in the energy-engineering analysis.
The results indicate that on average the life cycle cost minimum in the USA occurs for an EER of 11.5 Btu-
hr/W (=3.37 W/W) for both the 26 and 52 kW units. Data has been gathered in the EU on: the average
efficiency of packaged AC systems; the typical price and installation costs of packaged systems as a function
of their cooling capacity; typical building load factors and marginal electricity prices, that are all necessary
inputs if the life cycle costs of European packaged systems are to be determined. However, in order to
129
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. The assumptions used in this calculation imply that not only the trend in relative
manufacturing cost vs. efficiency for packaged AC units is the same in the EU as in the USA, but also that
the relative trends in distribution, installation and maintenance costs are the same. However, when the life
cycle cost results produced in this manner were compared with those produced for large packaged systems
with an EER of 2.25W/W derived from the values quoted in the tables of section 2.2, the results were found
to agree to within 0.3%! This implies that the adapted US equipment cost versus efficiency relationships are
reliable for use in the EU.
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. The results of the analysis taking these factors
into account is shown in Figure 6.13 for the 26kWc unit, which is most representative of the EU market.
They show that the life cycle cost minimum occurs for large packaged units with an EER of 3.22 W/W when
a 6% real discount rate is applied.
The comparable results for the 52kWc unit are shown in Figure 6.14. Although the overall life cycle cost per
kW are lower for the 52 kW unit the minimum still occurs for an EER of 3.22 W/W.
Figure 6.13: Estimated average life cycle cost per m2 of cooled space per year vs. 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 2.4 2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
Figure 6.14: Estimated average life cycle cost per m2 of cooled space per year vs. EER for large packaged
air conditioners on the EU market (based on a 52kWc unit)
130
€17
€16
€15
ALLC (€/m2/yr
€14
€13
€12
€11
2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
6.3 Energy Efficiency of Air Handling Units seen as tradable goods

Major European manufacturers, 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, air conditioning and
refrigeration. The fact is that the energy consumption (operating) costs represent more than 80% of the total
cost during the lifetime of the unit. Investment and maintenance together make only about 20% of the Life
Cycle Costs. Decreasing the Life Cycle Cost means in fact in the same time an important energy saving.
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,
essentially because large differences exist between European countries concerning climate conditions and
also energy prices (electricity, energy for heating and cooling). 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.

In many cases it may be better to select a larger unit in order to reduce the operating costs. The internal
pressure drop will decrease and efficiency of other functioning parts will be better. Even if the investment
cost increases with a larger unit, the pay back will be better.
Fans integrated in AHU

A well designed duct system with low pressure losses will greatly reduce the electrical consumption of fans.
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 ).
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. 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;
in London the figures become smaller but the difference larger (25% in cooling demand).
Table 6.3 indicates the minimum efficiency levels for fans recommended by Eurovent ( Efficiency figures
valid for most common operation hours (around 3000 hours/year). Higher efficiency values are strongly
recommended for longer operation periods.
131
Table 6.3. 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
1,00 3600 30% 31% 32% 33% 35% 36% 37% 39% 39% 40% 40%
1,25 4500 31% 32% 33% 34% 35% 37% 38% 39% 40% 40% 41%
1,60 5760 32% 33% 34% 35% 36% 38% 39% 40% 41% 42% 42%
2,00 7200 33% 34% 35% 36% 38% 39% 40% 41% 42% 43% 43%
2,50 9000 34% 35% 36% 37% 39% 40% 42% 43% 43% 44% 45%
3,15 11340 35% 36% 38% 39% 40% 42% 43% 44% 45% 45% 46%
4,00 14400 37% 38% 39% 40% 42% 43% 45% 46% 46% 47% 48%
5,00 18000 39% 40% 41% 42% 43% 45% 46% 47% 48% 48% 49%
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 36000 43% 44% 45% 47% 48% 49% 51% 52% 53% 53% 54%
12,50 45000 44% 45% 47% 48% 49% 51% 52% 53% 54% 54% 55%
16,00 57600 45% 46% 48% 49% 50% 52% 53% 54% 55% 55% 56%
20,00 72000 46% 47% 48% 49% 51% 52% 53% 54% 55% 56% 56%
25,00 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, cold or hot climates.
A certain level of heat recovery is recommended by Eurovent, Table 6.4, by taking advantage of the results
of the Eurovent LCC study.
132
Table 6.4. Levels of heat recovery as recommended by Eurovent, 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* 65% 70% 75%
- max. pressure drop 200 Pa 150 Pa 125 Pa
Plate heat exchanger without bypass
Plate heat exchanger with bypass
Heat pipe
Run around coil loop
* 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. Fans
using the inverter, give possibility to maintain the optimum speed for different air flow rate demands during
the day.
In the Eurovent Recommendation there are many examples for various European conditions. 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 0.6. Fan consumption remains constant. Heating saving potential is
very important in every climate and it is over 40%. Cooling savings are less important and only significant in
hot summer locations (3% in Seville). The combination between heat recovery and free cooling has been
proven by simulation to be a simple addition of savings.
133
7. 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, b) evaluate the relative energy
performance of different systems to enable fair comparisons to be made between systems.
7.1 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, noise,
geometry of building) and the comparisons could in principle only take place between 2 or 3 systems, once
the building is designed. 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. To give an idea of the cost of comfort,
we computed here the ALCC of a few systems installed in an hypothetical 2000 m2 building (800 hours
equivalent, optimised SSEER).
Table 7.1 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 2,25 248000 128000 20,75
Packaged systems 2,25 188000 128000 16,77
(under windows)
CAC - Central Air

Conditioners
TAC
Large packages (Roof 2,25 130000 128000 12,91
tops...)
Large splits with primary 2,25 274000 128000 22,48
CAV 1,30 336000 221538 34,76
VAV 1,70 352000 169412 34,33
2P FCU 2,00 318000 144000 27,53
4P FCU 2,00 326000 144000 28,10
WLHP 2,40 200000 120000 17,30
VRF with primary air 2,80 348000 102857 31,78
Note: Sizing = 120W/m² for CAC, 240 for RAC
134
Table 7.2 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 2,25 220000 128000 16,69
Packaged systems 2,25 160000 128000 13,30
(under windows)
CAC - Central Air

Conditioners
TC
Large splits 2,25 186000 128000 14,77
2P FCU 2,00 220000 144000 17,23
4P FCU 2,00 228000 144000 17,68
WLHP 2,40 100000 120000 9,65
VRF without primary air 2,80 260000 102857 22,01
Note: Sizing = 120W/m² for CAC, 240 for RAC
Comparison of costs and sensitivities
Note that the solutions not providing total air conditioning but just cooling, cost 5-6 Euros less than the
others. 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 sensitivity to electricity price is significant : –2/+ 3.5 Euros per year and square meter if we consider the
extremes of the prices on the EU market. Energy represents 30 to 50% of total expenditures. The values
obtained here seem very high compared with what we can read in some places. All investment has been
taken as amortised over 15 years which is a heavy assumption, specially if we think of the systems providing
ventilation. They have two functions and we could as well decide to make economic calculations on the
cooling function alone.
7.2 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. The overall efficiency of
the air-handling system relates in principle to the efficiency of each of its consituent components, the
operating mode, system configuration and operating conditions, the efficiency of air diffusers. in addition the
efficiency depends on ducting: specific pressure drop rating, the air tightness rating and the influence of
thermal insulation. Fan specific consumption depends not only on fan efficiency, but also on pressure drop in
the distribution system. 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, duct
sizing can be very important (and of course, filter selection). Moreover the importance of full and part-load
operation was stressed and led us to detailed computer simulation (DOE software).
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), 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
135
comes first and leads to more air changes with the outside, and to controlled ventilation. As a result Local
extraction (LE) is the dominant feature in some countries, while V (central ventilation) dominates others.
Obviously V allows a better air quality (dust, temperature, etc.) but is more costly. The energy impact of the
two philosophies is large but not really part of our study, but the capacity of heat recovery being very
different in the two situations, we have to consider them. Central ventilation (V) has been the base of our
study.
There are also in some countries obligations of ventilation in cascade, the exhaust taking place in the “very
polluted” rooms. We have not investigated further this option.
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. The flow rate is lower and you have to create the exhaust
air circulation to the heat recovery station at a cost.
Motors and fans efficiency

Overall Fans efficiency can be treated with a power/flow ratio like 0.25 W/(m3/h) for improved systems
against 0.75 W/(m3/h) for the worst. Note that W/(m3/h) is a measure of Delta P/efficiency; it includes the
fan as a component and the design effort on the air circuit. This is the normative parameter in the USA. You
cannot install air distribution systems if SPF is not under a certain value : 0.47 W/(m3/h) for CAV (SPF =
1.7 W/(l/s)) and 0.57 W/(m3/h) for VAV (SPF = 2.05 W/(l/s)). A similar rule exists in the USA for the
ventilation aspect of rooftops, etc. More speculatively, parameters like the power/flow ratio or the combined
efficiencies of motors/fans used by Eurovent in AHU should - in principle - be applied to the whole air
supply (and extract) system in a proper building thermal regulation.
For existing CAV system type, we used an average specific consumption of 0.47 W/m³/h. In fact, we assume
that a 15% reduction of SC (0.4 W/m³/h) may be achieved by the use of high efficiency fans with an overcost
of 2 Euros/fan kW. Conclusions of DOE simulations: logically, fan consumption suffers a 15% reduction but
also there is a decrease in cooling due to fan heat that ranges from 4 to 12.5% and that we have taken as 8%
on average.
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), even if its market share is now declining. The direct
comparison with DOE simulation of the effects in Europe of variable flow (option called VAV, Variable Air
Volume, in American English) has been possible. However fans are different in the two systems, as in the
US regulations (CAV=0.47 W/(m3/h) and VAV=0.57 W(/m3/h). 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. In this section we study the sole active unit of HVAC air-side.
As previously commented, every system type but VAV is equipped with constant air volume fans. Efficiency
index is specific consumption and energy use will be expressed in kWh/m².
This end-use consumption is almost constant along the year for CAV fans, see figure 7.1. The only
difference is due to the number of working days per month. This is not the case for VAV systems where
differences are due to flow regulation. As an example we represent monthly fan consumption for CAV and
VAV system types.
Figure 7.1 Fans energy over months in a CAV and in a VAV system
136
CAV and VAV fan consumption for SEVILLE
5
kWh/m²
3
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
CAV 4,39 3,79 4,39 4,19 4,39 4,19 4,19 4,59 3,79 4,39 3,99 3,99
VAV 0,57 0,5 0,54 0,53 0,66 0,98 1,63 1,93 1,11 0,78 0,49 0,5
VAV consumption reduction percentage is higher during winter because of lower loads, VAV boxes operate
at minimum flow, which corresponds to outdoor ventilation air ratio. During the summer there is also a large
saving potential that is always higher than 60 %.
Conclusions: VAV saving percentage varies from 75 to 80 % despite specific consumption for fans (0.57
W/(m³/h)) is higher than that for CAV (0.47 W(/m³/h)). 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. On average we could say that VAV save 80% of fan energy in Europe. There are
discussions about the reality of those results on the field due to O&M issues : balancing, controllers tuning,
etc. Conservatively, 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. Europe should take
such a measure when applying its “Energy Performance of Buildings” new directive.
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. It seems not to present any specific
potential for our study because they are very uncommon.
Terminal reheat issues

On air-side it’s also possible to provide evidence (simulation or literature) about the order of magnitude of
terminal reheat ; some think it should be completely banned in regulations (except when provided from some
renewable source of energy like condenser heat).
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. The cost and performance
of heat recovery from condenser being known, this option could enter the C/B-analysis.
Air Side Free Cooling (Economiser)

Air side Free Cooling (FC) is the key option in a cost benefit analysis : small cost, large potential. Free
cooling simulations in DOE have been complemented with cost functions. The present use of this feature in
EU systems is far from 100%. Many regulations make it more or less compulsory (USA, Portugal). It is
related also with the choice of an optimal blowing temperature : the higher , the better for EER and for FC.
Of course, free cooling is very climate and control dependent. We are using by default the classic
temperature control. There is a very limited overcost of 2% of overall system costs (ducting, new
connections, control) for a reduction of chiller electricity by 20% (no auxiliaries increase). The potential
given by simulation is far higher but we have taken this conservative assumption due to potential O&M
uncertainties.
137
The real figures given by DOE show that the cooling saving potential is strongly dependent on climate.
Oceanic weathers like London offer a more than 80 % reduction of cooling consumption. In any case, and
even for very hot climates like Seville, saving potential is over 20%. Pumping consumption is also reduced
due to cooling load decrease and pumps "on demand" control.
Quality of Air Diffusion

We have in the Italian UNI standard a set of interesting figures, but no cost data. The option of
“displacement” allows to lower significantly the loads. In reality a large portion of the loads doesn’t impact
comfort because they are too high in the room. 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, we have a very energy efficient
treatment of demand. Research has not yet produced perfectly consistent values for this approach which
already works in practice (a few percent of new installations). Displacement is now reported to be less costly
by 15% in installation costs but, as in the case of chilled ceilings, the system is unable to answer to the total
heating needs in winter and an overcost appears somewhere else.
AHU improvement
We should promote the investigation on the development of low pressure drop AHU components (filters,
coils, heat exchangers, sound traps, etc.) since the larger part of the fan pressure is dissipated in the AHU.
The size of the AHU should be determined after a LCC analysis considering the annual operating hours, the
fan energy consumption and the unit cost of the options for the air speed in the unit (2.0 m/s, 2.5 m/s, 3.0
m/s and 3.5 m/s),
The quality of the duct system should be controlled in some way, for example;
• Mandatory air leakage tests (air leakages of more than 20% are common),
• Use of low pressure drop connection pieces,
• Air balance criteria and equipment,
• Velocity or pressure drops limits,

