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 HOUDANTEnergie 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

11

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

12

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.

13

Figure 1. CAC system description tree showing the most common CAC systems.

‘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

LOCAL OR CENTRAL

LOCAL CENTRAL

ROOM BY ROOM

Roof Top-Splits

SERIES OF ROOMS BUILDING

R A C Other CAC

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

A.H.U.s and DUCTWOR

INDUCTION UNITS

FAN-COIL UNITS

2 pipes 3 pipes 4 pipes

MULTI-SPLIT

V R F

14

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(YEARLY MARKET)

EU-15 added (or replaced.) m2

0,00

20,00

40,00

60,00

80,00

100,00

120,00

1975 1980 1985 1990 1995 2000 2005

Mm2

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

Germany11%

Greece5%

Spain24%

France12%

Portugal2%

Italy25%

UK8%

Others13%

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.

15

Figure 4. AC market share by AC type expressed in terms of newly installed cooled-area in EU buildings in 1998

Splits >12kW7%

chillers45%

Packages5%

Rooftops5%

VRF2%

RAC< 12 kW36%

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

USA (EIA)

chillers

packages

all RAC

Figure 5b. The share of cooled-floor area by AC type in non-residential buildings in the EU for 1999-2000

16

EUR (EECCAC)

chillers

packages

all RAC

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)

R2 = 0.0003

R2 = 0.0073

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0 100 200 300 400 500 600 700 800 900 1000 Capacity kW

EER

air cooled water cooled Regression (air cooled) Regression (water cooled)

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

17

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.

18

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 3.29reversible air conditioning 1.9 2.48 2.96

Floor 3.31 3.34 3.39Cooling only water conditioning 2.9 3.73 4.09reversible 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.

19

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

0,0%

5,0%

10,0%

15,0%

20,0%

25,0%

30,0%

35,0%

40,0%

A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)

R407C

R134a

R22

HFC

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%.

20

Figure 8. Average cooled-floor area per inhabitant for EU countries and the EU as a whole in 2000.

Total A/C in 2000

0

1

2

3

4

5

6

7

B DK D GR E F IRL I L NL A P FIN S UK EU-15

m2/inhabitant

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

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

21

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

0,0

20,0

40,0

60,0

80,0

100,0

120,0

140,0

Air Coo

led w

ith w

ater d

istrib

ution

Air Coo

led w

ith ai

r dist

ributi

on

Air Coo

led w

ith ai

r +hu

midity

contr

ol

Wate

r Coo

led +

water d

ist.(c

oolin

g)

Water C

ooled

with

air d

ist.(c

oolin

g)

Water C

ooled

+air +

hum.(c

oolin

g)

Outside

wate

r + w

ater d

ist

Outside

wate

r + ai

r dist

Outside

wate

r +air

+hum

TWO LO

OPS + CHILL

ERVRF

PACK&SPlarge

Rtops

RACs on o

ne lo

op

MSSpli

ts

PACKsmall

Single

Ducts

CAC Systems

kWh/

m2 London

MilanSeville

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

0

50 000

100 000

150 000

200 000

250 000

300 000

1990 1995 2000 2005 2010 2015 2020

GW

h

RACPACKFCUCAVtotal conventionnal heating

22

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 (Electricity only)

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

Heating function Without REV.

51,598 74,442 111,084 164,517 203,330 236,765 250,844

Heating function With present REV. (El.)

7,374 11,495 18,894 28,913 35,875 42,333 45,040

Table 5 Cooling only energy consumption by country and year (for comparison with national projections)

Total Cooling GWh/ year

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

23

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

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

Optimisation of Cost/kW final

80

90

100

110

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

EER

Euro

/kW

24

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

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

€0

€200

€400

€600

€800

€1,000

€1,200

€1,400

2.6 2.8 3.0 3.2 3.4 3.6EER (W/W)

Life

cyc

le c

ost (

€/kW

c

3

4

5

6

7

8

9

10

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

SEER

Tota

l cos

t (Eu

ros/

m2)

ALCC17-800hALCC10-800hALCC17-400hALCC6-800hALCC10-400hALCC6-400h

25

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

SPECIFIC CONSUMPTION

ALCC Euros/m2/year

kWh/m2/YEAR

airwaterpackagesrac

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

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

27

29

31

33

35

37

39

0,00% 20,00% 40,00% 60,00% 80,00% 100,00% 120,00%% of reference

Euro

s/m

2 ALCC17ALCC10ALCC6

26

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

27

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

Circuit 1 Compressor 1 0 0 0 1 Compressor 2 0 1 1 1

Circuit 2 Compressor 3 0 0 1 1 Compressor 4 1 1 1 1

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.

28

Figure 16 Reduced efficiency while decreasing part load ratio (same source temperatures) for the testedwater cooled chillers

The overall performance improvement (or degradation) at part load (temperature effects being substracted) is given on figure 17 for the five water cooled tested units. Figure 17 Reduced efficiency while decreasing part load ratio (same source temperatures) for the water cooled chillers

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.

Reduced efficiency of the part load stages for air cooled chillers

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

0 0.2 0.4 0.6 0.8 1 1.2

Part load ratio

N° 5N° 7N° 8N° 9N° 2

Reduced efficiency of the part load stages for water cooled chillers

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

0 0.2 0.4 0.6 0.8 1 1.2Part load ratio

N° 4N° 1N° 3N° 6

29

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

Part load ratio Temperatures Weighting coefficients Temperatures Weighting

coefficients Temperatures Weighting 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) Weighting coefficients Temperatures Weighting

coefficients Temperatures Weighting coefficients

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).

30

Figure 17. Comparison of HSEER with EER on the tested chillers

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 EER

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5

EER

HSE

ER

HSEER versus IPLV

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4 4,5

IPLV

HSE

ER

31

Figure 19 comparison of EMPE with HSEER for the tested chillers

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

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

HSEER versus EMPE

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4

EMPE

HSE

ER

HSEER versus ESEER

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4

ESEER

HSE

ER

32

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.

Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON FULL LOAD INFO

On average packages and chillers reach in 2005 the EER level corresponding to the minimum LCC (Best Available Technologies with present information); 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.

33

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

34

• 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).

35

36

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

37

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

38

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.

39

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

40

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

ACsystem

Pe energy input

Pc cooling effect control of: Temperature Humidity IAQ

air conditioned space

Pr heat rejected

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.

41

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.

42

• 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:

43

• 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.

