Escolar Documentos
Profissional Documentos
Cultura Documentos
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/274048785
CITATIONS READS
6 452
24 authors, including:
Some of the authors of this publication are also working on these related projects:
The inspection of building services through continuous monitoring and benchmarking –The
iSERVcmb project View project
All content following this page was uploaded by Philippe Rivière on 25 March 2015.
Contract DGXVII-4.1031/P/00-009
CO-ORDINATOR: Jérôme ADNOT, ARMINES, France
Assisted by Paul WAIDE
PW Consulting, UK
PARTICIPANTS
Jérôme ADNOT, Philippe RIVIERE, Dominique MARCHIO,
Martin HOLMSTROM, Johan NAESLUND, Julie SABA
Centre d’Energétique, Ecole des Mines de Paris, France
Sule BECIRSPAHIC
Eurovent Certification
Carlos LOPES
ADENE-CCE, Portugal
Isabel BLANCO
IDAE, Spain
Luis PEREZ-LOMBARD, Jose ORTIZ
AICIA, Spain
Nantia PAPAKONSTANTINOU, Paris DOUKAS
University of Athens, Greece
Cesare M. JOPPOLO
Politecnico di Milano, Italy
Carmine CASALE
AICARR, Italy
Georg BENKE
EVA, Austria
Dominique GIRAUD
INESTENE, France
Nicolas HOUDANT
Energie Demain, France
Philippe RIVIERE, Frank COLOMINES
Electricité de France
Robert GAVRILIUC, Razvan POPESCU, Sorin BURCHIU
UTCB, Bucharest
Bruno GEORGES
ITF, France
Roger HITCHIN
BRE, UK
With the additional participation of experts from Eurovent Cecomaf
© 2003 ARMINES
ARMINES
60, bd St Michel
75272 Paris Cedex 06
France
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
1. INTRODUCTION ....................................................................................... 36
Selection of technical experts ...................................................................................................... 36
Participation of energy agencies, utilities, manufacturers and national experts ............... 37
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
5.1 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
6.3 Energy Efficiency of Air Handling Units seen as tradable goods .......................................... 131
Fans integrated in AHU .................................................................................................................... 131
Heat recovery section of AHU ......................................................................................................... 132
7.2 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
7.4 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
8.2 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
8.5 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
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
SUMMARY OF RESULTS
Air-conditioning constitutes a rapidly growing electrical end-use in the European Union (EU), yet the
possibilities for improving its energy efficiency have not been fully investigated. As opposed to room air-
conditioners (RAC) central air conditioning (CAC) systems, which are defined as air conditioning systems
with more than 12kW of cooling capacity in the EU, are not bought or selected in a shop. They may be
selected by an installer of packaged units. They are usually designed by an AC engineer and the components
selected following the engineers’ recommendations.
The definition of CAC applied in the EU does not correspond to the definition used in the USA where a
“Central Air Conditioner” is a ducted package AC system, which are relatively infrequent in Europe and
sized often under 12kW, a piece of equipment that we would call a RAC in Europe. European CAC systems
are commercial AC systems usually specified by engineers or technicians, who choose the system
technology without any direct influence from the customer, except for the specification of the desired
11
environment and other conditions such as maximum overall price, etc.. There are a large variety of systems
and technical options (regarding system structure and control) in use while there is also a large variation in
comfort conditions (not only in terms of the set-point, like space heating, but also in the nature of comfort)
obtained.
Within the EECAC study twelve participants from eight countries including the EU manufacturers'
association, Eurovent, engaged in identifying the most suitable measures to improve the energy efficiency of
commercial chillers and AC systems. This study benefited from the co-operation between laboratories,
consultants and Eurovent, which was established during the conduct of the SAVE sponsored EERAC study
concerned with room air conditioners (EERAC 1999).
It was made easier by the existing information scheme by a subsidiary of the manufacturers’ association
called Eurovent-Certification. However the existing information scheme in Europe is based on testing at
nominal operating conditions and lags behind the information available in some foreign countries for the
same type of equipment (such as the ARI certification programme in the USA). To be really effective,
energy efficiency options have to be defined not on the basis of nominal operating conditions but at a variety
of part load conditions, which better reflects the CAC operating modes that occur in real use. The energy
efficiency options list has also to cover secondary systems (distribution) which were found of equal
importance for reaching the minimum cost of service.
Definitions of all CAC systems found on the EU market have been given.
The structure of a CAC system and consequently its name results from the accumulation of a number of
decisions on the choice of essential components. The first choice determining a system is the type of the fluid
being centrally refrigerated and circulated. The most frequent (and really dominant option) is the use of a
chiller, which generates cold water (typically at 7°C) and which is used to transfer "cold" to the building
space in part via a water distribution network and in part via a centrally treated airflow. To transfer the
“cold” to air, Air Handling Units (AHU) are used. In the majority of situations however, chilled water is
circulated up to the rooms and the air of the room comes directly in contact with it. Even in this predominant
CAC system a choice must be made on how to transfer “cold” to the air of the room. The two most common
approaches are an “induction system” and a “Fan-Coil Unit” (FCU) and these both of these systems can
operate with a water distribution network having two or three or four pipes.
Other AC systems are applicable to a series of rooms or spaces such that their applicability is dependent on
the number of rooms and the general situation of the building. In many cases large Unitary Air Conditioners
(or Packaged), which are self-contained direct-expansion (i.e. without using water as an intermediate heat
transfer vector) apparatus can be applied as can Multi-Split systems, which are a particular assembly of small
“split systems” and were originally investigated under the EERAC study. In addition a new variant of the
“split system” concept most commonly known as the VRF (Variable Refrigerant Flow) system, but more
generally as a modulated capacity system, is capable of significant energy savings and has occupied a market
segment. These system descriptions lead to the idea of a CAC system description tree of which a non-
exhaustive set of branches is presented in Figure 1.
12
Figure 1. CAC system description tree showing the most common CAC systems.
LOCAL OR CENTRAL
LOCAL CENTRAL
MULTI- VRF
SPLIT
‘Fluid’ refers to the primary heat transfer fluid from the building to the refrigeration system
All CAC equipment test standards have been reviewed and studied to assess their
suitability to represent energy efficiency under real operating conditions.
Eurovent-Certification, a branch of Eurovent has defined test conditions at which equipment energy
performance is to be reported by European industry, based upon performance testing at full load, in
accordance with CEN standards. The American Refrigeration Institute (ARI) and American Society of
Heating, Refrigerating and Air Conditioning Engineers (ASHRAE) have defined US-national test standards
in a similar way; except these involve a mixture of testing at full and part-load conditions so that the results
can be extrapolated to provide the average annual performance of CAC equipment. This has enabled US
legislative bodies to readily establish minimum performance criteria which are based upon comparative
performance under representative operating conditions. Thus in some way, the US standards have progressed
futher than the European ones, although it has been established in this study that they are not suitable for use
in European conditions without modification. There are ISO efforts for the testing aspect but the respective
standards are not all available.
CAC market and stock data have been assembled for the first time.
National surveys of the CAC market, usage and regulatory environment were conducted by the EECCAC
study participants for their country. This took advantage of each participants national contacts including
assembling and synthesising rough data supplied by local manufacturers’ or importers’ associations or even
involved subcontracting national consultants. It received a very significant help from the manufacturers
13
associations. The resulting set of country reports for: Austria, France, Germany, Greece, Italy, Portugal,
Spain, United Kingdom (with special thanks to the BRE) provides a unique set of data at the national level.
The CAC market is expanding rapidly in Europe, as can be seen from the additional cooled building floor-
area installed from 1980 to 2000 for the EU-15, Figure 2 (including new systems and refurbishment).
Figure 2. Annual addition of building cooled-floor area by CAC in the EU (either really added or replaced)
100,00
80,00
M
m2 60,00
40,00
20,00
0,00
1975 1980 1985 1990 1995 2000 2005
Because of a strongly differential growth rate across EU Member States, the relative share of the total EU
cooled floor-area of countries such as France or Germany, which was large in the 1980’s has become small
in the 1990’s. The high growth in CAC installed in Italy and Spain means that these countries now account
for more than 50% of the EU market, as is apparent from the CAC floor-area installation figures for 1998 by
country shown in Figure 3.
Figure 3. National shares of installed CAC-cooled floor area in EU buildings in 1998
Others Germany
13% 11%
Greece
5%
UK
8%
Spain
24%
Italy
25%
Portugal France
2% 12%
The market shares for all competing AC systems, have been determined in all usage sectors and for all years
between 1990 and 2020, see Figure 4 for example based on year 1998.
14
Figure 4. AC market share by AC type expressed in terms of newly installed cooled-area in EU buildings in
1998
Splits >12kW
7%
RAC< 12 kW
36%
chillers
45%
VRF
2%
Rooftops
5%
Packages
5%
Similar data to the European data is available for the world’s largest market, the USA, from the CBECS
programme of the US Department of Energy’s Energy Information Administration. The figures cover the
same years (1999-2000) and the same type of building stock (non residential buildings in use); however, the
choice of AC equipment is very different. Packaged AC accounts for the majority of cooled floor area in the
USA while chiller based CAC systems dominate in Europe, Figure 5.
Figure 5a. The share of cooled-floor area by AC type in non-residential buildings in the USA for 1999-2000
US A (E IA )
chillers
pack ages
all RA C
Figure 5b. The share of cooled-floor area by AC type in non-residential buildings in the EU for 1999-2000
15
E UR (E E CCA C)
c hillers
pack ages
all RA C
Despite this difference in technical preferences, the US market is so large in absolute terms that for every
CAC type there are more square metres of cooled floor space in the USA than in the EU.
Figure 6. Chiller energy efficiency (EER) at full load as a function of cooling capacity for chillers available on
the EU market in 1999. There are two groups of chillers, with distinct testing conditions (water cooled and air
cooled, that cannot be compared)
4.5
4.0
2
R = 0.0073
3.5
3.0
2.5
2
R = 0.0003
EER
2.0
1.5
1.0
air cooled
water cooled
0.5
Regression (air cooled)
Regression (water cooled)
0.0
0 100 200 300 400 500 600 700 800 900 1000
Capacity kW
The average EER is indeed 3.57 with water whereas it is 2.52 for the systems with air as a rejection
medium under the conditions of the testing Standard, see Table 1. One could conclude in favour of
a clear superiority of Water cooled systems over Air cooled systems. Nevertheless, water-cooled
systems are expensive (for both when a cooling tower or natural ground water is used), and will
1
Energy efficiency ratio, which is measured under full load conditions
16
therefore only be common among large capacity systems. The operating conditions on the field may
be very different from the testing conditions and reverse premature conclusions.
17
Table 1. The range of chiller energy efficiency (EER) for different types of chiller systems found on the EU
market
EER
Categories Type Condenser Application min ave max
Complete unit Cooling only air conditoining 1.9 2.53 3.29
reversible air conditioning 1.9 2.48 2.96
Floor 3.31 3.34 3.39
Cooling only water conditioning 2.9 3.73 4.09
reversible water conditioning 2.9 3.57 4.09
Condenserless Cooling only water conditioning 2.76 3.21 3.69
Interestingly, the reversible systems, which have an average EER of 2.48 W/W, have an almost identical
energy efficiency to the cooling-only systems which have an average EER of 2.53 W/W.
A number of countries outside the EU have implemented market transformation policy measures to raise the
energy efficiency of CAC systems installed in their markets, including: the USA, Canada, Mexico, Korea,
Japan, Australia and New Zealand. Most policy measures have been aimed at packaged units, which are not
so important on the European market; however, the US IPLV approach, which has been found to not be
directly applicable in the EU, was -before this study and the AICARR’s EMPE proposal, the only attempt to
address the specific issues of chiller energy performance.
18
All the elements of a possible grading of Cooling market have been assembled
Tables have been produced on the basis of the technical findings and of the market statistics that allow to
define a grading scale for each segment of the market, allowing a fair comparison of equipment despite of
testing conditions and technical differences. Statistics show for each segment (like figure 7 for the largest
segment) which part of the market falls in each grade.
Figure 7 Air cooled chillers, Cooling, below 750kW , statistics with proposed grading and for each refrigerant
40,0%
35,0%
30,0%
25,0%
R407C
R134a
20,0%
R22
HFC
15,0%
10,0%
5,0%
0,0%
A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)
Splits and Packages are grouped in one single category (Air-cooled air conditioners) for presentation of
results in table 2.
Table 2 Example of proposed grading in Cooling function -mixed statistics (splits and packages mixed) ;
present market average = 2.46
19
Figure 8. Average cooled-floor area per inhabitant for EU countries and the EU as a whole in 2000.
m2
/in
4
ha
bit
ant
0
B DK D GR E F IRL I L NL A P FIN S UK EU-15
The evolution of the various economic sectors and their demand for comfort vary a lot. Only trade and
offices really grow in relative terms and they may reach 70% of stock by 2020.
Table 3 Area conditioned in each country and year (such areas can be compared with national statistics)
Years
Country 1990 1995 2000 2005 2010 2015 2020
AU Mm2 cooling 12,01 15,68 20,06 26,29 30,29 33,01 33,95
Mm² reverse 1,45 2,06 2,74 4,83 5,57 6,08 6,27
BE Mm2 cooling 4,03 8,98 20,36 32,41 42,77 52,09 54,29
Mm² reverse 0,84 1,84 4,03 6,46 8,43 10,24 10,73
DE Mm2 cooling 3,78 6,62 11,30 19,92 29,24 37,57 42,30
Mm² reverse 0,70 1,35 2,50 4,12 6,01 7,72 8,69
FI Mm2 cooling 15,88 24,06 36,43 43,28 47,28 50,19 50,99
Mm² reverse 1,35 2,28 3,71 7,49 8,21 8,74 8,89
FR Mm2 cooling 93,40 129,39 180,37 293,24 390,57 472,24 502,39
Mm² reverse 32,79 45,84 64,98 106,59 141,52 171,24 182,61
GE Mm2 cooling 34,07 66,29 127,64 216,74 298,51 365,63 400,13
Mm² reverse 4,88 9,54 18,81 30,61 41,65 51,09 56,23
GR Mm2 cooling 11,04 23,06 48,23 80,47 108,97 140,88 145,99
Mm² reverse 5,29 11,17 23,65 40,07 54,24 70,12 72,68
IR Mm2 cooling 5,03 6,81 9,37 13,84 17,07 19,39 20,37
Mm² reverse 0,75 1,08 1,78 2,30 2,83 3,22 3,41
IT Mm2 cooling 130,85 175,63 258,76 368,74 414,88 450,33 467,85
Mm² reverse 29,22 43,81 73,26 106,86 120,93 132,38 138,18
2
LU Mm cooling 0,25 0,43 0,87 1,34 1,76 2,07 2,20
Mm² reverse 0,07 0,10 0,17 0,26 0,35 0,40 0,43
NE Mm2 cooling 22,25 39,02 66,88 87,71 101,28 110,49 113,62
Mm² reverse 1,84 3,55 6,50 12,17 14,03 15,38 15,89
PO Mm2 cooling 8,46 12,51 18,73 34,84 52,08 68,41 78,27
Mm² reverse 4,67 7,27 11,25 18,47 27,53 36,11 41,31
SP Mm2 cooling 64,24 102,68 172,69 248,07 295,71 342,20 352,20
Mm² reverse 34,61 56,66 97,11 136,02 161,33 186,01 191,57
SW Mm2 cooling 38,41 53,26 69,38 78,17 83,23 87,28 88,21
Mm² reverse 4,08 6,14 8,74 14,90 15,88 16,68 16,92
UK Mm2 cooling 94,29 127,63 173,15 248,36 294,19 326,80 340,28
Mm² reverse 14,17 20,41 31,06 43,81 51,73 57,87 61,07
Total Mm² cooling 538,01 792,07 1214,23 1793,42 2207,83 2558,59 2693,04
Total Mm² reverse 136,71 213,10 350,28 534,96 660,23 773,29 814,88
20
Then we had to move from area statistics to energy use statistics. We computed the electricity consumption
of a square meter for AC depending on its location, its economic sector (typical use) and on the AC system.
In other words, we have obtained (through DOE simulation and physical extrapolation) energy consumption
figures for each system , each building use and each climate as shown in figure 9 under the form of a specific
value : consumption per square meter.
Figure 9 Consumption of the 18 systems in three climates as simulated with DOE software
140,0
120,0
100,0
kWh/m2
80,0 London
Milan
60,0 Seville
40,0
20,0
0,0
S
F
ps
l
st
um
ts
op
t
n
g)
al
tro
is
g
ng
lit
rg
VR
io
tio
M
LE
uc
di
in
in
m
rd
to
lo
Sp
ut
+h
la
n
i
bu
ol
Ks
D
er
R
IL
co
oo
oo
e
ai
SP
rib
co
on
at
tri
H
ir
le
C
.(c
.(c
+
st
ity
+a
K&
w
t.(
is
ng
PA
di
on
st
um
rd
id
er
is
+
+
Si
C
di
er
m
er
rd
at
PS
ai
s
PA
+h
r
hu
at
AC
at
er
w
ai
h
at
O
w
w
at
ir
r+
t
de
i
R
w
ith
LO
+a
w
de
ith
si
ai
de
w
d
si
w
ut
d
O
le
ith
ed
si
le
ut
d
TW
d
oo
le
ut
oo
w
O
le
l
oo
oo
rC
O
oo
C
le
C
C
rC
Ai
er
oo
er
er
at
Ai
at
at
W
W
r
W
Ai
CAC Systems
The three main sections of our BAU scenario predictions relate with : the actual cooling demand, the winter
demand of the cooled areas if no reversible use took place, the winter demand of the cooled areas with the
reversible use presently estimated. Figure 10 shows the first two values (cooling and associated heating
consumption by technical type) for the BAU.
Figure 10 Energy for cooling consumption split by technical type of cooling and related conventional heating
300 000
250 000
200 000
RAC
PACK
GWh
50 000
0
1990 1995 2000 2005 2010 2015 2020
21
The tables 4, 5 and 6 give the main values (EUR15) for the three functions. Note that gas is accounted for as
a secondary energy, with the same value as electricity.
Table 4 Total energy demand generated by AC (TWh either electric or gas or added)
Energy demand (TWh) 1990 1995 2000 2005 2010 2015 2020
Table 5 Cooling only energy consumption by country and year (for comparison with national projections)
22
UK 826 1 129 1 339 1 484 1 540
Total 18 073 27 336 33 154 38 371 40 103
We can keep in mind a 51 TWh consumption estimate for all AC in 2000 (18 MtCO2) becoming 95 TWh in
2010 (33 MtCO2). Such impacts are not small, but limited if we compare them with other uses in buildings
(heating, home electronics, better lighting, etc.). Remember the figures given correspond to BAU, and that
there is no significant EE measure on that market. So the next question is : how far can we improve the
balance? What is the potential of improvements paying for themselves but not realised by the present market
structure? This question can be tackled at three levels : the most frequent cold generating equipments
(chillers, packages), the cold generating plant (depending on its number of hours of operation, and climate),
and the full system, including distribution.
Optimisation of a chiller to improve its EER on the basis of capacity cost only
We have performed some engineer economic calculation and compared the technical improvements
proposed in the study with the diversity found on the market. We introduce one by one the possible
improvements (better compressor, better evaporator, etc.) and we see how the price of the service rendered
(the kW of cooling capacity) varies. For a given electrical power the capacity varies proportionally to EER;
for a given capacity, the compressor can be reduced when EER increases. So the cost per kW decreases with
the first steps of performance and only increases later (see figure 11).
110
100
Euro/kW
90
80
2 2,1 2,2 2,3 2,4 2,5 2,6 2,7 2,8 2,9
EER
Conclusion : the best chiller having the same cost (assumed here 100 Euros/kW) as the present “worst
performer” has an EER around 2.80. The range from 2.00 to 2.80 shows reasonable prices for a chiller
judged only on capacity. It corresponds exactly to the present market. The minimum cost chiller according to
our analysis has the same EER as the average market (EER 2.50), which may be considered as a validation
of our cost reconstruction.
2
Seasonal energy efficiency ratio, the energy efficiency ratio which reflects the real usage conditions of the equipnent
over the year
23
to 17 cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe), and equivalent
usage durations (at full load) taken as 400 or 800 hours/year .
Figure. 12 The annual cost of the service rendered by a chiller in terms of SEER
10
8
Total cost (Euros/m2)
ALCC17-800h
7
ALCC10-800h
ALCC17-400h
ALCC6-800h
ALCC10-400h
6
ALCC6-400h
3
2,00 2,20 2,40 2,60 2,80 3,00 3,20 3,40 3,60 3,80 4,00
SEER
The optimal level of performance for the screw chiller considered is about 40% more efficient than the
present « bottom » of the market : it has an SEER between 3.00 and 3.50. It may correspond to a chiller with
a correct EER around 2.46 (enhanced evaporator and condenser, improved compressor) and a capacity split
between 3 or 4 compressors. For manufacturers, there are other ways to reach 3.25 SEER, less expensive, but
our objective was to find out if there is a margin for improvement.
€1,400
€1,200
€1,000
Life cycle cost (€/kWc
€800
€600
€400
€200
€0
2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
24
System optimisation : all air systems
We have concentrated our efforts on air based distribution systems which show presently the most
consumption and the highest cost. The designers need the whole range of AC solutions to cover the domain
of geometries and air quality requirements. So the bottleneck to the expression of a global reduction in
consumption will be the point (shown on figure 14 by an array) where the improved air based solutions start
not to pay for themselves : the designers will find it is too heavy a constraint to get under this value.
Figure 14 The key point of energy efficiency : the best attainable air based system
ALCC Euros/m2/year
air
water
rac packages
SPECIFIC CONSUMPTION
kWh/m2/YEAR
The search for the optimum has been done in the same way as previously, through successive additions,
including part load options (Figure 12), with a 6% discount rate, electricity prices ranging from 6 to 17
cEuros/kWh (the most frequent being 10 cEuros for this type of customer in Europe), and equivalent usage
durations (at full load) taken as 400 or 800 hours/year. After sorting options and combinations, the optimal
trajectory of improvement of the annualised cost of ownership (ALCC) is given in figure 15.
Figure 15 Optimising with 6, 10 and 17 cEuro/kWh a full all air system in Seville for lowest ALCC
39
37
35
33
ALCC17
Euros/m2
ALCC10
31 ALCC6
29
27
25
The optimum is very flat, specially if we consider the highest cost of electricity. The regulatory measure
could be taken anywhere between a 0% and a 60% reduction without generating overcosts (in the LCC
definition) in Seville.
25
Part load performance has been quantified for the first time and the methods have been
tested
The report shows how in the case of most chillers, the part capacity performance can be better than full load
at the same temperature. This results from the reduction in refrigerant flow –and consequent improvement in
heat transfer efficiency at part load. The compression ratio is decreased so that compressor isentropic
efficiency increases. There is much progress being made in the control of these part load phenomena. The
issue is : how to represent, certify and translate in a single figure those improvements. There was on the table
the original US-IPLV method and a European version called EMPE.