The piping systems should be designed so that all AHUs should not have balance valves and flow control
valves. Every AHU should be equipped with a small variable speed pump that delivered the correct water
flow, at each moment (the control signal that normally goes to the control valve would go to the variable
speed pump). This measure eliminates the energy pressure dissipated in the balance valves and the flow
control valves.
7.3 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. For instance moving from the classic 7/12 °C regime to 8/13 °C will increase average from 9.5 °C to
10.5 °C. It will at the same time reduce temperature difference between room and FCU from 12.5 K to 11.5
K (with 22 °C inside) and increase requested area. 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.
Radiant panels or beams are one step further in the same direction. They allow an increase in distribution
temperature by using large areas for heat exchange. How to consider the option of radiant panels? since the
benefit claimed comes from the change of temperature regime in the water distribution equipment, costs and
benefits have been evaluated by extrapolating the temperature effect to the new data both in summer and
winter. There is a second order effect coming from the fact that part of the cooling is radiant, entering in a
different manner the equation of human comfort. There is also the possibility to operate them directly from a
cooling tower where cold is generated by evaporation. We have investigated only the first phenomenon
(temperature regime) and only in one case.
138
In a special configuration of the ceiling panels, they are mounted under a plenum with injected air, likely for
increasing the relatively small cooling capacity. If the cooling capacity of the primary air and the capacity of
the chilled ceiling are enough together to ‘treat’ the load, the cost seems the same as the one with FCU. 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.
In another realisation of the radiant system, recently patented by the firm Van Holsteijn en Kemna (VHK), 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. When radiant cooling is not enough, ventilation is
started. Entering water temperatures under room temperatures by only 10K would be enough to deliver 200
W/m2 of radiator. This supposes a decrease in Chiller consumption and an increase in Pump consumption.
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.
We have simulated with DOE software this option. The assumption was to determine performance
improvement with a small ΔT difference in a FCU4P system (8/13°C instead of 7/12), since the relative
returns will diminish if we move further. The electricity saving is very low (0.1 to 0.6%) and we abandoned
this solution in the study.
Design of flow in water circulation

Efficiency of pumps for water circulation is the first obvious issue. Pump efficiency is the result of motor
and mechanical effects, and present average values are 0.8 and 0.62, respectively. If we improve pumps
efficiencies, we will save pumping energy. High performance values are 0.85 and 0.67 for motor and
mechanical efficiencies, respectively. Pumping saving potential of this measure is 13%. Total saving
potential is low (from 0.7 to 1.7%) due to the limited importance of pumping energy.
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. On the examples taken from company documents, there is a significant benefit at system level
from doing this.
The piping systems could also be designed so that all equipment should not have balance valves and flow
control valves. Every equipment should be equipped with a small variable speed pump that delivered the
correct water flow, at each moment (the control signal that normally goes to the control valve would go to
the variable speed pump). This measure eliminates the energy pressure dissipated in the balance valves and
the flow control valves.
Our base case assumption is that every system has three-way valves and for this reason circulation loop and
pump flow are constant. In a few DOE simulations, we valued the saving potential of using two-way valves
and variable flow pumps, that is, pump is controlled by a variable frequency drive. Decrease in cooling
consumption due to pump heat ranges from 3.6 to 5.1% (less demand since the heat is not dissipated by the
pumps). High pumping savings range from 60 to 72% (relative to pumps consumption).
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. It may have a low energy penalty if done properly :
this is a behavioural more than technical issue, difficult to regulate because very dependent on experience
gained with a specific building and climate. In principle it could be simulated by generating extreme
behaviours, but we decided this was too uncertain and we abandoned the idea of simulating or regulating this
type of intervention.
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. Even if temperature cannot be increased at design
conditions, it’s a good practice to “reset” it off design. Two types of information are available : load and
outdoors temperature. The electricity saving is very low (0.1 to 1.3%) and we abandoned this solution in the
study.
Influence of terminal equipment
139
Improvement of flow efficiency in FCU and AHU water circuits is possible in various ways : Variable flow
pump, better Pump efficiencies, adequate Pump performance curve. A remarkable option at system level is
to go for “local pumping systems”. Terminal units demand what they need and the evaporator ‘sees’ a
variable flow. Evaporators can accept flow rates decreased by 40% typically. The gains in pumping cost
become significant.
We have considered the basic Air and Water System types (Two pipes fan-coil, Four pipes fan-coil, Water
loop heat pump). 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, one of the reversibility approaches.
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- can be estimated. 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; 4P is around 15%). Also we suspect that the electrical resistance is used for main space heating not just
for this adjustment…..so it’s the main enemy of reversibility.
Simultaneous demand of heating and cooling
Some systems have a capacity to transfer heat from one zone to another. Such advanced multizone systems
can be justified by its actual benefits. We have gathered some elements on advanced multizone strategies
(WLHP, TWL, VRF,…). The system using RAC on a water loop (WLHP) is relatively frequent (1.5 % of
total cooled area) and presents specific energy conservation features : transfer from one zone to another,
high EER and COP year round, etc. It seems a relatively frequent solution in commercial malls because it
allows individual metering of consumption by each user.
In the same way, 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. Finally
VRF is one step further in the same direction. It is one way of operating at variable speed (see part on
packaged systems). But it is also an interesting system for transfer between zones demanding heat and cold
(but this not always realised).
The DOE simulation allowed us to understand the real order of magnitude of simultaneous heating and
cooling. In this relatively complex office building where internal and external zones are treated separately,
where various facades receive differently the sun, the effect corresponds to only a few percent of the
demand. More precisely, we have computed for each hour with simultaneity the lowest of the two quantities :
cooling demand, heating demand and expressed it in percent of demand, either cooling or heating, table 7.3.
Table 7.3 Importance of simultaneous heating and cooling
In % of heating
CAC system LO MI SE
CAV -0.26% -0.23% -4.90%
VAV -0.92% -0.92% -3.03%
FC2P -6.24% -5.89% -17.61%
FC4P -7.45% -6.16% -23.62%
PACK -5.33% -4.67% -18.17%
WLHP -6.19% -5.47% -21.49%
In % of cooling
CAC system LO MI SE
CAV 0.59% 0.15% 0.21%
VAV 5.91% 1.24% 0.50%
FC2P 26.88% 7.42% 3.62%
FC4P 29.66% 7.73% 4.78%
PACK 32.22% 7.41% 4.08%
WLHP 31.06% 7.74% 4.53%
140
The economics of the transfer are as favourable as expected. The homogeneity of figures between air
systems on one hand and all other systems on the other hand is interesting.
Heat rejection
Cooling tower fans should have variable speed drives and, in systems with more than one cooling tower, all
cooling towers should work simultaneously at all times (this strategy reduces drastically the cooling tower
fan energy consumption) Temperature control should be modified.
Using natural water –river, ground water, etc.- 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. In some
circumstances the chiller becomes useless and the natural water can cool directly the building (see system
TWL as an example). Control is easy since underground temperatures are constant. Costs and administrative
problems are reported as enormous in Italy and Spain while France maintains such a policy (Aquapac). There
is also the possibility to generate DHW (Domestic Hot Water) at a small cost from condensing heat.
7.4 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. The designers need the whole range of solution to cover the domain of
geometries and air quality requirements. 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.
Figure7.1 Bottleneck of air systems
ALCC Euros/m2/year
air
packages water
rac
SPECIFIC CONSUMPTION
kWh/m2/YEAR
The result of optimisation
Table 7.2 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
pumps (MWh) (0,10 (0,06 (0,17
(MWh) E/kWh) E/kWh) E/kWh)
0 – CAV 121,67 37,76 159,43 1 053 640 30,89 29,84 32,73
1 – VAV-50%/ 121,67 18,88 140,55 1 090 090 31,56 30,63 33,17
+6 E/m2
2 – FC –20% 97,34 37,76 135,10 1 074 710 31,05 30,16 32,61
/+2 Euros/m2
3 – Fans –8% 111,94 32,10 144,03 1 054 140 30,65 29,70 32,31
;–15 %; +2
E/kW
4 – HR 60% - 118,02 37,76 155,78 1 090 090 31,81 30,78 33,60
200Pa; -3%
+6 E/m2
5 –Lower HL 121,67 35,12 156,79 1 126 540 32,80 31,77 34,61
in AHU –
7%/+12E/m2
6 –Optimised 73,00 37,76 110,76 1 062 390 30,32 29,59 31,60
chiller – see
chapter 6
After sorting and combinations, the optimal trajectory of improvement is given in figure 7.3.
Figure 7.3 Optimising with 6, 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% 20,00% 40,00% 60,00% 80,00% 100,00% 120,00%

% of reference
The optimum if very flat, specially if we get interested with the highest cost electricity. The regulatory
measure could be taken anywhere between a 0% and a 60% reduction without generating overcosts (in the
LCC definition) in Seville. The optimal regulation would request a 50% cut in electricity consumption
compared with a standard CAV design.
142
8. EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE
8.1 The importance and nature of part-load management measures
Importance of establishing a EU method about part load

Until now, as described in chapter 2.5, the chillers and other vapour compression cycles have been
tested using the full load standard [CEN, 1997] for heating and [CEN, 1998] for cooling. This couple of
standards has to be replaced soon, by its own revision [CEN, 2002]. Consumption is not governed by full
load EER given in such a standard but by the average part load EER, called often a Seasonal EER (SEER).
Such a figure is largely available on the US market and is called there an IPLV (Integrated Part Load Value).
The one we are looking for should represent the universe of EU buildings and climates, hence the proposed
name, the ESEER (European SEER). Only a ESEER largely agreed can be accepted as a basis for comparing
chillers, or grading chillers from A to G in an undiscussable manner. We shall propose such an ESEER in the
coming pages.
An extension to the new full load EU standard contains some information about part load testing, it
is an EnV, 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. In Italy, however, a part load performance
standard has already been defined and accepted [UNI, 2002]. Part load performance is tested at different part
load ratios, defined as the ratio of the cooling capacity of one stage to the full load capacity stage. The
evolution of the efficiency with the part load ratio is still a subject for research and we have proposed here
original results.
Given that the US-IPLV climatic conditions are not relevant for Europe, 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. The resulting index is called EMPE. The EMPE methodology is not
different from the IPLV one. For a large set of modelled chillers, a comparison is drawn between EMPE and
IPLV figures, 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). The constraint is to minimize
the testing time while ensuring maximum precision, 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 new ESEER method is compared
with the US-IPLV and EMPE proposal under both respects : time spent and accuracy.
The potential gain associated with part load management is high (for instance +30 % in EER, i.e. -
30 % in electricity consumption in our chapter 6.1 optimisation exercise). As long as a good method is not
agreed, 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. A good SEER definition is the essential tool for achieving actual and comparable gains not artefacts.
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. We need to describe the means of
reducing the capacity to understand the reduced temperature and the part capacity behaviours. Both
depend on the kind of compression circuit. The compressors treated hereafter are of three types,
reciprocating, screw or scroll compressors. About part load behaviour of centrifugal chillers we give only
qualitative indications.
Centrifugal compressors at part load
143
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). In a centrifugal compressor, before the impeller, the inlet
vane guide enables to create more or less swirl to reduce capacity to match the load. This is a mechanical
type of unloading. However, the two stages compressor enable to unload at a lower step. Speed of rotation :
centrifugal chillers cannot be operated for small flow rates since the rotation speed needed would be too
high. Part load : the surging phenomena occurs at low part loads, 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. There are different modes for unloading the centrifugal chillers : prerotation inlet vane guides,
variable speed. These two are the more common ones. However, when the load becomes inferior to the
surging load, to enable cutting off the compressor, a hot gas bypass strategy is adopted leading to still poorer
performances at very low loads. By associating in series two compressors the surge limit goes under the
single compressor one (10% instead of 20 or 25% load). The part load performances seem always less than
full load ones in what we have investigated.
Reciprocating compressors at part load

The reciprocating compressor owns a spring valve at inlet and outlet. At suction, the inlet valve remains open
as long as the pressure in the chamber is lower than the suction pressure. The valve at leaving opens only
when the pressure in the chamber reaches the pressure of condensation. At that time, the end of the piston
allows gas into the high pressure side. Thus, when external conditions vary, this compressor adapts its
discharge and evaporative pressure to external conditions. This is the temperature aspect of part load.
The other treatment of part load is through capacity reduction. Mechanical realization of part capacity for a
compressor with four pistons and two stages of compression is :
• only two pistons compress the refrigerant, the valves of both the others remaining open; the fluid
which passes in the pistons in open position is pumped. This induces pumping losses.
• or only two pistons compress the refrigerant, the valves of both the others being closed; this induces
a heating of the engine.
Screw compressors at part load

In its basic configuration, contrary to the reciprocating compressor, the compressor has no means of adapting
the pressure of exit of compression to the pressure of condensation. Thus, any difference between the
discharge pressure and the condensation pressure is synonymous of energy losses.
For part capacity behavior, a slide valve or an equivalent steps system (corresponding to discrete steps for
the bypass) is used to control the capacity. 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. This slide valve (Figure 8.1) is
generally controlled by discrete steps. In practice, the following stage is called variable Vi. 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. Hence, 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, and thus the cooling capacity of the
chiller. The last option should be to use a variable speed drive for the motor. Some solutions with variable
drive speed exist on the market. But the option is scarcely expanded because of its cost, given that reducing
the speed of this compressor decreases the tightness of the lobes of the screw compressors, forcing the
manufacturers to increase the full load speed to be able to reduce it at part capacity operation.
Figure 8.1 : slide valve position experiencing low Vi and high Vi (a), and Vi variable (b), from [PILL85].
144
Scroll compressors at part load
The rationale is about the same one as for the basic screw : nor unloading is available for chiller applications,
neither adaptation to varying condensing pressure. Generally, one uses several compressors in parallel on the
same circuit and make them cycle. For two scroll compressors on the same circuit, two capacity steps plus
the full load are available if nominal capacities of each of the two compressors differ.
Thus, different technologies are used to control part capacity stages. Depending on the kind of compressor
circuit, one can find unloading by varying the number of available circuits or by varying the flow rate in one
circuit. To perform this latter control of the refrigerant flow in the cycle, one can use variable speed drive
(for screw chillers only in our scope), variable Vi unloading (for screw chillers only), unloading of
multistage compressor (screw or reciprocating), or shutting down a compressor over two or more (sole
option for scroll, available on screw and reciprocating as a supplementary mean).
Staging of Part capacity (control issues)