44

Figure 2.5. CAC system description tree showing the most common CAC systems

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

LOCAL OR CENTRAL

LOCAL CENTRAL

ROOM BY ROOM

Roof Top-Splits

SERIES OF ROOMS BUILDING

R A C Other CAC

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

A.H.U.s and DUCTWORK

INDUCTION UNITS

FAN-COIL UNITS

2 pipes 3 pipes 4 pipes

MULTI-SPLIT

V R F

45

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

Water to airHeat exchanger

Circulationfan

Air filterOutdoorair supplyIndoor

air return

46

Outdoor air

Return air

variable flow boxes

S

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.

47

Figure 2.9 Induction Unit

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

ZONES

AIR

DISTRIBUTION

AIR

VOLUME

AIR

TEMPERATURE

A.H.U.

CONFIGURATION

SINGLE-ZONE SINGLE DUCT FIXED VARIABLE CAV with terminal re-heat

with by-pass on re-

Air from central system

Indoor air

CoilFilter

Diffusion

Nozzles

48

circulation

with by-pass on mixed air

with control on coil capacity

MULTI-ZONE SINGLE DUCT

PER EACH ZONE

FIXED

PER ZONE

VARIABLE

PER ZONE

Multi-zone CAV

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

PRIMARY AIR

SECONDARY AIR

FIXED

VARIABLE

VARIABLE

CONSTANT

High pressure systems

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

TEMPERATURE

AIR DISTRIBUTION

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

49

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

50

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

51

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 full load)

ARI Standard 550/590 – 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

52

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 temperature

45°C

Entering condenser hot water temperature

40°C

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 (DB)

Water entering condenser

Operating hours %

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

53

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

54

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.

55

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

Dt < 0°C

ev

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

Dt < 0°C

ev

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-

56

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.

57

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

wet-bulb

27

19

21

15

29

19

Temperature of air entering outdoor side (°C)

dry-bulb

wet-bulb1)

35

24

27

19

46

24

Condenser water temperature2) (°C)

inlet

outlet

30

35

22

27

30

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.

58

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

59

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.

60

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.

61

Manufacturers directories

Nationalconfidential

sales data

EUROVENT confidential

sales data

EUROVENT Directories

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.

62

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

0,00

20,00

40,00

60,00

80,00

100,00

120,00

1975 1980 1985 1990 1995 2000 2005

Mm

2

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)

Source: EECCAC from Eurovent Experts and country reports

0,00

5,00

10,00

15,00

20,00

25,00

30,00

1975 1980 1985 1990 1995 2000 2005

Mm2addedorreplaced

ItalySpain

GermanyFranceOthersUKGreecePortugal

63

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

Germany11%

Greece5%

Spain24%

France12%

Portugal2%

Italy25%

UK8%

Others13%

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

0102030405060708090

100

Aus

tria

Fran

ce

Por

tuga

l

Italy

UK

defa

ult

OthersTradeOffices & work placesHotels / restaurant / barHospitals

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).

64

Figure 3.5. Share of conditioned floor space by building type for each AC system type across the EU

0,00%10,00%20,00%30,00%40,00%50,00%60,00%70,00%80,00%90,00%

100,00%

RA

C

chille

rs

pack

ages

&spl

its

Roo

ftops

VR

F

educationhousestradeofficeshotels&barhospitals

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

Splits >12kW11%

chillers71%

Packages8%

Rooftops7%

VRF3%

Evolution of market shares of techniques Figure 3.7 indicates which systems and segments are experiencing the largest growth (from 1996 to 2000).

65

Figure 3.7. Average annual rate of growth in conditioned floor area by type of CAC for the period 1996-2000

00,020,040,060,080,1

0,120,140,16

LargeSplits+14%

Chillers+8,5%

Packages+2,5%

Rooftops+9%

VRV+13%

SmallA/C

+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

0,00%

20,00%

40,00%

60,00%

80,00%

100,00%

120,00%

hosp

itals

hote

ls&

bar

offic

es

trade

hous

es

educ

atio

n

VRFRooftopspackages&splitschillersRAC

Figure 3.9 shows the growth in conditioned floor area by each type of CAC&RAC system from 1980 to 2000 across the EU.

66

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

0

20

40

60

80

100

120

140

160

1980 1985 1990 1995 1998 2000

Mm2

RAC<12 kWVRFRooftopsPackagesSplits >12kWchillers

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 loops2%

Air (AHU)39%

Nat, Water 1%

Classic(FCU)

58%

Two loopsAir (AHU)Nat, Water Classic(FCU)

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.

67

Figures 3.11. The share of non-residential conditioned building floor area cooled by each primary AC type in the USA in 1999-2000

USA (EIA)

chillers

packages

all RAC

Figure 3.12. The share of non-residential conditioned building floor area cooled by each primary AC type in the EU in 2000

EUR (EECCAC)

chillers

packages

all RAC

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.

68

Figure 3.13. Conditioned non-residential building floor area by AC type in the EU and in the USA in 2000

0100020003000400050006000700080009000

USA (EIA) EUR (EECCAC)

all RACpackageschillers

Mm2

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

0%20%40%60%80%

100%120%

Austria

Belgium

...Fran

ce

German

y

Greece Ita

ly

Netherl

ands

Portug

alSpa

in

Ukingd

om

69

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.

0%10%20%30%40%50%60%70%80%90%

100%

Reversibility ofchillers

Total reversibility %age of watersystems

SpainFranceItalyUK

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.

70

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.

0% 20% 40% 60%

Large Splits

chillers

Packages

Rooftops

VRV

RAC

UKItalyFranceSpain

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

0

50

100

150

200

250

UK Spain France Italy

Coo

ling

capa

city

(kW

)

Size of large Splits(kW )Average size ofchillers (kW )

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.