The percentage of operating hours assigned at each part load condition (in the US-IPLV) is intended to be
representative of the US climate and buildings but not of the European ones. Further to this, an analysis of
the method shows that the ARI part-load temperature testing points are "sized" to be "representative" of
US buildings (cooling until in negative Celsius temperatures, for instance- as can be shown by drawing the
loads in terms of outside temperatures).
The first remark in the Italian proposal is that the operating conditions are rather different from Southern
Europe conditions. And even, if Northern Europe countries may need air conditioning in summer, it cannot
be said that Italy would need air conditioning at 12.8°C as normal operating conditions. Therefore, AICARR
proposed a new energy index, named EMPE (Average Weighed Efficiency in Summer regime in Italian)
directly deriving from IPLV, with different energy weights and, in particular, with different temperatures at
the condenser inlet, more suitable for the European climate and requirements in the air conditioning field.
The AICARR proposal, EMPE was not based on a sufficiently large climatic and technical investigation. Its
strength (being very close to the existing US method, which aggregated many factors) was also its weakness.
We had the opportunity to go further by constructing a data base of EU chillers at part load, understanding
better part load, and proposing two separate methods, one for part load reporting and certification, the other
one for the computation of SEER.
We have been able to define a new method called ESEER that enables to calculate the seasonal efficiency
for all European chillers (centrifugal units are not treated completely in this document by lack of specific
information but seem likely to be covered by the proposed method, due to the Us experience). The constraint
was to minimize the testing time while ensuring maximum precision, it is to say that the error coming from
the reduction of the data to single points should be inferior to the testing uncertainty. The new ESEER
method has been compared with the US-IPLV and EMPE proposal under both respects : time spent and
accuracy.
Original knowledge has been generated during the “Joint project” of EDF R&D facility and manufacturers
from Eurovent wanting to promote part load performance. The main tool used was actual testing of EU
equipment but a number of group meetings allowed to build a common thinking frame. The technical
description of the chillers tested follows on tables 7 and 8, split by condensation type.
26
For all the tested chillers, some common testing points were made according to either the US-IPLV
or the EMPE conditions depending on the manufacturer will. For all chillers, a supplementary point was
added to fulfil the CEN EnV requirement : nominal inlet condensing temperature (35°C for air and 30°C for
water) and 50% load ratio referred at this nominal inlet condensing temperature For chillers n° 2, 3, 4 and 8,
only IPLV or EMPE points plus the CEN one were available. For the others as many testing points as
desirable have been obtained. In all circumstances a simple model has been used to draw the performance
maps from existing testing points.
For discrete stages chillers, it would be easier to describe performance at a given stage not at a given
percentage. For the very few continuously controlled chillers, fours stages can be defined in terms of input.
Since temperature and load can be tested independently and recombined, there is no need for combined
testing & weighting (like IPLV).
About certifying Part Load : what the manufacturers give to their customers is a « map » of performance,
not only values at the four arbitrary percentages and temperatures, plus the final Eurovent grading when it is
available, based on a SEER. The customer can rely on the Eurovent SEER computed from this map … or
compute its specific SEER for its specific demand. No need to test every condition reported in the “map”: the
benefit of Eurovent is the fair and independent choice of a few points on the map, as usual, and the
associated independent testing.
We arrived also at applicable conclusions on the way to report the SEER in the Eurovent directory. We
started from HSEER, the DOE reference that we generated. It is proven that each set of outside conditions
(for each sector, climate, type of chiller, type of secondary system) can be reduced to four or five external
conditions without loss of accuracy. The ESEER index proposed here is a set of 4 conditions given for
E.U. as a whole, but there can be as many similar indices as specific demands: sector, country, etc.
We have introduced a format for the description of the stages of a chiller, like in table 9 and following,
suitable for Eurovent specification. For each stage, the manufacturer has only to declare which piece of its
equipment is operating and to indicate CC , the cooling capacity and EP, the electric power absorbed. The
certifying body has only to check a few of the values, selected in the same conditions as usual. Note that this
procedure is in fact already used for some chillers with various speeds, namely “low noise” chillers with the
possibility of reduced fan speed.
Table 9 : Part load performance of water cooled scroll chiller N°4, as could be reported in Eurovent part load
certification scheme
N° 4 // WT : 30°C STAGES 1 2 3 4
Compressor 1 0 0 0 1
Circuit 1
Compressor 2 0 1 1 1
Compressor 3 0 0 1 1
Circuit 2
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
27
Figure 16 Reduced efficiency while decreasing part load ratio (same source temperatures) for the
testedwater cooled chillers
Reduced efficiency of the part load stages for water cooled chillers
1.2
1.1
1
N° 4
0.9
N° 1
N° 3
0.8
N° 6
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
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
Reduced efficiency of the part load stages for air cooled chillers
1.3
1.2
1.1
N° 5
1
N° 7
0.9 N° 8
N° 9
0.8
N° 2
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
28
EECCAC final figures for a European SEER method (ESEER)
Our work clearly shows also that the methodology for air and water cooled chillers enabled to extract
seasonal operating temperature conditions with errors on the seasonal efficiencies that are inferior to the
experimental uncertainties, for all chillers, included single compressor units. However, it also shows that the
experimental uncertainty is quite high. It mainly comes from the uncertainty measurement on the
temperature difference at the evaporator. In order to simplify the application of the index, some rounding can
be done without modifying noticeably the ESEER figures obtained, largely under the experimental
uncertainty. A comparison of the conditions of the 3 available indexes is proposed Table 10 for air cooled
chillers.
Table 10. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers
ESEER ARI EMPE
Weighting Weighting Weighting
Part load ratio Temperatures Temperatures Temperatures
coefficients coefficients coefficients
100 35 3% 35 1% 35 10 %
75 30 33% 26.7 42 % 31.3 30 %
50 25 41% 18.3 45 % 27.5 40 %
25 19 23% 12.8 12 % 23.8 20 %
Temperatures of the ESEER are comprised between EMPE temperatures above and ARI temperature under.
ESEER weighting coefficients give more weight to the 25% point load than both index. For 50 and 75%,
coefficients are nearer to the EMPE index. The 100% coefficient is 3%, nearer from the IPLV one. A
comparison of the conditions of the 3 available indexes is proposed Table 11 for water cooled chillers.
Table 11. Comparison of the ESEER conditions with the EMPE and IPLV indexes for water cooled chillers
ESEER ARI EMPE
Weighting Weighting Weighting
Part load ratio Temperatures (°C) Temperatures Temperatures
coefficients coefficients coefficients
100 30 3% 29,4 1% 29.4 10%
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.
29
Figure 17. Comparison of HSEER with EER on the tested chillers
HSEER versus EER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5
EER
IPLV and EMPE are more accurate than EER for classification but do not give enough
accuracy for comparison of chillers
A classification based on US-IPLV or EMPE would be largely false but would not distort completely the
market (figures 18 and 19).
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
IPLV
30
Figure 19 comparison of EMPE with HSEER for the tested chillers
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
EMPE
Based on similar assumptions, the two methods, IPLV and EMPE have at the end the same advantages and
disadvantages.
The newly proposed ESEER method allows grading and ranking of chillers by order of merit
We see on figure 20 that the new method has the most important characteristic expected from a market
transformation tool : almos no misclassification; a piece of equipment graded better than an other one is
better or equivalent.
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
ESEER
Conclusion : the differences are relatively large between existing methods and reality, and not always in the
same direction. The newly proposed ESEER method is more accurate in a noticeable manner and satisfies
the needs of Eurovent certification process as well as the expectations of the DGTREN in a market
transformation.
Energy efficiency options have been defined for each system configuration
Th possibilities are so numerous, and so system dependent that the process of “filtering” the most promising
was difficult. More than 20 basic systems and 100 variations were considered on a qualitative basis and then
"filtered". A number of systems and energy efficiency options do not need detailed quantitative simulation
because one or more of the following apply: they are infrequent; their use is are not expanding; the literature
31
is already sufficient to enable the savings potentials and costs to be defined; or simulation is impossible. For
the others a detailed simulation has been organised.
High energy savings are possible, as well as significant CO2 emissions reductions, at a net negative cost,
such that all parties (manufacturers, consumers and utilities) would find a benefit in the deployment of
efficient CAC. However the chain going from the manufacturer (most have already improved equipment in
their directories together with simpler one) to the final consumer is distorted by a number of factors
including: the consideration of initial cost as the only decision criterion by most designers and installers; the
problem of certifying something built on site and only once; the lack of incentives for operators to optimise
efficiency; the absence or inadequacy of building codes addressing this new and rather complex equipment
segment, which is very difficult to model; different incentives caused by the separation between the plant
owner and the building renter, between the building renter and the CAC operator, etc. As a result the policy
measures to be proposed in the EECCAC study should not only address the offer of more efficient CAC
equipment, which in fact seems to be the smaller difficulty, but should also address all the factors which
have to reshape and activate the chain relating the energy used to the final service: the conditioned square
meter.
The French measure is basically a minimum efficiency threshold above which AC systems can benefit
from EDF’s marketing and financial support. The Portuguese measures cover a number of sizing
obligations, and require the use of a central AC system with a number of energy saving features such as ’heat
recovery’, ’part load staging’, and ’free cooling’ over a certain cooling capacity threshold (usually 25kW).
Monitoring and maintenance are also included in the Portuguese RSECE the local regulation applicable to
CAC). The UK Market Transformation Programme is a comprehensive set of possible policy measures
including: obligations written into building codes (i.e. the imposition of limits on cooling demand);
information (through a national release of the Eurovent directory database and the later elimination of the
least efficient equipment) and voluntary "best practice " initiatives. The US ASHRAE 90.1 standard is a fully
integrated set of policy measures which include minimum energy performance thresholds, design guidelines
and specific system requirements. It is fully described in the report.
At the European level a draft Directive on Energy Efficiency of Buildings is currently under transposition.
This requires the calculation of building energy performance, which itself demands knowledge of Air
Conditioning (and other) system efficiencies. This will help the adoption of the best technical solution in new
buildings. Article 8 requires that central air-conditioning systems of greater than 12 kW cooling capacity
shall be regularly inspected. On the basis of this inspection, which shall include an assessment of the air-
conditioning efficiency and the sizing compared to the cooling requirements of the building, the competent
authorities shall provide advice to the users on the possible improvement or replacement of the air-
conditioning system and on alternative solutions. So there will be a movement towards improvement also in
the existing buildings.
32
Scenario 3 BAT- THE BEST CONSUMER CHOICE WITH PROPER PART LOAD INFO
All packages and chillers reach in 2005 the SEER level with the minimum LCC (BAT with upcoming
information given by part load testing). Part load IS taken into account in Eurovent grading and so the
corresponding improvement is obtained. The policy measure associated is to ban many classes of equipment
or a negociated agreement on average part load performance like ACEA agreement for cars
Scenario 4 FREE COOLING IS USED AT ITS MAXIMUM POTENTIAL
Obligation of introducing free cooling on air side of air based distribution systems at a certain value of air
flow (Portuguese regulation and Ashrae) even for primary air (which is the case of our simulations, at
comfort level TC). There is a reduction in cooling demand which is climate dependant and has been
expressed here by country and system.
Scenario 5 VARIABLE FLOW COMPULSORY IN ALL AIR BASED SYSTEMS
There is a reduction in cooling demand which is climate dependant but has been applied here on auxiliaries
consumption in Air based systems with national values.
Scenario 6 BRITISH REGULATION ON AC – HEATING, COOLING AND AIR MOVEMENT-
ADAPTED FOR EACH EU CLIMATE
Introduction of a MEPS on total electricity used for Heating ventilating and AC in kWh/ m2; to know the
cost we have to evaluate the less costly options, which may be on either side, primary or secondary; national
values are different and have been derived from UK with corrections for DD and fitted to each country. The
policy instrument would be a strong and harmonised implementation of EPB directive. The less expensive
way of attaining the objective is the improvement of chillers. Starting from their present averages of EER
and SEER, this policy induces almost no extra cost for any stakeholder, and absolutely no cost provided it’s
applied to all manufacturers (and so that they all pass on the costs to the customer). To obtain this “free”
market transformation a prescriptive minimum should be applied to local manufacturers and importers at the
same time.
All the elements for an action plan on Air Conditioning are available in the full report
The analysis presented in this study has shown that there is a significant variation in energy efficiency for all
types of CAC equipment that have been investigated when tested under standard test conditions. The
measures which can be considered to encourage the higher energy efficiency levels for new CAC equipment
are:
• 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)
• the development of an EU model building code that addresses air conditioning amongst other energy
end-uses. (an EU equivalent to ASHRAE 90.1 and which like ASHRAE 90.1 is subject to
continuous improvement)
• The development of practical public domain CAC system design tools which: a) can aid system
designers to develop energy efficient CAC designs, b) can enable to compare of the relative benefits
of different system designs, c) can be used in building thermal regulations to demonstrate
compliance with requirements
33
• The development of EU benchmarks for CAC system efficiency expressed in terms of: building
function and size; occupancy and purpose; quality of comfort provision and climate (e.g. cooling and
heating degree days)
Further to this, Member States should undertake a revision of their building thermal regulations to
address a number of specific issues aimed at reducing CAC energy consumption which are described
in the report.
The European Commission and/or a coalition of willing Member States should also consider:
• Making efforts to define best practices in operation and maintenance performance contracting
With an aim of informing national building thermal regulations and the implementation of the Energy
Performance in Building Directive.
Member States could also consider the development of low cost mechanisms to encourage the adoption of
good practice for CAC operation and maintenance (namely by ESCOs).
34
35
1. INTRODUCTION
SAVE II is an EU programme to promote the rational use of energy within the European Community.
The EECCAC working group began a study in April 2001 to investigate the technical and economic
potential of measures to raise the energy efficiency of Central Air-Conditioners (CAC). The EECCAC
study follows-on from the EERAC SAVE study which investigated the potential for measures to raise
the energy efficiency of room air-conditioners (RAC) and which is available from the same co-ordinator
(EERAC 1999). Since the EERAC study covered all types of AC of under 12 kW cooling capacity, the
present study is concerned with air-conditioning systems over 12 kW and will eventually integrate the
two segments in a common picture of the European industry and market.
The objectives of the study are:
• to estimate the electric power consumption of CAC,
• to estimate potential energy savings deriving from the use of more efficient CAC,
• to investigate ways in which these savings can be realised,
• to make appropriate recommendations, on the basis of a cost–benefit analysis.
The working party has been gathered and co-ordinated by Jérôme Adnot from Armines. The work has
been organised in tasks for which the best experts have been chosen as task leaders. A broad coverage of
energy agencies, utilities, technical experts and manufacturers’ representatives are involved in the work
in order to be sure that state of the art knowledge is available for each aspect of the study.
• 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.
• INESTENE and –later- Energie Demain are leading consultancies on demand-side management
(DSM) in France,
• Eurovent Certification was established by the Eurovent/Cecomaf manufacturers’ association for the
certification of performance of air-conditioning and ventilation equipment.
• The University of Athens, in particular the Group of Building Environmental Studies, is very active
in the field of solar cooling and energy conservation in buildings; the group carries out research,
specialised studies, application projects, education, and dissemination of information.
• Politecnico di Milano, and namely the Department for Energy studies is the main supporting
laboratory for the HVAC engineers gathered in AICARR,
• AICIA supports the research by ETSIS, the famous HVAC engineering school in Sevilla,
• UTCB is the Romanian Technical University for building sciences,
• ITF is an HVAC consultancy in the region of Chambery, France
36
Participation of energy agencies, utilities, manufacturers and national experts
The coverage of national and industrial expertise in the study was ensured through the participation of
the following bodies:
• EdF (Electricité de France), the French electricity utility, who brings an important contribution to
the study of this growing electrical end-use. EDF are represented by Pascal Dalicieux and Frank
Colomines
• ADENE-CCE, the Portuguese energy-conservation agency, has a significant experience;
• AICARR, the Italian association of Air Conditioning, Heating and Refrigerating engineers, who
includes professionals working on international standards development, maintains special
Observatories on Hospital air conditioning technology and Refrigerant fluids and cycles.
• EVA, the Austrian energy research and policy institution in which the federal and provincial
administrations (‘Bund’ and ‘Länder’, respectively) and some 30 important institutions and
corporations from a variety of economic sectors co-operate.
• IDAE, the Spanish energy-conservation agency, has close relationship with all bodies having an
influence on CAC in Spain,
• BRE is the leading UK centre of expertise on buildings and construction. Its energy-related activities
include technical research and consultancy, managing the Government's buildings energy-efficiency
information programme, providing strategic analysis of energy-efficiency policy options and
modelling the energy performance of the UK building stock.
37
2. CENTRAL AIR-CONDITIONERS IN EUROPE: DEFINITIONS AND BASIC
DATA
2.1. Importance of AC for human health and productivity performance, link with
ventilation
The use of air conditioning is increasing rapidly in Europe, as a result of an increasing trend towards control
of the indoor environment and a wider diffusion of air-conditioning systems as a consequence of economic
growth, which has made them more affordable. However, this is also partly a consequence of a movement
towards higher economic productivity. Accurate figures have shown that a better indoor working
environment leads to less quality problems, higher productivity and less accidents in the workplace, provided
it does not create too much noise. We are not in a situation where air conditioning specialists and companies
generate an artificial need, but are rather in a situation where they offer new ways to answer existing needs or
decrease total costs. However there is a need for them to be able to prove case by case that their techniques
are cost effective for the intended purpose and that they have optimised their proposed solution, which is a
major reason why industry has been so active in co-operating with the EECCAC study.
What is "comfort"?
In any case, the comfort level to be reached should reflect the nature and quality of the activity which takes
place in the conditioned space. There is no value in “cooling buildings”, but there is in being able to establish
desired comfort levels in the internal spaces where people work or perform other tertiary activities.
To give a rough presentation of the range of comfort conditioning requirements and circumstances that can
be encountered, the following main cases are listed:
NC- Natural Cooling which is obtained, day or night time, by forced-ventilation, when outdoor conditions
permit, or by any other ways of transferring heat to the outside, provided they are not based on the operation
of a compressor. Natural cooling is usually insufficient to attain always and everywhere the required comfort
levels but can be found sufficient in most circumstances in climates like the UK or Northern Europe.
PC- Partial Cooling which is obtained with air conditioning equipment that provides partial control of the
temperature. For instance, the rooms are cooled but fresh air may be introduced without cooling, or the
installed cooling capacity of the AC system may be insufficient for all circumstances, as a result the internal
air cannot be kept at a constant temperature. This may be felt as comfortable in France or Germany.
TC- Total Cooling, wherein the AC system provides full temperature control and includes the provision of
the minimum rate of ventilation air changes required for hygienic purposes at an adequate temperature. This
type of equipment allows a degree of dehumidification consequent to the cooling effect– it is a very frequent
level of comfort today.
TAC- Total Air Conditioning, which includes full control of temperature and humidity as well as provision
of the minimum ventilation rate required for hygienic purposes but is not capable of attaining air purity
conditions for specific IAQ (Indoor Air Quality) levels.
AAC- Advanced Air Conditioning, same as TAC but with a full control of IAQ. These systems are
particularly applied in hospitals or clean rooms in the electronic industry.
The variation in comfort level changes the consumption of energy. If one wanted to make a complete
comparison, it would be necessary to give a monetary value or some other proxy quantity to discomfort and
to balance it with the total cost of the service. The question was less difficult for the RAC studied in the
EERAC study which were deemed to provide " equal service" at the PC or TC level of comfort. The
differentiation of comfort has substantial consequences for the systems’ energy consumption, but one should
not regard a decrease in the quality of the indoor environment as a means of saving energy without being
conscious of the trade-offs involved.
38
Comfort level, Ventilation: our assumptions for the study
From one extreme to the other both the cost of installing the initial AC equipment purchase and of operating
it can vary by a factor 10, but consideration of the choice of who or what really needs a certain level of
comfort is outside the scope of the present project. Most comparisons between systems are thus to be done
internally to a given assumed comfort level even if it is possible to gain an idea of the relative cost of the
various possible comfort levels. Total Cooling will be used as the appropriate value for benchmarking. It is
more difficult to predict the future level of comfort expectations and even more the speed of change.
There is a strong interaction between two functions: ventilation (i.e. air changes) and cooling. The technical
systems used differ from one country to another depending on the basic philosophy embedded in the
regulations and building codes. Two philosophies of ventilation seem to exist in Europe : in the first one
(adopted by Northern countries), ventilation comes first as an hygienic necessity and then a further decision
leads to cool the space or not. In the second one (apparently Southern States), the decision of A/C comes first
and leads to more air changes with the outside, and to controlled ventilation. Central ventilation (with
cooled “primary air”) is the base of our technical study together with the TC comfort level.
1
Tertiary is a European word indicating all human activities and related buildings other than industry or households
39
and air quality. From an energy perspective, this situation is summarised by a “load” to be extracted or
cooling effect Pc (the minimum thermodynamic quantity necessary to maintain the defined comfort
conditions). In fact the desired comfort conditions may include thermal comfort (which is expressed in
terms of a mix of convective and radiative temperatures), humidity control and indoor air quality (IAQ),
which is usually obtained through ventilation, i.e. by the change of indoor air, and filtration components.
Figure 2.1 Essential quantities in the process of air-conditioning in summer, seen from an energetic perspective
Pr heat rejected
air conditioned space
AC Pc cooling effect
system control of: Temperature
Humidity
IAQ
Pe energy input
The accepted energy performance index is called the ‘energy efficiency ratio’ (EER) and is defined as:
EER = Pc / Pe
Cooling only systems (not including ventilation, or air quality, or humidity control) extract heat in
summer from inside the room (Pc), approximately equivalent to the value of the “load”, through the use
of electricity (Pe). Usually the heat rejected outside (Pr) has an energetic value equivalent to Pe + Pc.
There are also some cooling systems which offer the possibility to produce heat instead of cold by
reversing their refrigeration cycle: such systems are called ‘reversible’. A similar index to the EER, the
coefficient of performance (COP), is applied to indicate the performance of reversible AC in the heating
mode. It is the ratio of the heat input into the conditioned space and the electric power consumed to
transfer it.
• air-cooled evaporators, or direct expansion evaporators consisting of a pack of finned tubes through
which the air is forced;
• liquid-cooled evaporators, or flooded evaporators consisting of a tabular shell in which the refrigerant
expands and cools a fluid circulating in a bundle of tubes inserted in the shell.
After its full evaporation the refrigerant vapour is compressed using a compressor for which the following
main technologies are used:
• reciprocating compressor
• screw compressor
• scroll compressor
• centrifugal compressor.
40
The centrifugal compressors will not be studied in details in this report. They are relatively infrequent
industrial compressors of a large size and very efficient2 and we can relate them more with “district cooling”
or “block cooling” that really with “Air Conditioning”.
A wide range of technologies are used to couple the compressor to the electric motor:
• open type or accessible compressors, presenting detacheable parts to access the compressor’s main
components and coupled to separate electric or thermal engines. They can be used with any
refrigerant but are generally employed in systems with medium to high cooling capacity.
• "Semi-hermetic" compressors that are similar to the open type compressors but have a common
casing with the electric motor; they are generally used for systems with medium cooling capacity.
Hermetic compressors, which have their body directly coupled to an electric motor cooled by the
refrigerant and enclosed in a totally sealed shell; these are generally used for systems with a small to
medium cooling capacity.