Generally, capacity staged chillers are controlled using a water inlet or water outlet set point control with a
dead band. Stages are successively triggered when water temperature increases and moves apart from the set
point. 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. Figure 8.2 shows the control scheme for a five step capacity chiller on the inlet
water temperature.
Figure 8.2. Typical control of a 5 capacity stages chiller in the cooling mode.
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, the water temperature varies more or less between 6 and 8°C for all stages,
depending on the water loop inertia and on the condensing temperature that modifies the cooling capacity of
stages.
We will assume here perfect control, the one represented by the scheme Figure 8.2, even if some
experimental testing of dynamic capabilities of chillers have shown that chillers did not always behaved this
way [AFCE, 2002]. However, dynamic testing installations are not available and would need long debates to
be specified and then adapted by certifying laboratories.
It has also been observed that most of the time, set point control temperatures were not respected, but
differed by more or less 1°C and sometimes even more from user selected values. For correct measurements
of inlet and outlet water temperature (like the ones used in standardised testing), either very long straight
pipes are needed so that a homogeneous flow may be reached before measurements or, pieces of equipment
have to be installed at the inlet and outlet of the evaporator to enhance the turbulence. On installed units
such equipments are not installed, water temperature measurements do not correspond to real temperatures
and the control behaviour can be far different from Figure 8.2.
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, the chiller will operate on each one of the two
neighbouring steps ; the cooling load is the weighted average of the two steps cooling capacities CC for the
same inlet condensing and outlet temperature. The corresponding operating times for each capacity step
145
enable to determine the electric power absorbed EP and thus the efficiency for each hour. Yearly efficiency
is calculated with Equation (1).
(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.
When the cooling load is lower than the smallest capacity the equipment can deliver, the chiller operates
only part of the time, thus fitting its cooling capacity to the load. In that case, each starting is an energy loss.
At each starting, the compressor has to establish the pressure difference between low and high pressure sides.
The unit only begins to cool water when the average refrigerant evaporating temperature is lower than the
average water temperature. Then, the superheating of the refrigerant has to stabilize : only at that time the
full capacity of the step is reached. On the contrary, establishing the full electric power is quite
instantaneous. This leads to an energy loss at the starting of the chiller. As a consequence of a review of all
existing experimental evidence, we selected (figure 8.3) the Italian standard [UNI, 2002] Equation named
(2) hereunder.
Figure 8.3. 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.
Degradation of the reduced efficiency versus the part load ratio

(same sources temperatures )
Ratio of part load efficiency to
1
0.9
full load efficiency
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Ratio of part load capacity to full load capacity (same source
temperatures )
The corresponding curve is represented on the figure 8.3, it corresponds to the following formula :
CC
EER = CC FL (2)
EER FL CC .C cyc + 1 − C cyc
CC FL
With Ccyc=0.9.
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.
Ccyc=0.9 is a proposed default coefficient. 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.
High pressure control at part load
The high pressure cannot go too low. It is generally maintained high enough by controlling the air flow rate
for air cooled chillers. Classical control consists of maintaining the high pressure above 15 bar by cycling the
146
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. Decreasing the flow rate at the condenser increases the high
pressure and thus decreases the performance.
The impact of this phenomena has been measured while carrying over a test campaign on a scroll unit. The
unit is divided into 2 distinct and symmetrical refrigerating circuits. Each circuit has a tandem scroll
compressor, which means two steps by circuit. Then, the capacity steps available are 100%, 75%, 50% and
25% ; in fact, due to the mechanical flow rate reduction, the 75%, 50% and 25% capacity steps, are slightly
higher than this theoretical staging. Each circuit has 3 fans on its condenser (line configuration).
Figure 8.4. Evolution of the chiller performances when reducing the condenser air flow rate presented
under a reduced form (EER/EERnom, EP/Epnom, CC/Ccnom in terms of reduced flow rate)
Reduced performances while varying flow rate

1.2
1.15
1.1 EER/EERno
m
1.05
1 EP/EPnom
0.95
0.9 CC/CCnom
0.85
0.8
0.75
0.7
0.5 0.6 0.7 0.8 0.9 1
% of nominal flow rate
The decrease of the efficiency while varying the flow rate is reported Figure 8.4. The chiller was operated at
50% load, one compressor in operation on each circuit. The 70% flow situation corresponds to 2 fans among
3 being ON for each circuit. Precise measurement of the flow rate was not available. The 50% flow rate
corresponds to one fan functioning on one circuit and two on the second. The efficiency decreases with a
square tendency when the air flow rate is reduced.
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, at reduced flow rate, the increase of the efficiency with the outside
air temperature decreases faster than at full flow rate. This fact is observed in reality, but in a more complex manner. At
full load, the high pressures are higher than at 50% load.
Figure 8.5. Condensation pressure control effect on reduced EER (EER/EERnom) in terms of outside air
temperature
147
Reduced EER versus OAT for different load ratios DOE2 curve
1,70
FL curve
1,60
1,50 PL curve, single

circuit 50 %
1,40
PL curve, single
1,30 circuit 75%
1,20 PL curve, single

circuit 25 %
1,10
PL curve, single
circuit VSDF
1,00
15 17 19 21 23 25 27 29 31 33 35 Without high
pressure control
OAT (°C)
On figure 8.5 we can understand that the high pressure control will impact differently single circuit and
double separated circuits units. For single circuit units, the 25% load would correspond to still higher
ambient temperature for triggering fan cycling or reduction speed. Thus, the efficiency variation with outside
air temperature is shown Figure 8.5 for 25% and for 75% load on a single circuit. On the contrary, for that
double circuit unit, the 25% load point does not differ from the 50% since one compressor is in operation on
one circuit, the 50% being strictly symmetrical. For double circuit and 75% load operation, one circuit is
operated at full load while the other is operated at 50% load.
8.2 Is the IPLV approach directly applicable to European conditions?

The percentage of operating hours spent at each part load condition, given in our description of US-IPLV,
chapter 2.5, (1% etc…) is intended to be representative of the US climate and buildings but not of the
European ones. Further to this, 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, for
instance- as can be shown by drawing the loads in terms of outside temperatures). 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.
Buildings used in deriving the US-IPLV

This standard covers all the tertiary sector buildings for air conditioning application on the whole US
territory. 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.
Four building groups have been identified depending on the occupation scenario and the possibility to use
free-cooling or not :
• Group 1, occupation 24h/day : 7days/week, cooling above –17.2 °C,
• Group 2, occupation 24h/day : 7days/week, cooling above 12.8 °C, free-cooling between –17,2 and
12,8°C.
• Group 3, occupation 12h/day : 5days/week, cooling above –17.2 °C,
• Group 4, occupation 12h/day : 5days/week, cooling above 12.8 °C, free-cooling between –17,2 and
12,8 °C.
Climate used in IPLV derivation
3
[ARI98] states that these cities represent 80% of the installed chillers in the US.
148
The mean climate used to perform the bin method is an average of the 29 cities climates. The climatologic
data were averaged without any weighting of energy or capacity installed in each city. Coming from this
average climate, an occurrence curve of dry bulb and wet bulb air temperature is drawn by 5 °F (2.8 °C) bins,
between –17,2 °C and 35 °C. The water temperature is deduced from the wet bulb temperature using an
added 8 °F (4.4 °C) approach.
Building cooling load calculation in US-IPLV

The details of the calculation are not explained. The only indications are given in the text. Internal loads
represents 38% of the total load above 12.8°C. The building load experiments either a 20% of maximum load
at 0°F (–17.2°C) for groups 1 and 3 either is null under 12,8°C for groups 2 and 4. The nominal full load
sizing for the four groups correspond to the highest bin, temperature higher than 95°F, or 35°C. The load
curve of group 1 is given figure 8.6. Each group is weighted in function of its relative weight in the US from
a statistical study.
Calculating US weighing coefficients

The following calculations depend now on the group chosen. Then, the results are weighed by the
representative coefficients of the groups amongst US building studied.
• Taking group 1 as an example, the load curve is multiplied bin by bin by the number of hours
experienced in each bin considering the average climate.
Then, one obtains the energy needs curve (figure 8.6). The ARI 550/590 unit for energy is the ton-hours.
What is shown is actually the product of the hours by bin and of the normalized capacity in %. 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.
Figure 8.6 : annual cooling needs as a function of outside air temperature, group 1, from [ARI98].
• On this curve, four integration intervals are then defined:

The sizing interval (for temperature higher than 95 °F : 35 °C) and the three other intervals : [0,55], [55,75],
[75,95]. For each interval, one integrates the energy curve that will give the energy weight of the four testing
points called respectively A, B, C and D values.
Interpolation scheme needed to reduce testing time

If the unit cannot unload to one of the specificied steps, two possibilities exist : either it can unload at a lower
capacity than the missing step or it cannot. If it can, 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.
149
Figure 8.7 illustrates the interpolation procedure for unit N°7. The graph differs from the representation used
in the IPLV standard for part load. The load ratio in % is related to full load and 35°C inlet air temperature.
The solid and very black line is the ARI assumed load/temperature curve. The less solid black line is
determined by the capability of the four steps of the chiller to reach the four specified temperatures. Three
straight lines representing the capacity stages relative to full load 35°C capacities are drawn for the stages 1,
2 and 3).
Figure 8.7 : 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 8.7 : when the outside air
temperature is 18°C and the load between about 30% and 62%, the chiller will cycle between the 50% and
25% stages. Thus the efficiency of the 50% point 18°C should be calculated as the weighted average of the
two points: [18, first stage] and [18, second stage]. 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.
To avoid the multiplication of the number of testing points, the ARI procedure uses only the testing points
(full circles) to perform interpolation of the efficiencies for specified load points. Thus instead of weighting
the two previously mentioned points, the standard proposes to weight the [18, first stage] and the [26, second
stage] points. Thus, the 50% point efficiency is underestimated in that case since efficiency of the [26,
second stage] point is lower than the [18, first stage] due to temperature decrease.
The consequence is that for that unit number 7, the global seasonal figure is underestimated by the
interpolation procedure whereas, for a continuous control screw unit, the IPLV exact figures would be
obtained, the capacity step chillers being penalized for not supplying continuous unloading. The method is
interesting since it enables to reduce the testing points number and its effect will be discussed further
hereunder.
EMPE: an answer to a need for a European weighting with IPLV-like testing

The first remark that led to the EMPE [AICARR, 2001] Italian proposal is that the operating conditions are
rather different from Southern Europe conditions. And even, if Northern Europe countries may need air
conditioning in summer, it cannot be said that Italy would need air conditioning at 12.8 °C as normal
operating conditions.
150
Therefore, AICARR proposed a new energy index, named EMPE (Average Weighed Efficiency in Summer
regime in Italian) directly deriving from IPLV, with different energy weights and, in particular, with
different temperatures at the condenser inlet, fitter for the European climate and requirements in the air
conditioning field.
The EMPE formula is absolutely similar to IPLV, but the values of energy weights and inlet temperature to
the evaporator and the condenser are those indicated in table 8.1:
Table 8.1: calculation conditions for EMPE
The AICARR proposal, EMPE was not based on a sufficiently large climatic and technical investigation. Its
strength (being very close to the existing US method, which aggregated many factors) was also its weakness.
We had the opportunity to go further by constructing a data base of EU chillers at part load, understanding
better part load, and proposing two separate methods, one for part load reporting and certification, the other
one for the computation of SEER.
Reduction of EMPE or IPLV to 2 points with extrapolation