71

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)

0,0020,0040,0060,0080,00

100,00120,00140,00160,00180,00

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Inde

x re

lativ

e to

199

6

France

UK

Italy

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)

Carrier (FR-IT)

York (UK-DK-FR) +Daikin

35 % 12 %

B (7) Climaveneta (IT)

Clivet (IT)

Mc Quay (IT)

RC (IT)

Lennox (FR)

CIAT (FR)

AERMEC (IT)

40 % 6 %

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

72

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 EUR

% import for production

Balance net Capacity MW/1998

Value MEuros/

/MW

Estimated Income (MEuros)

chillers

&CT

95% 5% 90% 6.8 0.25 1700

splits

> 12kW

90% 5% 85% 1.1 0.2 200

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 weighted in income

87.7% - 84.2% 21.6 - 4100

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

73

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)

R2 = 0.0003

R2 = 0.0073

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0 100 200 300 400 500 600 700 800 900 1000

Puissance frigorifique kW

EE

R

air cooledwater cooledLinéaire (air cooled)Linéaire (water cooled)

Source: Eurovent directory

74

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 in kW

Number of models

EER min

EER ave

EER max

Packaged, cooling only, air cooled, conditioning

≤50kW 50kW<CC≤100kW 100kW<CC≤300kW 300kW<CC≤500kW >500kW

174 102 99 4 6

1.9 1.93 2.12 2.52 2.41

2.55 2.49 2.53 2.56 2.57

3.3 3.04 2.85 2.59 2.66

Packaged, reverse cycle, air cooled, conditioning

≤50kW 50kW<CC≤100kW 100kW<CC≤300kW 300kW<CC≤500kW

105 35 14 5

1.9 1.99 2.1 2.56

2.48 2.43 2.49 2.73

2.96 2.84 2.73 2.86

Packaged, reverse cycle, air cooled, floor cooling/heating

≤50kW 6 3.31 3.34 3.39

Packaged, cooling only, water cooled, conditioning

≤50kW 50kW<CC≤100kW 100kW<CC≤300kW 300kW<CC≤500kW >500kW

8 10 31 20 15

3.31 3.55 2.9 3.16 2.9

3.75 3.77 3.72 3.79 3.62

4.06 3.96 4.05 4.04 4.09

Packaged, reverse cycle, water cooled, conditioning

≤50kW 50kW<CC≤100kW 100kW<CC≤300kW 300kW<CC≤500kW >500kW

8 5 3 5 7

2.99 2.9 2.9 3.85 3.84

3.28 3.18 3.45 3.94 3.98

3.5 3.5 3.8 3.98 4.09

Remote condenser, cooling only, water cooled, conditioning

≤50kW 50kW<CC≤100kW 100kW<CC≤300kW 300kW<CC≤500kW >500kW

6 3 14 6 7

3.13 3.16 2.96 2.87 2.76

3.32 3.2 3.27 3.18 3.03

3.53 3.25 3.7 3.46 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 3.29reversible air conditioning 1.9 2.48 2.96

Floor 3.31 3.34 3.39cooling water conditioning 2.9 3.73 4.09reversible 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

75

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

1.58

5.3

11.6

13.04

21.0620.34

13.47

11.17

2.44

0

5

10

15

20

25

75-80 81-85 86-90 91-95 96-100 101-105 106-110 111-120 121-130

EER/EER aver

% of models

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.

76

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 EERScroll air 178 12.2 49.8 158 1.9 2.5 3.39

wat 24 13.7 64 163.6 3.11 3.51 4Screw 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 4.09Reciprocating 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 EERRefrigerant Min. Ave. Max. Min. Ave. Max.

R22 12.1 143.03 921 1.9 2.79 4.09R407C 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.

77

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:

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 ;

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 ;

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);

the competition between manufacturers is only expressed in terms of Euro/kW, not Euro/EER or Euro/kWh consumed later ;

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;

the problem for consultants of completely specifying and certifying the quality of something so complex, built on site and only once,

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

78

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.

79

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.

80

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

air / water

Outdoor Unit Water Coefficient of performance

minimum

Dry Bulb temperature(°C)

Wet Bulb temperature(°C)

Inlet temperature (°C)

Outlet temperature (°C)

Cooling mode

35 ** 12 7 EER>2,2

Heating mode

7 6 40 45 COP>2,5

-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

81

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.

82

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%

83

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:

84

- 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.

85

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

86

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

87

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

88

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:

89

(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

90

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, 15tons

8.5EER 9.7EER ARI340/360

Water-source heat pump, 4tons (cooling mode)

9.3EER (85°FEWT) 12.0 EER (86°FEWT) ARI320d(ARI/ISO-13256-1after29-Oct-2001)

Water-cooled screw chiller, 125tons

3.80 COP 3.90 IPLV 4.45 COP 4.50 IPLV ARI590

Water-cooled centrifugal chiller, 300tons

5.20 COP 5.30 IPLV 6.10 COP 6.10 IPLV ARI550

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.

91

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-

92

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 Weighting

Electricity Gas Site EUI Source EUI $UI (USD)

Assembly 0.068 9.5% -5.3% 4.4% 7.2% 7.5

93

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 (kW)

Minimum cooling 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

94

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 equipment

Comfort level

System

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

95

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 RAC+1air TC RACs on one loop

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

96

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

97

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

98

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

99

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.

100

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.

101

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

0

500

1000

1500

2000

2500

3000

1985 1990 1995 2000 2005 2010 2015 2020 2025

Mm2 cooled

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

0

1

2

3

4

5

6

7

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.

102

Figure 5.3. Cooled area per European in 2020 per Member State and for the EU as a whole

m2/inhabitant

0

2

4

6

8

10

12

14

16

18

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

0

100

200

300

400

500

600

1985 1990 1995 2000 2005 2010 2015 2020 2025

Years

Mm2 cooled

Spain

Italy

France

GermanyGreecePortugal

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)

103

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

104

0

500

1000

1500

2000

2500

Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020

Mm2

educationresidencestradeofficeshotelsbarshospitals

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

0

500

1000

1500

2000

2500,

3000,

Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020

RAC Rooftops PACK & Large SplitVRF Chillers

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.

105

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.

106

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.

107

• 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 Per sq. meter (kWh/m2)

Needs SCL kWh/m2

Electricity to compressor SEC kWh/m2

Electricity total SSEC kWh/m2

SEER

SSEER

Initial Cost

kEuros

ALCC In Euros

(0,10 E/kWh)

ALCC In Euros

(0,06 E/kWh)

ALCC In Euros

(0,17 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

108

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

109

Reference office building unitary cooling consumptions

0,0

20,0

40,0

60,0

80,0

100,0

120,0

140,0

Air Coo

led w

ith w

ater d

istrib

ution

Air Coo

led w

ith ai

r dist

ributi

on

Air Coo

led w

ith ai

r +hu

midity

contr

ol

Wate

r Coo

led +

water d

ist.(c

oolin

g)

Wate

r Coo

led w

ith ai

r dist

.(coo

ling)

Wate

r Coo

led +a

ir +hu

m.(coo

ling)