Figure 2.2 An aircooled chiller (courtesy Climaveneta)
After its compression the refrigerant vapour is condensed while evacuating the heat corresponding to the
one absorbed at evaporator level and the thermal equivalent of the work of the compressor. The
condenser technology depends primarily on the required application and the heat source. Condensers
used in CAC systems are divided into three categories:
• air-cooled condensers consisting of a finned tube heat exchanger (figure 2.3). The primary factor
which influences the performance of the condenser, is the outside air temperature.
• water-cooled condensers consisting of finned tubes with internal grooves to increase the heat
transfer surface area and the overall heat transfer coefficient. The temperature and flow rate of water
have the greatest influence on the condensing temperature. The water used as the coolant may be
from a natural water source (such as a river or aquifer) or from re-circulated water that’s been
cooled in a cooling tower.
2
They enable a high pressure ratio because of the absence of alternative compression. The compression ratios can vary
between 2 and 30. The turbine is called the impeller. If the fluid enters the impeller with a tangential component or
swirl, that would occur only at non-nominal or bad designed points, the speed of the refrigerant would be consequently
reduced as related to the speed of the impeller. The ratio of pressure producing work to kinetic energy output is known
as the impeller reaction and ranges from 0.4 to 0.7. That’s why, after the impeller, one can find a diffuser that ends
converting kinetic energy into pressure lift. The observed performance is around 6.00 in terms of EER and 5.7 in terms
of SEER, between 50 and 100 % better than the chillers of smaller size.
41
• Evaporative condensers, which are used in industrial applications and combine a condenser and a
cooling tower in a single apparatus.
Figure 2.3 Inside an air cooled chiller (Courtesy Airedale)
A water cooled chiller (figure 2.4) is generally used with a cooling tower. Among cooling towers there
are three principal systems:
• Indirect contact (Dry) cooling towers where there is no contact between the cooling fluid (air) and the
fluid to be cooled (water)
• Direct contact (Wet) cooling towers where there is a direct contact between the two fluids thus providing
better heat transfer
• Wet-dry towers, which contain a conventional wet type tower in combination with an air-cooled heat
exchanger. They are especially used to reduce water vapour plumes and hence water consumption.
A wet cooling tower (which displays better energy performance) is more at risk of cultivating the
legionella bacillus and consumes water.
Figure 2.4 A water cooled chiller (Courtesy Carrier)
After condensation the refrigerant is expanded by an expansion valve, which is used to throttle the
refrigerant fluid back to the evaporator pressure and to control the refrigerant flow. Three systems are
used:
42
• expansion devices with a constant pressure difference.
• thermostatic expansion valves that are controlled via the superheating.
• electronic expansion valves that are also controlled via the superheating.
The consequences in terms of energy consumption of all these technical choices will be investigated in
the rest of the study.
• Distribution structure including networks of fans, ducts and pumps for refrigerated air and water
circulation,
• 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.
43
Figure 2.5. CAC system description tree showing the most common CAC systems
LOCAL OR CENTRAL
LOCAL CENTRAL
MULTI- VRF
SPLIT
44
Water to air
Heat exchanger
Circulation
fan
Air filter
Outdoor
Indoor air supply
air return
There are various kinds of FCU:
• a 2 pipe fan coil (2P) for cooling only; the heat exchanger comprises one supply and one return pipe only
for refrigerated water;
• 2 pipe fan coil with change-over (2PR). The same system is used in all zones and comprises one supply
and one return pipe as the previous one but circulation of which can be of either hot or cold water. A
reversible chiller supplies either cooling or heating and changeover from one mode to the other is
centrally regulated according to season. This system cannot heat and cool simultaneously in two
different rooms and hence is used when the summer-winter transition is easily distinguishable;
• 2 pipe fan coil with electric heating (2PE). This system may be reversible or not, according to needs. An
additional electric resistance heater can be complementary to the reversible heating mode or can be the
main heating source for weak loads during the winter period. This system can heat or cool
simultaneously in two different rooms;
• 4 pipe fan coil (4P) with two coils frequently assembled together The same system can be used in all
zones, and comprises a supply and return for both hot and cold water and can thus heat or cool
simultaneously different rooms of the same building. A 3 pipes system as well as a 4 pipes system with
only one coil existed and disappeared.
There are also a wide-variety of air handling units (AHU –see figures 2.7 and 2.8) used for the remote
preparation of cold air.
Figure 2.7 Air Handling Unit with heating, cooling and variable air flow distribution
45
Return air
Outdoor air
Figure 2.8 The « coil » providing heating and cooling inside an AHU (Courtesy Trane)
Induction units (IU), less frequent nowadays, are used when centrally distributed air is further cooled at the
local level through thermal contact in the IU with refrigerated water circulated in a central water loop, Figure
2.9.
46
Figure 2.9 Induction Unit
Diffusion
Filter Coil
Indoor air
Nozzles
47
circulation
with by-pass on mixed air
with control on coil
capacity
MULTI-ZONE SINGLE DUCT FIXED VARIABLE Multi-zone CAV
PER EACH ZONE PER ZONE PER ZONE
VARIABLE CONSTANT VAV systems
VARIABLE VARIABLE VVT systems
DUAL DUCT FIXED VARIABLE with terminal re heat
first duct - cold air
second duct – hot air
DUAL CONDUIT High pressure systems
PRIMARY AIR FIXED VARIABLE
SECONDARY AIR VARIABLE CONSTANT
2. AIR-AND-WATER SYSTEMS
refrigeration: chiller (air or water cooled with/out cooling tower)
Air treatment: primary air – central station type, air handling units
secondary air – room treatment
TYPE OF AIR AIR VOLUME AIR AIR DISTRIBUTION
TEMPERATURE
PRIMARY AIR FIXED VARIABLE SINGLE DUCT std velocity for systems
with fan coils
high velocity for systems
with induction units
SECONDARY AIR Treatment by room terminals: fan coils, induction units, radiant panels
48
Number of water loops connected with the chiller
The simplest solution is to have one single water loop in the conditioned space. It can be used for cooling
when connected to a water chiller and for heating, when connected to a boiler, the transition being called
“change over”. The same loop can be served by a reversible chiller provided refrigerated waterin cooling
mode or hot water in the heat pump mode. For more comfort, two water loops are installed (hot and cold)
with distinct generators, or alternatively to each side of the chiller (reversible heating and cooling). The
chiller is still only connected to one loop at the same time. In a further refinement the chiller is installed
between the cold and the hot loop, taking complete advantage of COP and providing at each instant a
reversible solution.
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.
49
Should we neglect RAC or include them here? To answer this question, figures about the market and level of
cooling provided by RACs drawn from the EERAC study have been utilised in the present study. RAC are
used for 78% in the economic sectors and for 22% in households. It is better to compare the shares of all
competing AC types in a single scenario for a given economic sector and country. This also helps to
understand the real basis of competition between RAC and CAC systems; and should thereby avoid
potentially distorting impacts from isolated policy actions.
Four types of air-cooled RAC are widely used:
• Split-packaged units; consisting of two sections (one indoor and one outdoor unit) connected only by
two pipes that transfers the refrigerant and a cable for the electric power. The indoor unit includes the
evaporator and a fan, while the outdoor unit has a compressor and a condenser. There is a range of “large
split systems” over 12 kW of cooling capacity and therefore are usually classified separately from the
residential splits.
• Multi-split-packaged units; consisting of several indoor units (usually four or more) connected to one
outdoor unit. This family of AC equipment is partly under 12 kW of cooling capacity and partly over.
VRF (Variable Refrigerant Flow) systems can be considered as a version of multi-split systems but are
always over 12 kW in cooling capacity.
• Single-packaged units; are commonly known as ‘window’ or ‘wall’ RACs wherein one side of them is in
contact with the outdoor air for condensation, while the other provides direct cooling to indoor space by
means of an air circulation fan.
• Single-duct units; which are packaged AC appliances that are kept inside the room while cooling the
space, and reject hot air from the condenser to the exterior space through a duct.
Water-cooled units of any type under 12 kW cooling capacity were part of the previous EERAC study. The
water used in RAC could in principle be drawn from a natural water source, but this is seldom available; the
main use of water-cooled RAC is therefore limited to closed-loop heat pumps as previously described in this
section as one of the CAC systems with a comparatively high system efficiency.
Sizing issues
In general all the related issues within the borders of CAC energy efficiency have been treated as diligently
as possible. In particular, the potential problem of equipment over-sizing had to be treated explicitly. Since a
CAC system is designed and sized by a professional engineer who is usually concerned to minimise the
50
initial investment cost, we can expect that on average there is no over-sizing in this case. There are however
many examples of oversizing, which should be related to some specific factor (type of contractor, level of
expertise, time allocated for sizing, etc). We made the assumption of a fair and economic sizing. In the case
of RAC, which are bought directly without any advice or even from a retailer interested achieving the
highest purchase value or from an installer who enjoys a margin on the equipment sold, it is assumed that the
average European RAC is over-sized by a factor of 2 (i.e. 100% over sizing). For consistency reasons we
have chosen a single value for sizing all CAC systems and converting capacities into areas and later areas
into capacities (120 W/m2) and another one for all RAC systems (240 W/m2). This value, being used twice
in opposed ways, has no influence on our statistics.
The ratios of consumption per square meter and the cooled areas are all presented as “standardised” area
(based on 120 W/m2) but on one occasion figures have been produced with a variable sizing depending on
location, building, system type to allow national comparison. Note that the notion of conditioned area is
uncertain in national statistics : not the gross built area, not the strict area of activity rooms; conventions may
vary from one country to another, a fact that gives interest to our repeatable “standardised” area.
Free cooling
At some time during the year, outside air can be used directly to cool the space without any special thermal
treatment. There are control issues associated (flow rate, movement of dampers, nature of control : based on
temperature only or on enthalpy), etc.
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
Entering chiller water temperature 12°C About the same (the water
flow is fixed by the
standard)
Entering condenser water temperature 30°C (water cooled) 29.4°C (85°F) (water
51
cooled)
Entering condenser air temperature 35°C (air cooled) 35°C (95°F) (air cooled)
-- Heating operation
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.
This load profile is the basis used to calculate the integrated part load value (IPLV), the seasonal average
efficiency of a chiller.
The Integrated Part Load Value is thus calculated using the following equation:
IPLV = 0.01A + 0.42B + 0.45C + 0.12D
Where A = EER at 100% of full load
52
B = EER at 75% of full load
C = EER at 50% of full load
D = EER at 25% of full load
The suitability of this index for European operating conditions is considered in Chapter 9. Before our study a
first proposal by the Italian AICARR was made, called EMPE.
53
units. In the USA the following test standard is used: ARI 365 – 1987 ‘Commericial and industrial unitary air
conditioning condensing units’ .
ISO TC 86/SC 6/WG 8 has developed a proposed working draft PWD 16345 for ‘Cooling Towers – Testing
and Rating for Performance’. The Cooling Tower Institute in the USA has also issued a test standard to rate
the performance of Cooling Towers, CTI 201-1986 ‘Certification Standard for Commercial Water Cooling
Towers’.
54
In particular, two different categories of air diffusion system can be identified:
- Mixing system with the air introduced above the occupied zone.
- Displacement systems or mixing systems with the air introduced below, or inside, the occupied
conventional space.
The following tables 2.4 and 2.5 indicate the ventilation efficiency values for each system category, in the
hypothesis that every system is projected, realized and installed in conformity with the manufacturer
directions and that we remain at nominal conditions.
Table 2.4 – ventilation efficiency for mixing system with the air introduced above the occupied zone.
Diffuser typology ev ev
Dt < 0°C Dt ≥ °C
Helicoidal effect diffuser 1,00 1,00
Cones diffuser 0,90 0,75
Linear slot diffuser 0,75 0,60
Outlet with single or double fin rows 0,70 0,60
Table 2.5: ventilation efficiency for Displacer systems or mixing systems with the air introduced below, or
inside, the occupied conventional space.
Diffuser typology ev ev
Dt < 0°C Dt ≥ 0 °C
Floor helicoidal diffuser 1.2 1.1
Underseat diffuser or similar 1.3 1.3
Displacement diffuser 1.3 0.8
The efficiency value characterizes the share of the flow rate introduced that actually reaches the occupied
zone, contributing to cut the heat loads and the air pollutants down. There is much to gain by this approach,
embedded into a UNI standard but not yet quantified here.
55
load and with a single set of stable environmental conditions, thus the part-load performance of variable or
multiple speed drive units is not reflected.
56
Table 2.6: Test conditions for the determination of cooling capacity, ISO
inlet 30 22 30
outlet 35 27 35
The USA has an extensive program for air conditioners and heat pumps, that includes the following
product types:
57
Similarly all central heat pumps with capacities up to 65000 Btu/h are rated using a similar approach to
produce a heating seasonal performance factor, HSPF. The commercial (small and large) air conditioners and
heat pumps can be air, water or evaporatively cooled.
For small-commercial air conditioners and heat pumps with capacities above 65000 Btu/h but below 135
000Btu/h, the cooling performance is regulated for the EER (static test conditions) and Integrated Part-Load
Value (IPLV), which is a measure designed to reflect performance under part-load conditions. The heating
performance is measured and regulated using a static COP test.
For large commercial air conditioners and heat pumps, defined as those with capacities above 135 000 Btu/h
(40 kW), a distinction is made depending on whether the capacity is above or below 760 000 Btu/h (205 kW)
and whether the unit is an air cooled air conditioner, an air cooled heat pump or a water/evaporatively cooled
air conditioner. Efficiency is measured and regulated in terms of the EER, COP and IPLV.
Korea is unique in having devised AC efficiency standards and targets that treat constant speed air
conditioners (those using a single-speed compressor) differently from those using a variable speed. Both the
fixed and variable-speed air conditioners (either room or unitary) must satisfy MEPS (Minimum Efficiency
Performance Standards), energy labelling and are also subject to aspirational efficiency targets. The variable
speed units are tested and rated using a seasonal energy efficiency ratio (SEER). Reversible units are not
subject to COP requirements but are required to satisfy the cooling-mode performance requirements.
In Japan a central air conditioner would be classified as a unitary air conditioner. Larger room air
conditioners, of a packaged type, are classified as unitary air conditioners and are subject to energy
efficiency targets, not MEPS and labelling. The existing targets differentiate depending on whether an
appliance is integral (windows) or split-type and whether it is cooling-only or reversible. The targets for
reversible units are a combination of EER and COP targets. Multi-split systems are currently excluded but
are about to be included in new energy efficiency target regulations, due to come into effect between 2004
and 2007. This implies that either the draft international test procedure for multi-split units is to be adopted
or that a new unique Japanese test procedure will be created.
Definitions
To facilitate understanding AC system energy performance, a common vocabulary has been defined based
upon the specification of seasonal or annual quantities. In establishing this balance, you separate the
auxiliaries according to the function being performed, that we consider being only two : heating with
ventilation, cooling with ventilation.
To apply this strict definition, we should come back to each time step and make some sophisticated
computation on the effects of the auxiliaries, mostly the ventilation auxiliaries at that time : do they act in the
58
direction of heating or in the direction of cooling? We have done something far simpler : when there is
heating only or cooling only, ventilation is integrated in the demand ; at the end of the year the unaffected
auxiliaries (floating or simultaneous heating and cooling) are allocated to heating and to cooling in
proportion of the total yearly demand.
The quantities of interest are :
1. SCL: the Summer Cooling Loads or "cooling energy"; it is assumed that the total summer cooling load,
including the energy for cooling and humidity treatment, is completely satisfied by the AC system. The
SCL takes into account zone loads, outdoor air load, the heating of air that is passed through fans and the
real system operating schedule and thermostatic control. It is also sometimes called the ”coil cooling
energy” (kWh) or ”coil cooling load” (kW).
2. SEC: the Summer Electricity Consumption for cooling; SEC = the electricity consumption of the
compressor of the cooling equipment (e.g. the chiller, package, etc).
3. SSEC: the System Summer Electric Consumption for cooling; SSEC = electricity consumption of the
whole system i.e. that of the: Fans + Pumps + Primary equipment.
4. WHL: “Winter Heat Loads" represent the heat demand of the building , the equivalent of SCL in winter
season.
5. WHG: “Winter Heat Gains" represent the heat generated by the equipment and effectively used (used in
substitution of normal heating mode in winter for reversible equipment); reversible heating has some
limitations and not all the load can be satisfied by the cooling equipment.
6. WED: the Winter Electricity Demand (Consumption) for reversible use of air-conditioning, including all
auxiliaries.
7. WEC: the Summer Electricity Consumption for reversible cooling; WEC = the electricity consumption
of the compressor of the cooling equipment used to heat (e.g. the chiller, package, etc).
8. SEER: Seasonal energy efficiency ratio during the summer: the ratio of SCL to SEC. This parameter
could be miscalculated due to the large preponderance of electricity consumption by auxiliary
applications in most systems; the same may be said about the SCOP (the seasonal coefficient of
performance during the winter for reversible systems;
9. SSEER: System Seasonal EER for summer: the ratio of SCL to SSEC, the real index of performance;
10. SCOP the Seasonal COP for winter: the ratio of WHG to WEC.
11. SSCOP the System Seasonal COP for winter: the ratio of WHL to WED, the equivalent of SSEER for
winter.
This study is concerned with identifying beneficial means to increase the SEER, SCOP, SSEER, SSCOP and
partly WHL. As far as the cooling loads are concerned this study is confined to an investigation of the ways
of decreasing the SCL that are related with equipment and control choices (i.e. addressing the family of “free
cooling” options). It is well known that with a suitable budget it is possible to deploy passive measures
which can provide an elevated degree of comfort without the use of AC; however, this is not the subject of
the present study.
Note that the use of the terms Summer and Winter in relation to AC comfort loads refers to the typical
seasonal period during which they apply and in some cases there will be parts of the building that have to be
cooled all year round. To give a clear definition of quantities we have called Winter the total of time periods
during which the heating demand is higher than the cooling demand and Summer the sum of all times where
the opposite occurs. The study does not address issues concerning the total heating load or the nature of the
back-up heating system because it is only concerned with that part of the heating which is provided by the
AC equipment.
59
Some countries, such as Switzerland, have targeted their CAC energy efficiency efforts purely on the
reduction of cooling loads through measures to improve the thermal performance of the building shell in
conjunction with electro-mechanical ventilation; however, this is made possible mainly because of the milder
climate. In some other countries, such as France and Portugal, a combination of policy measures have been
deployed that combine passive measures to increase comfort in unconditioned spaces (and thereby resulting
in a lower demand for AC) with those that are intended to raise the efficiency of AC equipment used in
conditioned spaces. The Mediterranean countries, because of their geographical position, have to rely in
many cases on air conditioning to attain desired comfort levels in the existing building stock.
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.
60
Manufacturers EUROVENT
directories Directories
National EUROVENT
confidential confidential
sales data sales data
Two database are permanently maintained by Eurovent at EU level. The first one is public: the directory of
certified products. The second one ("sales"), which is a data base on total units sold without reference to
performance, is confidential although some segments have been made available for use in this study, under
the condition that any data pertaining to a specific manufacturer is strictly anonymous. At the time of the
EECCAC study Eurovent had also assembled a temporary data base on the numbers of AC units sold in 2001
that included their efficiency although this was not made available for use in the study. There are also
directories and technical literature of some manufacturers that constitute by themselves a data base that can
be considered a good representation of the market, with the additional benefit of giving an idea of public
prices, at least in relative terms hence allow an indication of the relationship between cost and efficiency to
be established. A data base similar to the Eurovent sales database exists in some countries within the national
associations and some of these have been made available to the study.
We used some estimated ratios to correct for these issues when no better data were available. Obviously real
national data from country reports have replaced these ratios whenever they became available from the
countries with a national participant. Checks on the consistency of various sources of information have found
them to be reasonably high (i.e. with discrepancies of a few tens of percent only). Obtaining good market
figures for the year 1998 was not the real objective but rather to gather enough data that would allow
reasonable projections of the AC stock to be assessed. Projections on the size of the stock going back in
time were made using rates based on measured data from 1996 to 2000, and estimated data from prior to
1996.
It should be stressed that all percentages are based on weighed statistics, with no figure being an arithmetic
average. The statistics are intended to give the right weight to the country, climate, type of AC system, etc.
61
3. MAIN FIGURES OF AIR-CONDITIONING IN EUROPE
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)
120,00
100,00
80,00
Mm2
60,00
40,00
20,00
0,00
1975 1980 1985 1990 1995 2000 2005
Figure 3.2. Apparent annual additional building floor area conditioned by CAC from 1980 to 2000, by EU
Member State (apparent means additions and replacements)
30,00
25,00 Italy
M
m
Spain
2
ad 20,00
de
d
or
re
pl
ac 15,00
ed Germany
France
10,00
Others
UK
5,00 Greece
Portugal
0,00
1975 1980 1985 1990 1995 2000 2005
62
As a result of different growth, the relative weight of some countries like France or Germany as a proportion
of total installed AC within the EU, which was large in the 1980s has become small in the 1990s. Today just
two countries, Spain and Italy, account for more than 50% of the entire EU market Figure 3.3.
Figure 3.3. Apparent additional building floor area conditioned by CAC in 1998, by country
Others Germany
13% 11%
Greece
5%
UK
8%
Spain
24%
Italy
25%
Portugal France
2% 12%
It is also pertinent to consider what type of buildings AC is being installed in. Figure 3.4 shows the share of
conditioned floor area by type of tertiary activity and country for CAC systems alone.
Figure 3.4. Share of CAC installed by tertiary sector for six European countries
100
90
80
70
60 Others
50 Trade
40 Offices & work places
30 Hotels / restaurant / bar
20
Hospitals
10
0
default
France
Portugal
UK
Austria
Italy
However CAC is also in competition with RAC so it is relevant to examine the type of building where each
type of system are installed, (Figure 3.5).
63
Figure 3.5. Share of conditioned floor space by building type for each AC system type across the EU
100,00%
90,00%
80,00%
70,00%
60,00% education
50,00% houses
40,00%
30,00% trade
20,00% offices
10,00%
0,00% hotels&bar
Rooftops
RAC
VRF
chillers
packages&splits
hospitals
This market is centred on offices and trade. It is shared between CAC and RAC technique for economical
reason (compared price) but also because various building sizes lead to the choice of distinct solutions. The
only exception are VRF type systems, maybe due to their flexibility of use and installation, corresponding to
hotels, bar and existing medium office buildings.
Figure 3.6. Share of installed conditioned space by CAC system in the EU in 1998
VRF
Rooftops 3% Splits >12kW
7% 11%
Packages
8%
chillers
71%
64
Figure 3.7. Average annual rate of growth in conditioned floor area by type of CAC for the period 1996-2000
0,16
0,14
0,12
0,1
0,08
0,06
0,04
0,02
0
Large Chillers Packages Rooftops VRV Small
Splits +8,5% +2,5% +9% +13% A/C
+14% +10,5%
The average growth rate for large splits of 14%, for VRF of 13% and small AC for 10.5% are very different
from the overall average growth rate of 9%. The competition is focused on the "new" segment of smaller
buildings (trade, small offices, etc), which have correspondingly smaller average loads. Figure 3.8 illustrates
for instance the importance of decentralised AC solutions in the trade sector while over the longer term the
increased share of RAC sales within the total AC market corresponds to the same phenomenon. It should be
remembered that "Splits" refers to large split systems of over 12 kW in cooling capacity and that smaller
ones are included in the term "RAC". We see a growing competition of RAC against chiller based solutions
and an adaptation of solutions for the treatment of smaller sites.