The following method has been proposed by one manufacturer. It is an amendment to IPLV or EMPE. Two
testing points only are performed, the first one at 100% load and design temperature and the second one at
50% of the load and the associated reduced temperature. The EER values at 75% and at 25% of the load,
necessary to define the index, would be obtained by interpolations and extrapolations on the IPLV or EMPE
curve. 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 manufacturer could then decide if making 4
tests or only 2.
The EER value at 75% of the load is calculated as follows:
151
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%. We have tested this simplification
extensively on a set of chillers (table 8.2).
Table 8.2 : 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 3,35 3,22 3,50 3,59 3,32 3,39 2,80 3,39
2 points method 3,82 3,42 3,92 3,85 3,16 2,91 3,26 3,48
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. Moreover a bias is introduced in the classification : some always loose, some always benefit.
This methodology suffers the same bias that was introduced by the interpolation process, again increased.
The evolution of part load efficiency at reduced temperature cannot be modelled simply by a linear
regression : it depends on the unit, even for the very commune air scroll range. Moreover, practical limits as
well as the impossibility to predict cycling keep us from recommending that method.
8.3. 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. The main tool used was actual testing of EU
equipment but a number of group meetings allowed to build a common thinking frame. The technical
description of the chillers tested follows on tables 8.3 and 8.4, split by condensation type.
Table 8.3. 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 8.4. 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, some common testing points were made according to either the ARI or the
EMPE conditions depending on the manufacturer will. For all chillers, 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, 2002]. For chillers n° 2, 3, 4
152
and 8, only IPLV or EMPE points plus the CEN one were available. For the others as many testing points as
desirable have been obtained. In all circumstances a simple model has been used to draw the performance
maps from existing testing points.
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. It is so when there is one single compressor per chiller, or various identical circuits. A significant
market share of chillers have various compressors and a complex circuiting, leading to improved part load
performance. But a given part load regime has to be defined by the actual status of each piece of equipment.
For discrete stages chillers, it would be easier to describe performance at a given stage not at a given
percentage. For the very few continuously controlled chillers, fours stages can be defined in terms of input.
Since temperature and load can be tested independently and recombined, there is no need for combined
testing (like IPLV).
About certifying Part Load : what the manufacturers give to their customers is a « map » of
performance, not only values at the four arbitrary percentages and temperatures, plus the final Eurovent
grading when it is available, based on a SEER. The customer can rely on the Eurovent SEER computed from
this map … or compute its specific SEER for its specific demand. 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, as usual,
and the associated independent testing.
We arrived also at applicable conclusions on the way to report the SEER in the Eurovent directory.
We started from HSEER, the DOE reference that we generated. It is proven that each set of outside
conditions (for each sector, climate, type of chiller, type of secondary system) can be reduced to four or five
external conditions without loss of accuracy. The ESEER index proposed here is a set of 4 conditions given
for E.U. as a whole, but there can be as many similar indices as specific demands: sector, country, etc.
We have introduced a format for the description of the stages of a chiller, like in table 8.5 and following, suitable for
Eurovent specification. For each stage, the manufacturer has only to declare which piece of its equipment is
operating and to indicate CC , the cooling capacity and EP, the electric power absorbed. The certifying body
has only to check a few of the values, selected in the same conditions as usual. Note that this procedure is in
fact already used for some chillers with various speeds, namely “low noise” chillers with the possibility of
reduced fan speed.
Table 8.5 : Part load performance of water cooled scroll chiller N°4, as could be reported in Eurovent part load
certification scheme
N° 4 // WT : 30°C STAGES 1 2 3 4
Circuit 1
Circuit 2
EP (kW) 8,80 17,60 27,17 38,27
CC (kW) 37,50 78,00 112,50 150,00
EER 4,27 4,47 4,12 3,92
Now we shall present examples of the proposed procedure. Let’s note that it is far easier to analyse
the part load behaviour of water cooled than it is for air cooled chillers. Indeed, the air cooled chiller stage
efficiencies can suffer different fan pattern and/or circuit separation that do not infer for water cooled
chillers. The chiller part load behaviour is described first for water cooled units, then air cooled units.
Water cooled chillers –experimental results
153
given on figure 8.8 for the four tested units.
Figure 8.8. 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
1.2
1.1
1
N° 4
0.9
N° 1
N° 3
0.8
N° 6
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
Let’s give more explanation about two of the tested chillers, N°4 and N°1, as examples of real life issues.
Chiller N°4 is a two circuit four scroll compressor chiller, with the same symmetrical tandem on each circuit.
The efficiency increase show that at 50%, one compressor by circuit is activated. At 75%, one circuit is at
full load and the other at half load. At 25%, only one circuit works at half load. Logically, at 25% and at
50%, the symmetry of the chiller would impose identical performances. The bias can come from many
causes, 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.
This complex part load behaviour can be summed up under one form per temperature – see Table 8.6, or one
single table with all temperatures –table 8.23.
Table 8.6 : Part load performance of water cooled scroll chiller N°4
N° 4 // WT : 30°C STAGES 1 2 3 4
Circuit 1
Circuit 2
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.
Now let’s consider chiller N°1 (Table 8.7) : it is a double circuit water screw chiller with one compressor by
circuit. One can see easily the difference between the two type of unloading, symmetrical, for higher than 50
part load ratios and on a single circuit, reducing the refrigerant flow rate at the minimum for part load ratio
smaller than 50%. In that case, the efficiency decreases somehow faster. Stages configurations and
performances are gathered Table 8.5 under the form proposed to Eurovent.
Table 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
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, 56 and 50) and 2 only for compressor 1 (100
and about 60). 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).
Air cooled chillers –experimental results
given on figure 8.9 for the five tested units.
Figure 8.9. 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
1.3
1.2
1.1
N° 5
1
N° 7
0.9 N° 8
N° 9
0.8
N° 2
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
Let’s give more explanation about two of the tested chillers, N°5 and N°2, as examples of real life issues.
Table 8.8 gives testing results on chiller N°5 : it is a single circuit unit with an asymmetrical scroll
compressor tandem, 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.
Table8.8: 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 8.9 : it is a screw double circuit unit. Refrigerant circuits are separated.
There is one screw compressor by circuit. 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.
Only the four tested stages are reported. The 50% and 75% capacity stages cannot be allocated to each
circuit. Only the 25% can. 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.
155
Table 8.9 : 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%
8.4 Derivation of a new SEER method (ESEER)
Given the complexity of the subject, the EECCAC group adopted the building simulation tool DOE2 to
simulate representative buildings of the European stock market. 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. However, when multiplying the simulation cases including building types (offices,
malls, hostels, hospitals, administration …), the climatic conditions and the different systems, it led to an
incommensurable number of simulation, without mentioning the number of buildings to be entered in the
used building code. We had to make some decisions.
The simulations leading to the reference values of SEER (HSEER)

Two buildings were simulated on computer, but buildings that do exist : an office and a commercial mall.
For each one, three climates have been simulated, adopting different envelope characteristics when moving
the building around Europe. The different systems identified in the stock and market study have been
simulated. CAC air and water distribution equipments have been simulated using the European average
efficiency values.
Hour after hour, the simulation uses then the characteristics of the real chillers modelled to compute the exact
yearly performance index : the HSEER (Hourly SEER), used then as a reference for other methods. At each
hour the outside enable to calculate all known stage capacities and respective electric powers, including the
high pressure control impact on each stage. Then the load is compared to each stage capacity. If the load is
lower than the smallest available capacity step, the cycling formula enables to calculate the electric power.
Otherwise, the weighting of electric power of each stage is found by the expression of the weighted average.
In fact this computation has been performed in two steps : first, chiller hourly load curves were extracted as
well as the coincident hourly OAT and the specific humidity. Then those conditions are used a number of
times with different chillers’ and equipment quality. The different climatic conditions and central air
conditioning systems available are described Table 8.10.
Table 8.10. 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 8.10. The tendency that links load and temperature is
very clear even if an important scatter is observed.
Figure 16. Milan office building CAV system hourly load curve
156
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, two problems appear that
will show up two significant limitations of the ARI method.
The load curves must enable to calculate the consumptions for all chillers. Thus, the load must be divided by
the sizing load, so that all the chillers may be compared on the same load and temperature repartition. In the
ARI standard, 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.
It clearly appears Figure 8.10 that the ARI hypothesis is not verified. The explanation is that even if
temperature is correlated to the load, other load pattern intervene as the solar loads, the thermal inertia and
the dehumidification loads that are the sources of the non explained variance by the load and temperature
correlation. This is the first limitation of ARI sizing assumptions.
So the real optimal design rule is : the maximal chiller capacity and the corresponding temperature
corresponds to the (load, OAT) couple that enables the chiller to cover all the cooling needs. The capacity
variation with OAT of a perfectly sized (500 kW, 30°C) chiller with this simple law has been drawn Figure
8.10 and shows sizing is correct. Of course, if maximal load were at lower temperature, it could happen that
the sizing could lead to non-satisfied needs ; 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.
The sizing realized for the three climates led to 30% constant oversizing for the 3 climates for the office
building. For real world installations, security coefficients are generally applied to the simplified sizing
method leading to huge oversizing up to 100%. Given the part load characteristics of the chillers, it seems
obvious that consequences for the seasonal efficiency will also be very important.
The seasonal efficiencies for the MILAN CAV load curve are presented Table 8.11 for 30% and 60%
oversizing for the air cooled chillers N°7 and N°2 .
Table 8.11. Impact of oversizing on seasonal performances for Milan CAV hourly load curve
SEER values
Oversizing N°7 N°2
0% 3.81 3.12
30% 3.83 2.81
60% 3.76 2.60
The differences are limited for the air scroll chiller N°7, because the reduced part load efficiency is still
higher than 1 at 25% and hardly lower than 1 for the 50% load reduced efficiency. On the contrary, for the
screw chiller N°2, a sharp efficiency degradation with the load had been noted. These results confirm that the
157
sizing is a key factor for seasonal performances analysis. And it also shows that no optimum sizing can be
done without the exact knowledge of the chiller part load performances.
In the ARI methodology, for reducing the load curve and temperature occurrences to 4 points, only the 100%
and more than 35°C OAT couples are kept. Whereas in our calculation methodology, all the points between
the full load stage and the step immediately inferior are shared between the two steps, giving weight to the
full load. The ARI methodology of reduction would lead almost to a null energy share for the full load stage,
which is not true for staged chillers but approaches the truth for continuous control chillers. This is the
second limitation of ARI sizing assumptions.
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, it is supposed that a virtual chiller, with 4 capacity steps at 25%, 50%,
75% and 100% of the design load is operating. The Milan-CAV load curve is used to illustrate the
methodology. But this methodology can –and will- be applied to all conditions obtained by simulation,
leading to the possibility of a four points representation of any condition or of the EU average of operating
conditions.
STEP 1 of the load curves reduction process

The sizing described above enables to transform Figure 8.10 with % load instead of kWh on the Y axis.
Then, the load curve is put under a grid format. On the X axis, the outside air temperature is binned. On the
Y axis, the load ratio is binned. In each rectangle, the relative kWh are added, giving a repartition of the
cooling energy needs on the grid (Figure 8.11).
Figure 8.11. Grid representation of the Milan-CAV load curve (load ratio bin length : 0.05, temperature bin
length : 2°C)
The average temperature is kept for each column and called hereafter binT(k). For each line, the average load
ratio is kept and called here after binL(i), where k varies between 1 and K, the number of temperature bins
and i varies between 1 and I, the number of % load bins.

A statistical reduction method enabled us to get, from the hourly load curve, the representation of a discrete
load curve and a discrete weighting curves.
Figure 8.12 gives for instance the discrete load curve for the hourly simulation of the office building in
Milan. It shows that it is not far from a straight line, as assumed in the ARI standard. However there would
be no weight for the 100% stage applying the ARI methodology from now on, as already explained.
Figure 8.12. Reduced load curve for the office building, Milan-CAV
158
Figure 8.13 gives the representation of the energy weight of each class as a function of temperature classes. It
is equivalent to the ton hour curve given in the ARI standard for Group 1 (Figure 8.7). It can be seen that
even for Milan, the chiller load and the associated weighting energy coefficients are still not null at
temperatures as low as 15°C.
Figure 8.13. Reduced weighting curve for the office building, Milan-CAV

The ARI methodology load curve reduction would not give any weight for the full load point. 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.
At this level, for each capacity step, we have obtained weighting coefficients that represent the energy
weights to be associated to the % load required (25, 50, 75 and 100) at each temperature bin. Each stage
weight and OAT repartition is represented Figure 8.14.
Figure 8.14. Reduced weighting by stage under the PT format, method 3rd Step, for the Milan-CAV office
building hourly load curve
159
So now, the weighting coefficients are known :
SEERP= [0.2024 0.4272 0.3369 0.0335]

The average temperature for each stage is calculated and the following operating temperatures are found :
SEERT=[17.9997 23.0611 28.0627 31.0514]
The results for the reduction for the Milan office building load curve are presented Table 8.12.
Table 8.12. 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, B, C, D
(nominal full load)
100 31.2 3%
75 28.0 34%
50 23.1 43%
25 18.1 20%
Results for more extreme weather conditions (London, Seville, different distribution
systems)
The hourly calculation methodology has been applied to two different load curves among the twelve
available (Table 8.10) :
• CAV-FC system in Seville
• CAV system in London
However, in order to separate the quality of the reduction by itself from the non linear models representing
the chillers, the seasonal performances are calculated successively for the 4 following configurations :
• Without cycling, without high pressure condensation control
• Without high pressure condensation control, with cycling
• Without cycling, with high pressure condensation control
• With both phenomena
The reduction results of the two specified load curves are presented Table 8.13. The two extreme load curves
reduction presented show that both weighting coefficients and temperature conditions greatly vary with
climatic conditions and systems.
160
Table 8.13. 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, using the CAV system
Seville CAV-FC London CAV
Energy weighting Energy weighting

Part load (%) Reference
Inlet air temperature (°C) coefficients A, B, C, D Inlet air temperature (°C) coefficients A, B, C, D
(nominal full load)
100 36.7 4% 27.6 1%

75 32.1 48 % 24.8 10 %
50 27.4 37 % 20.9 42 %
25 22.8 11 % 17.1 47 %
The results of the reduction methodology are presented Table 8.14. 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), the reduced index figure (noted ESEER) and the relative efficiency
difference between the two seasonal figures. We translate the information in terms of ranking : chiller ranked 1 is better than chiller ranked
2, and so on.
Table 8.14. 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 2.18 2.59 2.51 2.47 2.93 2.18 2.59 2.51 2.47 2.93
EER ranking 5 2 3 4 1 5 2 3 4 1
HSEER 3.01 3.38 2.73 3.09 3.13 4.08 4.59 3.30 4.21 3.19
No cycling, no fan cycling ESEER 3.05 3.44 2.73 3.14 3.15 4.08 4.60 3.30 4.22 3.31
Relative
1.2% 1.8% 0.0% 1.6% 0.7% 0.0% 0.2% 0.0% 0.2% 3.9%
deviation
HSEER 2.93 3.34 2.69 3.06 3.11 3.70 4.43 3.08 4.11 3.14
Cycling only ESEER 3.00 3.43 2.72 3.14 3.15 3.88 4.58 3.20 4.22 3.31
Relative
2.4% 2.6% 1.1% 2.4% 1.3% 4.7% 3.3% 3.7% 2.7% 5.5%
deviation
HSEER 2.99 3.35 2.71 3.05 3.00 3.80 4.26 3.05 3.91 2.96
Fan cycling only ESEER 3.03 3.40 2.73 3.11 3.10 3.72 4.31 3.01 3.90 3.07
Relative
1.4% 1.4% 0.8% 1.9% 3.4% -1.9% 1.4% -1.4% -0.4% 3.5%
deviation
HSEER 2.90 3.31 2.66 3.03 2.98 3.46 4.12 2.86 3.82 2.92
Cycling and fan cycling ESEER 2.98 3.39 2.71 3.11 3.10 3.56 4.31 2.93 3.90 3.07
Relative
2.7% 2.3% 2.0% 2.6% 4.0% 2.9% 4.5% 2.4% 2.0% 5.1%
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 8.14 :