Outside

wate

r + w

ater d

ist

Outside

wate

r + ai

r dist

Outside

wate

r +air

+hum

TWO LO

OPS + CHILL

ERVRF

PACK&SPlarge

Rtops

RACs on o

ne lo

op

MSSpli

ts

PACKsmall

Single

Ducts

CAC Systems

kWh/

m2 London

MilanSeville

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

Total Cooling Load

0,0

20,0

40,0

60,0

80,0

100,0

120,0

140,0

System

Air Coo

led w

ith w

ater d

istrib

ution

Air Coo

led w

ith ai

r dist

ributi

on

Air Coo

led w

ith ai

r +hu

midity

contr

ol

Water C

ooled

+ wate

r dist

.(coo

ling)

Water C

ooled

with

air d

ist.(c

oolin

g)

Water C

ooled

+air +

hum.(c

oolin

g)

Outside

wate

r + w

ater d

ist

Outside

wate

r + ai

r dist

Outside

wate

r +air

+hum

TWO LO

OPS + CHILL

ERVRF

PACK&SPlarge

Roof to

ps

RACs on o

ne lo

op

Multi S

plits

Splits

Small pa

ckag

es

kWh/

m2

LONDON MILAN SEVILLE

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.

110

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 %

0%20%40%60%80%

100%

Air Coo

led w

ith w

ater d

istrib

ution

Air Coo

led w

ith ai

r dist

ributi

on

Air Coo

led w

ith ai

r +hu

midity

contr

ol

Wate

r Coo

led +

water d

ist.(c

oolin

g)

Wate

r Coo

led w

ith ai

r dist

.(coo

ling)

Wate

r Coo

led +a

ir +hu

m.(coo

ling)

Outside

wate

r + w

ater d

ist

Outside

wate

r + ai

r dist

Outside

wate

r +air

+hum

TWO LO

OPS + CHILL

ERVRF

PACK&SPlarge

Roof to

ps

RACs on o

ne lo

op

Multi S

plits

Splits

Small pa

ckag

es

Single

Ducts

CAC systems

As

% o

f tot

al k

Wh/

m2

Pump %

Fan %

Compressor%

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 SSEER11 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

111

Comfort System Compressor Fan Pump T Cool SEER SSEER12 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 14 TAC RACs on one loop 7,48 2,56 0,06 10,1 2,08 1,54 11 TAC VRF 3,87 7,86 0 11,73 4,03 1,33 10 TAC TWO LOOPS + CHILLER 4,13 7,86 4,94 16,93 5,04 1,23 4 TAC Water Cooled + water dist.(cooling) 7,81 8,07 2,28 18,16 2,67 1,15 7 TAC Outside water + water dist 6,18 7,86 4,94 18,98 3,37 1,1 1 TAC Air Cooled with water distribution 10,02 7,86 1,35 19,23 2,08 1,08 13 TAC Roof tops 7,83 22,27 0 30,09 1,99 1,99 5 TAC Water Cooled with air dist.(cooling) 7,49 22,55 1,73 31,78 2,78 0,66 8 TAC Outside water + air dist 5,91 22,3 4,39 32,6 3,53 0,64 2 TAC Air Cooled with air distribution 9,67 22,3 0,8 32,77 2,15 0,64 6 TAC Water Cooled +air +hum.(cooling) 9,74 24,81 1,73 36,29 2,14 0,57 9 TAC Outside water +air +hum 7,68 24,53 4,39 36,6 2,71 0,57 3 TAC Air Cooled with air +humidity control 12,57 24,53 0,8 37,9 1,66 0,55

Table 5.5c SEER and SSEER, specific consumption in kWh/m2 in Milano, ranked by order of merit

Comfort System Compressor Fan Pump T Cool SEER SSEER12 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 14 TAC RACs on one loop 22,92 7,21 0,88 31,02 2,38 1,76 18 TAC Single Ducts 30,94 2,09 0,04 33,08 1,77 1,65 4 TAC Water Cooled + water dist.(cooling) 21,79 17,42 5,97 45,18 3,37 1,63 1 TAC Air Cooled with water distribution 27,12 14,35 3,87 45,35 2,71 1,62 7 TAC Outside water + water dist 19,36 14,35 11,98 45,69 3,8 1,61 10 TAC TWO LOOPS + CHILLER 20,39 14,35 11,98 46,73 3,61 1,57 13 TAC Roof tops 29,61 36,98 0 66,59 1,85 1,85 5 TAC Water Cooled with air dist.(cooling) 24,84 40,05 4,74 69,64 2,96 1,06 8 TAC Outside water + air dist 21,93 36,98 10,75 69,66 3,35 1,06 2 TAC Air Cooled with air distribution 30,86 36,98 2,65 70,49 2,38 1,04 9 TAC Outside water +air +hum 28,51 40,68 10,75 79,94 2,58 0,92 6 TAC Water Cooled +air +hum.(cooling) 32,29 44,06 4,74 81,09 2,28 0,91 3 TAC Air Cooled with air +humidity control 40,12 40,68 2,65 83,45 1,83 0,88 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.

112

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

or fans pumps Boiler fans pumps Cooling Heating Cooling & Heating

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.

113

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

0

50 000

100 000

150 000

200 000

250 000

300 000

1990 1995 2000 2005 2010 2015 2020

GW

h

RACPACKFCUCAVtotal conventionnal heating

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)

114

Total consumption by country - BAU

0

10 000

20 000

30 000

40 000

50 000

60 000

AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK

GW

h

CO 2000CO 2020CH 2000CH 2020RH 2000RH 2020

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) Electricity and Gas

1990 1995 2000 2005 2010 2015 2020

Cooling function (Electricity only)

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

Heating function Without REV.

51,598 74,442 111,084 164,517 203,330 236,765 250,844

Heating function With present REV. (El.)

7,374 11,495 18,894 28,913 35,875 42,333 45,040

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

115

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

0

5 000

10 000

15 000

20 000

25 000

30 000

35 000

40 000

45 000

AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK

GWh

TRAOFFHOUHOSEDUCAH

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

116

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).

117

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.