Figures 3.8. The percentage of AC supplied by each AC type by user sector for the EU in 1998
120,00%
100,00%
80,00%
VRF
60,00%
Rooftops
40,00% packages&splits
chillers
20,00%
RAC
0,00%
offices
trade
education
hotels&bar
houses
hospitals
Figure 3.9 shows the growth in conditioned floor area by each type of CAC&RAC system from 1980 to 2000
across the EU.
65
Figure 3.9. Total conditioned floor area provided by each type of AC in the EU tertiary and industrial sectors
from 1980 to 2000
160
140
120
100
RAC<12 kW
VRF
M
m2 80 Rooftops
Packages
Splits >12kW
chillers
60
40
20
0
1980 1985 1990 1995 1998 2000
CAC systems based on chillers account for the majority of the CAC market, but among these there are
two dominant subsystems with market shares of the same order of magnitude: chiller systems using
AHU and those using FCU, Figure 3.10.
Figure 3.10.The share of chiller CAC systems (based on installed conditioned floor area) by sub-system type
across the EU for 1998
Two loops
2%
Two loops
Air (AHU)
Air (AHU)
Classic(FC 39%
U) Nat, Water
58% Nat, Water Classic(FCU)
1%
66
Figures 3.11. The share of non-residential conditioned building floor area cooled by each primary AC type in
the USA in 1999-2000
US A (E IA )
chillers
packages
all RA C
Figure 3.12. The share of non-residential conditioned building floor area cooled by each primary AC type in
the EU in 2000
E UR (E E CCA C)
chillers
packages
all RA C
However the US market is so large in absolute terms that for every segment there is more conditioned floor
area in the USA than in Europe, Figure 3.13.
67
Figure 3.13. Conditioned non-residential building floor area by AC type in the EU and in the USA in 2000
Mm2
9000
8000
7000
6000
all RAC
5000
packages
4000
chillers
3000
2000
1000
0
USA (EIA) EUR (EECCAC)
Figure 3.14. The share of chiller systems using water distribution systems (based on installed conditioned
area) in the EU
120%
100%
80%
60%
40%
20%
0%
ly
n
y
l
ria
m
ga
an
nd
ai
nc
ec
Ita
do
...
st
Sp
rtu
m
a
re
la
m
Au
ng
Fr
er
er
Po
G
iu
ki
h
G
lg
U
et
Be
68
Reversible use of Air Conditioning
One important aspect of this study is the reversible use of the cooling equipment for heating, Figure 3.15. On
a packaged or split product, it's easy to see if it is reversible (the owner may use the reversibility feature or
not) but for chillers tighter definitions are required. The statistics on chiller reversibility are derived from
data on a number of system sub-types:
water cooled chillers including water-to-water heat pumps
air-cooled chillers including condenserless water-cooled systems
air-to-water heat pumps with reverse cycle
water-to-air heat pumps with reverse cycle on a water loop
centrifugal chillers either hermetic or open type according to connection between the motor and the
compressor.
It has been assumed here that reversibility is a feature of 10% of the water cooled chillers and all the air-to-
water heat pumps. It is further assumed that the pure air cooled and the centrifugal chillers are none
reversible.
Figure 3.15. The share of conditioned non-residential building area provided by reversible CAC (for chillers
only and for all CAC) and by water-based (using FCUs) distribution systems for four EU countries in 1998.
100%
90%
80%
70% Spain
60% France
50%
40% Italy
30% UK
20%
10%
0%
Reversibility of Total reversibility %age of water
chillers systems
69
Figure 3.16. Market shares of AC technical solutions in four European countries (based on installed
conditioned floor area in non-residential buildings) in 1998.
RAC
VRV
UK
Rooftops Italy
Packages France
Spain
chillers
Large Splits
Chillers are predominant in France while RAC Italy is equally divided between RACs and chillers. Packages
have a comparatively large market share in Spain as do VRV systems in the UK. The average size (cooling
capacity) of chillers is smaller in Italy and the other Mediterranean countries, which corresponds to the
importance in the AC market of small trading enterprises and small offices compared with the larger tertiary
building complexes found in the UK, Figure 3.17.
Figure 3.17. Average cooling capacity of chillers in four EU countries, based on 1998 data
250
Cooling capacity (kW)
200
50
0
UK Spain France Italy
A number of "mini-chillers" with a small capacity are more popular solution in Italy than in other European
countries (Figure 3.18). As a result, the Italian market of chillers when expressed in terms of the number of
chillers sold is growing rapidly whereas some other national markets have risen smoothly or have even
stagnated. Competition between “local” systems, VRV and mini-chillers for the medium-size building
market is the dominant issue for the future.
70
Figure 3.18. Growth rates are high in countries with both small and large systems and smaller in countries
with only large hydronic systems (UK)
180,00
160,00
Index relative to 1996
140,00 France
120,00
100,00 UK
80,00
60,00 Italy
40,00
20,00
0,00
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
20
The value and nature of the European CAC market
According to the information gathered for this study, it appears that many manufacturers operate on an EU-
wide level. The largest are usually foreign owned companies, resulting from the fact that a number of the
countries where they originate have a large and mature internal market (e.g. Japan and the USA) which
results in a transfer of technology and experience to their European branches. This does not mean that these
local companies of foreign corporations have no technical autonomy, but it partly explains the operation of
the market.
The European CAC equipment industry is self-sufficient within Europe and is fairly concentrated although
less than the car industry. In terms of market share the manufacturers can be categorised into three broad
groups, Table 3.1.
Table 3.1. Market share of the "Top Ten" European chiller manufacturers
Even if the most significant European manufacturers belong to American groups, their centres of
development in Europe constitute a rather autonomous technical base of the European AC industry, which is
71
the third largest in the world after the USA and Asia. This base, joined to the existence of average-sized
manufacturers which are 100% European, ensures a great autonomy of supply for Europe. It results in the
independence and equality of the European manufacturers’ association, Eurovent, compared with its
American equivalent, ARI.
Table 3.2 shows the estimated value of the European CAC market without taking into account the imported
contents of the equipment (i.e. without considering the value of components imported from outside Europe
compared with those produced in Europe).
Table 3.2. Estimated value of the European CAC market by CAC equipment type (source: coordinator)
Segments % production % import for Balance net Capacity Value MEuros/ Estimated
EUR production MW/1998 Income (MEuros)
/MW
chillers 95% 5% 90% 6.8 0.25 1700
&CT
splits 90% 5% 85% 1.1 0.2 200
> 12kW
Packages 90% 10% 80% 0.9 0.2 200
rooftops 95% 5% 90% 0.7 0.15 100
VRF 50% 10% 40% 1.25 1 250
FCU/AHU 100% 0% 100% 6.8 0.1 650
RAC 75% 0% 75% 5.0 0.2 1000
Total 87.7% - 84.2% 21.6 - 4100
weighted
in income
These figures do not include the value of installation but do include the profit margin of the equipment
suppliers. The US and Japanese markets are worth about 10 000 MEuros per annum on the same basis.
Other stakeholders
Installers, designers and operators all have to adapt to the customer demands. They have to display a
competitive cost, or be able to guarantee a high reliability (better servicing, better contracts) in order to
compete. There is almost no freedom for installers and designers to be rewarded for the extra energy
efficiency of the systems they may promote although operators can be reimbursed through performance
contracting.
Utilities are important stakeholders. Summer peaking may be a problem for some Southern European utilities
but is often seen as a market opportunity for Northern European utilities.
Governmental agencies and ministries are responsible for the development of building codes. Thermal
insulation, which is often introduced into building codes to limit heating requirements, very often also helps
lower cooling needs; however, in some cases increased insulation can aggravate summer discomfort and
increase the need for AC.
Building thermal regulations usually aim to minimise AC energy demand but often "don't know how” it
should be done. There is a hesitation between a pure prescription on some elements (an obligation of means)
and a global limitation of demand, leaving the designer free to choose the elements and to assemble them to
reach the target (the obligation of results). The problem arises from the lack of energy consumption
calculation methods that are applicable to a wide range of systems. European countries cope with this
problem in different ways, but nobody appears to be happy with their current regulations.
Extension to EU accession states of the CAC market is already a reality. An indication of the problems and
opportunities of CAC in the EU Accession States has been gained through a detailed study of the situation in
72
Romania. This has given some insights into how the findings may be applicable in the rest of the CEEC. The
methodology applied regarding the creation of national CAC stock statistics from an analysis of export and
import figures can be applied in other CEEC countries in the same way as for the EU countries and be used
to project CAC energy consumption and identify nationally specific issues.
EER as a function of capacity and cooling medium for a chiller under 750 kW
Figure 3.19 shows the EER as a function of capacity for chillers on the EU market according to their mode of
condensation.
Figure 3.19. Chiller EER as a function of cooling capacity for 1998. There are two groups of chillers, with
distinct testing conditions (water cooled and air cooled, that cannot be compared)
4.5
4.0
R2 = 0.0073
3.5
3.0
2.5
R2 = 0.0003
EER
2.0
1.5
1.0
air cooled
water cooled
0.5
Linéaire (air cooled)
Linéaire (water cooled)
0.0
0 100 200 300 400 500 600 700 800 900 1000
Puissance frigorifique kW
73
Tables 3.3 and 3.4 indicate the average and range of EER values for CAC systems found on the European
market.
Table 3.3. Average and extreme EER values for chillers on the EU market, split according to Eurovent
internal categories, for year 1998
Table 3.4. Summary of average and extreme EER values by chiller category on the EU market
EER
Categories Type Condenser Application min ave max
Complete unit cooling air conditioning 1.9 2.53 3.29
reversible air conditioning 1.9 2.48 2.96
Floor 3.31 3.34 3.39
cooling water conditioning 2.9 3.73 4.09
reversible water conditioning 2.9 3.57 4.09
Condenserless cooling water conditioning 2.76 3.21 3.69
Statistically there is no relationship between chiller EER and its cooling capacity; however, on average there
is significantly higher EER for chillers which are cooled with water compared with those that use air. In fact
this improvement is not inherent to the chiller, but rather represents the temperature regime found in cooling
towers. The values used for testing the two types of system are somehow arbitrary and it may be that the
74
apparent benefit from water cooling is not fully realised in practice. Based on the standard test data the
average EER for water-cooled chillers is 3.57 W/W whereas for air-cooled chillers it is 2.52 W/W.
Nevertheless, water-cooled systems are relatively expensive (because of the additional cost of using either a
cooling tower or of accessing a natural water supply) and will therefore only tend to found for larger capacity
systems. Interestingly, the average EER of the reversible systems is almost the same as for the cooling-only
systems .
Figure 3.5. Distribution of EER/EERave (where the average EER is average for the same product category)
for chillers on the EU market in 1998
25
21.06
20.34
20
15
% 13.47
13.04
of models 11.6
11.17
10
5.3
5
2.44
1.58
0
75-80 81-85 86-90 91-95 96-100 101-105 106-110 111-120 121-130
EER/EER aver
It is interesting to consider to what extent this difference can be attributed to differences in the type of chiller
compressor used. From a total of 698 chillers in the Eurovent database the type of compressor was known for
304 of them. The following compressor types can be distinguished:
- "scroll" (orbital)
- "screw"
- "reciprocating" (i.e. with pistons)
Table 3.6 shows the proportion of chillers by compressor type within this subset of models as a function of
the condensing medium. Table 3.7 gives additional data that enable a comparison of chiller efficiency as a
function of compressor type and condensing medium.
The average performance of chillers with air-cooled condensers is almost independent of the compressor
type at ~2.5 W/W. The only significant performance difference on overall averages is seen for the chillers
with water-cooled condensers that use screw compressors who have an average EER of ~3.9 W/W compared
to ~3.5 W/W for those using scroll or reciprocating compressors. We note also that the best air cooled
chillers (the top runner, not the best on average) are the ones with scroll, then reciprocating, then screw.
75
Table 3.6. partial statistics (not from Eurovent) on compressor type
Table 3.7. Comparison of chiller full-load nominal performance values depending on the type of compressor
and type of condensing medium
Type Cond. number Min.kW ave.kW Max.kW Min EER ave EER Max EER
Scroll air 178 12.2 49.8 158 1.9 2.5 3.39
wat 24 13.7 64 163.6 3.11 3.51 4
Screw air 14 196.1 451.3 789.1 2.35 2.5 2.66
wat 36 132 478.6 920.7 3.65 3.91 4.09
Reciprocating air 26 24.2 163.5 350 2.16 2.52 2.74
wat 26 136 407.2 847 2.99 3.54 4.06
It is also interesting to consider to what extent chiller EER depends on the choice of refrigerant. The data
shown in Table 3.8 suggests there is a small variation but perhaps not as much as had been expected.
Table 3.8. Chiller energy efficiency at full-load as a function of the type of refrigerant used
capacity EER
Refrigerant Min. Ave. Max. Min. Ave. Max.
R22 12.1 143.03 921 1.9 2.79 4.09
R407C 12.5 106.125 782.28 2.1 2.68 4.06
R22/R407C 12.2 62.68 163.6 1.9 2.6 4
Reciprocating compressors were prevalent for small cooling capacity systems but tend to give way to orbital
(scroll) compressors at medium cooling capacities, (because of multiple advantages: lower noise, wear, etc.),
and to screw compressors for the larger capacities.
76
4. FACTORS GOVERNING THE DESIGN, SELECTION, INSTALLATION AND
OPERATION OF CAC SYSTEMS
4.1 Actors involved with CAC systems
The main barriers to efficiency
As opposed to RAC, CAC are usually designed and specified by a chain of engineers or technicians, who
define the system assembly for a given building without the direct influence of the customer. Everybody
could gain something from the marketing of efficient CAC, but each actor has a limited vision of the chain:
¾ the consideration of initial cost as the only decision criterion by most designers and installers,
because this is almost the only way the customer judges them ;
¾ the separation between the plant owner and the renter, between the renter and the operator, etc. ,
nobody being ready to pay for the other’s benefit ;
¾ the customer only judges the initial cost because he is not aware of the other aspects (no coded
information on other aspects) or not interested (owner and occupant having distinct interests);
¾ the competition between manufacturers is only expressed in terms of Euro/kW, not Euro/EER or
Euro/kWh consumed later ;
¾ the absence and/or intrinsic difficulty of developing building codes for this relatively new source of
energy consumption which is very complex to model and characterise;
¾ the problem for consultants of completely specifying and certifying the quality of something so
complex, built on site and only once,
¾ the lack of incentives for energy efficiency for most system operators, except in case of a good EPC
(Energy Performance Contract),
As a result of these complex factors the measures to be proposed in EECCAC cannot only address the
efficiency of the CAC equipment supplied to the market, which in many ways is the smallest problem, but
should also aim to reshape and activate the chain leading to the final service: the conditioned square meter.
Hence the necessity to develop a better description of the chain.
77
standardisation body. The technical options related with sizing have to be defined in such a way as to leave
degrees of freedom to the designer that will bear this responsibility.
78
Figure 4.1 Label stamped on Eurovent certified appliances.
79
Manufacturers who participate in the EdF scheme must supply Eurovent with more test-point data than the
minimum which is required to be quoted in the Eurovent directory (which is currently the T3 test results in
the heating mode and the T1 test results in the air-conditioning mode). EdF and Eurovent have an agreement
wherein Euroevent will do additional certification testing for the RACs that are in the EdF programme.
Supplying the data to Eurovent means that it can be independently verified, through their certification
process, before being used by EdF. EdF also obliges products promoted within the directory to attain a
minimum energy efficiency value under each test condition. An example of these requirements are given in
Table 4.1. but there are many sets of requirements for the various equipment types. All values used are
Eurovent certified values, subject to independent testing.
Table 4.1. Minimum energy performance requirements for reversible heat pumps (or chillers) of the air /water
type with a capacity higher than 30 kW (Application in fan coil or hybrid systems (FC or radiators + h&cfloor))
that are included in EdF’s promotion
Cooling 35 ** 12 7 EER>2,2
mode
Heating 7 6 40 45 COP>2,5
mode
-7 -8 * 45 COP>1,5
* temperature function of the flow and identical to the one in heating mode at à +7°C outside
** non-controlled
The commercial importance of being included in EdF’s scheme has resulted in these efficiency requirements
having a substantial impact on the efficiency of the market in France.
80
4.4Existing national regulations within the EU (which apply at the system level)
In general building thermal regulations "want" to reduce the energy demand associated with air conditioning
but usually they "don't know how”. There is a hesitation between a pure prescription on some elements
(obligation of means) and a global limitation of demand, leaving the designer free to choose the elements and
to assemble them to reach the target (obligation of results). The problem arises from the lack of consumption
calculation methods applicable to a wide range of systems. Different countries cope with this problem in
different ways.
Among EU Member States only Portugal and the UK have significant measures in their building regulations
designed to limit the energy consumption of air conditioning systems. These are described in detail below.
81
Table 4.2. minimum number of stages required in the RSECE
Heating Cooling
Power Nº Stages Power Nº Stages
<100 kW 1 <40 kW 1
100 to 1000 kW 2 40 to 200 kW 2
1001 to 4000 4 201 to 500 4
>4000 kW 6 >500 kW 6
The 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.
The general guidance is that naturally ventilated spaces should not overheat and that cooled spaces
should not require excessive cooling plant capacity. This may be satisfied by limiting glazing area,
providing adequate shading, or designing for night cooling operation.
For spaces with glazing facing only one orientation, the requirement will be satisfied by limiting glazing
area to a percentage of the internal area:
North 50%
NE/NW/S 40%
E/SE/W/SW 32%
82
Horizontal 12%
Alternatively, it is acceptable to show either that:
- the solar heat load per unit floor area averaged between the hours of 7:30 and 17:30 would not be
greater than 25 W/m2 with the solar irradiances for the location that are not exceeded more than 2.5% of
occasions for July (between 1976 and 1995)
- showing by (acceptable) calculation that, in the absence of mechanical cooling or mechanical
ventilation, the space will not overheat when subjected to an internal heat gain of 10 W/m2
Elemental Method: Heating efficiencies
The carbon intensity of the heat generating equipment at maximum output and 30% (system) output
should be no higher than:
Maximum output 30% output
Natural gas 0.068 0.065
Other fuels 0.091 0.088
(all figures in kgC/kWh)
However, these figures may be exceeded if building insulation levels are increased beyond the minima.
Carbon emission factors for different fuels are tabulated: significant figures are
natural gas 0.053
oil 0.074
electricity 0.113
In effect, these figures taken together define the minimum CoP that is required for heating by reverse-
cycle operation of air-conditioning
Elemental Method: Air conditioning efficiency.
For offices, air conditioning systems should have a satisfactory "Carbon Performance Rating" (CPR).
Notionally, this is a limit on the carbon emissions per m2 of floor area from air conditioning or
mechanical ventilation systems under standard operating conditions. In practice, it operates as a limit on
the installed cooling power and fan power per m2 of floor area, since the calculation procedure
prescribes standard figures for equivalent hours of full load operation. Benefit can be claimed for a
number of control and other features.
The maximum allowable ratings (all in kgC/m2/year) for new installations are:
New building Existing building
Air conditioning 10.3 11.35
Mechanical ventilation 6.5 7.35
For substantial modifications to existing systems, the performance must be the least demanding of either
these values or a 10% improvement on the original value.
Key calculation parameters:
Equivalent hours of full load operation:
83
- mechanical ventilation or air conditioning fans 3700 hours per year
- cooling plant 1000 hours per year
The calculated CPR may be reduced by applying factors to the fan or cooling capacity to reflect energy-
saving design features. These factors depend on the level of plant monitoring provided. They are
multiplicative.
Column C figures apply if no plant monitoring is provided; column B when any of: energy metering,
run-hour metering, internal zone temperature monitoring are provided; column A when, in addition, the
monitoring system has the ability to draw attention to "out of range" values.
Air distribution systems:
A B C
Operation in mixed mode with natural ventilation 0.85 0.90 0.95
Controls which restrict the hours of operation of the system 0.90 0.93 0.95
Efficient means of controlling air flow rate 0.75 0.85 0.95
"Mixed mode" requires the provision of sufficient opening windows and an interlock to prevent air-
conditioning operating when windows are open. It is only permissible when the perimeter zone of the
space accounts for more than 80% of the floor area.
"Air flow rate control" requires variable speed drives or variable pitch fan blades: throttling or inlet
guide vanes do not qualify for the allowance.
Refrigeration plant:
A B C
Free cooling from cooling tower 0.90 0.93 0.95
Variation of fresh air using economy cycle or mixed mode 0.85 0.90 0.95
Controls which restrict the hours of operation of the system 0.85 0.90 0.95
Controls which prevent simultaneous heating and cooling in the same zone 0.90 0.93 0.95
Efficient control of plant capacity, including modular plant 0.90 0.93 0.95
Partial ice storage 1.80 1.86 1.90
Full ice thermal storage 0.90 0.93 0.95
"Efficient control of capacity" requires good part load efficiency (without defining this): hot gas bypass
does not qualify.
"Full ice storage" requires chillers to operate only at night: if day and night chiller operation is intended,
this is "partial ice storage".
For buildings other than offices, there are no explicit requirements for air conditioning system
efficiency. This causes some difficulties when applying the Carbon Emissions Calculation Method (see
later) and guidance on the calculation of SSEER is being prepared.
However, there is a general requirement that components such as fans, pumps and refrigeration
equipment are reasonably efficient and appropriately sized, so there is scope to introduce performance
requirements.
84
There are requirements for SPF, Specific Fan Power, the electricity demand (of motor rating) per
airflow unit (expressed as W/litre/second), and encouragement of variable flow control.
For ACMV systems in new buildings, the SPF should not exceed 2.0, and preferably 1.5, and for new
systems in existing buildings or substantial alterations to existing systems should not exceed 3.0. There
are exceptions for non-comfort applications
Whole-building method
For offices, as an alternative to the air-conditioning CPR, a similar calculation may be calculated for the
combined emissions from the heating, lighting and air-conditioning systems. This provides greater
building services design flexibility, but the building envelope requirements of the elemental method
must also be complied with.
The required values of whole-building CPR are:
Building type New office refurbished office
Naturally ventilated 7.1 7.8
Mechanically ventilated 10.0 11.0
Air-conditioned 18.5 20.4
Calculation procedures are explained in BRE Digest 457.
Carbon Emissions Calculation Method
This route permits the greatest design flexibility. It requires the designer to show that the carbon
emissions from the proposed building are not greater than those for a notional building of the same
shape and size, designed to comply with the Elemental Method. There are, however, still some
constraints on acceptable values for some parameters - for example, air leakage of the building.
There is no prescriptive list of acceptable calculation methods, but a completed copy of Appendix B of
CIBSE AM11 "Building Energy and Environmental Modelling" is an acceptable demonstration.