• The full load nominal efficiency appears to be a poor indicator of the seasonal efficiency. No policy
measure should be based on it only as far as capacity staged or continuous capacity control chillers
are concerned.
• For both conditions, the ESEER follows the HSEER classification of seasonal efficiencies. However,
the ESEER values are always higher than the HSEER values by 2% to 5.1%.
• For the two different load curves, it appears that the chiller classification is largely modified ; it
shows the important impact of load on the seasonal efficiency. The chiller N°2 is the first for the
EER sequence, the third for the Seville sequence and the last for the London sequence.
• The relative maximal bias created for Seville by the methodology reduction is [+ 2%, + 4%]. The
relative maximal bias created for Seville by the methodology reduction is [+ 2%, + 5.1%].
• Cycling and fan cycling impacts cannot be neglected for the London curve conditions as it is shown
by the HSEER evolution. However, both impacts are logically very low for the CAV-FC load curve
in Seville, conformingly to table 9 reduction results.
161
Table 8.15 below enables to check that the load curves selected for the sensitivity analysis represent indeed
extreme conditions for temperature and load. For the London CAV system, the temperatures are the lower
(except for the FC4P system) and weighting coefficients maximum at low loads. For the Seville CAV-FC
system, respective reverse conclusions on temperatures and load can be made.
Table 8.15. 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 24.8 20.9 17.1 0.7% 9.7% 42.5% 47.1%
CAV-FC 27.6 26.1 24.3 22.4 4.2% 26.7% 37.6% 31.5%
FC4P 27.6 24.6 20.1 16.1 0.5% 8.7% 48.5% 42.3%
VAV 27.6 25.6 22.4 17.6 1.1% 7.7% 29.1% 62.1%
Milan CAV 31.2 28.0 23.1 18.1 3.6% 33.9% 41.7% 20.8%
CAV-FC 31.2 28.0 24.8 22.0 5.7% 54.4% 31.1% 8.8%
FC4P 31.4 28.1 23.1 17.7 3.1% 32.0% 40.5% 24.3%
VAV 31.6 28.9 24.5 19.1 2.6% 30.7% 39.5% 27.2%
Seville CAV 36.7 32.1 26.3 19.8 3.5% 38.2% 39.1% 19.2%
CAV-FC 36.7 32.1 27.4 22.8 4.4% 47.5% 37.3% 10.7%
FC4P 36.9 32.3 26.5 19.2 2.8% 35.3% 40.2% 21.7%
VAV 37.2 33.4 28.0 21.1 1.6% 30.7% 43.9% 23.8%
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.
Extrapolating to the European stock of chillers in use

Results of this weighting (described in chapter 5) are given Table 8.16 for air cooled chillers.
Table 8.16. 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 0.8% 0.3% 0.1% 0.2% 9.9% 3.2% 5.4% 0.1% 38.0% 0.0% 0.7% 1.1% 37.3% 0.4% 2.5%
WC
FCU 0.8% 0.4% 0.1% 0.3% 9.0% 3.2% 4.9% 0.1% 34.6% 0.0% 0.8% 1.0% 41.4% 0.6% 2.8%
Southern Europe country visibly represent most of the cooling energy needs in Europe. We can reduce this information to a set of 6
coefficients that will be used to weight the SEER obtained from London, Milan and Seville load curves. Final weighting coefficients for the
available load curves types are presented Table 8.17.
Table 8.17. Cooling energy needs hourly load curves weighting coefficients for the air cooled chillers, for CAV
and FCU, free cooling and VAV specific applications (WC : weighting coefficient)
Climate London Milan Seville
CAV 5.3% 16.1% 18.4%
FCU 8.2% 22.2% 29.8%
WC
Free Cool. 6.5% 41.6% 51.9%
VAV 10.6% 40.3% 49.1%
CAV (air distribution) and FCU (water distribution) systems enough to represent the present stock. Figures
of the two first lines of Table 8.17 for CAV and FCU are to be understood as a complete set of coefficients
for 6 load curves. The sum of the 6 weighting coefficients equals 1. For the values for CAV+Free Cooling
and VAV applications, only the air distribution systems are concerned . The 3 weighting coefficients issued
from CAV market shares are here to generate scenarios of improved efficiency.
162
At last, these coefficients enable to weight the results of Table 8.15 and to present final ESEER average
European conditions in Table 8.18. The values of the two first arrays are our proposal for EU chillers rating.
The other results show how free cooling and VAV application could change those recommended values.
Table 8.18. Application of the method to derive a ESEER for air cooled chillers used with the free cooling or
VAV options
ESEER Free-cooling VAV
100 34.0 3% 33.9 4% 34.6 2%
75 30.1 33% 30.1 45% 31.2 30%
50 24.7 41% 25.5 35% 26.3 41%
25 18.6 23% 21.0 15% 20.1 28%
The differences for the VAV application for weighting coefficients can be neglected while each stage temperature increases by about 1°C. For
free cooling, the lower stage temperature increases while weightings move towards higher loads. For both options, the average coefficients
only slightly moves, 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.
In a similar way the values have been defined and validated for water cooled chillers. However, for water
cooled units, the water temperature at the condenser inlet depends not only on the OAT but also on :
o the condensing water flow rate,
o the tower performance curves,
o the specific humidity.
Only the open type towers have been considered here, since they represent 80% of the European stock,
despite of the Legionella disease that certainly greatly modified the sales. Within the ARI standard, 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,
which is false. Here, real towers have been sized using the cooling towers electronic catalogue of the leading
manufacturer, completed with the NUT-epsilon heat exchanger theory, considerations on control of
equipment and recent correlations. On Figure 8.14, for a specific screw chiller extracted from the
manufacturer catalogue, the part load ratios versus the condenser inlet water hourly temperatures are drawn.
The load curve is the Milan CAV one. 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.
Figure 8.14. 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).
163
It clearly appears Table 8.19 that the temperature dependency is not linked first to climatic conditions but to the sizing and to the cooling
tower choice. It has to be recalled that the cooling towers have been sized for maximum wet bulb temperature with 1% of yearly occurrence.
The lower temperature results in Seville as compared to Milan show that on average wet bulb temperatures are higher in Milan. The main
differences amongst climates and systems are found for weighting coefficients differences.
Table 8.19. 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 28.3 24.8 21.7 18.5 0.38% 10.96% 41.53% 47.13%
CAV-FC 27.2 26.1 22.3 20.3 2.23% 22.34% 44.54% 30.89%
FC4P 28.3 24.5 21.5 18.3 0.32% 10.38% 46.35% 42.96%
VAV 26.8 26.2 22.4 18.3 0.07% 9.42% 27.23% 63.28%
Milan CAV 28.8 25.9 21.9 18.4 2.96% 35.53% 39.66% 21.85%
CAV-FC 30.4 26.5 23.2 19.8 0.04% 43.26% 42.65% 14.05%
FC4P 27.5 25.3 21.6 18.0 1.87% 32.39% 43.29% 22.45%
VAV 28.7 26.1 22.5 18.5 2.90% 32.51% 34.75% 29.84%
Seville CAV 29.0 26.2 22.6 18.6 3.31% 38.69% 38.16% 19.85%
CAV-FC 29.0 26.2 22.8 19.5 4.15% 45.90% 40.06% 9.89%
FC4P 28.9 26.2 22.8 18.6 2.30% 36.20% 38.97% 22.54%
VAV 28.9 26.5 23.2 18.8 2.03% 30.28% 42.67% 25.02%
The same market shares can be used for water cooled chillers than for air cooled chillers. because a constant
share of water cooled and air cooled systems has been used, the same for all countries. However, since Table
8.19 exhibits different load weighting coefficients, the final weighting for the ESEER for water cooled
chillers slightly differs from the air cooled chillers coefficients, Table 8.20.
Table 8.20. 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 28.6 2% 29.4 2% 28.7 2%
75 25.8 34% 26.2 44% 26.2 30%
50 22.3 40% 22.9 41% 22.8 39%
25 18.5 24% 19.6 12% 18.6 29%
8.5 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.
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, for all chillers, included single compressor units. However, it also shows that the
experimental uncertainty is quite high. It mainly comes from the uncertainty measurement on the
temperature difference at the evaporator. In order to simplify the application of the index, some rounding can
be done without modifying noticeably the ESEER figures obtained, largely under the experimental
uncertainty.
A comparison of the conditions of the 3 available indexes is proposed Table8.21 for air cooled chillers.
164
Table 8.21. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers
ESEER ARI EMPE
100 35 3% 35 1% 35 10 %
75 30 33% 26.7 42 % 31.3 30 %
50 25 41% 18.3 45 % 27.5 40 %
25 19 23% 12.8 12 % 23.8 20 %
Temperatures of the ESEER are embedded by EMPE temperatures above and ARI temperature beneath.
ESEER weighting coefficients give more weight to the 25% point load than both index. For 50 and 75%,
coefficients are nearer to the EMPE index. The 100% coefficient is 3%, nearer from the IPLV one.
A comparison of the conditions of the 3 available indexes is proposed Table 8.22 for water cooled chillers.
Table 8.22. Comparison of the ESEER conditions with the EMPE and IPLV indexes for water cooled chillers
ESEER ARI EMPE
Part load ratio Temperatures (°C) Temperatures Temperatures
100 30 3% 29,4 1% 29.4 10%
75 26 33% 23,9 42% 26.9 30%
50 22 41% 18,3 45% 23.5 40%
25 18 23% 18,3 12% 21.9 20%
Temperatures of the ESEER are embedded by the EMPE ones above and ARI temperature beneath except
for the 25% point. The ESEER weighting coefficients give more weight to the 25% point load than both
index. For 50 and 75%, coefficient are nearer to the EMPE index. The ESEER 100% weighting coefficient is
nearer from the IPLV one.
Classification : who is right?
We shall compare now four classifications : according to EER, US-IPLV, EMPE, ESEER, using as a
reference the actual EU values obtained by simulation in the three locations and properly weighted. We take
the point of view of a user of the Eurovent certification system : by selecting a “better” chiller, am I really
selecting a better chiller?
EER is a poor selection tool

Figure8.15. Comparison of HSEER with EER on the tested chillers
165
HSEER versus EER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5
EER
IPLV and EMPE are more accurate than EER for classification but do not give enough
accuracy for comparison of chillers
Figure 8.16. comparison of US-IPLV with HSEER for the tested chillers
HSEER versus IPLV
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
IPLV
Figure 8.17. comparison of EMPE with HSEER for the tested chillers
166
HSEER versus EMPE
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
EMPE
Based on similar assumptions, the two methods, IPLV and EMPE have the same advantages and disadvantages.
The proposed ESEER method allows grading and ranking of chillers by order of merit
Figure 8.18 . comparison of ESEER with HSEER for the tested chillers
HSEER versus ESEER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
ESEER
Conclusion : the differences are relatively large between existing methods and reality, and not always in the
same direction. 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.
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. Since there
is no European specific standard to perform part load testing, the analysis is based on :
• the full load testing definition [CEN, 1998],
• the IPLV standard [ARI, 1998], that contains some remarks about testing,
• the experience gained during the “Joint project” at EDF R&D facility, DMT and manufacturer
laboratories visited.
167
The testing must fulfil the following associated constraints :
• it must enable to start the part load certification next year,
• it must not require the manufacturer presence for the testing,
• it must respect the chiller ESEER sequence,
• it must minimize the number of testing points, it is to say, the time needed to perform all tests, the
real cost factor.
First, 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, so as to minimise interpolation and iterations;
the computation of ESEER will be only a weighting/interpolation of testing results. In case we want a
“blind” checking, more time is needed to guess out the proper conditions for testing in such an ARI
approach. We shall present afterwards another way of doing, which seems to us more in the “spirit” of
Eurovent certification : once the manufacturer gives the full table of performance, three points are taken
“randomly” and checked; it is enough to allow the use of the full table and the calculation of the ESEER.
In a certification approach, the manufacturers must give to Eurovent the cooling capacities, 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”. Table 28 gives an example, for chiller number
7, that will be explained just after. Under this format, the table allows to answer directly to the question :
what happens to performance when the load (resp. the temperature) decreases.
Table 8.23. For chiller N°7, stage capacities, part load ratio (% of full load, OAT = 35°C), electric power, and
efficiencies.
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% 153.7 116.8 81.5 38.4
EP P1 60.0 43.8 26.9 14.2
EER 2.6 2.7 3.0 2.7
30 CC 166.3 82% 126.4 57% 88.1 41.6
EP 53.9 P2a 39.4 P2’ / P2b 24.2 12.7
EER 3.1 3.2 3.6 3.3
25 CC 176.3 134.0 61% 93.4 28% 42.9
EP 48.9 35.8 P3a 22.0 P3’ (P3b) 11.9
EER 3.6 3.7 4.3 Fan cycling 3.6
19 CC 185.0 137.0 93.0 29% 43.9
EP 43.8 32.8 20.7 P4 10.9
EER Fan cycling 4.2 Fan cycling 4.2 Fan cycling 4.5 Fan cycling 4.0
1ST TESTING POINT