118

CO2 emissions for cooling function in Europe

0

5 000

10 000

15 000

20 000

25 000

30 000

35 000

40 000

45 000

2000 2005 2010 2015 2020

kt C

O2

UKSWSPPONELUITIRGRGEFRFIDEBEAU

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

H1 TAC Air Cooled with water distribution 25,37 5,99 1,97 10,15 33,33 18,11 98,53 83,312 TAC Air Cooled with air distribution 6,06 5,76 0,23 2,43 12,05 8,42 98,03 94,393 TAC Air Cooled with air +humidity control 7,28 6,34 0,23 2,91 13,84 9,48 115,0

4110,6

74 TAC Water Cooled + water dist.(cooling) 25,37 5,99 1,97 10,15 33,33 18,11 96,19 80,975 TAC Water Cooled with air dist.(cooling) 6,06 5,76 0,23 2,43 12,05 8,42 96,87 93,236 TAC Water Cooled +air +hum.(cooling) 7,28 6,34 0,23 2,91 13,84 9,48 111,5

0107,1

37 TAC Outside water + water dist 25,37 5,99 1,97 10,15 33,33 18,48 93,58 78,73

119

8 TAC Outside water + air dist 6,06 5,76 0,23 2,43 12,05 8,79 96,86 93,599 TAC Outside water +air +hum 7,28 6,34 0,23 2,91 13,84 9,85 110,3

1106,3

110 TAC TWO LOOPS + CHILLER 5,82 5,99 0,23 1,16 12,04 9,50 71,24 68,7011 TAC VRF 26,93 5,99 0,23 6,72 33,15 12,71 88,14 67,6912 TAC PACK&SPlarge 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,3013 TAC Roof tops 5,91 5,75 0,23 3,03 11,90 8,78 106,2

2103,1

114 TAC RACs on one loop 26,86 4,64 6,85 10,87 38,35 22,36 98,66 82,6715 TAC Multi Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,3016 TAC Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,3017 TAC Small packages 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,3018 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+RH1 TAC Air Cooled with water distribution 96,23 20,38 4,84 38,49 121,4 63,71 136,5 78,802 TAC Air Cooled with air distribution 54,53 39,57 1,71 21,81 95,81 63,09 124,4 91,763 TAC Air Cooled with air +humidity control 65,43 43,53 1,71 26,17 110,6 71,41 143,2 103,94 TAC Water Cooled + water dist.(cooling) 96,23 20,38 4,84 38,49 121,4 63,71 135,9 78,235 TAC Water Cooled with air dist.(cooling) 54,53 39,57 1,71 21,81 95,81 63,09 124,0 91,326 TAC Water Cooled +air +hum.(cooling) 65,43 43,53 1,71 26,17 110,6 71,41 142,3 103,07 TAC Outside water + water dist 96,23 20,38 4,84 38,49 121,4 69,16 137,1 84,848 TAC Outside water + air dist 54,53 39,57 1,71 21,81 95,81 68,54 125,1 97,919 TAC Outside water +air +hum 65,43 43,53 1,71 26,17 110,6 76,86 143,2 109,410 TAC TWO LOOPS + CHILLER 61,51 20,38 1,71 14,45 83,60 45,12 98,10 59,6211 TAC VRF 108,3 20,38 1,71 39,35 130,4 59,73 142,1 71,4612 TAC PACK&SPlarge 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,4813 TAC Roof tops 54,93 39,60 1,71 38,78 96,24 78,38 126,3 108,414 TAC RACs on one loop 106,4 9,44 7,00 47,17 122,8 63,60 132,9 73,7115 TAC Multi Splits 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,4816 TAC Splits 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,4817 TAC Small packages 108,3 7,93 1,71 64,16 117,9 72,09 126,3 80,4818 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+RH1 TAC Air Cooled with water distribution 88,83 14,94 4,48 35,53 108,2 54,95 144,4 91,172 TAC Air Cooled with air distribution 54,77 28,30 1,64 21,91 84,72 51,85 144,8 112,03 TAC Air Cooled with air +humidity control 65,73 31,13 1,64 26,29 98,50 59,07 168,5 129,04 TAC Water Cooled + water dist.(cooling) 88,83 14,94 4,48 35,53 108,2 54,95 144,6 91,315 TAC Water Cooled with air dist.(cooling) 54,77 28,30 1,64 21,91 84,72 51,85 144,3 111,56 TAC Water Cooled +air +hum.(cooling) 65,73 31,13 1,64 26,29 98,50 59,07 166,6 127,27 TAC Outside water + water dist 88,83 14,94 4,48 35,53 108,2 58,54 145,1 95,398 TAC Outside water + air dist 54,77 28,30 1,64 21,91 84,72 55,44 144,5 115,39 TAC Outside water +air +hum 65,73 31,13 1,64 26,29 98,50 62,65 166,1 130,210 TAC TWO LOOPS + CHILLER 62,11 14,94 1,64 14,67 78,69 37,68 116,2 75,2211 TAC VRF 99,77 14,94 1,64 42,20 116,3 57,14 146,6 87,3912 TAC PACK&SPlarge 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,513 TAC Roof tops 55,81 28,30 1,64 48,84 85,75 77,14 152,3 143,714 TAC RACs on one loop 117,97 7,84 8,16 42,93 133,9 58,92 164,9 89,9415 TAC Multi Splits 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,516 TAC Splits 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,517 TAC Small packages 99,77 6,26 1,64 66,79 107,6 73,05 136,1 101,518 TAC Single Ducts 99,77 6,26 1,64 75,23 107,6 81,48 140,7 114,5

120

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 system 6.6 22.7 20.5

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 tops...) non rev 3.4 25.7 21.4

Large packages (Roof tops...) reversible 1.3-2.7 3.5 28.8 30.8

Large splits with primary air Non reversible 6.6 22.7 20.4

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.

121

122

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. Figure 6.1. Chiller cost versus cooling capacity, as a function of the refrigerant

R2 = 0.7823

R2 = 0.7719

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

0 20 40 60 80 100 120 140 160Cooling capacity kW

Prixce euro

R22R407CRegression(R22)Regression (R407C)

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

y = 230.98x + 5130.2 R2 = 0.7835

y = 122.46x + 4944.3 R 2 = 0.9228

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

0 20 40 60 80 100 120 140 160 180Cooling capacity kW

Price euro

Air cooled water cooled Regression(Air

l d)Regression (water l d)

123

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 onlyCoolingCap. kW

Price(euro)

EER Price(euro)

EER Price differencein %

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

124

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).

125

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

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

Optimisation of Cost/kW final

80

90

100

110

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

EER

Euro

/kW

3

4

5

6

7

8

9

10

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

SEER

Tota

l cos

t (Eu

ros/

m2)

ALCC17-800hALCC10-800hALCC17-400hALCC6-800hALCC10-400hALCC6-400h

126

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 with the CAV system TAC

“poorest” WC chiller with CT

“best” WC chiller with CT

“poorest” WC chiller with natural water

“best” WC chiller with natural water

“poorest” AC chiller

“best” AC 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 with the CAV system TAC

“poorest” WC chiller with CT

“best” WC chiller with CT

“poorest” WC chiller with natural water

“best” WC chiller with natural water

“poorest” AC chiller

“best” AC 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!