Although this route is expected to be used only for a few, probably high-profile, buildings, it has already
generated some criticism. These relate to its implementation, rather than the principle. The main
problems are:
- the designer is required to design two buildings and their systems in order to demonstrate compliance
- some design parameters (notably air-conditioning system efficiency) are not defined in the elemental
method but are required to enable the comparison to be carried out.
Guidance on both these points, in the form of a simplified calculation for the notional building, has been
prepared (CIBSE,2004). Of particular relevance to the EECCAC study is the need to make explicit
assumptions about - or calculations of - about the seasonal system energy efficiency ratio (and seasonal
system coefficient of performance).
Information Provision: metering and logbooks
The Regulations require that the owner and/or occupier of the building be provided with a logbook that
contains, amongst other things, the design assessment for CPR or other benchmarks, commissioning
details, operating instructions, and details of all meters provided. A recommended template for this has
been published by CIBSE.
There is also a requirement that sub-metering be provided. This includes separate metering for tenancies
of more than 500 m2 (though for tenancies below 2500 m2, proportioning of cooling may be
85
acceptable). Generally, any chiller installation (which may include more than one chiller) of greater than
20 kW input power should be separately metered, and any motor control centre providing power to fans
and pumps of more than 10 kW input power.
86
application of requirements for boiler inspection, air conditioning system inspection and building
certification (see below) may not be delayed beyond January 2009.
The Directive aims at reducing the consumption of energy in new and renovated buildings, excluding
industrial buildings, through the following:
A) Establishment of a general framework and common methodology for calculating the integrated
energy performance of buildings.
B) The development and application of minimum energy performance standards to new buildings and to
certain existing buildings when they are renovated.
C) Certification schemes for new and existing buildings on the basis of the above standards and public
display of energy performance certificates and recommended indoor temperatures and other relevant
climatic factors in public buildings and buildings frequented by the public.
D) Specific inspection and assessment of boilers and heating/cooling installations.
All these measures are to be taken nationally, before January,4 2004. The field of application of (D) for
CAC i.e. > 12 kW is exactly that of this study. Buildings of more 1000 m2 in floor area are to
addressed in A, B and C. The method of calculation to be established in (A) is the responsibility of each
Member State and thus is not unified across Member States. It should include the energy consumption
of AC and ventilation. There could be obvious advantages in harmonising such national building codes,
namely in the very technical field of CAC. The measures (B) and (C) envisaged in Article (6) may
require the modification of the insulation, lighting and ventilation requirements existing in some
countries and thus have an indirect influence on air-conditioning. Indeed, heating remains the essential
concern of EU building codes, even after the harmonisation.
Article 8 requires central air-conditioning systems over 12 kW to be regularly inspected. Article 9
requires Member States to put in place a system that ensures that certification of buildings and
inspection of equipment are carried out by qualified and independent personnel. An Annex to the
proposal contains the main aspects to be taken into account when calculating the energy performance of
buildings and requirements for inspection of boilers and central air conditioning systems. It also creates
an EU-wide technical committee comprised of representatives from Member States that will be
responsible for the development and maintenance of the inspection rules.
The draft Framework Directive for “Eco-design of End-Use Equipment” (to be adopted)
The European Commission has developed a draft proposal for a new Directive, which amongst other
measures would give the Commission the right to establish mandatory minimum energy performance
standards (MEPS) for end-use equipment. The annex of the Directive stipulates that the level of energy
efficiency used in the standards will be set aiming at the least life cycle cost for the final users using a
real discount rate of 5% and realistic assumptions about product lifetime. The determination of this is to
be based on the results of a technical-economical analysis. As yet there is no clear time line regarding
when this draft will be submitted to the council of ministers and parliament for approval.
87
Amended Rational Planning Techniques Directive proposal from March 1997.
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.
88
(a) minimum energy-efficiency requirements for the design and construction of; 1. New buildings and their
systems, 2. New portions of buildings and their systems, and 3. New systems and equipment in existing
buildings.
(b) criteria for determining compliance with these requirements.
The provisions of the standard apply to:
(a) the envelope of buildings provided that the enclosed spaces are: 1. heated by a heating system whose
output capacity is greater than or equal to 3.4 Btu//h*ft2 (10W/m2), or 2. cooled by a cooling system whose
sensible output capacity is greater than or equal to 5 Btu/h * ft-2 (15 W/m2);
(b) the following systems and equipment used in conjunction with buildings: 1. heating, ventilating, and air
conditioning, 2. service water heating, 3. electric power distribution and metering provisions, 4. electric
motors and belt drives, and 5. lighting.
Moreover, Standard 90.1 focuses on comfort conditioning rather than industrial, manufacturing, or
commercial processes. Note, too, that the stated purpose of the standard is to provide minimum
requirements; a designer or owner can always exceed these basic conditions for compliance.
The latest version of Standard 90.1 which was issued in 1999 has several differences from the previous 1989
version. It has been reorganised for ease of use, such that the new standard clarifies requirements and
provides a simplified compliance path for small commercial buildings. More importantly, the 1999 edition
expands the standard's scope to include both new and existing buildings and building systems. For alterations
and additions, the 90.1 User's Manual notes that, “In general, the Standard only applies to building systems
and equipment…that are being replaced.” A life-cycle-cost analysis was used to define the criteria in the
1999 edition and thereby balance energy efficiency with economic reality.
Standard 90.1—1999 addresses building components and systems that affect energy usage. The technical
sections of the standard, Sections 5 through 10, specifically address components of the building envelope,
HVAC systems and equipment, service water heating, power, lighting, and motors. Each technical section
contains general requirements and mandatory provisions; some sections also include prescriptive and
performance requirements.
To comply with Standard 90.1—1999, the prospective design must first satisfy the general requirements and
mandatory provisions of each technical section. But that's not all. The design must either (a) fulfil additional
prescriptive and performance requirements described in each technical section or (b) satisfy the energy cost
budget (ECB) method.
The ECB method permits tradeoffs between building systems (lighting and fenestration, for example) if the
annual energy cost estimated for the proposed design does not exceed the annual energy cost of a base design
that fulfils the prescriptive requirements. Using the ECB approach requires simulation software that can
analyse building energy consumption and model the energy features of the proposed design. Standard 90.1
sets minimum requirements for the simulation software. Suitable programs include BLAST, DOE-2, and
TRACE™.
89
3. A Simplified Approach option, which consists of a subset of all mandatory provisions and prescriptive
requirements
For small buildings, the “simplified approach” consolidates the provisions on roughly two pages so that
design professionals can quickly locate all applicable requirements. The difference lies in ease of use and the
degree of flexibility allowed. Eligibility for this approach requires that the building occupy less than 25000
sq ft of gross floor area and not more than two stories. Another prerequisite (there are others) is that each air-
cooled or evaporatively-cooled HVAC system serves only one zone.
Table 4.3. Summary of the revised of ASHRAE 90.1 energy performance requirements
90
The standard also addresses balancing for air systems larger than 1 hp and for hydronic systems larger than
10 hp. It further requires control elements to be calibrated, adjusted and in proper working condition for
buildings that exceed 50000 sq ft.
91
speed motors, or variable-speed drives. If the cooling tower has three cells, at least two of them must be
equipped with speed control.
Energy recovery. Systems larger than 5000 cfm that bring in lots of outdoor air (at least 70 percent of design
airflow) must include energy recovery; the means of recovery must be at least 50-percent effective. This
proviso will probably lead to the increased use of energy recovery in air handlers dedicated to ventilation,
particularly in retrofit applications in which ventilation airflow is brought into compliance with
ANSI/ASHRAE 62.1.
Exceptions to this airside requirement include (but are not limited to) series-style energy recovery and
systems in which the largest exhaust air stream is less than 75 percent of design outdoor airflow.
Heat recovery for service water heating is required in facilities that operate 24 hours a day, where the heat
rejection capacity exceeds 6 million Btu/h, and where the service-water heating load exceeds 1 million Btu/h.
Table 4.4. Percentage Change in Whole-Building Energy Use Intensity (EUI) and Dollars Use Intensity ($UI) through application of
ASHRAE/IESNA 90.1—1999
Building Type Floor Area Electricity Gas Site EUI Source EUI $UI (USD)
Weighting
Assembly 0.068 9.5% -5.3% 4.4% 7.2% 7.5
92
Education 0.218 11.4% -6.3% 5.2% 8.6% 9.0
Food 0.027 -1.2% 1.7% -0.4% -0.8% -0.9
Lodging 0.079 0.2% -6.5% -1.7% -0.6% -0.5
Office 0.190 10.6% -12.7% 8.2% 9.7% 9.8
Retail 0.246 15.7% -30.7% 12.7% 14.7% 14.9
Warehouse 0.173 -71.6% -11.3% -45.1% -58.7% -59.7
National 1.000 7.3% -8.6% 3.9% 5.9% 6.2
Table 4.5. Minimum energy performance requirements for packaged air conditioners with a cooling capacity
between 7.5 and 65 kW in Australia
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
93
These requirements apply to all unitary (i.e. packaged) AC equipment within the specified cooling capacity
range and hence applies to large room AC units, multi-splits, VRF units and classical packaged systems
(rooftops and cabinets). Japanese test conditions for packaged AC units are mostly compatible with ISO and
European test standards. As yet there are no measures for chillers and there are no labelling requirements for
this kind of AC equipment.
Korea
Korea only has MEPS in place for unitary split-packaged AC units of between 10 and 17.5 kW in cooling
capacity. These are required to attain a mandatory minimum EER of 2.25 W/W, but in addition the
government expects manufacturers to attain a minimum sales-weighted efficiency level of 2.93 W/W. As yet
there are no measures for larger packaged units or chillers and there are no labelling requirements for this
kind of AC equipment.
Taiwan
Taiwan has implemented the following MEPS for chillers since January 1st 2003 (Table 4.7). Taiwanese test
conditions for chiller units are compatible with ISO and European test standards.
4.6 Choices and measures which could increase the efficiency of CAC systems
Measures which could increase globally the efficiency of CAC
Among the dominant types of CAC, some are known for their better Energy Efficiency. One way of
improving EE is to promote some specific system types. We have discarded such an option, but we have
assembled the information about the performance difference among the dominant 18 types :
N° Type Observation Terminal Comfort System
equipment level
1 CAC Air cooled chillers FCU+1air TC Air Cooled with water distribution
2 CAC Air cooled chillers CAV/AHU TC Air Cooled with air distribution
3 CAC Air cooled chillers CAV/AHU TAC Air Cooled with air +humidity control
4 CAC Cooling towers FCU+1air TC Water Cooled + water dist.(cooling)
5 CAC Cooling towers CAV/AHU TC Water Cooled with air dist.(cooling)
6 CAC Cooling towers CAV/AHU TAC Water Cooled +air +hum.(cooling)
7 CAC Natural Water FCU+1air TC Outside water + water dist
8 CAC Natural Water CAV/AHU TC Outside water + air dist
9 CAC Natural Water CAV/AHU TAC Outside water +air +hum
94
10 CAC Natural Water FCU+1air TC TWO LOOPS + CHILLER
11 CAC Natural Water Package TC VRF
12 CAC Natural Water Package TC PACK&SPlarge
13 CAC Natural Water Package TC Roof tops
14 CAC Water cooled RAC+1air TC RACs on one loop
RAC
15 RAC Air cooled RAC RAC+1air TC Multi Splits
16 RAC Air cooled RAC RAC+1air TC Splits
17 RAC Air cooled RAC RAC+1air TC Small packages
18 RAC Air cooled RAC RAC+1air TC Single Ducts
Primary air (1air) has been added here to each non air based system, for equality of comfort. In fact it is only
installed in some situations.
95
D&S7 Regulation is expressed in terms of carbon intensity, giving to electricity a weight close to its GW
impact
D&S8 Careful organisation of set points and control dead bands since design
D&S9 Automatic adjustment of pressures planned from design
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)
96
AS1 Central Free cooling
AS2 Smaller % of outside air
AS3 VAV
AS4 Better fans in AHU
AS5 Application of Eurovent specifications for AHUs (less leakage, more insulation)
AS6 Cost effective AHU
AS7 Optimised blowing temperature (10-16 °C)
AS8 Quality of the moisture control system Seems critical for energy consumption
AS9 Correction of the poor multi-zone efficiency of Air Systems
AS10 Sensor of occupancy and other "demand controlled" ventilation
AS11 Central heat/cold Recovery within the HVAC system
AS12 "Displacement" strategy by use of stratification of the rooms (low inlet speeds) or by other
displacement strategies
AS13 The air flow follows the hygienic demand and has not a minimum value over the minimal hygienic
demand
AS14 Prohibition and successive cooling & reheating
AS15 Ducts insulation and leakage limitation
AS16 Existence of an A/C stopping & controlling possibility in each zone
AS17 Ventilation should be in cascade among rooms
AS18 Reversibility by use of the same chiller as a heat pump
AS19 Recovery of heat for DHW
97
E1 District or block cooling
E2 Absorption or mixed strategies
E3 Cool storage
E4 Use of condenser to heat DHW
E5 "double dividend strategies" based on a higher efficiency of office equipment and lighting leading to
lower AC loads
E6 Evaporative cooling
E7 Dessicant cooling
E8 Natural cooling from cooling tower
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).
98
The gains reported in the previous types will only be achieved over the long term therefore the maintenance
or improvement of performance, by technical measures or contractual means (such as Energy Performance
Contracting) or by periodic audit, is an interesting family of options.
99
5. PROJECTIONS TO YEARS 2010 AND 2020 (BAU SCENARIO)
Table 5.1. AC saturation coverage levels used in the EECCAC study (cooled floor area (m2)/total building
floor area (m2))
CAC extrapolation is performed with an empirical function giving the yearly market in terms of the ratio:
x = Stock at present time /stock at time(infinity)
This choice makes it possible to represent the takeoff and the saturation of the market by one single equation.
In fact the AC saturation levels are very different from one sector to another: for example, in Austria the
hospitals are air-conditioned but the houses are not, etc. However there is not enough data to model each
sector. The residential sector, which was already treated in the EERAC study, has not been remodelled here
and the other economic sectors (hospitals, hotels & bars, offices, trade, houses, education) are modelled as if
they followed the same "learning curve", even if the starting point and the final saturation level (at time
infinity) are different from one sector to another. The distribution of AC by sector is estimated by the
reconciliation of the projected data with the national data and then applying time invariant AC stock sharing
coefficients by sector and by AC type.
For the past the model extrapolates back to 1970 to generate stocks for the periods 1970-75, 1975-80, etc...
which will be renewed 15 years later when the AC equipment is removed. The overall growth in the size of
the stock is the difference between the apparent market and level of renewal.
100
The stocks are obtained by integration of the apparent markets extrapolated (into the past) and by simulation
and integration of the annual true markets (apparent market minus renewal of existing AC) into the future.
Figure 5.1. Evolution of the total cooled floor area in Europe from 1985 to 2020
2500
2000
1500
1000
500
0
1985 1990 1995 2000 2005 2010 2015 2020 2025
The AC stock values can be expressed in many ways, for instance in terms of the cooled area (m2) per
inhabitant as in Figure 5.2.
Figure 5.2. Cooled area per European in 2000 by Member State and for the EU as a whole
m2/inhabitant
7
0
B DK D EL E F IRL I L NL A P FIN S UK EU-15
The RAC stock figures obtained by this method differ from those given in the EERAC study except for the
residential sector where the EERAC figures are used (in terms of kW not m2). The estimated cooled area per
European in the future is shown in Figure 5.3.
101
Figure 5.3. Cooled area per European in 2020 per Member State and for the EU as a whole
m2/inhabitant
18
16
14
12
10
0
B DK D EL E F IRL I L NL A P FIN S UK EU-15
The cooled area "per European" is projected to rise from 3 to 6 square meters over the next 20 years.
Figure 5.4. Evolution of cooled-floor area from1985 to 2020 at the national level
Mm2 cooled
600
500
400
Spain
Italy
France
300
Germany
Greece
Portugal
200
100
0
1985 1990 1995 2000 2005 2010 2015 2020 2025
Years
We propose to the reader Table 5.2 with the most important national values, because this may be a useful
tool in some national frameworks. However we have to comment accounting methods. The stock has been
estimated through national surveys and by integration over time (with due replacement rates) of a few
Eurovent market data available to us. All conditioned areas hereunder are « standardised » areas,
corresponding to the typical European sizing ratio (120 W/m2 for CAC, 240 W/m2 for RAC). No country
has exact statistics of conditioned areas, but some are close to it. The values hereunder cannot be compared
directly to such « national » statistics for two reasons:
1- the sizing ratios vary according to climate, national habits, sector, etc. (this has been partly corrected for)
102
2- which area is conditioned when a building is air conditioned is usually uncertain : the gross area of the
building? Certainly less! The strict area of activity rooms? Certainly more! (lobbies, adjacent rooms, etc.); so
both our “standardised” figures and the national figures have a margin of uncertainty and should be
compared with caution.
Table 5.2 Area conditioned in each country and year (such areas can be compared with national statistics)
Years
Country 1990 1995 2000 2005 2010 2015 2020
AU Mm2 cooling 12,01 15,68 20,06 26,29 30,29 33,01 33,95
Mm² reverse 1,45 2,06 2,74 4,83 5,57 6,08 6,27
BE Mm2 cooling 4,03 8,98 20,36 32,41 42,77 52,09 54,29
Mm² reverse 0,84 1,84 4,03 6,46 8,43 10,24 10,73
DE Mm2 cooling 3,78 6,62 11,30 19,92 29,24 37,57 42,30
Mm² reverse 0,70 1,35 2,50 4,12 6,01 7,72 8,69
FI Mm2 cooling 15,88 24,06 36,43 43,28 47,28 50,19 50,99
Mm² reverse 1,35 2,28 3,71 7,49 8,21 8,74 8,89
FR Mm2 cooling 93,40 129,39 180,37 293,24 390,57 472,24 502,39
Mm² reverse 32,79 45,84 64,98 106,59 141,52 171,24 182,61
GE Mm2 cooling 34,07 66,29 127,64 216,74 298,51 365,63 400,13
Mm² reverse 4,88 9,54 18,81 30,61 41,65 51,09 56,23
GR Mm2 cooling 11,04 23,06 48,23 80,47 108,97 140,88 145,99
Mm² reverse 5,29 11,17 23,65 40,07 54,24 70,12 72,68
IR Mm2 cooling 5,03 6,81 9,37 13,84 17,07 19,39 20,37
Mm² reverse 0,75 1,08 1,78 2,30 2,83 3,22 3,41
IT Mm2 cooling 130,85 175,63 258,76 368,74 414,88 450,33 467,85
Mm² reverse 29,22 43,81 73,26 106,86 120,93 132,38 138,18
LU Mm2 cooling 0,25 0,43 0,87 1,34 1,76 2,07 2,20
Mm² reverse 0,07 0,10 0,17 0,26 0,35 0,40 0,43
NE Mm2 cooling 22,25 39,02 66,88 87,71 101,28 110,49 113,62
Mm² reverse 1,84 3,55 6,50 12,17 14,03 15,38 15,89
PO Mm2 cooling 8,46 12,51 18,73 34,84 52,08 68,41 78,27
Mm² reverse 4,67 7,27 11,25 18,47 27,53 36,11 41,31
SP Mm2 cooling 64,24 102,68 172,69 248,07 295,71 342,20 352,20
Mm² reverse 34,61 56,66 97,11 136,02 161,33 186,01 191,57
SW Mm2 cooling 38,41 53,26 69,38 78,17 83,23 87,28 88,21
Mm² reverse 4,08 6,14 8,74 14,90 15,88 16,68 16,92
UK Mm2 cooling 94,29 127,63 173,15 248,36 294,19 326,80 340,28
Mm² reverse 14,17 20,41 31,06 43,81 51,73 57,87 61,07
Total Mm² cooling 538,01 792,07 1214,23 1793,42 2207,83 2558,59 2693,04
Total Mm² reverse 136,71 213,10 350,28 534,96 660,23 773,29 814,88
The differences are very small on total between the two types of figures (standardised or not) but not for a
specific country. Only the “standardised” values are used in the rest of the present report.
Sectoral market
The evolution of the various economic sectors and their demand for comfort vary a lot (figure 5.5). Only
trade and offices really grow in relative terms and they may reach 70% of stock by 2020.
Figure 5.5. The evolution of cooled floor-area by EU economic sector from 1985 to 2020
103
Mm2
2500
2000
1500 education
residences
trade
offices
hotelsbars
1000 hospitals
500
0
Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020
3000,
2500,
2000
RAC
Rooftops
1500 PACK & Large Split
VRF
Chillers
1000
500
0
Stock 1990 Stock 1995 Stock 2000 Stock 2005 Stock 2010 Stock 2015 Stock 2020
104
We have chosen the American DOE model for generating consumption estimates and a summary
explanation of how it is used to conduct the EECCAC system simulations is given now.
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
105
Envelope has been adapted to each country construction habits. Insulation cannot be the same in the various
countries. The final decision regarding this matter has been taking the building’s envelope from the
“TRIBU” study of Building Codes (TRIBU, 1994). Basic differences between climates are the following: for
exterior walls insulation thickness varies from 4 cm in Seville to 8 cm in Milan; we have used double glazing
in every climate, but clear in London, low emissive in Milan and low emissive and reflective in Seville.
Thermal comfort and indoor air quality have been guaranteed for every system to allow comparisons.
The simulation covers seven different system types: Constant air volume (CAV), variable air volume (VAV),
Roof-top units (RT), fan-coil four pipes (FC4P) and two pipes (FC2P), package terminal air conditioners
(PTAC) and water to air heat pumps connected to a close condensation loop (WLHP). Every HVAC end use
is covered, namely, fans, pumps, cooling and heating.
Lighting and plug equipment, despite being non HVAC uses, have been also considered because they
generate a large share of the cooling load.
After an exhaustive filtering process, six WS EEO have been studied and ranked. They regard to chilled
water temperature control and water transport. Basic AS EEO have been valued, including air transport
efficiency, air side economiser and exhaust air heat recovery. The results of those simulations is given in the
next chapter.
Some basic remarks have to be kept in mind for central systems:
• They have been designed assuming the same zoning and air distribution. So there is the same number of
AHU's for CAV, VAV and RT. This implies that zone supply air flows are also the same for each
climate.
• Same ventilation level (that is same zone outdoor air flow rate) has been considered. For VAV systems,
the minimum supply setting of each VAV box equals the design outdoor ventilation rate, and, at AHU
level, outdoor flow rate is always maintained constant. This supposes that each central system is always
handling the same amount of outdoor air, however VAV handle a variable supply air flow rate.
• Air transport efficiencies, expressed in terms of specific consumption (W/m3/h) is equal to 0.47 for
constant volume fans (SPF = 1.7 W/(l/s)) and 0.57 for variable volume ones (SPF = 2.05 W/(l/s)). A
variable speed motor is used to control supply flow for VAV.
• Air side economizer and exhaust air heat recovery are not installed when describing the stock CAC
market.
With regards to zonal systems, the following issues should be pointed out:
• One (or some of equal size) terminal unit is installed for each thermal zone.
• Ventilation is guaranteed using a primary air AHU that provides neutral (22 ºC) outdoor air directly to
every building zone. Heat recovery is not used for this AHU.