The testing begins with the full load nominal temperature, the inlet evaporator water temperature is 12°C and
the outlet is 7°C. (100%,T1) noted after P1.
2ND TESTING POINT

Then, both the outside condensing fluid and the inlet evaporator temperatures are decreased to the 75% load
point. It means the inlet water temperature at the evaporator is decreased until 10.75°C. A % tolerance on the
cooling capacity is calculated and consequently, the tolerance on the inlet water temperature at the
evaporator is known.
For continuous control chillers : the chiller can supply the adapted capacity within the allowed tolerance.
Point 2, (75%,T2) noted after P2.
168
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
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
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
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
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
show that the free cooling and VAV options should be differentiated from typical CAV installations, mainly
because of the differences in the weighting coefficients.
As given, the coefficients are nearer from Southern Europe operating temperatures and weightings (Italy,
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. Temperatures can be shifted easily
of 1°C if needed (as practised to round the operating temperatures).
This index could be used for single circuit units. However, a method must be adapted to determine the
cycling degradation versus the load. The default cycling law could then be revised for all chillers.
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.
• Then, the reduction methodology enables to characterize 4 capacity step points and operating
temperatures.
• The presentation of part load results by the manufacturers enables to select the chiller on 4 efficiency
points.
Certifying seasonal performances for chillers means indicating an average efficiency generally higher than
nominal efficiency as has been largely figured. But it also means to avoid the efficiency competition may be
based on non-representative indicators, as nominal full load EER is. Thus, the seasonal performance index is
thought to be a huge and necessary progress to strengthen energy efficiency of chillers.
Future versions should consider the extension to single stage units that generally operate at different
conditions. It has been shown however that the methodology could be applied to these units for the load
curves treated.
Similar work has to be performed in the heating mode since air to water reversible chillers is an increasing
end use in Europe. The same philosophy could also be applied to ground water condensing chillers and heat
pump, also a developing market in France and Germany.
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.
In order to approach field reality, the integration of dynamics into the part load testing index should also be
considered. Nevertheless, supplementary work has to be performed to reach a such far away goal from actual
chillers characterization.
This work could also serve as a basis for developing a seasonal index for room air conditioners, the more
developing end-use in Europe nowadays.
172
9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY CAC
SCENARIOS
In this section, efficiency policy and technology scenarios which project energy consumption, peak power demand and
CO2 emissions will be produced and the results compared with the base case for period of 1990 to 2020. 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.
9.1-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. The policy measure associated is banning some classes of
equipment either directly (Directive ) or by voluntary agreement. We can also expect that a certain number of
years of labelling and communication by energy agencies reaches the same point, nobody wanting to buy a
« poor » image equipment.
Chillers AC : the average being 2.50, the classes E F and G should be banned, the weighted gain is 0.23 on
full market average and the factor is 96.2% to be applied to compressor consumption. Chillers WC : the
average being 3.85, the classes E F and G should be banned, the weighted gain is 0.175 on full market
average and the factor is 93.0% to be applied to compressor consumption. Packages and large splits the
average being 2.46, the classes E F and G should be banned, the weighted gain is 0.05 on full market average
and the factor is 98.0% to be applied to compressor consumption. RAC classes E F and G should be banned
and the average gain corresponding is 0.10 on an EER around 2.50 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. The policy measure associated is banning many classes of
equipment or a negotiated agreement on average full load performance like ACEA agreement for cars.
Chillers either the average moves from 2.50 (present market) to 2.80 or the classes D E F and G are banned,
the factor is 89.3%.Packages and large splits either the average moves from 2.46 (present market) to 3.22 by
a voluntary agreement or the classes B to G are banned, the gain is 0.78 on full market average so the factor
is around 76.4% There will be a consequence for reversible winter heating, that will be provisionally taken as
the same factor RAC either the average moves from 2.50 (present market) to 3.20 by a voluntary agreement
or the classes B to G are banned, the gain is 0.7 on full market average so the factor is around 78.1% There
will be a consequence for reversible winter heating, that will be provisionally taken as the same factor.
All packages and chillers reach in 2005 the SEER level with the minimum LCC (BAT with upcoming
information given by part load testing). Part load is taken into account in Eurovent grading and so the
corresponding improvement is obtained. 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 3.00 (present market) to 3.64 and the gain is 0.24 on full market
average (+18% of which 12% may be obtained as well with scenario2) ; the factor is 82.4%
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. 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
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, 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 8.5
Table 8.5 Consumption of each system, relative to present, 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, 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; to know the
cost we have to evaluate the less costly options, which may be on either side, primary or secondary; national
values are different and have been derived from UK with corrections for DD.
The impact has been calculated with the assumption of a weighted mix of both situations : new buildings and
new installations. The overall reduction being a 12% reduction, we apply then -12% to each item of AC
(fans, pumps etc). The policy instrument would be a clear and harmonised implementation of EPB directive.
The less expensive way of attaining the objective is the improvement of chillers. Starting from their present
averages of EER and SEER, this policy induces almost no extra cost for any stakeholder, and absolutely no
cost provided it’s applied to all manufacturers (and so that they all pass on the costs to the customer). To
obtain this “free” market transformation a prescriptive minimum should be applied to local manufacturers
and importers at the same time.
9.2 Results of scenarios

General Evolution
The scenarios are ranked in the following order : 1,4,5,2,6,3. The maximum flexibilty in demand is around
20TWh and can be obtained by introducing policy measures related with part load efficiency of chillers.
However a strong commitment on full load achieves about the same gains.
174
Figure 9.1 Demand evolution, depending on scenarios
140,00
120,00
100,00
BAU
80,00 Scenario 1
TWh for AC
Scenario 4
Scenario 5
Scenario 2
60,00 Scenario 6
Scenario 3
40,00
20,00
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. We can also expect that a certain number of years of labelling and communication by energy
agencies reaches the same point, nobody wanting to buy a « poor » image equipment. The effects are strong
on trade and offices as shown on figure 9.2
Figure 9.2 Savings in 2020 (CAHORE stands for Cafes, Hotels, Restaurants)
Scenario 1
4000
3500
3000
2500
GWh saved in AC
2000
1500
1000
500
0
CAHORE Education Hospitals Households Offices Trade Total
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. Figure 9.3 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.
Fig 9.3 savings in 2020 in scenario 2 2020
175
Scenario 2 in 2020
16000
14000
12000
10000
GWh
8000 Série1
6000
4000
2000
0
CAHORE Education Hospitals Households Offices Trade Total
Part load is taken into account in Eurovent grading and so the corresponding improvement is obtained. The
policy measure associated is banning many classes of equipment or a negotiated agreement on average part
load performance like ACEA agreement for cars. We see on table 9.1 a large influence on chiller based
systems due to part load and also on the other systems even if they operate at full load.
Table 9.1
Savings (TWh) 2005 2010 2015 2020
Chiller based 4,32 6,93 9,86 10,45

systems in
SC3
Others in SC3 10,21 12,53 14,77 15,70
Total savings 14,53 19,46 24,64 26,15

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. The results of table 9.2 show a real potential for this improvement.
Table 9.2
Savings (TWh) 2005 2010 2015 2020
Savings due to Free 2,00 3,75 5,68 6,15

Cooling
Scenario 5 Variable Air Flow

Obligation of variable air flow in air distribution systems translated in national values for savings in table 9.3
176
Table 9.3
Savings (TWh) 2005 2010 2015 2020
Savings due to 3,89 7,85 12,28 13,45

VAV
Scenario 6 British regulation on AC – heating, cooling and air movement- adapted for each
EU climate
The gains are expressed nationally (table 9.4) for a consistent effort of all countries in the same way as the
UK is doing. The gains are represented by the figures in table xx
Table 9.4
kWh/m² Present average New buildings New
installations
Austria 225,0 183,6 202,4
Belgium 200,2 163,4 180,1
Denmark 224,9 183,5 202,3
Finland 253,7 207,0 228,2
France 198,1 161,6 178,2
Germany 225,0 183,6 202,4
Greece 201,6 164,5 181,3
Ireland 182,4 148,8 164,1
Italy 201,6 164,5 181,4
Luxembourg 200,2 163,4 180,1
Netherlands 196,0 159,9 176,4
Portugal 201,6 164,5 181,4
Spain 165,3 134,9 148,7
Sweden 253,7 207,0 228,2
UK 182,4 148,8 164,1
Table 9.5
Savings (TWh) 2005 2010 2015 2020
Savings due to 4,74 9,62 15,07 16,45

scenario 6
Such a regulation provides a gain not so high as a very strong prescription, but is safer because the designer
can put the effort on any segment of the project.
177
10. POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC ENERGY
PERFORMANCE
10.1 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. As CAC systems are diverse in
nature, are often designed on-site rather than simply being factory made packaged-systems and are installed
and operated in diverse circumstances, viable policy measures will need to take account of the diverse
circumstances which apply to them in order to realise the potential energy savings. CAC equipment, 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. These type of measures include certification, energy labelling,
and minimum energy performance requirements (either mandatory or voluntary in nature). Proposals for
each of these are made in section 10.3. In the case of minimum energy performance requirements, these
could be developed within the mandate of the proposed Directive on Ecodesign of End-Use Equipment;
however, 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, 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.
Important as these measures are, 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, installation and operation of specific CAC systems. Policy measures
which can realise these savings at the design and installation stage are typically provided through building
thermal regulations. 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; however,
there are many more areas for action than are specified within it. The most advanced national building
thermal regulations addressing the energy consumption of central air conditioning systems are in the UK and
Portugal; yet even these are not as mature or as encompassing as the US ASHRAE 90.1-1999 standard.
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. Specific proposals regarding this are made in sections 10.2, 4 and 5.
10.2 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. The
measures which can be considered to encourage a higher energy efficiency levels for new CAC equipment
are:
• Provision of information (labelling, grading, efficiency ratings)
• Removing less efficient models from the market (MEPS or voluntary agreements)
• Encouraging higher sales-weighed average efficiency levels through negotiated agreements (e.g.
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. Manufacturers generally test the efficiency of their
equipment and report the results in their product literature and catalogues. Without independent verification
178
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. 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, thus the user of the
information is required to have a high degree of expertise to be able to interpret the reported efficiency level.
An effective way of indicating the relative energy efficiency of products in a simplified manner is to use a
categorical energy label, i.e. 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. The current EU energy label
applied to household appliances, which grades appliance energy efficiency using an A to G rating scale, is an
effective and widely known example of this.
CAC equipment are generally large and are not generally ‘on display’ in a shop window at the moment of
purchase. It is therefore debatable whether there is much advantage in applying a removable energy label to
the product itself; however, 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. 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. 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. 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.
The current EU energy labelling Directive is restricted to household appliances hence would require
amendment to address this issue. 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. However, 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, 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. As mentioned a
removable paper label as currently applied to household appliances is not likely to satisfy these requirements.
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, including on-line sales, 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.
For the time being there is no such scheme in place and it may be some time before one is formally
developed; however, industry associations, such as Eurovent, have expressed an interest in adopting such a
comparative grading approach, which could be applied by them on a voluntary basis (e.g.. 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). By so doing it would allow manufacturers to express the
relative energy performance of their products and, as has been seen for household appliances, would allow
greater differentiation of products within the market place. 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. In the spirit of aiding the rapid adoption of an efficiency grading scheme for central air
conditioning equipment, 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.
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. 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.
179
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.
Full or part-load? The first issue to consider is whether the scale should be based on full or partial load
ratings. 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; however, at the present time
there is no accepted part-load test and only full load efficiency ratings are generally available. The analysis
of Chapter 8 has led to a proposal for an EU IPLV for chillers based upon specific rating conditions;
however, 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. 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. The results of
Chapter 8 have indicated that commonly, but not always, the relative efficiency of equipment determined at
full load is indicative of its relatively efficiency at part-load, therefore there is little risk of misleading the
public by adopting an initial grading system based on full load performance. Furthermore, the part-load
ratings are usually equivalent to or slightly higher than the full load ratings, which suggests that the same
efficiency range may be applicable to both the full and part-load efficiency grades. Accordingly this report
presents proposals for A to G efficiency grades based on the analysis of full-load performance data, which
are intended for use in the near term. In the future, at such a time when EU part-load ratings are widely
available, it would be appropriate to review the appropriateness of these grades for translation into part-load
grades.
Heating and cooling-modes. 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. Therefore the
same approach is followed here.
Product categories. 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.e. were the energy consumed by the cooling tower
to be included) it would be possible to use the same scale for both; however, 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.
The basic chiller types to be separated correspond at least to the different testing conditions applied in the
test standard, which inherently generate incomparable figures as follows:
• water cooled (in cooling and/or heating-mode for reversible units),
• air cooled (in cooling and/or heating-mode for reversible units),
• floor-feed systems (in cooling and/or heating-mode for reversible units);
• condenser-less units.
The efficiency ratings are not directly comparable between these 7 distinct sets of test conditions.
The chillers could also be separated into two categories as a function of cooling capacity say, those with a
capacity less than 750kW and those with a capacity greater than 750kW. 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. Some centrifugal units
are already integrated in the catalogue; although it seems they can be graded on the same scale.
Some markets require ducted condensers, which degrade the EER. It is proposed to introduce a specific
classification for them. 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 0.3) and in fact heat
180
exchange drops in this case, and the correction of the standards cannot translate the reality. Establishing a
classification based on specific statistics for ducted units will eliminate the two problems. However, when
testing ducted units, the available static pressure is set by the design of the manufacturer. For this reason, the
testing does not give comparable figures. Therefore it is recommended that manufacturers should supply a
common static pressure for testing ducted units.
Proposed grading scale
In order to require the same effort for each type of chiller, 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. The average EER and COP values are of course based on data which include units using
HFCs. The following classes and values regarding R22 are given as an information only since the European
market will no longer have R-22 chillers.
The construction of the scale for the different categories, intends to respect the following rules, classified by
order of importance:
• use of the same classes width (for simplicity),
• use of limits of classes ending by 0.05 or 0.1,
• adjustment of the extremes (G of about 10 %, A of about 1 %),
• centering on the average of the catalogue (equal treatment between types),