127

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)

128

0%

5%

10%

15%

20%

25%

30%

35%

40%

0 50 100 150 200 250Cooling capacity (kW)

Sha

re o

f mod

els EU 1998

US 2003

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)

129

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)

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

-$500

$0

$500

$1,000

$1,500

$2,000

8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5

EER (kW/ton)

EquipmentCostDelta($)

-$250

$0

$250

$500

$750

$1,000

8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5

EER (kW/ton)

EquipmentCostDelta($)

130

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)

€10

€11

€12

€13

€14

€15

€16

€17

2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6EER (W/W)

ALL

C (€

/m2/

yr

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)

131

€11

€12

€13

€14

€15

€16

€17

2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6EER (W/W)

ALL

C (€

/m2/

yr

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.

132

Table 6.3. Minimum efficiency levels of fans recommended by Eurovent

[m3/s] [m3/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%

MINIMUM TOTAL EFFICIENCY* FAN/MOTOR COMBINATION [%]

Air flow rate Available static fan pressure [Pa] **

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.

133

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* - max. pressure drop

65%

200 Pa

70%

150 Pa

75%

125 Pa Plate heat exchanger without bypass - min. dry efficiency* - max. pressure drop

45%

250 Pa

50%

200 Pa

55%

150 Pa Plate heat exchanger with bypass - min. dry efficiency* - max. pressure drop

40%

300 Pa

45%

250 Pa

50%

200 Pa Heat pipe - min. dry efficiency* - max. pressure drop

45%

300 Pa

50%

250 Pa

55%

200 Pa Run around coil loop - min. dry efficiency* - max. pressure drop

40%

300 Pa

45%

250 Pa

50%

200 Pa * Dry efficiency based on a mass ratio of 1

Running the Air Handling Unit at a speed which is needed for the actual demand will also save energy. 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.

134

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

TAC

Hypothetical

SSEER

Multi Split systems 2,25 248000 128000 20,75

Packaged systems (under windows)

2,25 188000 128000 16,77

CAC - Central Air Conditioners

TAC

Large packages (Roof tops...)

2,25 130000 128000 12,91

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

135

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 Air

TC

Hypothetical

SSEER

Multi Split systems 2,25 220000 128000 16,69

Packaged systems (under windows)

2,25 160000 128000 13,30

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

136

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

137

CAV and VAV fan consumption for SEVILLE

0

1

2

3

4

5

kWh/

m²

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

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

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.

138

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.

139

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

140

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%

141

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

SPECIFIC CONSUMPTION

ALCC Euros/m2/year

kWh/m2/YEAR

airwaterpackagesrac

The result of optimisation

Table 7.2 OPTIMISATION OF AIR BASED IN SEVILLE

For

cooling and

For secondary

fans

Total electricity

Initial Cost ALCC of

system

ALCC of

system

ALCC of

system

142

pumps (MWh)

(MWh) (0,10 E/kWh)

(0,06 E/kWh)

(0,17 E/kWh)

0 – CAV 121,67 37,76 159,43 1 053 640 30,89 29,84 32,73 1 – VAV-50%/ +6 E/m2

121,67 18,88 140,55 1 090 090 31,56 30,63 33,17

2 – FC –20% /+2 Euros/m2

97,34 37,76 135,10 1 074 710 31,05 30,16 32,61

3 – Fans –8% ;–15 %; +2 E/kW

111,94 32,10 144,03 1 054 140 30,65 29,70 32,31

4 – HR 60% -200Pa; -3% +6 E/m2

118,02 37,76 155,78 1 090 090 31,81 30,78 33,60

5 –Lower HL in AHU –7%/+12E/m2

121,67 35,12 156,79 1 126 540 32,80 31,77 34,61

6 –Optimised chiller – see chapter 6

73,00 37,76 110,76 1 062 390 30,32 29,59 31,60

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

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.

25

27

29

31

33

35

37

39

0,00% 20,00% 40,00% 60,00% 80,00% 100,00% 120,00%% of reference

Euro

s/m

2 ALCC17ALCC10ALCC6

143

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

144

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].

145

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. Part load (%)

Inlet water temperature °C

0 12 11 10 9 8 7 6

40

80

Dead-band 1°C

Dead-band 1°C

20

60

100

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

146

enable to determine the electric power absorbed EP and thus the efficiency for each hour. Yearly efficiency is calculated with Equation (1).

∑

∑

=

== 8760

1

8760

1

ii

ii

EP

CCSEER

(1)

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.

The corresponding curve is represented on the figure 8.3, it corresponds to the following formula :

cyccycFL

FLFL C1.CCC

CCCC

CC

EEREER

−+= (2)

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

Degradation of the reduced efficiency versus the part load ratio (same sources temperatures )

00.10.20.30.40.50.60.70.80.9

1

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 )

Rat

io o

f par

t loa

d ef

ficie

ncy

to

full

load

effi

cien

cy

147

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)

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

Reduced performances while varying flow rate

0.7

0.75

0.8

0.85

0.9

0.95

1

1.05

1.1

1.15

1.2

0.5 0.6 0.7 0.8 0.9 1% of nominal flow rate

EER/EERnom

EP/EPnom

CC/CCnom

148

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.

Reduced EER versus OAT for different load ratios

1,00

1,10

1,20

1,30

1,40

1,50

1,60

1,70

15 17 19 21 23 25 27 29 31 33 35

OAT (°C)

DOE2 curve

FL curve

PL curve, singlecircuit 50 %

PL curve, singlecircuit 75%

PL curve, singlecircuit 25 %

PL curve, singlecircuit VSDF

Without highpressure control

149

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.

150

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

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.