• Air transport efficiencies, expressed in terms of specific consumption (W/m3/h) equals to 0.15 for FC
and WLHP terminal units. PTAC fan consumption is considered as cooling consumption since
manufacturers' data include this consumption in EER.
For hydronic systems, main remarks follow:
• Chilled water loops provide water at 7 ºC to cooling coils while hot water is supplied at 60 ºC. Water
delta T for cooling and heating are 5 and 10 ºC respectively.
• An air-condensed screw chiller (EER = 2.6) is used to provide chilled water to cooling coils, and a gas
standard hot water boiler (Eff = 0.88) as heat source.
106
• Each primary and secondary water loop is equipped with a constant flow circulation pump. Efficiency
figures may be found in the technical detailed report of Task 5.
• Hydronic system except FC2P supplies chilled and hot water using independent circulation loops (four
pipes facility). The FC4P system has been kept to represent the typical fan coil system.
Adjustment for chiller quality and options not covered in DOE software
We post-processed the results given by DOE2 to make them flexible in terms of selection of a chiller. We
ran a specific program in which the load and climatic conditions remain the same but the quality of the
chiller can be adjusted according to the findings of the techno-economic analysis (next chapter). The post-
processing consists in keeping all the auxiliaries given by DOE, to cover the cooling load with any
alternative chiller, given the outside conditions extracted from the simulation. This also allowed us to
consider the case of wet cooling towers which represent a limited but non negligible share of the market.
Finally we could also consider less frequent solutions that are presented as very efficient as VRV, chiller on
natural water, etc. by extending the post processing.
The sizing of the system has been adapted to each simulated climate (Seville, London, Milano) and so the
three locations do not display the same installed capacity for the same building shape. The building
conditioned area is about 4800 m2 under the form which has been simulated. Note that depending on our
objective we have used the nominal square meter of the building (the one known in national statistics), and
sometimes the standardised m2 when it’s related with consumption (standardised sizing of 120 W/m2).
The two objectives don’t give the same results : nominal conditioned area is 4800 m2 while standardised
areas for cost calculation and stock modelling are respectively 6200, 3200 and 5200 m2 for this same
building in the three climates( SE: 160 W/m2, LO: 80 W/m2, MI : 130 W/m2 –more detailed figures have
been used by system types).
Table 5.3 Consumption per physical square meter, cooling demand, efficiencies and total cost of cooling one
square meter of the office building.
Electricity Needs Electricity to Electrici Initial ALCC ALCC ALCC
SSEER
SEER
Per sq. meter SCL compressor ty total Cost In Euros In Euros In Euros
(kWh/m2) kWh/m2 SEC kWh/m2 SSEC kEuros (0,10 (0,06 (0,17
kWh/m2 E/kWh) E/kWh) E/kWh)
Seville-CAV 115,05 59,25 99,26 1,94 1,16 1008 34,86 31,73 40,34
London-CAV 20,82 10,87 32,77 1,92 0,64 528 16,77 15,73 18,60
Milan-CAV 73,53 36,73 70,49 2,00 1,04 848 28,31 26,09 32,20
Seville-RAC 104,02 54,40 58,52 1,91 1,78 382 14,86 13,02 18,09
London-RAC 15,59 7,97 8,39 1,96 1,86 202 6,08 5,82 6,55
Milan-RAC 54,63 26,39 28,53 2,07 1,92 322 11,60 10,42 13,67
The cooling loads are different from one place to another but, since the sizing is not the same, the equivalent
number of hours of operation (load in kWh divided by sizing in W, for one square meter) is less variant for
this same building in the three climates (SE: around 700 hours, LO: around 200 hours, MI : around 500
hours). More detailed figures could be defined by system types. The figures given here are still consistent
with the ones used for those places in EERAC, while the meteorological data and the software, as well as the
level of definition have been largely improved. The EERAC figure did include a penalty for degradation of
107
performance over time, but we assumed here that CAC were perfectly maintained at their initial performance
due to the larger building size.
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.
Fig 5.8a Total cooling consumption for CAC systems per square meter
108
Reference office building unitary cooling consumptions
140,0
120,0
100,0
kWh/m2
80,0 London
Milan
60,0 Seville
40,0
20,0
0,0
S
F
ps
l
R
ts
op
st
um
n
g)
al
tro
is
g
ng
lit
rg
VR
io
tio
M
LE
uc
di
in
in
m
rd
to
lo
Sp
ut
+h
la
n
i
bu
ol
Ks
D
er
R
IL
oo
co
oo
e
SP
ai
rib
co
at
on
tri
H
ir
le
C
.(c
.(c
+
st
ity
+a
K&
w
is
t.(
ng
PA
di
on
st
um
rd
id
er
is
+
+
Si
C
di
er
m
er
rd
at
PS
ai
s
PA
+h
r
hu
at
AC
at
er
w
ai
h
at
O
w
w
at
ir
r+
t
de
i
R
w
ith
LO
+a
w
de
ith
si
ai
de
w
d
si
w
ut
d
O
e
ith
ed
si
le
ut
d
ol
TW
d
le
ut
oo
w
O
le
l
oo
oo
rC
O
oo
C
le
C
C
rC
Ai
er
oo
er
er
at
Ai
rC
at
at
W
W
W
Ai
CAC Systems
The main consumption factor is clearly the climatic area. The second important parameter comes from the
system type. Depending on each CAC system type, the load is treated either completely centrally or only
partially ; thus, the total cooling load will differ because of the weight of supplementary fan energy released
in the air to be treated and supplementary pumping energy released on the water loop when cooling. It is
shown on Figure 5.8b.
Figure 5.8b. Total cooling load for each system type for the 3 climates
80,0
60,0
40,0
20,0
0,0
n
t
n
st
um
g)
g)
op
ps
s
em
g)
s
e
s
R
ro
is
tio
tio
lit
lit
ge
rg
VR
di
LE
lin
lin
lin
rd
nt
lo
Sp
Sp
+h
st
la
bu
ft
ka
r
co
oo
IL
te
b
ai
e
Sy
SP
co
oo
o
tr i
tri
ac
on
ti
ir
H
a
.(c
.( c
+
ity
ul
+a
is
t.(
s
R
C
K&
lp
i
M
st
on
m
d
id
rd
er
is
al
di
C
hu
m
er
er
rd
at
ai
PS
Sm
er
PA
u
at
at
er
AC
w
+
ai
+h
h
at
w
w
O
at
ir
t
de
i
ith
R
+a
w
LO
w
de
ith
r
ai
si
de
w
d
+
w
si
ut
d
le
O
ith
le
si
d
ut
O
d
oo
le
TW
e
ut
w
oo
le
O
l
oo
oo
C
O
oo
C
le
C
r
C
rC
Ai
er
oo
er
er
at
Ai
rC
at
at
W
W
W
Ai
The difference between consumption and load representation Fig 5.8a and 5.8b enables to separate the part of
the consumption differences between CAC coming from efficiency.
The first part of these difference comes from the repartition of energy between fan, pumps and cooling that is
presented hereafter Fig 5.9.
109
Figure 5.9. Contribution of each piece of equipment in % of total consumption (per standardised square
meter for cooling, fans, pumps) in Seville.
100%
80%
Pump %
60%
40% Fan %
20%
0% Compressor
%
SP F
AC Ro ge
on ps
um
es
ts
op
R
st
s
st
n
oo l
g)
n
g)
)
ro
l p plit
g
lit
K& VR
di utio
id utio
LE
uc
di
di
lin
o n f to
lin
Si kag
r
in
lo
nt
Sp
+h
la
D
er
de +hu ool
ir
IL
ut ter coo
co
e
rib
o
a
ac
at
ti
PS air
le
tri
ai t.(c
ul
+
at oled ate ity
st
C
w
ng
.(
s
t.(
M
di
TW e w er
+
is
+
is
al
C
LO er
m
er
ir
rd
si wat
rd
Sm
PA
a
at
hu
at
er with ith
O
a
w
ir
+
O id e
R
w
o o ith
+a
w
w
r C Co ith
oo air
d
s
w
d
+
w
O
e
si
le
ut
ed
ol
O
d
ut
le
O
oo
W led
C
o
C
C
Ai Air
C
er
oo
er
r
Ai
at
at
W
W
CAC systems
It is very impressive to see that the auxiliaries can reach the same order of magnitude than the real chiller
consumption for central systems, and even larger in the case of London. The improvement in Air based
systems should come from the improvement of secondary equipment and control.
The preceding analysis is based on the bare figures summarized in the 3 following tables respectively for
Seville, London and Milan. The very high figures for SEER in London are partly due to climate and partly to
the assumption that the square meter considered is a “standardised” square meter.
Table 5.5a Results for all systems, per physical square meter : SEER and SSEER, specific consumption in
kWh/m2 for cooling in Seville, ranked by order of merit
Table 5.5b SEER and SSEER, specific consumption in kWh/m2 in London, ranked by order of merit
110
Comfort System Compressor Fan Pump T Cool SEER SSEER
12 TAC PACK&SPlarge 7,97 0,42 0 8,39 1,96 1,86
15 TAC Multi Splits 7,97 0,42 0 8,39 1,96 1,86
16 TAC Splits 7,97 0,42 0 8,39 1,96 1,86
17 TAC Small packages 7,97 0,42 0 8,39 1,96 1,86
18 TAC Single Ducts 9,51 0,42 0 9,93 1,64 1,57
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
Interestingly, the results are similar from one location to another. The decentralised systems have a large
benefit, despite the fact that we have added to them a primary air system to bring them up to the same
comfort level than the other ones. Even Single Ducts are better than any collective system. VRF are a good
system, in the middle of decentralised systems, except for Seville where they show a benefit. Wet cooling
towers and systems with two water loops display the same performance as the best centralised system, but
not overpass them. SEER and SSEER do not follow the ranking based on consumption, but this is due to
problems in the definition of “load” in DOE2 software with distinct systems. A regulation should be simply
based on electricity consumption if we want to avoid such misunderstandings. Our feeling is that the
designers should keep the right to use whatever system they need to cope with the project specificities but
that they should be obliged to improve the system chosen to reach a certain level of consumption.
111
The simulations have then been extended to each of the 15 Member States by use of outside climatic
information from an extended data base. For instance, while the most extreme cooling loads are covered
through Seville, the winters and summers in London are milder than those in a large part of central and
northern continental Europe hence it is not possible to cover many climates using this station mixed with the
more southerly locations.
An analysis had to be done based on cooling and heating degree day data (CDD & HDD). Searching on
www.i-wex.com produced cooling and heating degree day data for a number of EU locations (CDD
threshold was 15.5 °C, HDD threshold temperature was 18.5 °C). Using the office simulation results a linear
relationship was established to predict energy consumption by cooling or heating equipment as a function of
HDD and CDD and used to generate results fitting exactly with the borders of each country Using these
combinations with the linear equations applied to the CDD and HDD data gives the following annual
average unit energy consumption for the same office building by EU country, Table 5.5.
Table 5.5d Annual average energy consumption per m2 by EU country (kWh/m2/year), weighted for systems
and sectors
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.
• The winter demand of the cooled areas if no reversible use took place
• The winter demand of the cooled areas with the reversible use presently estimated.
Figure 5.12 shows the first two values (cooling and associated heating consumption by technical type) for the
future.
112
Figure 5.12 Energy for cooling and conventional heating associated with the cooled area -consumption by
technical type for cooling
300 000
250 000
200 000
RAC
PACK
GWh
50 000
0
1990 1995 2000 2005 2010 2015 2020
Figure 5.13. Total energy consumption by AC type in Europe in 2000 and 2020 for the three main quantities :
cooling, heating (if no reversibility); heating (with present reversibility rate)
113
Total consumption by country - BAU
60 000
50 000
40 000
CO 2000
CO 2020
GWh
CH 2000
30 000
CH 2020
RH 2000
20 000 RH 2020
10 000
0
AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK
Tables 5.6 and 5.7 give the main values (EUR15) for the three functions. Note that gas is accounted for as a
secondary energy. (CO : Cooling, CH : Conventional heating, RH : Reverse Heating)
Electricity demand (TWh) 1990 1995 2000 2005 2010 2015 2020
Electricity and Gas
Cooling function 22,879 33,683 51,636 78,103 94,727 109,631 114,579
(Electricity only)
Heating function 51,598 74,442 111,084 164,517 203,330 236,765 250,844
Without REV.
Heating function 7,374 11,495 18,894 28,913 35,875 42,333 45,040
With present REV. (El.)
Table 5.7 Cooling only results by country and year (for comparison with national statistics)
114
PO 1 020 2 049 3 072 4 039 4 621
SP 19 689 28 333 33 573 38 719 39 915
SW 391 378 403 421 425
UK 2 359 3 227 3 826 4 241 4 401
Total 51 636 78 103 94 727 109 631 114 579
45 000
40 000
35 000
30 000
TRA
OFF
G 25 000 HOU
W
h 20 000 HOS
EDU
15 000 CAH
10 000
5 000
0
AU BE DE FI FR GE GR IR IT LU NE PO SP SW UK
Let's consider first the issue of refrigerants. R22 is the most commonly used A/C refrigerant; however, as this
fluid has an ozone-depletion and a global warming potential its production is prohibited in developed
countries. A/Cs can use alternative refrigerants such as R290, R407C, R-134a and R-410A (these refrigerants
are more or less compatible with the operating parameters of a traditional R22 unit). R407C, R134a and
R410A are the only refrigerant largely used in CAC directories for substituting R22. TEWI (total equivalent
warming impact) is the integrated index used to measure the global-warming impact of all gaseous
emissions, including those from direct and indirect sources.
What are the interactions between refrigerant change and energy efficiency? For an optimist, a higher energy
efficiency and a more environmentally benign refrigerant will both result in a lower contribution to global
115
warming. For a pessimist, there is a trade-off between choosing a better fluid for direct emissions (leaks to
the atmosphere) and a better fluid for indirect emissions (lower electricity use). Fortunately, our study group
is not in charge of the issue of change of refrigerants. The direct contribution of A/C to global warming will
drop independently from our action, due to other policies on which we do not interfere.
Why aren't we considering other atmospheric emissions than global warming by CO2. Atmospheric pollution
from power plants is composed of dust, NOx and SO2, which all have a regional impact, and CO2 with a
world wide impact. We have assumed here that checking European regulation on acid pollution was not part
of our objectives. On this subject, the "ExternE" study gave recently values of the external costs of power
plant pollution that we could use in case of necessity. The consideration for CO2 is different; the Kyoto
protocol has been made recently, its full implementation in Europe is not yet achieved and the market has not
yet taken it into account; furthermore there is a European bubble and the trends or measures considered here
can gave directly positive or negative consequences on the achievement of the European objectives.
So we have decided to adapt our environmental considerations to our designated range of actions: energy
consumption changes resulting in a lower indirect CO2 release. Since there is a European bubble, one can
assume one average CO2 content of the European kWh, set here at 350 gCO2/kWh, the marginal rate with
Combined Cycles which are likely to be installed in Summer peaking countries to cope with the new
demand. In fact, the average for OECD (440) and the exact figure for CO2 content per kWh are available for
each country, this can be taken into account in details if needed. Here we forget about other environmental
effects : radio-elements, accidents in the case of nuclear plants, etc because we have a marginal approach and
nuclear is not the marginal energy. Note that the external cost of the CC plants are among the lowest, except
nuclear plants. Its order of magnitude being 10-20 % the external cost effects can be estimated by computing
the potential impact of electricity costs rising by 20%, a trend which may have other causes or never happen.
Water use is another environmental impact of Air Conditioning to be taken into account. It's mostly the case
for water cooled chillers using cooling towers. They are used in the about 12% of cooled area, and consume
about 3-4 kg of fresh water per kWh rejected (not only the part evaporated but also the poorly controlled
salts (de-concentration). We will not devote much time to the issue, simply estimate the total quantity and
take it into the cost (1-3 euros/m3) in the optimisation.
TEWI (Total Equivalent Warming Impact) and leak rates of CAC systems
The greenhouse effect contributions by the installation of refrigerant fluids should be evaluated on the total
of their life cycle. The principal contribution to the greenhouse effect of the air conditioners comes from their
energy consumption. Actually, each kWh of electricity consumed implies a CO2 emission depending on the
specific utility plant in that country. The TEWI index was introduced to compare the direct additional
contributions of refrigerant system emissions and the indirect contributions due to the energy consumption of
these systems.
The uncertainty in the evaluation of TEWI are the same as that for the GWP (+/-35%), for which we add the
uncertainty in the evaluation of the emissions and energy consumption. These evaluations strongly depend
on the quality of the data.
Two formulas can be used to evaluate the contributions. The most simple is written:
116
Some people make a difference between the annual emission rate and the fluid recuperation at the end of the
appliance's life cycle, taking into account the fact that the recuperation has become mandatory in numerous
countries. Practically, leaks during recovery are close to leaks during one year of operation, and the simplest
equation is enough.
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.
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.
Figure 5.15 : National cooling CO2 emissions of AC by country for 2000, 2005, 2010, 2015 and 2020.
117
CO2 emissions for cooling function in Europe
45 000
40 000 UK
SW
35 000 SP
PO
30 000 NE
LU
25 000 IT
kt CO2
IR
20 000 GR
GE
15 000 FR
FI
10 000 DE
BE
5 000
AU
0
2000 2005 2010 2015 2020
Such impacts are not small, but limited if we compare them with other uses in buildings (heating, home
electronics, better lighting, etc.).
Use of water
A wet cooling tower (which displays better energy performance) is more at risk of cultivating the legionella
bacillus and consumes water (through evaporation, formation of droplets and desalination) at the rate of 3-4
kg per kWh of heat rejected, which is equivalent to requiring a few hundred of litres of water per year for
each cooled square metre in a typical southern European office building. Nonetheless the evaporation of this
water produces a small improvement in energy performance, of the order of a few kWh/year per square
metre cooled.
The water consumption should be taken into account to perform cycle cost analysis of CAC systems cooled
by water condensing chillers.
Table 5.7a specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in Seville; final
(commercial) energies are added without conversion
118
8 TAC Outside water + air dist 6,06 5,76 0,23 2,43 12,05 8,79 96,86 93,59
9 TAC Outside water +air +hum 7,28 6,34 0,23 2,91 13,84 9,85 110,3 106,3
10 TAC TWO LOOPS + CHILLER 5,82 5,99 0,23 1,16 12,04 9,50 1
71,24 1
68,70
11 TAC VRF 26,93 5,99 0,23 6,72 33,15 12,71 88,14 67,69
12 TAC PACK&SPlarge 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
13 TAC Roof tops 5,91 5,75 0,23 3,03 11,90 8,78
106,2 103,1
14 TAC RACs on one loop 26,86 4,64 6,85 2
10,87 38,35 22,36 98,66 1
82,67
15 TAC Multi Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
16 TAC Splits 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
17 TAC Small packages 26,93 4,50 0,23 11,29 31,66 15,79 90,17 74,30
18 TAC Single Ducts 26,93 4,50 0,23 11,68 31,66 16,17 99,10 83,62
Table 5.7b specific consumption in kWh/m2 for Heating (H) and reversible heating (RH) in London; final
(commercial) energies are added without conversion
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)
119
A policy interpretation of the figures is only possible if something is assumed about the competition between
gas and electricity, either their cost, or their CO2 content or their “primary” energy value. The coefficient 2.2
is sufficient to represent about all aspects and gives the exact CO2 value (the ratio 400:180). If we
summarise the comparison to SCOP and CO2 emissions, and if we consider only the most frequent systems
inherited from the past in the present stock, table 5.8 gives us interesting indications.
Table 5.8 SCOP and CO2 emissions of the heating function with comfort level TAC
Very often more CO2 is emitted with a reversible system than with a non reversible one. The best heating
system is almost always a classic independent one. This is not a set of values against reversibility. Simply, a
system which is structured and sized to face the very demanding conditions of Summer will consume more
in Winter than the simplest systems used for heating only. It bears the weight of the auxiliaries, and becomes
a less efficient realisation of electric heating. It’s an invitation to research : how to make air conditioning
systems – a growing social demand- sober in Winter? Reversibility is not an easy task, it’s one of the
challenges of the next chapter.
120
121
6. TECHNICAL AND ECONOMIC EVALUATION OF THE ELEMENTARY EQUIPMENT
USED IN CAC
6.1 Energy-engineering analysis of chillers
Chiller prices as a function of the refrigerating fluid and EER
Chillers using the refrigerant R407C, which has been developed as a zero ODP substitute to R22, on average
have an identical energy performance and do not appear to be any more expensive to purchase, judging from
an analysis of their publicly quoted prices. Figure 6.1 shows the price of the equipment as a function of its
refrigerating power and refrigerant. From this it appears that there is no additional cost for chiller equipment
that uses R407C compared with those which use R22.
45000
Figure 6.1. Chiller cost versus cooling capacity, as a function of the refrigerant
40000
y = 230.98x + 5130.2
2
R = 0.7835
35000
45000
30000
y = 122.46x + 4944.3
2
R = 0.9228
40000
25000
Price
euro
20000
35000
15000
R 2 = 0.7719
10000
Air cooled
30000
5000
water cooled
Regression(Air
l d)
Regression
l d)
(water R 2 = 0.7823
0
0 20 40 60 80 100 120 140 160 180
25000 Cooling capacity kW
Prixce
euro
20000
15000
10000
R22
R407C
5000
Regression(R22)
Regression (R407C)
0
0 20 40 60 80 100 120 140 160
Cooling capacity kW
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
122
The cost of a reversible chiller is on average 10% higher than the cost of a traditional cooling-only chiller.
The data in Table 6.1 show a comparison of prices and EER for a sample of 89 cooling-only models with 44
reversible models.
123
behaviour of the equipment. This would occur at a relatively low cost as the chiller is already being installed
on the test bench. Second: some chillers (one of which will be a reversible air to water heat pump) will be
fully tested (i.e. with full performance mapping) using EDF’s facilities.
• 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.
124
Figure 6.3 The cost of a chiller at nominal capacity according to its EER
110
100
Euro/kW
90
80
2 2,1 2,2 2,3 2,4 2,5 2,6 2,7 2,8 2,9
EER
Conclusion : the best chiller having the same cost (100 Euros/kW) as the present “worst” has an EER around
2.80. The range from 2.00 to 2.80 shows reasonable prices for a chiller judged only on capacity. It
corresponds exactly to the present market. The minimum cost chiller according to our analysis has the same
EER as the average market EER 2.50), which may be considered as a validation of our cost reconstruction.
Figure 6.4. The cost of the service rendered by a chiller in terms of SEER
10
8
Total cost (Euros/m2)
ALCC17-800h
7
ALCC10-800h
ALCC17-400h
ALCC6-800h
ALCC10-400h
6
ALCC6-400h
3
2,00 2,20 2,40 2,60 2,80 3,00 3,20 3,40 3,60 3,80 4,00
SEER
The optimal level of performance for a screw chiller is about 40% more efficient than the present « bottom »
of the market. We have seen that the chiller chosen in chapter 5 to represent the stock has a SEER around
2.00. The optimal chiller is already present on the market. It has a SEER between 3.00 and 3.50. One way to
reach this performance is an EER around 2.46 (enhanced evaporator and condenser, improved compressor)
and a splitting in 3 or 4 scroll units of the capacity of the compressor. We have estimated the associated
125
overcost at 12.3 Euros/kW (+12.3%). Once again, manufacturers engineers may have other ways to reach
3.25 SEER, less expensive, but our objective was to find out if there is a margin for improvement. There is a
large margin for improvement and central solutions are not condemned in comparison with packaged units if
they improve their performance.