From these basic rules come the following proposals (Tables 10.1 and 10.2).
Table 10.1 Proposed efficiency grades for chillers in the cooling-mode
EER boarders Air Cooled Air Cooled, Water Cooled Water Cooled, Remote condenser
Floor heating and
Floor heating and cooling
cooling
A/B 3.10 3.65 5.05 4.75 3.55
B/C 2.90 3.50 4.65 4.60 3.40
C/D 2.70 3.35 4.25 4.45 3.25
D/E 2.50 3.20 3.85 4.30 3.10
E/F 2.30 3.05 3.45 4.15 2.95
F/G 2.10 2.90 3.05 4.00 2.80
Note: for borders, A, for air-cooled units, is strictly superior to 3.10.
Table 10.2 Proposed efficiency grades for chillers in the heating-mode
COP boarders Air Cooled Air Cooled, Water Cooled Water Cooled,
Floor
Floor
A/B 3.25 4.20 4.45 4.50
B/C 3.05 4.05 4.15 4.25
C/D 2.85 3.90 3.85 4.00
D/E 2.65 3.75 3.55 3.75
E/F 2.45 3.60 3.25 3.50
F/G 2.25 3.45 2.95 3.25
Note: for borders, A, for air-cooled units, is strictly superior to 3.25.
Impact of the proposed grading on the chiller cooling market
181
Figure 10.1. Air cooled chillers, cooling-mode, < 750kW
40.0%
35.0%
30.0%
25.0%
R407C
R134a
20.0%
R22
HFC
15.0%
10.0%
5.0%
0.0%
A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)
refrigerant R407C 2.41

refrigerant R134a 2.55
refrigerant HFC 2.42
refrigerant R22 2.59
A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1) 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% 8.0% 28.3% 19.6% 21.7% 19.6% 2.9% 100.0%
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
Figure 10.2. Water cooled chillers, cooling-mode, < 750kW
45%
40%
35%
30%
R407C
25%
R134a
R22
20%
HFC
15%
10%
5%
0%
A (>5,05) B (>4,65) C (>4,25) D (>3,85) E (>3,45) F (>3,05) G (<3,05)
refrigerant R407C 3.79

refrigerant R134a 4.34
refrigerant HFC 3.85
refrigerant R22 3.88
A (>5,05) B (>4,65) C (>4,25) D (>3,85) E (>3,45) F (>3,05) G (<3,05) 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 0.0% 21.4% 39.3% 25.0% 14.3% 0.0% 0.0% 100.0%
SUM R22 5 3 15 19 28 14 10 94
% R22 5.3% 3.2% 16.0% 20.2% 29.8% 14.9% 10.6% 100.0%
%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 10.3 and 10.4.
183
Figure 10.3. Sales and Market statistics
Sale statistics R407C Eurovent catalog values R407C
1,6 1,6
3,3 2000 1,7 3,3 250 1,7
3,2 1500 1,8 3,2 200 1,8
150
3,1 1000 1,9 3,1 1,9
100
500
3 2 3 50 2
0 0
2,9 2,1 2,9 2,1
2,8 2,2 2,8 2,2
2,7 2,3 2,7 2,3

2,6 2,4 2,6 2,4
2,5 2,5
Sales and Eurovent catalogue statistics for air condensing units are convergent: sales statistics show a better
efficiency.
Figure 10.4. Sales and Market statistics
Sale statistics R407C Eurovent catalog values R407C
2,7 2,7
5,5 500 2,9 20
5,5 2,9
400
5,3 3,1 5,3 15 3,1
300
10
200
5,1 3,3 5,1 3,3
100 5
0 0
4,9 3,5 4,9 3,5
4,7 3,7 4,7 3,7
4,5 3,9 4,5 3,9

4,3 4,1 4,3 4,1
Sales and Eurovent catalogue statistics for water condensing units are convergent: sales statistics show a
better efficiency..
184
Proposal of grading of packaged AC in Europe (extension of the RAC labelling scheme)
(1) Air-cooled air conditioners - cooling mode
Table 10.3 Proposed efficiency grades for large-split packaged AC in the cooling-mode
Energy efficiency class Energy efficiency ratio
A 3.20 < EER

B 3.20 > EER > 3.00
C 3.00 > EER > 2.80
D 2.80 > EER > 2.60
E 2.60 > EER > 2.40
F 2.40 > EER > 2.20
G 2.20 > EER
Table 10.4 Proposed efficiency grades for large unitary packaged AC in the cooling-mode
A 3.00 < EER

B 3.00 > EER > 2.80
C 2.80 > EER > 2.60
D 2.60 > EER > 2.40
E 2.40 > EER > 2.20
F 2.20 > EER > 2.00
G 2.00 > 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
A 3.60 < EER

B 3.60 > EER > 3.30
C 3.30 > EER > 3.10
D 3.10 > EER > 2.80
E 2.80 > EER > 2.50
F 2.50 > EER > 2.20
G 2.20 > EER
Table 10.6 Proposed efficiency grades for unitary packaged AC in the cooling-mode
A 4.40 < EER

B 4.40 > EER > 4.10
C 4.10 > EER > 3.80
D 3.80 > EER > 3.50
E 3.50 > EER > 3.20
F 3.20 > EER > 2.90
G 2.90 > EER
185
(3) Air-cooled air conditioners – heating mode
Table 10.7 Proposed efficiency grades for large-split packaged AC in the heating-mode
Energy efficiency class Coefficient of performance
A 3.60 < COP

B 3.60 > COP > 3.40
C 3.40 > COP > 3.20
D 3.20 > COP > 2.80
E 2.80 > COP > 2.60
F 2.60 > COP > 2.40
G 2.40 > COP
Table 10.8 Proposed efficiency grades for large unitary packaged AC in the heating-mode
A 3.40 < COP

B 3.40 > COP > 3.20
C 3.20 > COP > 3.00
D 3.00 > COP > 2.60
E 2.60 > COP > 2.40
F 2.40 > COP > 2.20
G 2.20 > COP
(4) Water-cooled air conditioners – heating mode

Table 10.9 Proposed efficiency grades for split and multi-split packaged AC in the heating-mode
A 4.00 < COP

B 4.00 > COP > 3.70
C 3.70 > COP > 3.40
D 3.40 > COP > 3.10
E 3.10 > COP > 2.80
F 2.80 > COP > 2.50
G 2.50 > COP
Table 10.10 Proposed efficiency grades for unitary packaged AC in the heating-mode
A 4.70 < COP

B 4.70 > COP > 4.40
C 4.40 >COP > 4.10
D 4.10 > COP > 3.80
E 3.80 > COP > 3.50
F 3.50 > COP > 3.20
G 3.20 > COP
186
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). Market
average efficiency = 2.46 W/W.
Class definition % on market Grade % with

equal class
width
3.20 < EER 2% A 2%
3.20 > EER > 3.00 5% B 5%
3.00 > EER > 2.80 7% C 7%
2.80 > EER > 2.60 15% D 15%
2.60 > EER > 2.40 22% E 22%
2.40 > EER > 2.20 26% F 26%
2.20 > EER 23% G 11%
Figure 10.5. Air-cooled air conditioners - 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 10.12 Air-cooled air conditioners - Heating function -mixed statistics (splits and packages mixed).
Market average efficiency = 2.87 W/W.
Class definition % on market Grade % with equal

class width
3.60 < COP 3% A
3.60 > COP > 3.40 5% B 5%
3.40 > COP > 3.20 6% C 6%
3.20 > COP > 2.80 46% D 23%
2.80 > COP > 2.60 17% E 17%
2.60 > COP > 2.40 11% F 11%
2.40 > COP 8% G 5%
187
Figure 10.6. Air-cooled air conditioners - Heating function -mixed statistics
% of market(density)
0,25
0,2
0,15
0,1
0,05
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. 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. Note also that we have here only full load COP and EER.
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. This
implies the introduction of energy efficiency requirements for new equipment, which could be mandatory
(MEPS) or voluntary in nature but attaining the same goal. Table 10.13 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. The
adoption of policy measures which would move the average new equipment efficiency levels to those
indicated in Table 10.3 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. 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. 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.
Table 10.13 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 4.50
Water-cooled chillers centrifugals 750 upwards 5-6
Air-cooled chillers 12 to 750 3.00
Large packaged AC (cooling mode) 12 to 75 3.22
RAC 3 to 12 3.20
An alternative approach to MEPS setting is to harmonise levels with those applied internationally. From a
commercial perspective there is some logic to this approach, because most companies supplying the EU
188
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. From a
programme management perspective this has the advantage that the MEPS efficiency thresholds are tried and
tested having been successfully applied elsewhere. Furthermore in the only case where the life cycle cost
analyses have been directly compared, for packaged AC units, the least life cycle cost efficiency level has
been found to be the same, which suggests that the US efficiency thresholds might be appropriate for
adoption in the EU.
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 10.14 ASHRAE 90.1 MEPS levels for CAC equipment (W/W)
Equipment Type MEPS level (W/W) Capacity range

Packaged AC (cooling-only) 3.02 19.5 to 39.5
Packaged AC (cooling-only) 2.84 39.5 to 70.3
Packaged AC (reversible) 3.02 19.5 to 39.5
Packaged AC (reversible) 2.84 39.5 to 70.3
Water-source heat pump, 3.51 Values about 14.1
4tons (cooling-mode)
Water-cooled screw chiller, 4.45 COP (4.50 US IPLV) Values about 440
125tons
Water-cooled centrifugal 6.10 COP (6.10 US IPLV) Values about 1056
chiller, 300tons
Of course there are other options than applying a mandatory minimum energy performance threshold, which
could achieve similar objectives. One approach would be to negotiate fleet-average efficiency targets with
the European industry. This would have the benefit that it would not have to wait that the Ecodesign of End-
Use Equipment Directive be implemented.
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, which could take advantage of any efficiency grading system
that is introduced. Some of potential measures are as follows:
• Establish public sector procurement guidelines (e.g. only class A equipment should be procured for
use in the public sector)
• Develop corporate procurement guidance documentation, 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.g. as in the UK web-site
www.ukepic.org)
• Develop low cost credit lines for more efficient equipment (e.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.g. as in the Dutch rebate scheme for class A household
appliances)
189
• Create favourable tax differentials for efficient equipment (e.g. lower VAT levels, or corporate tax
breaks for manufacturers and/or corporate procurers)
10.3 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, as outlined in section 10.2, will
only realise some of the potential to save energy for CAC systems. Such measures are necessarily focused on
the individual components and not on the performance of the system as a whole, therefore they do not
encompass the freedoms and constraints applying to the system designer in trying to design an efficient yet
effective CAC system. Furthermore they do not address the activities of the installation engineer who is
responsible for executing the system design and commissioning the system. The comparison of the energy
performance of eighteen different CAC systems, each designed to provide total cooling in a typical EU
office, 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. 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, 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. 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, the building type and of the cooling and heating degree days.
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, environmental and air quality
requirements of the client within acceptable cost boundaries.
As such building codes are the primary policy measure which promote the adoption of more efficient system
types; however, 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. One difference
regards how the requirements for energy using systems in building codes should be expressed. 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. In some Member States the building codes are articulated in
terms of limits regarding the maximum allowable energy consumption per m2. Once the fuel of the heating
and cooling system has been fixed these two approaches are effectively transposable; however, 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. 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. In Portugal, however, there are no requirements on overall annual
energy or carbon but rather a set of simple prescriptions to follow, which are designed to save energy. Both
approaches have their merits and indeed can even be integrated as is the case in the US ASHRAE 90.1-1999
standard. 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”, method which requires the use of one of a number of designated detailed building thermal
simulation tools. Following the simple prescriptive measures is an easy way for a designer to demonstrate
190
their compliance with the standard; however, the prescriptions are relatively rigorous compared with the
requirements when a detailed simulation tool is used. Therefore the standard creates an incentive for
designers to use detailed building thermal simulation software. The combination of a prescriptive compliance
pathway and a pathway based on meeting overall energy limits demonstrated, through the use of detailed
simulation software, simultaneously meets the needs of designers “in a hurry” dealing with standard design
briefs and those who have specific and complicated design briefs; who may need more freedom to meet the
same energy goals than would otherwise be allowed through application of a set of simple prescriptions.
Legal basis for policy measures targeting more efficient system structures
As mentioned, building codes are the primary policy measure available to encourage the adoption of more
efficient system types. 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. In the case of the UK this Directive may simply
require minor revisions of the existing regulations for building cooling energy performance, but for most
other Member States it will require completely new regulations to be developed addressing the cooling
system. 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.
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.
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. The intention has been to systematically collect and
summarise the different approaches and to develop suggestions for a European 'model code'. Ultimately such
a code could be the EU equivalent to the US ASHRAE 90.1 standard, 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. Sharing development and analytical effort makes considerable sense for such a major
undertaking and is a key recommendation from this study. 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.
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. Moving from G to F or E might be based only on full-load
ratings i.e. readily available EER values. Higher grades such as E or D could be reached on the basis of a
certain value of SEER, taking part load optimisation efforts into account. Part load is not only a phenomenon
to be computed, it's by itself one of the most important energy saving features.
Following these measures the importance of system design cannot be neglected. A designer could refer to a
design procedure proving he has considered all cost effective options of the project. and reached a certain
performance level like C (- 25 %) or B (-50 %) or even A (-75%) compared with the average European
performance level, just as is currently the case in the US Energy Star for buildings scheme.
In absence of CEN standardisation on many subjects, existing methods in some countries could be
provisionally approved as ways to reach an A or a B. 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,
here by the will of the households, 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.
The satisfaction of the requirements would need to be determined at the design/installation stage through
verification using simulation tools. This implies that public domain simulation tools would need to be
developed to support the model building code. 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.
191
Specific recommendations
• the development of an EU model building code that addresses air conditioning amongst other energy
end-uses. (an EU equivalent to ASHRAE 90.1 and which like ASHRAE 90.1 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, b) can enable the comparison of the relative
benefits of different system designs, 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; occupancy and purpose; quality of comfort provision and climate (e.g. 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.g. 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.g. the integration of airside and waterside economisers
which are capable of operating in conjunction with mechanical cooling). 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.g. variable speed drives or variable pitch fan
blades
• Proper sizing of components such as fans (e.g. 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.g. that the SPF
should not exceed 1.5)
• Limits on the maximum SPF of mechanical ventilation systems in existing buildings (e.g. that the
SPF should not exceed 3.0)
• 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. simultaneous heating & cooling)
• Stipulations for off-hour controls including:
192
1. 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.)
2. Zone isolation capabilities that permit areas of the building to continue operating while others are
shut down
3. Automatic shutdown
4. Setback controls
5. 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, including modular plant (i.e. good part-load efficiency) (e.g. the
use of power stages to allow output to be adapted to the demand)
• Efficient control of heat rejection equipment capacity, including modular plant (e.g. good part-load
efficiency for cooling towers)
• Full cold thermal storage (i.e. 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.g. electric heating power provided by the Joule effect should not
exceed 5% of the total heating power installed, nor 25kW by independent zone).
• Limits on terminal re-heating. (e.g. terminal re-heating is allowed for cooling-only systems but can
not exceed 10% of the installed cooling power).
For the installation and commissioning of CAC systems introduce building code measures which
encourage:
• commissioning tests to be conducted for boilers, chillers (power and efficiency), cooling towers,
pumps, hydraulic tests, heat exchangers, controllers, 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
• 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. Certificates shall not be valid for more than 10 years.
• The certificate shall include reference values such as legal standards and benchmarks. It shall be
accompanied by recommendations for the cost-effective improvement of the energy performance .
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, 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.
10.4 Policies and measures to improve system maintenance and operation