10

15

20

25

30

35

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Part load ratio (%)

Outside air temperature (°C)

ARI aircondensationcurve75

50

25

151

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:

152

2%50%100

%75EEREER

EER+

=

The EER value at 25% of the load can be calculated as follows:

)( %75%50%50%25 EEREEREEREER −+=

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

153

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 Compressor 1 0 0 0 1 Compressor 2 0 1 1 1

Circuit 2 Compressor 3 0 0 1 1 Compressor 4 1 1 1 1

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

154

The overall performance improvement (or degradation) at part load (temperature effects being substracted) is 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

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 Compressor 1 0 0 0 1 Compressor 2 0 1 1 1

Circuit 2 Compressor 3 0 0 1 1 Compressor 4 1 1 1 1

EP/EPFL 23% 46% 71% 100%CC/CCFL 25% 52% 75% 100%

EER/EERFL 109% 114% 105% 100% Let’s note here that the presentation under this format ensures that the manufacturer has consciously chosen this staging as optimum and hopefully that it has been prioritised as factory default control parameters. 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%

Reduced efficiency of the part load stages for water cooled chillers

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

0 0.2 0.4 0.6 0.8 1 1.2Part load ratio

N° 4N° 1N° 3N° 6

155

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 The overall performance improvement (or degradation) at part load (temperature effects being substracted) is 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

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.

Reduced efficiency of the part load stages for air cooled chillers

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

0 0.2 0.4 0.6 0.8 1 1.2

Part load ratio

N° 5N° 7N° 8N° 9N° 2

156

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

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

Office building Climate

System type London Milan Seville CAV

CAV-FC VAV

FCU 4P

157

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

Hourly load curve

050

100150200250300350400450500550

10 15 20 25 30 35OAT (°C)

kWh CAV-MILAN

158

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.

STEP 2 of the load curves reduction process 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

159

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

STEP 3 of the load curves reduction process 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

160

So now, the weighting coefficients are known :

SEERP= [0.2024 0.4272 0.3369 0.0335]

STEP 4 of the load curves reduction process The average temperature for each stage is calculated and the following operating temperatures are found :

SEERT=[17.9997 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

Part load (%) Reference (nominal full load) Inlet air temperature (°C)

Energy weighting coefficients A, B, C, D

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.

161

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

Part load (%) Reference (nominal full load) Inlet air temperature (°C)

Energy weighting coefficients A, B, C, D Inlet air temperature (°C)

Energy weighting coefficients A, B, C, D

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

No cycling, no fan cycling

HSEER 3.01 3.38 2.73 3.09 3.13 4.08 4.59 3.30 4.21 3.19 ESEER 3.05 3.44 2.73 3.14 3.15 4.08 4.60 3.30 4.22 3.31 Relative deviation 1.2% 1.8% 0.0% 1.6% 0.7% 0.0% 0.2% 0.0% 0.2% 3.9%

Cycling only

HSEER 2.93 3.34 2.69 3.06 3.11 3.70 4.43 3.08 4.11 3.14 ESEER 3.00 3.43 2.72 3.14 3.15 3.88 4.58 3.20 4.22 3.31 Relative deviation 2.4% 2.6% 1.1% 2.4% 1.3% 4.7% 3.3% 3.7% 2.7% 5.5%

Fan cycling only

HSEER 2.99 3.35 2.71 3.05 3.00 3.80 4.26 3.05 3.91 2.96 ESEER 3.03 3.40 2.73 3.11 3.10 3.72 4.31 3.01 3.90 3.07 Relative deviation 1.4% 1.4% 0.8% 1.9% 3.4% -1.9% 1.4% -1.4% -0.4% 3.5%

Cycling and fan cycling

HSEER 2.90 3.31 2.66 3.03 2.98 3.46 4.12 2.86 3.82 2.92 ESEER 2.98 3.39 2.71 3.11 3.10 3.56 4.31 2.93 3.90 3.07 Relative deviation 2.7% 2.3% 2.0% 2.6% 4.0% 2.9% 4.5% 2.4% 2.0% 5.1%

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.

162

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

WC 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%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

WC

CAV 5.3% 16.1% 18.4%

FCU 8.2% 22.2% 29.8%

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.

163

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

Part load ratio Temperatures Weighting coefficients Temperatures Weighting

coefficients Temperatures Weighting coefficients

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).

164

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

Part load ratio Condenser inlet temperatures (°C)

Weighting coefficients

Condenser inlet temperatures (°C)

Weighting coefficients

Condenser inlet temperatures (°C)

Weighting 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.

165

Table 8.21. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers

ESEER ARI EMPE

Part load ratio Temperatures Weighting coefficients Temperatures Weighting

coefficients Temperatures Weighting 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 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) Weighting coefficients Temperatures Weighting

coefficients Temperatures Weighting coefficients

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

166

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

Figure 8.17. comparison of EMPE with HSEER for the tested chillers

HSEER versus EER

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5

EER

HSE

ER

HSEER versus IPLV

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4 4,5

IPLV

HSE

ER

167

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

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.

HSEER versus EMPE

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4

EMPE

HSE

ER

HSEER versus ESEER

0

0,5

1

1,5

2

2,5

3

3,5

4

0 0,5 1 1,5 2 2,5 3 3,5 4

ESEER

HSE

ER

168

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 OAT (°C) % of FL

at 35°C % of FL at 35C % of FL

at 35°C % of FL 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.

169

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 (+)

ST, PID 2 hours (P2’,P3’,P4’) 2 hours

(P2ab,P3ab,P4ab)

FC, PID 1 hour (P3’,P4’) 1 hour

(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.

170

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

OAT (°C)

% (full load

35°C)

% (full load

35°C) % (full load

35°C) % (full load 35°C)

19 C1 C2 CC 29% 43.9 C1 C2 CC 61% 93.0 C1 C2 CC 89% 137.0 C1 C2 CC 120% 185.0

Fan 66% 0% EP 18% 10.9 66% 66% EP 35% 20.7 66% 66% EP 55% 32.8 66% 66% EP 73% 43.8

Comp 50% 0% EER 157% 4.0 50% 25% EER 175% 4.5 50% 100% EER 163% 4.2 100% 100% EER 165% 4.2

25 C1 C2 CC 28% 42.9 C1 C2 CC 61% 93.4 C1 C2 CC 87% 134.0 C1 C2 CC 115% 176.3

Fan 66% 0% EP 20% 11.9 100% 100% EP 37% 22.0 100% 100% EP 60% 35.8 100% 100% EP 82% 48.9

Comp 50% 0% EER 141% 3.6 50% 25% EER 166% 4.3 50% 100% EER 146% 3.7 100% 100% EER 141% 3.6

30 C1 C2 CC 27% 41.6 C1 C2 CC 57% 88.1 C1 C2 CC 82% 126.4 C1 C2 CC 108% 166.3

Fan 100% 0% EP 21% 12.7 100% 100% EP 40% 24.2 100% 100% EP 66% 39.4 100% 100% EP 90% 53.9

Comp 50% 0% EER 128% 3.3 50% 25% EER 142% 3.6 50% 100% EER 125% 3.2 100% 100% EER 120% 3.1

35 C1 C2 CC 25% 38.4 C1 C2 CC 53% 81.5 C1 C2 CC 76% 116.8 C1 C2 CC 100% 153.7

Fan 50% 0% EP 24% 14.2 100% 100% EP 45% 26.9 100% 100% EP 73% 43.8 100% 100% EP 100% 60.0

Comp 25% 0% EER 106% 2.7 50% 25% EER 118% 3.0 50% 100% EER 104% 2.7 100% 100% EER 100% 2.6

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

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

171

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

172

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.