An example can be find in table 6.1 hereunder. For the bottom of the market we have used the same chiller
as for the stock : 2.5 EER with a poor part load control (uncontrolled screw). For the best range of products
on the market we consider the same nominal EER but the best part load behaviour we found experimentally
on a 4 scrolls chiller. Part load optimisation would bring 10 to 20% decrease of the total bill of the office
building simulated (SSEER from moving from 1,16 to 1,34, from 0,64 to 0,73, from 1,04 to 1,22). The
relative change at that bill level is half of what it is at chiller level due to the weight of auxiliaries remaining
unchanged.
The energy benefits of water compared with air appear small, once you include all the auxiliaries needed to
reach natural water or to run a cooling tower and the climatic differences don’t change that comparison.
What is important is the part load behaviour of the chiller not its type!
126
6.2 Engineering approach of the performance of Packaged units
For the most part the large packaged air conditioners, such as ‘roof tops’, found in the European market are
either identical to, or share the same technology as, models available for sale in wider international markets,
such as the USA. Bearing in mind this technological similarity and the resource constraints applying to the
current study a decision was made to adapt the results of existing techno-economic energy engineering
analyses conducted for this type of equipment in the USA for use in Europe rather than conduct a fresh
European analysis.
The US Department of Energy imposed minimum energy performance requirements for large packaged air
conditioners (known as ‘unitary air conditioners’ in the USA) through the EPCA in 1992. As recently as
1999 the non-binding ASHRAE 90.1 standard proposed minimum energy performance requirements for the
same appliances and these have since been made mandatory requirements at the state level by a large
majority of US states. In 2000 the US DOE launched a revision process for the existing EPCA MEPS which
aims to set more stringent MEPS from 200X. Following the US MEPS development process a full techno-
economic energy engineering analysis has been conducted for large packaged air conditioners, which forms
the basis for the results reported in this section. An analysis of the US market for large packaged central air
conditioners established that the market could be adequately represented by a techno-economic energy
engineering analysis of two fundamental models: 1) a roof-top unitary air conditioner having a cooling
capacity of 7.5 tons (26 kW), and 2) a roof-top unitary air conditioner having a cooling capacity of 15 tons
(52 kW).
A parallel analysis of the European market shows that the average cooling capacity of large packaged AC
units in the EU is 28.9 kW while that in the USA is 36.2 kW. Figure 6.5 shows the distribution of models by
cooling capacity in the two markets. As a result the smaller 26 kW base case unit is much more
representative of the type of models found on the EU market than the 52 kW unit.
The only other significant difference in the products found on the two markets concerns the average energy
efficiency levels. As a result of the existing US regulations the minimum permissible EER for packaged AC
units with a cooling capacity between 19 and 39.5 kW is 2.61 W/W and for those with a cooling capacity
between 39.5 and 70.3 kW is 2.41 W/W. In 2003 the lowest efficiency unit which was active on the US
market had an EER of 2.5 W/W and the average efficiency was 2.9 W/W. The maximum EER level found on
the US market in 2003 was 4 W/W. The lowest EER considered in the US energy engineering analysis is
2.78 W/W for both the 26 kW and 52 kW units. By contrast the average efficiency of packaged units in the
EU market and within the Eurovent database was 2.46 W/W in 1998, the minimum EER was 1.78 W/W and
the maximum EER was 3.58 W/W. The large difference in the lower and average efficiency levels can be
ascribed to the impact of the US policy measures and the absence of equivalent measures in the EU.
Figure 6.5. Share of large packaged air conditioners as a function of cooling capacity in the EU and US
markets (source: Eurovent and ARI databases)
127
40%
35%
30%
EU 1998
Share of models
25%
US 2003
20%
15%
10%
5%
0%
0 50 100 150 200 250
Cooling capacity (kW)
The goal of the US energy engineering analysis was to develop cost versus efficiency curves of large
packaged AC units to guide policy development. In particular the intention was to determine the life cycle
cost of packaged AC units as a function of their EER. The methodology used was as follows:
A total of eighteen large packaged AC units, representing several manufacturers and a wide range of
efficiency levels were examined and four units chosen. The selected units were broken down (physically or
using catalog/design data) to create a bill of materials that was fed into a cost model.
The cost model itemises ‘fixed’ factory expenses such as: equipment and plant depreciation, tooling
amortisation, equipment maintenance, utilities, indirect labour, cost of capital and overhead labour and
‘variable’ factory expenses such as: manufactured materials, purchased materials, fabrication labour,
assembly labour, shipping and indirect materials. It also itemises corporate expenses such as: research and
development, net profits, general & administration costs, warranty costs, taxes and sales and marketing costs.
The inputs to the cost model were reviewed by individual manufacturers and the values adjusted if
appropriate. The cost efficiency relationships established in this way were found to follow an exponential
growth curve, thus the data for each manufacturer was regressed to a exponential curve.
Each of these curves was in turn regressed to a single market-average exponential curve to give the results
shown in Figures 6.6 and 6.7 below. These figures also show the upper and lower 95% confidence intervals
as broken dotted lines about the average line (solid).
Figure 6.6. The incremental cost of 26 kWc packaged air conditioners as a function of their efficiency on the
US market in 2001 (US$) (source: TIAX 2002)
128
$1,000
Eq $750
ui
p
m
$500
en
t
Co
st $250
De
lta
($) $0
-$250
8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5
EER (kW/ton)
Figure 6.7. The incremental cost of 52 kWc packaged air conditioners as a function of their efficiency on the
US market in 2001 (US$) (source: TIAX 2002)
$2,000
Eq $1,500
ui
p
m $1,000
en
t
Co
$500
st
De
lta
($) $0
-$500
8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5
EER (kW/ton)
In addition to using this market-based reverse engineering approach a classic design option analysis was also
conducted to explore the cost efficiency relationships of products with potentially higher efficiency levels
than those found on the existing US market .
129
establish the relationship between these life cycle costs and the nominal efficiency of the packaged system it
is necessary to adapt the US cost-efficiency data to reflect the circumstances in the EU in the absence of
equivalent European data. The assumptions used in this calculation imply that not only the trend in relative
manufacturing cost vs. efficiency for packaged AC units is the same in the EU as in the USA, but also that
the relative trends in distribution, installation and maintenance costs are the same. However, when the life
cycle cost results produced in this manner were compared with those produced for large packaged systems
with an EER of 2.25W/W derived from the values quoted in the tables of section 2.2, the results were found
to agree to within 0.3%! This implies that the adapted US equipment cost versus efficiency relationships are
reliable for use in the EU.
For 26kWc units the US analysis implied an average equivalent of 2097 hours of full load operation per year
while 800 hours per year is deemed more likely for the EU. The results of the analysis taking these factors
into account is shown in Figure 6.13 for the 26kWc unit, which is most representative of the EU market.
They show that the life cycle cost minimum occurs for large packaged units with an EER of 3.22 W/W when
a 6% real discount rate is applied.
The comparable results for the 52kWc unit are shown in Figure 6.14. Although the overall life cycle cost per
kW are lower for the 52 kW unit the minimum still occurs for an EER of 3.22 W/W.
Figure 6.13: Estimated average life cycle cost per m2 of cooled space per year vs. EER for large packaged
air conditioners on the EU market (based on a 26kWc unit)
€17
€16
€15
ALLC (€/m2/yr
€14
€13
€12
€11
€10
2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
Figure 6.14: Estimated average life cycle cost per m2 of cooled space per year vs. EER for large packaged
air conditioners on the EU market (based on a 52kWc unit)
130
€17
€16
€15
ALLC (€/m2/yr
€14
€13
€12
€11
2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6
EER (W/W)
131
Table 6.3. Minimum efficiency levels of fans recommended by Eurovent
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%
132
Table 6.4. Levels of heat recovery as recommended by Eurovent, depending on the number of hours of
operation per year (h/a)
Running the Air Handling Unit at a speed which is needed for the actual demand will also save energy. Fans
using the inverter, give possibility to maintain the optimum speed for different air flow rate demands during
the day.
In the Eurovent Recommendation there are many examples for various European conditions. It is also
possible to see how different parameters influence the Life Cycle Cost and the consultant or purchaser may
look at the special conditions that are valid just for his particular system for the case of a cross flow sensible
heat recover device with an efficiency of 0.6. Fan consumption remains constant. Heating saving potential is
very important in every climate and it is over 40%. Cooling savings are less important and only significant in
hot summer locations (3% in Seville). The combination between heat recovery and free cooling has been
proven by simulation to be a simple addition of savings.
133
7. TECHNICAL & ECONOMIC EVALUATION OF CAC SYSTEM PERFORMANCE AS A
FUNCTION OF THE DESIGN OF THE AC SYSTEM
We need to simulate different CAC systems in order to a) understand the relative importance of each aspect
of the CAC system in influencing overall CAC energy consumption, b) evaluate the relative energy
performance of different systems to enable fair comparisons to be made between systems.
Table 7.1 Hypothetical ALCC of a few air conditioning systems for a 2000 m2 building with comfort level TAC
Investment Energy
Euros/15 y Euros/15 y Euros/m2/y
RAC with Primary Air Hypothetical
TAC SSEER
Multi Split systems 2,25 248000 128000 20,75
Packaged systems 2,25 188000 128000 16,77
(under windows)
134
Table 7.2 Hypothetical ALCC of a few air conditioning systems for a 2000 m2 building with comfort level TC
Investment Energy
Euros/15 y Euros/15 y Euros/m2/y
RAC without Primary Hypothetical
Air
SSEER
TC
Multi Split systems 2,25 220000 128000 16,69
Packaged systems 2,25 160000 128000 13,30
(under windows)
135
comes first and leads to more air changes with the outside, and to controlled ventilation. As a result Local
extraction (LE) is the dominant feature in some countries, while V (central ventilation) dominates others.
Obviously V allows a better air quality (dust, temperature, etc.) but is more costly. The energy impact of the
two philosophies is large but not really part of our study, but the capacity of heat recovery being very
different in the two situations, we have to consider them. Central ventilation (V) has been the base of our
study.
There are also in some countries obligations of ventilation in cascade, the exhaust taking place in the “very
polluted” rooms. We have not investigated further this option.
136
CAV and VAV fan consumption for SEVILLE
5
kWh/m²
3
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
CAV 4,39 3,79 4,39 4,19 4,39 4,19 4,19 4,59 3,79 4,39 3,99 3,99
VAV 0,57 0,5 0,54 0,53 0,66 0,98 1,63 1,93 1,11 0,78 0,49 0,5
VAV consumption reduction percentage is higher during winter because of lower loads, VAV boxes operate
at minimum flow, which corresponds to outdoor ventilation air ratio. During the summer there is also a large
saving potential that is always higher than 60 %.
Conclusions: VAV saving percentage varies from 75 to 80 % despite specific consumption for fans (0.57
W/(m³/h)) is higher than that for CAV (0.47 W(/m³/h)). The main reason for this large saving potential is the
adaptation of zone flow rate to real load conditions which are normally under design values due to load
calculation oversizing. On average we could say that VAV save 80% of fan energy in Europe. There are
discussions about the reality of those results on the field due to O&M issues : balancing, controllers tuning,
etc. Conservatively, we made calculations with a 50% saving value.
The Ashrae approach requests variable speed in all air ducts not only blowing through a target of
consumption to be reached at full load and another one at a specified part load value. Europe should take
such a measure when applying its “Energy Performance of Buildings” new directive.
Dual Duct final distribution systems (rather inefficient by principle) are only allowed with variable flow in
some countries and this measure could be extended to all countries. It seems not to present any specific
potential for our study because they are very uncommon.
137
The real figures given by DOE show that the cooling saving potential is strongly dependent on climate.
Oceanic weathers like London offer a more than 80 % reduction of cooling consumption. In any case, and
even for very hot climates like Seville, saving potential is over 20%. Pumping consumption is also reduced
due to cooling load decrease and pumps "on demand" control.
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),
138
In a special configuration of the ceiling panels, they are mounted under a plenum with injected air, likely for
increasing the relatively small cooling capacity. If the cooling capacity of the primary air and the capacity of
the chilled ceiling are enough together to ‘treat’ the load, the cost seems the same as the one with FCU. In
fact the difference appears in winter because the ceiling looses part of its heating capacity compared with a
regular FCU and another system is needed in addition.
In another realisation of the radiant system, recently patented by the firm Van Holsteijn en Kemna (VHK), a
local unit (having the appearance and the cost of a local radiator or FCU of good quality) combines
ventilation (controlled room by room) and radiant cooling. When radiant cooling is not enough, ventilation is
started. Entering water temperatures under room temperatures by only 10K would be enough to deliver 200
W/m2 of radiator. This supposes a decrease in Chiller consumption and an increase in Pump consumption.
Fans situation and performance is quite unclear and we decided to wait for the next step of development of
the system to include it in the analysis.
We have simulated with DOE software this option. The assumption was to determine performance
improvement with a small ΔT difference in a FCU4P system (8/13°C instead of 7/12), since the relative
returns will diminish if we move further. The electricity saving is very low (0.1 to 0.6%) and we abandoned
this solution in the study.
139
Improvement of flow efficiency in FCU and AHU water circuits is possible in various ways : Variable flow
pump, better Pump efficiencies, adequate Pump performance curve. A remarkable option at system level is
to go for “local pumping systems”. Terminal units demand what they need and the evaporator ‘sees’ a
variable flow. Evaporators can accept flow rates decreased by 40% typically. The gains in pumping cost
become significant.
We have considered the basic Air and Water System types (Two pipes fan-coil, Four pipes fan-coil, Water
loop heat pump). Using four pipes FCU or using the two pipes system based on “change over” will allow to
feed the four pipes FCU from heat recovered from the condenser, one of the reversibility approaches.
The importance of additional consumption generated by the 2 pipes FCU with electrical heating versus the 4
pipes –if we want to insure the same temperature all the time- can be estimated. Note that we can estimate
that on the market those two solutions are still frequent (2PE was 25% a decade ago and seems now at 10%
only; 4P is around 15%). Also we suspect that the electrical resistance is used for main space heating not just
for this adjustment…..so it’s the main enemy of reversibility.
Simultaneous demand of heating and cooling
Some systems have a capacity to transfer heat from one zone to another. Such advanced multizone systems
can be justified by its actual benefits. We have gathered some elements on advanced multizone strategies
(WLHP, TWL, VRF,…). The system using RAC on a water loop (WLHP) is relatively frequent (1.5 % of
total cooled area) and presents specific energy conservation features : transfer from one zone to another,
high EER and COP year round, etc. It seems a relatively frequent solution in commercial malls because it
allows individual metering of consumption by each user.
In the same way, the uncommon TWL (a promising two water loops system experimented in France and in
the UK allowing simultaneous heating and cooling) can provide simultaneous heating and cooling. Finally
VRF is one step further in the same direction. It is one way of operating at variable speed (see part on
packaged systems). But it is also an interesting system for transfer between zones demanding heat and cold
(but this not always realised).
The DOE simulation allowed us to understand the real order of magnitude of simultaneous heating and
cooling. In this relatively complex office building where internal and external zones are treated separately,
where various facades receive differently the sun, the effect corresponds to only a few percent of the
demand. More precisely, we have computed for each hour with simultaneity the lowest of the two quantities :
cooling demand, heating demand and expressed it in percent of demand, either cooling or heating, table 7.3.
Table 7.3 Importance of simultaneous heating and cooling
In % of heating
CAC system LO MI SE
CAV -0.26% -0.23% -4.90%
VAV -0.92% -0.92% -3.03%
FC2P -6.24% -5.89% -17.61%
FC4P -7.45% -6.16% -23.62%
PACK -5.33% -4.67% -18.17%
WLHP -6.19% -5.47% -21.49%
In % of cooling
CAC system LO MI SE
CAV 0.59% 0.15% 0.21%
VAV 5.91% 1.24% 0.50%
FC2P 26.88% 7.42% 3.62%
FC4P 29.66% 7.73% 4.78%
PACK 32.22% 7.41% 4.08%
WLHP 31.06% 7.74% 4.53%
140
The economics of the transfer are as favourable as expected. The homogeneity of figures between air
systems on one hand and all other systems on the other hand is interesting.
Heat rejection
Cooling tower fans should have variable speed drives and, in systems with more than one cooling tower, all
cooling towers should work simultaneously at all times (this strategy reduces drastically the cooling tower
fan energy consumption) Temperature control should be modified.
Using natural water –river, ground water, etc.- as a heat rejection medium is very beneficial in energy terms
because the high heat exchange coefficients and low temperature at the condenser improve EER. In some
circumstances the chiller becomes useless and the natural water can cool directly the building (see system
TWL as an example). Control is easy since underground temperatures are constant. Costs and administrative
problems are reported as enormous in Italy and Spain while France maintains such a policy (Aquapac). There
is also the possibility to generate DHW (Domestic Hot Water) at a small cost from condensing heat.
7.4 The possible strength of regulatory efforts and the minimum LCC solutions
Concentration of efforts on Air based systems
We have concentrated our efforts on the air system which show presently (under the CAV form) the most
consumption and the highest cost. The designers need the whole range of solution to cover the domain of
geometries and air quality requirements. So the bottleneck to the expression of a global reduction in
consumption will be the point (shown hereunder by an array) where the improved air based solutions start
not to pay for themselves : the designers will find it is too heavy a constraint.
ALCC Euros/m2/year
air
packages water
rac
SPECIFIC CONSUMPTION
kWh/m2/YEAR
141
pumps (MWh) (0,10 (0,06 (0,17
(MWh) E/kWh) E/kWh) E/kWh)
0 – CAV 121,67 37,76 159,43 1 053 640 30,89 29,84 32,73
1 – VAV-50%/ 121,67 18,88 140,55 1 090 090 31,56 30,63 33,17
+6 E/m2
2 – FC –20% 97,34 37,76 135,10 1 074 710 31,05 30,16 32,61
/+2 Euros/m2
3 – Fans –8% 111,94 32,10 144,03 1 054 140 30,65 29,70 32,31
;–15 %; +2
E/kW
4 – HR 60% - 118,02 37,76 155,78 1 090 090 31,81 30,78 33,60
200Pa; -3%
+6 E/m2
5 –Lower HL 121,67 35,12 156,79 1 126 540 32,80 31,77 34,61
in AHU –
7%/+12E/m2
6 –Optimised 73,00 37,76 110,76 1 062 390 30,32 29,59 31,60
chiller – see
chapter 6
After sorting and combinations, the optimal trajectory of improvement is given in figure 7.3.
Figure 7.3 Optimising with 6, 10 and 17 cEuro/kWh a full all air system
39
37
35
33
ALCC17
Euros/m2
ALCC10
31 ALCC6
29
27
25
The optimum if very flat, specially if we get interested with the highest cost electricity. The regulatory
measure could be taken anywhere between a 0% and a 60% reduction without generating overcosts (in the
LCC definition) in Seville. The optimal regulation would request a 50% cut in electricity consumption
compared with a standard CAV design.
142
8. EFFICIENCY RATING AT PART LOAD: AN IPLV FOR EUROPE
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.
143
The efficiency of this kind of chillers at design point depends on : the size of the compressor (less stages =
better efficiency because of intermediary losses). In a centrifugal compressor, before the impeller, the inlet
vane guide enables to create more or less swirl to reduce capacity to match the load. This is a mechanical
type of unloading. However, the two stages compressor enable to unload at a lower step. Speed of rotation :
centrifugal chillers cannot be operated for small flow rates since the rotation speed needed would be too
high. Part load : the surging phenomena occurs at low part loads, the flow comes back through the impeller
leading to a cyclic phenomena badly known so that manufacturers forbid the chiller to work in these
conditions. There are different modes for unloading the centrifugal chillers : prerotation inlet vane guides,
variable speed. These two are the more common ones. However, when the load becomes inferior to the
surging load, to enable cutting off the compressor, a hot gas bypass strategy is adopted leading to still poorer
performances at very low loads. By associating in series two compressors the surge limit goes under the
single compressor one (10% instead of 20 or 25% load). The part load performances seem always less than
full load ones in what we have investigated.
• 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.
144
Scroll compressors at part load
The rationale is about the same one as for the basic screw : nor unloading is available for chiller applications,
neither adaptation to varying condensing pressure. Generally, one uses several compressors in parallel on the
same circuit and make them cycle. For two scroll compressors on the same circuit, two capacity steps plus
the full load are available if nominal capacities of each of the two compressors differ.
Thus, different technologies are used to control part capacity stages. Depending on the kind of compressor
circuit, one can find unloading by varying the number of available circuits or by varying the flow rate in one
circuit. To perform this latter control of the refrigerant flow in the cycle, one can use variable speed drive
(for screw chillers only in our scope), variable Vi unloading (for screw chillers only), unloading of
multistage compressor (screw or reciprocating), or shutting down a compressor over two or more (sole
option for scroll, available on screw and reciprocating as a supplementary mean).
80
60
Dead-band 1°C
40
20
0 6 7 8 9 10 11 12
Given that chillers generally operate at full load and nominal inlet condensing temperature with a 5°C
between inlet and outlet, the water temperature varies more or less between 6 and 8°C for all stages,
depending on the water loop inertia and on the condensing temperature that modifies the cooling capacity of
stages.
We will assume here perfect control, the one represented by the scheme Figure 8.2, even if some
experimental testing of dynamic capabilities of chillers have shown that chillers did not always behaved this
way [AFCE, 2002]. However, dynamic testing installations are not available and would need long debates to
be specified and then adapted by certifying laboratories.
It has also been observed that most of the time, set point control temperatures were not respected, but
differed by more or less 1°C and sometimes even more from user selected values. For correct measurements
of inlet and outlet water temperature (like the ones used in standardised testing), either very long straight
pipes are needed so that a homogeneous flow may be reached before measurements or, pieces of equipment
have to be installed at the inlet and outlet of the evaporator to enhance the turbulence. On installed units
such equipments are not installed, water temperature measurements do not correspond to real temperatures
and the control behaviour can be far different from Figure 8.2.
Cycling between stages at part load
The following representation of the chiller performance when load is higher than the smallest capacity step is
adopted : if the load lies between two capacity steps, the chiller will operate on each one of the two
neighbouring steps ; the cooling load is the weighted average of the two steps cooling capacities CC for the
same inlet condensing and outlet temperature. The corresponding operating times for each capacity step
145
enable to determine the electric power absorbed EP and thus the efficiency for each hour. Yearly efficiency
is calculated with Equation (1).
(1)
8760
∑ CC i
SEER = i =1
8760
∑ EP
i =1
i
where CC is the cooling capacity and EP the electric power absorbed in each operating
condition actually met.
When the cooling load is lower than the smallest capacity the equipment can deliver, the chiller operates
only part of the time, thus fitting its cooling capacity to the load. In that case, each starting is an energy loss.