Policy aims and potential measures targeting improved O&M
The maintenance or improvement of performance, by technical measures or contractual means (Energy
Performance Contracting) or by periodic audits can result in significant energy savings for CAC systems.
Optimal operation of the system can be encouraged through intelligent control regimes which in turn can be
encouraged through appropriate energy performance contracting. 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.
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.
It has been estimated that without proper maintenance the efficiency of CAC systems deteriorates by 2%
every 5 years. 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. In particular the service and cleaning of heat exchangers should be properly encouraged.
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. 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.
Member States are obliged to ensure that this inspection is conducted in an independent manner by qualified
and or accredited experts.
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.
194
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. However, 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. 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. Some aspects of these guidelines would necessarily be specific to the
situation applying in individual Member States, but some would be common to all Member States. 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, so that the findings
could be fed into the national provisions being drawn-up by 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.
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, especially as it is provided by utilities and service companies in the form of energy services. It is
envisaged that Member States will set targets to promote and support energy efficiency services, (e.g., third
party financing) and programmes, especially for smaller energy consumers such as SMEs. This could
certainly be used as an opportunity to improve the O&M of CAC if appropriate rules can be defined.
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 12.5 kW
should be independently metered.
The UK regulations also require that the owner and/or tenant of the building be provided with a logbook that
contains, amongst other things, the design assessment for CPR or other benchmarks, commissioning details,
operating instructions, and details of all meters provided. There is an additional requirement that sub-
metering should be provided. This includes separate metering for tenancies of more than 500 m2 (though for
tenancies below 2500 m2, proportioning of cooling may be acceptable). Generally, 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.
The US ASHRAE 90.1 standard requires that drawings, manuals, and a narrative of system operation
must be supplied to the building owner. This is a sensible provision because even if an engineer designs a
great system, it's unlikely that energy savings will accrue if the operator doesn't know how the system should
work. The standard also addresses balancing for air systems larger than 1 hp and for hydronic systems larger
than 10 hp. It also requires control elements to be calibrated, adjusted, and in proper working condition for
buildings that exceed 50,000 sq ft.
Specific recommendations
• 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.
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
• A regular maintenance and monitoring system be adopted for all CAC systems (e.g. 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.
196
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.
Categorical energy label An energy label that classifies product efficiency into one of several classes. Examples
include the EU’s energy labels, which rank efficiency from A to G, and Australia’s energy
label, which ranks efficiency from 1 to 6 stars. Korea, Thailand, Iran, Brazil, Mexico and
India have all developed categorical energy labels.
Control cycle The period between two successive starts or two successive stops of the compressor in a
refrigerating system.
Defrost cycle The period between two successive starts or two successive stops of a defrost heater in a
appliance with an automatic defrost system.
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’).
Montreal protocol The internationally binding agreement to phase out ozone-depleting substances such as
CFCs.
Net present value (NPV) The monetised value of future costs expressed in terms of their discounted value at the
present time.
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. The payback period can be
‘simple’ in that no discounting of future savings is applied, or it can be the converse in
which the future savings are discounted using a real discount rate.
Thermal bridge A high thermal conductivity pathway.
Top Runner The term applied to the Japanese appliance energy-efficiency policy, wherein MEPS have
been set at efficiency levels equivalent to those of the highest efficiency appliance on the
market.
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, Heating and Refrigerating Engineers.
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, Refrigerating and Air Conditioning Engineers
BAT Best Available Technology
BAU Business As Usual
BRE Building Research Establishment
197
CAC Central Air Conditioners
CAHORE Cafes, Hotels, 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)
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)
ESCO Energy Service Companies
EU European Union
Eurovent European association of refrigeration, air-conditioning and ventilation equipment manufacturers
FCU Fan Coil Unit
GB Great Britain (excludes Northern Ireland)
GEA Group for Efficient Appliances
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
RSECE The Portuguese building thermal regulations which include requirements for AC systems
SEER Seasonal Energy-Efficiency Ratio (W/W)
SPF Specific Fan Power, in Watts (of motor rating) per litre/second of airflow.
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, 2002, Adnot J., Becirspahic S., Marchio D., Colomines F., Rivière P., Seasonal efficiency of primary
air conditioning systems, Procedings of the AFCE conference, Ecole des Mines de Paris, Sept 2002.
AICARR,2001 Average weighed efficiency of compression chillers: AICARR’s proposal for a calculation
method
AICARR, 2001, E. Bacigalupo, C Vecchio, M. Vio, M. Vizzotto., 2001, Average weighed efficiency of
compression chillers: a proposal to AICARR for a calculation method, Permanent Technical Committee for
"Refrigeration" in AICARR's Technical Activity Commission.
ARI, Standard 550/590, Water Chilling Packages using the vapor compression cycle, 1998.
ASHRAE Handbooks Fundamentals and Systems. J.F Kreider, A.Rabl, 1994, Heating and Cooling of
Buildings, Design for Efficiency, Mc Graw-Hill Book Company.
J.Bouteloup, M.Le Guay, J.Liguen, 4 vol, 1996, 1997, 1998, 1999, Air-conditioning, Air Handling, “Les
Editions Parisiennes”, France.
CEN, 1997, « Air conditioners and heat pumps with electrically driven compressors – Heating mode”,
standardCenelec EN 255, 1997.
CEN, 1998, « Air conditioners and heat pumps with electrically driven compressors – Cooling mode”,
standardCenelec EN 12055, 1998.
199
CEN, 2002, prEN 14511 – 1 à 5, proposed revision of [CEN98] and [CEN97], 2002.
CIBSE, 2004, "Guidance for the use of the carbon emissions calculation method: CIBSE TM32:2003" (ISBN
1 903287 41 3)
Ph Davy de Virville, 1994, Control Guide for Ventilation and Air-conditioning, thermic school, Paris,
France.
EERAC [1999] Energy efficiency of room air-conditioners. Paris: Ecole des Mines de Paris et al, for DG-
TREN, the Commission of the European Communities , SAVE contract DGXVII4.103/D/97.026, May.
Eurovent, 1998 Eurovent-Certification. 1998, Annuaire des Produits Certifiés. www.eurovent-
certification.com.
LBNL (2002) ‘Commercial Unitary Air Conditioner & Heat Pump: Life-Cycle Cost Analysis -
Inputs and Results’, Lawrence Berkeley National Laboratory for the US Department of Energy, December
2002.
RHEVA, 1993, The international dictionary of heating, ventilating and air-conditioning, E&FN Spon,
England.
REHVA, 1997 , Preliminary European Guide on HVAC design, Dominique Marchio, Eric Auzenet, ECEEE
1997 summer study proceedings
Rivière, 2001 general presentation and use of a method of calculation of consumption of the RAC (Room Air
Conditioners) by Ph. RIVIERE, J. ADNOT, M. ORPHELIN
Tiax (2002) ‘Engineering-analysis cost-efficiency curves. Commercial unitary air-cooled air-conditioners and air-source
heat pumps’, TIAX LLC, Cambridge Massachusetts for the US Department of Energy, January 2002.
TRIBU, (1994). "Etude Comparative des Réglementations Thermiques Européennes des Bâtiments non
résidentiels", TRIBU, 1994 for Ademe
UNI, (2002) UNI Standard 10963 : Air conditioners , chillers and heat pumps- part load tests.
Westphalen, 1999) "Energy Consumption Characteristics of Commercial Building HVAC Systems Volume
II: Thermal Distribution, Auxiliary Equipment, and Ventilation" Prepared by Detlef Westphalen and Scott
Koszalinski, Arthur D. Little, Inc. For U.S. Department of Energy, October 1999
200
View publication stats

Eeccac

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Eeccac

Încărcat de

Drepturi de autor:

Formate disponibile

See

Energy Efficiency and Certification of Central

Technical Report · April 2003

Philippe Rivière Luis Pérez-Lombard

SEE PROFILE SEE PROFILE

José Ortiz Roger Hitchin

SEE PROFILE SEE PROFILE

IAQ & VENTILATION IN HEALTHCARE FACILITIES View project

The user has requested enhancement of the downloaded file.

Energy Efficiency and Certification of Central

Tel: (+33) 1 40 51 91 74 Fax: (+33) 1 46 34 24 91

SUMMARY OF RESULTS ............................................................................. 11

2. CENTRAL AIR-CONDITIONERS IN EUROPE: DEFINITIONS

2.2. Basic definitions ........................................................................................................................ 39

2.3. Description of other aspects of systems ....................................................................................... 44

2.4 Description of systems not using chillers ..................................................................................... 49

2.5. Testing standards and performance standards .............................................................. 51

2.6. Overall view of energy performance ........................................................................................... 58

2.7. Statistical databases used and information gathered ................................................................. 60

3. MAIN FIGURES OF AIR-CONDITIONING IN EUROPE ................ 62

3.2. Technical response to the demand...................................................................................... 64

3.3. A few technical trends on the market ................................................................................ 68

3.4. Statistics on present Energy Efficiency on the EU market ......................................... 73

4. FACTORS GOVERNING THE DESIGN, SELECTION, INSTALLATION

4.2 Practices and procedures adopted in CAC system design .......................................................... 77

market-transformation efforts within the EU (equipment) .............................................................. 78

5. PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO) .......... 100

5.2 Computation of energy consumption in European conditions ................................. 104

5.3 Energy consumption in 1990, 1998, 2010 and 2020..................................................... 112

5.4 Global warming and other environmental impacts .................................................... 115

5.5 Heating, reversible or not .................................................................................................... 118

6. TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY

6.2 Engineering approach of the performance of Packaged units .................................................. 127

7. TECHNICAL & ECONOMIC EVALUATION OF CAC SYSTEM PERFORMANCE AS A

7.1 Comparison of different CAC systems ....................................................................................... 134

7.3 Other cost & efficiency trade-offs ............................................................................................... 138

8. EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE ........ 143

8.4 Derivation of a new SEER method (ESEER) ............................................................................. 156

9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY

9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY

9.2 Results of scenarios....................................................................................................................... 174

10. POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC

10.1 Some fundamental considerations regarding policy measures .............................. 178

Definitions and general terms used in the study .............................................................................. 197

List of abbreviations ........................................................................................................................... 197

REFERENCES ................................................................................................................................... 199

ROOM BY ROOM SERIES OF ROOMS BUILDING

RAC Roof Top-Splits Other CAC

FLUID: AIR ONLY FLUID: AIR AND WATER FLUID: REFRIGERANT

A.H.U.s and INDUCTION FAN-COIL

2 pipes 3 pipes 4 pipes

AREA COOLED EU-15 added (or replaced.) m2

The present Energy Efficiency efforts have been reviewed

Class definition % on market Grade % with equal class

Total A/C in 2000

Reference office building unitary cooling consumptions

Total cooling consumption by subtype

150 000 FCU

Cooling function 22,879 33,683 51,636 78,103 94,727 109,631 114,579

Total Cooling Year

Table 6 Numerical results about CO2 emissions due to cooling in Europe

Kt CO2 2000 2005 2010 2015 2020

Figure 11 The cost of a chiller at nominal capacity according to its EER