173

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.

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 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)

174

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 system

Compressor demand 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.

175

Figure 9.1 Demand evolution, depending on scenarios

0,00

20,00

40,00

60,00

80,00

100,00

120,00

140,00

2000 2005 2010 2015 2020

TWh

for A

C

BAUScenario 1Scenario 4Scenario 5Scenario 2Scenario 6Scenario 3

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

0

500

1000

1500

2000

2500

3000

3500

4000

CAHORE Education Hospitals Households Offices Trade Total

GW

h sa

ved

in A

C

Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON FULL LOAD INFO The policy measure associated is banning many classes of equipment or a negotiated agreement on average full load performance like ACEA agreement for cars. 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

176

Scenario 2 in 2020

0

2000

4000

6000

8000

10000

12000

14000

16000

CAHORE Education Hospitals Households Offices Trade Total

GW

h

Série1

Scenario 3 BAT- THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO Part load is taken into account in Eurovent grading and so the corresponding improvement is obtained. 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 systems in SC3

4,32 6,93 9,86 10,45

Others in SC3 10,21 12,53 14,77 15,70

Total savings in SC3

14,53 19,46 24,64 26,15

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 Cooling

2,00 3,75 5,68 6,15

Scenario 5 Variable Air Flow Obligation of variable air flow in air distribution systems translated in national values for savings in table 9.3

177

Table 9.3

Savings (TWh) 2005 2010 2015 2020

Savings due to VAV

3,89 7,85 12,28 13,45

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 scenario 6

4,74 9,62 15,07 16,45

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.

178

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

179

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.

180

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

181

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,

Floor heating and cooling

Water Cooled Water Cooled, Floor heating and cooling

Remote condenser

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,

Floor

Water Cooled Water Cooled, 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

182

Figure 10.1. Air cooled chillers, cooling-mode, < 750kW

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)

R407CR134aR22HFC

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%

183

Figure 10.2. Water cooled chillers, cooling-mode, < 750kW

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

A (>5,05) B (>4,65) C (>4,25) D (>3,85) E (>3,45) F (>3,05) G (<3,05)

R407CR134aR22HFC

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.

184

Figure 10.3. Sales and Market statistics

Sale statistics R407C

0

500

1000

1500

20001,6

1,71,8

1,9

2

2,1

2,2

2,32,4

2,52,6

2,7

2,8

2,9

3

3,1

3,23,3

Eurovent catalog values R407C

050

100150200250

1,61,7

1,8

1,9

2

2,1

2,2

2,32,4

2,52,6

2,7

2,8

2,9

3

3,1

3,23,3

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

0100200300400500

2,72,9

3,1

3,3

3,5

3,7

3,94,14,3

4,5

4,7

4,9

5,1

5,3

5,5

Eurovent catalog values R407C

0

5

10

15

202,7

2,9

3,1

3,3

3,5

3,7

3,94,14,3

4,5

4,7

4,9

5,1

5,3

5,5

Sales and Eurovent catalogue statistics for water condensing units are convergent: sales statistics show a better efficiency..

185

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

Energy efficiency class Energy efficiency ratio

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

Energy efficiency class Energy efficiency ratio

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

Energy efficiency class Energy efficiency ratio

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

186

(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

Energy efficiency class Coefficient of performance

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

Energy efficiency class Coefficient of performance

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

Energy efficiency class Energy efficiency ratio

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

187

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)

2%

5%

7%

15%

22%

26%

11%

0%

5%

10%

15%

20%

25%

30%

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%

188

Figure 10.6. Air-cooled air conditioners - Heating function -mixed statistics

% of market(density)

0

0,05

0,1

0,15

0,2

0,25

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 range

(kW)

EER at least life cycle cost

(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

189

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, 4tons (cooling-mode)

3.51 Values about 14.1

Water-cooled screw chiller, 125tons

4.45 COP (4.50 US IPLV) Values about 440

Water-cooled centrifugal chiller, 300tons

6.10 COP (6.10 US IPLV) Values about 1056

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)

190

• 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

191

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.

192

Specific recommendations The European Commission and/or a coalition of willing Member States should consider:

• the development of an EU model building code that addresses air conditioning amongst other energy end-uses. (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:

193

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:

194

• 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.

195

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 The European Commission and/or a coalition of willing Member States should consider:

• Making efforts to define best practice in operation and maintenance

• Making efforts to define best practice in operation and maintenance performance contracting

With an aim of informing national building thermal regulations and the implementation of the Energy Performance in Buildings Directive.

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

196

• 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.

197

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

198

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)

CPR Carbon Performance Rating (UK regulation)

DD Degree Days EDF Electricité de France

EER Energy-Efficiency Ratio (W/W)

EMPE Italian Method for part load rating

EMS Energy Management Systems

EPB Energy Performance in Buildings (EU regulation)

CPR Carbon Performance Rating (UK regulation)

ESCO Energy Service Companies

EU European Union

Eurovent European association of refrigeration, air-conditioning and ventilation equipment manufacturers

FCU Fan Coil Unit

GB Great Britain (excludes Northern Ireland)

GEA Group for Efficient Appliances

CPR Carbon Performance Rating (UK regulation)

HFC Hydrofluorocarbons

IDAE Spanish energy-conservation agency

IEC International Electrotechnical Committee

IPLV Integrated Part-Load Value

ISO International Standards Organisation

LBL-MIM Lawrence Berkeley Laboratory – Manufacturer Impacts Model

LCC Life-Cycle Cost

LLCC Least Life-Cycle Cost

MEPS Minimum Energy Performance Standards (same as MEES, Minimum Energy Efficiency Standards)

NPV Net Present Value

ODP Ozone-Depletion Potential

RAC Room Air Conditioners (in the wide sense)

R&D Research and Development

199

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.

200

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