At each starting, the compressor has to establish the pressure difference between low and high pressure sides.
The unit only begins to cool water when the average refrigerant evaporating temperature is lower than the
average water temperature. Then, the superheating of the refrigerant has to stabilize : only at that time the
full capacity of the step is reached. On the contrary, establishing the full electric power is quite
instantaneous. This leads to an energy loss at the starting of the chiller. As a consequence of a review of all
existing experimental evidence, we selected (figure 8.3) the Italian standard [UNI, 2002] Equation named
(2) hereunder.
Figure 8.3. Ratio of part load efficiency to full load efficiency for the same inlet condensing
temperature and outlet water temperature as a function of part load capacity in the same conditions.
1
0.9
full load efficiency
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Ratio of part load capacity to full load capacity (same source
temperatures )
The corresponding curve is represented on the figure 8.3, it corresponds to the following formula :
CC
EER = CC FL (2)
EER FL CC .C cyc + 1 − C cyc
CC FL
With Ccyc=0.9.
It is used here to compute the part load performances of single circuit units and the performances of multi-
staged units when load is inferior to the smallest capacity step available.
Ccyc=0.9 is a proposed default coefficient. Supplementary work could be performed to set the experimental
testing conditions enabling the calculation of this coefficient from the manufacturers specifications of the
minimum water loop volumes and of the smallest capacity step available.
High pressure control at part load
The high pressure cannot go too low. It is generally maintained high enough by controlling the air flow rate
for air cooled chillers. Classical control consists of maintaining the high pressure above 15 bar by cycling the
146
fan or by switching one fan OFF (the first one being the better one since the whole area remains in use for
the heat exchange) with a fixed dead band. Decreasing the flow rate at the condenser increases the high
pressure and thus decreases the performance.
The impact of this phenomena has been measured while carrying over a test campaign on a scroll unit. The
unit is divided into 2 distinct and symmetrical refrigerating circuits. Each circuit has a tandem scroll
compressor, which means two steps by circuit. Then, the capacity steps available are 100%, 75%, 50% and
25% ; in fact, due to the mechanical flow rate reduction, the 75%, 50% and 25% capacity steps, are slightly
higher than this theoretical staging. Each circuit has 3 fans on its condenser (line configuration).
Figure 8.4. Evolution of the chiller performances when reducing the condenser air flow rate presented
under a reduced form (EER/EERnom, EP/Epnom, CC/Ccnom in terms of reduced flow rate)
The decrease of the efficiency while varying the flow rate is reported Figure 8.4. The chiller was operated at
50% load, one compressor in operation on each circuit. The 70% flow situation corresponds to 2 fans among
3 being ON for each circuit. Precise measurement of the flow rate was not available. The 50% flow rate
corresponds to one fan functioning on one circuit and two on the second. The efficiency decreases with a
square tendency when the air flow rate is reduced.
The following point was to determine how the efficiency varies with the outside air temperature when the
chiller operates at reduced flow rate : does the reduced efficiency increase with the same slope than with the
full air flow rate ?
Given the NUT-epsilon curve of the heat exchanger, at reduced flow rate, the increase of the efficiency with the outside
air temperature decreases faster than at full flow rate. This fact is observed in reality, but in a more complex manner. At
full load, the high pressures are higher than at 50% load.
Figure 8.5. Condensation pressure control effect on reduced EER (EER/EERnom) in terms of outside air
temperature
147
Reduced EER versus OAT for different load ratios DOE2 curve
1,70
FL curve
1,60
1,10
PL curve, single
circuit VSDF
1,00
15 17 19 21 23 25 27 29 31 33 35 Without high
pressure control
OAT (°C)
On figure 8.5 we can understand that the high pressure control will impact differently single circuit and
double separated circuits units. For single circuit units, the 25% load would correspond to still higher
ambient temperature for triggering fan cycling or reduction speed. Thus, the efficiency variation with outside
air temperature is shown Figure 8.5 for 25% and for 75% load on a single circuit. On the contrary, for that
double circuit unit, the 25% load point does not differ from the 50% since one compressor is in operation on
one circuit, the 50% being strictly symmetrical. For double circuit and 75% load operation, one circuit is
operated at full load while the other is operated at 50% load.
• Group 2, occupation 24h/day : 7days/week, cooling above 12.8 °C, free-cooling between –17,2 and
12,8°C.
• Group 4, occupation 12h/day : 5days/week, cooling above 12.8 °C, free-cooling between –17,2 and
12,8 °C.
3
[ARI98] states that these cities represent 80% of the installed chillers in the US.
148
The mean climate used to perform the bin method is an average of the 29 cities climates. The climatologic
data were averaged without any weighting of energy or capacity installed in each city. Coming from this
average climate, an occurrence curve of dry bulb and wet bulb air temperature is drawn by 5 °F (2.8 °C) bins,
between –17,2 °C and 35 °C. The water temperature is deduced from the wet bulb temperature using an
added 8 °F (4.4 °C) approach.
• 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].
149
Figure 8.7 illustrates the interpolation procedure for unit N°7. The graph differs from the representation used
in the IPLV standard for part load. The load ratio in % is related to full load and 35°C inlet air temperature.
The solid and very black line is the ARI assumed load/temperature curve. The less solid black line is
determined by the capability of the four steps of the chiller to reach the four specified temperatures. Three
straight lines representing the capacity stages relative to full load 35°C capacities are drawn for the stages 1,
2 and 3).
35 Outside air
temperature (°C)
ARI air
condensation
30 curve
75
25 50
25
20
15
10
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Part load ratio (%)
One example of the interpolation can be described with the help of figure 8.7 : when the outside air
temperature is 18°C and the load between about 30% and 62%, the chiller will cycle between the 50% and
25% stages. Thus the efficiency of the 50% point 18°C should be calculated as the weighted average of the
two points: [18, first stage] and [18, second stage]. The corresponding point is located at the crossing point
of the ARI curve and the horizontal plain arrow at 18°C and 50% load ratio.
To avoid the multiplication of the number of testing points, the ARI procedure uses only the testing points
(full circles) to perform interpolation of the efficiencies for specified load points. Thus instead of weighting
the two previously mentioned points, the standard proposes to weight the [18, first stage] and the [26, second
stage] points. Thus, the 50% point efficiency is underestimated in that case since efficiency of the [26,
second stage] point is lower than the [18, first stage] due to temperature decrease.
The consequence is that for that unit number 7, the global seasonal figure is underestimated by the
interpolation procedure whereas, for a continuous control screw unit, the IPLV exact figures would be
obtained, the capacity step chillers being penalized for not supplying continuous unloading. The method is
interesting since it enables to reduce the testing points number and its effect will be discussed further
hereunder.
150
Therefore, AICARR proposed a new energy index, named EMPE (Average Weighed Efficiency in Summer
regime in Italian) directly deriving from IPLV, with different energy weights and, in particular, with
different temperatures at the condenser inlet, fitter for the European climate and requirements in the air
conditioning field.
The EMPE formula is absolutely similar to IPLV, but the values of energy weights and inlet temperature to
the evaporator and the condenser are those indicated in table 8.1:
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.
151
EER100% + EER50%
EER75% =
2
The EER value at 25% of the load can be calculated as follows:
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
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
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.
For all the tested chillers, some common testing points were made according to either the ARI or the
EMPE conditions depending on the manufacturer will. For all chillers, a supplementary point was added to
fulfil the CEN EnV requirement : nominal inlet condensing temperature (35°C for air and 30°C for water)
and 50% load ratio referred at this nominal inlet condensing temperature [CEN, 2002]. For chillers n° 2, 3, 4
152
and 8, only IPLV or EMPE points plus the CEN one were available. For the others as many testing points as
desirable have been obtained. In all circumstances a simple model has been used to draw the performance
maps from existing testing points.
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
Compressor 1 0 0 0 1
Circuit 1
Compressor 2 0 1 1 1
Compressor 3 0 0 1 1
Circuit 2
Compressor 4 1 1 1 1
EP (kW) 8,80 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.
153
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
Reduced efficiency of the part load stages for water cooled chillers
1.2
1.1
1
N° 4
0.9
N° 1
N° 3
0.8
N° 6
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
Let’s give more explanation about two of the tested chillers, N°4 and N°1, as examples of real life issues.
Chiller N°4 is a two circuit four scroll compressor chiller, with the same symmetrical tandem on each circuit.
The efficiency increase show that at 50%, one compressor by circuit is activated. At 75%, one circuit is at
full load and the other at half load. At 25%, only one circuit works at half load. Logically, at 25% and at
50%, the symmetry of the chiller would impose identical performances. The bias can come from many
causes, one being the specific configuration of the plate heat exchanger : the two distinct refrigerant circuits
use the same brazed heat exchanger in order to cut the costs.
This complex part load behaviour can be summed up under one form per temperature – see Table 8.6, or one
single table with all temperatures –table 8.23.
Table 8.6 : Part load performance of water cooled scroll chiller N°4
N° 4 // WT : 30°C STAGES 1 2 3 4
Compressor 1 0 0 0 1
Circuit 1
Compressor 2 0 1 1 1
Compressor 3 0 0 1 1
Circuit 2
Compressor 4 1 1 1 1
EP/EPFL 23% 46% 71% 100%
CC/CCFL 25% 52% 75% 100%
EER/EERFL 109% 114% 105% 100%
Let’s note here that the presentation under this format ensures that the manufacturer has consciously chosen
this staging as optimum and hopefully that it has been prioritised as factory default control parameters.
Now let’s consider chiller N°1 (Table 8.7) : it is a double circuit water screw chiller with one compressor by
circuit. One can see easily the difference between the two type of unloading, symmetrical, for higher than 50
part load ratios and on a single circuit, reducing the refrigerant flow rate at the minimum for part load ratio
smaller than 50%. In that case, the efficiency decreases somehow faster. Stages configurations and
performances are gathered Table 8.5 under the form proposed to Eurovent.
Table 8.7 : Part load performance of water cooled scroll chiller N°1
N°1 // WT : 30°C STAGES 1 2 3 4 5 6
Circuit 1 Compressor 1 0% 0% 0% 59% 101% 100%
154
Circuit 2 Compressor 2 51% 58% 100% 56% 56% 100%
EP/EPFL 33% 36% 51% 70% 87% 100%
CC/CCFL 25% 28% 49% 58% 79% 100%
EER/EERFL 74% 78% 96% 82% 91% 100%
It clearly appears that 3 stages are used for compressor 2 (100, 56 and 50) and 2 only for compressor 1 (100
and about 60). The percentage for each compressor corresponds to the ratio of the cooling capacity of the
compressor to its full load capacity (half the chiller capacity).
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
Reduced efficiency of the part load stages for air cooled chillers
1.3
1.2
1.1
N° 5
1
N° 7
0.9 N° 8
N° 9
0.8
N° 2
0.7
0.6
0.5
0 0.2 0.4 0.6 0.8 1 1.2
Part load ratio
Let’s give more explanation about two of the tested chillers, N°5 and N°2, as examples of real life issues.
Table 8.8 gives testing results on chiller N°5 : it is a single circuit unit with an asymmetrical scroll
compressor tandem, which means three capacity steps. The increased efficiency with reducing the refrigerant
flow rate from stage 2 to stage 1 is counterbalanced by the relative increasing weight of the fan consumption.
Table8.8: Part load performance of air cooled scroll chiller N°5
N°3 // OAT : 35°C STAGES 1 2 3
Circuit 1 Compressor 1 1 0 1
Compressor 2 0 1 1
Fan 1 1 1
EP/EPFL 38% 52% 100%
CC/CCFL 46% 64% 100%
EER/EERFL 120% 124% 100%
Now let’s consider N°2 in table 8.9 : it is a screw double circuit unit. Refrigerant circuits are separated.
There is one screw compressor by circuit. Each compressor can unload partly continuously from about 100%
to 75% and then two supplementary stages at 66% and 33% are available for each one.
Only the four tested stages are reported. The 50% and 75% capacity stages cannot be allocated to each
circuit. Only the 25% can. This part load behaviour just confirms that unloading on a single circuit with a
slide valve is very inefficient as compared to the full load efficiency of the screw.
155
Table 8.9 : Part load performance of air cooled screw chiller N°2
N°2 // OAT : 35°C STAGES 1 2 3 4
Circuit 1 Compressor 1 30% ? ? 100%
Fans 3 3 3 3
Circuit 2 Compressor 2 0% ? ? 100%
Fans ? 3 3 3
EP/EPFL 28% 55% 64% 100%
CC/CCFL 15% 50% 71% 100%
EER/EERFL 52% 89% 111% 100%
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.
Sizing issues for chillers rating as shown by the simulation of the buildings
The Milan CAV hourly load curve is presented Figure 8.10. The tendency that links load and temperature is
very clear even if an important scatter is observed.
Figure 16. Milan office building CAV system hourly load curve
156
Hourly load curve
550
500
450
400
350
300
kWh
CAV-MILAN
250
200
150
100
50
0
10 15 20 25 30 35
OAT (°C)
When trying to calculate air cooled chiller performances from hourly load curve, two problems appear that
will show up two significant limitations of the ARI method.
The load curves must enable to calculate the consumptions for all chillers. Thus, the load must be divided by
the sizing load, so that all the chillers may be compared on the same load and temperature repartition. In the
ARI standard, that point is solved by assumption since all the straight line load curves have the same
maximal temperature of 35°C that also corresponds to the maximum load.
It clearly appears Figure 8.10 that the ARI hypothesis is not verified. The explanation is that even if
temperature is correlated to the load, other load pattern intervene as the solar loads, the thermal inertia and
the dehumidification loads that are the sources of the non explained variance by the load and temperature
correlation. This is the first limitation of ARI sizing assumptions.
So the real optimal design rule is : the maximal chiller capacity and the corresponding temperature
corresponds to the (load, OAT) couple that enables the chiller to cover all the cooling needs. The capacity
variation with OAT of a perfectly sized (500 kW, 30°C) chiller with this simple law has been drawn Figure
8.10 and shows sizing is correct. Of course, if maximal load were at lower temperature, it could happen that
the sizing could lead to non-satisfied needs ; an iteration process on hourly load and chiller capacity has been
adapted to make sure that the cooling capacity is enough all the year long.
The sizing realized for the three climates led to 30% constant oversizing for the 3 climates for the office
building. For real world installations, security coefficients are generally applied to the simplified sizing
method leading to huge oversizing up to 100%. Given the part load characteristics of the chillers, it seems
obvious that consequences for the seasonal efficiency will also be very important.
The seasonal efficiencies for the MILAN CAV load curve are presented Table 8.11 for 30% and 60%
oversizing for the air cooled chillers N°7 and N°2 .
Table 8.11. Impact of oversizing on seasonal performances for Milan CAV hourly load curve
SEER values
Oversizing N°7 N°2
0% 3.81 3.12
30% 3.83 2.81
60% 3.76 2.60
The differences are limited for the air scroll chiller N°7, because the reduced part load efficiency is still
higher than 1 at 25% and hardly lower than 1 for the 50% load reduced efficiency. On the contrary, for the
screw chiller N°2, a sharp efficiency degradation with the load had been noted. These results confirm that the
157
sizing is a key factor for seasonal performances analysis. And it also shows that no optimum sizing can be
done without the exact knowledge of the chiller part load performances.
In the ARI methodology, for reducing the load curve and temperature occurrences to 4 points, only the 100%
and more than 35°C OAT couples are kept. Whereas in our calculation methodology, all the points between
the full load stage and the step immediately inferior are shared between the two steps, giving weight to the
full load. The ARI methodology of reduction would lead almost to a null energy share for the full load stage,
which is not true for staged chillers but approaches the truth for continuous control chillers. This is the
second limitation of ARI sizing assumptions.
Reduction of European hourly load curves to a set of four conditions (based on the example
of Milano)
To reduce the load curve to 4 points, it is supposed that a virtual chiller, with 4 capacity steps at 25%, 50%,
75% and 100% of the design load is operating. The Milan-CAV load curve is used to illustrate the
methodology. But this methodology can –and will- be applied to all conditions obtained by simulation,
leading to the possibility of a four points representation of any condition or of the EU average of operating
conditions.
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.
Figure 8.12. Reduced load curve for the office building, Milan-CAV
158
Figure 8.13 gives the representation of the energy weight of each class as a function of temperature classes. It
is equivalent to the ton hour curve given in the ARI standard for Group 1 (Figure 8.7). It can be seen that
even for Milan, the chiller load and the associated weighting energy coefficients are still not null at
temperatures as low as 15°C.
Figure 8.13. Reduced weighting curve for the office building, Milan-CAV
Figure 8.14. Reduced weighting by stage under the PT format, method 3rd Step, for the Milan-CAV office
building hourly load curve
159
So now, the weighting coefficients are known :
SEERP= [0.2024 0.4272 0.3369 0.0335]
Table 8.12. Reduction of the chiller hourly load curve for the office building in Milan using the CAV system for
air cooled chillers
Energy weighting
Part load (%) Reference
Inlet air temperature (°C) coefficients A, B, C, D
(nominal full load)
100 31.2 3%
75 28.0 34%
50 23.1 43%
25 18.1 20%
Results for more extreme weather conditions (London, Seville, different distribution
systems)
The hourly calculation methodology has been applied to two different load curves among the twelve
available (Table 8.10) :
• CAV-FC system in Seville
• CAV system in London
However, in order to separate the quality of the reduction by itself from the non linear models representing
the chillers, the seasonal performances are calculated successively for the 4 following configurations :
• Without cycling, without high pressure condensation control
• Without high pressure condensation control, with cycling
• Without cycling, with high pressure condensation control
• With both phenomena
The reduction results of the two specified load curves are presented Table 8.13. The two extreme load curves
reduction presented show that both weighting coefficients and temperature conditions greatly vary with
climatic conditions and systems.
160
Table 8.13. Reduction of the chiller hourly load curve for the office building in Seville using the CAV-FC system
for air cooled chillers and in London, using the CAV system
Seville CAV-FC London CAV
The results of the reduction methodology are presented Table 8.14. The nominal full load efficiency (at 35°C) is reported for each chiller and
so is the hourly seasonal efficiency ratio (noted HSEER for hourly), the reduced index figure (noted ESEER) and the relative efficiency
difference between the two seasonal figures. We translate the information in terms of ranking : chiller ranked 1 is better than chiller ranked
2, and so on.
Table 8.14. Accuracy of the reduction for the tested air cooled chillers
Conditions Seville CAV-FC load curve London CAV load curve
Chillers N° 5 N° 7 N° 8 N° 9 N° 2 N° 5 N° 7 N° 8 N° 9 N° 2
EER 2.18 2.59 2.51 2.47 2.93 2.18 2.59 2.51 2.47 2.93
EER ranking 5 2 3 4 1 5 2 3 4 1
HSEER 3.01 3.38 2.73 3.09 3.13 4.08 4.59 3.30 4.21 3.19
No cycling, no fan cycling ESEER 3.05 3.44 2.73 3.14 3.15 4.08 4.60 3.30 4.22 3.31
Relative
1.2% 1.8% 0.0% 1.6% 0.7% 0.0% 0.2% 0.0% 0.2% 3.9%
deviation
HSEER 2.93 3.34 2.69 3.06 3.11 3.70 4.43 3.08 4.11 3.14
Cycling only ESEER 3.00 3.43 2.72 3.14 3.15 3.88 4.58 3.20 4.22 3.31
Relative
2.4% 2.6% 1.1% 2.4% 1.3% 4.7% 3.3% 3.7% 2.7% 5.5%
deviation
HSEER 2.99 3.35 2.71 3.05 3.00 3.80 4.26 3.05 3.91 2.96
Fan cycling only ESEER 3.03 3.40 2.73 3.11 3.10 3.72 4.31 3.01 3.90 3.07
Relative
1.4% 1.4% 0.8% 1.9% 3.4% -1.9% 1.4% -1.4% -0.4% 3.5%
deviation
HSEER 2.90 3.31 2.66 3.03 2.98 3.46 4.12 2.86 3.82 2.92
Cycling and fan cycling ESEER 2.98 3.39 2.71 3.11 3.10 3.56 4.31 2.93 3.90 3.07
Relative
2.7% 2.3% 2.0% 2.6% 4.0% 2.9% 4.5% 2.4% 2.0% 5.1%
deviation
HSEER ranking 4 1 5 2 3 3 1 4 2 5
ESEER ranking 4 1 5 2 3 3 1 4 2 5
• Cycling and fan cycling impacts cannot be neglected for the London curve conditions as it is shown
by the HSEER evolution. However, both impacts are logically very low for the CAV-FC load curve
in Seville, conformingly to table 9 reduction results.
161
Table 8.15 below enables to check that the load curves selected for the sensitivity analysis represent indeed
extreme conditions for temperature and load. For the London CAV system, the temperatures are the lower
(except for the FC4P system) and weighting coefficients maximum at low loads. For the Seville CAV-FC
system, respective reverse conclusions on temperatures and load can be made.
Table 8.15. Applying the reduction methodology to the set of available load curves
Temperatures (°C) Weighting coefficients
Load 100% 75% 50% 25% 100% 75% 50% 25%
Climate System
London CAV 27.6 24.8 20.9 17.1 0.7% 9.7% 42.5% 47.1%
CAV-FC 27.6 26.1 24.3 22.4 4.2% 26.7% 37.6% 31.5%
FC4P 27.6 24.6 20.1 16.1 0.5% 8.7% 48.5% 42.3%
VAV 27.6 25.6 22.4 17.6 1.1% 7.7% 29.1% 62.1%
Milan CAV 31.2 28.0 23.1 18.1 3.6% 33.9% 41.7% 20.8%
CAV-FC 31.2 28.0 24.8 22.0 5.7% 54.4% 31.1% 8.8%
FC4P 31.4 28.1 23.1 17.7 3.1% 32.0% 40.5% 24.3%
VAV 31.6 28.9 24.5 19.1 2.6% 30.7% 39.5% 27.2%
Seville CAV 36.7 32.1 26.3 19.8 3.5% 38.2% 39.1% 19.2%
CAV-FC 36.7 32.1 27.4 22.8 4.4% 47.5% 37.3% 10.7%
FC4P 36.9 32.3 26.5 19.2 2.8% 35.3% 40.2% 21.7%
VAV 37.2 33.4 28.0 21.1 1.6% 30.7% 43.9% 23.8%
Thus it can be concluded that the methodology proposed is a qualified tool to classify the air cooled chillers
at the condition to respect seasonal efficiency classes wide at least of 5% of the market average ESEER
absolute figure.
Table 8.16. Cooling energy needs national weighting coefficients for air cooled chillers for CAV and FCU (WC :
weighting coefficient)
Country Aus Bel Den Fin Fra Ger Gre Ire Ita Lux Neth Por Spa Swe UK
CAV 0.8% 0.3% 0.1% 0.2% 9.9% 3.2% 5.4% 0.1% 38.0% 0.0% 0.7% 1.1% 37.3% 0.4% 2.5%
WC
FCU 0.8% 0.4% 0.1% 0.3% 9.0% 3.2% 4.9% 0.1% 34.6% 0.0% 0.8% 1.0% 41.4% 0.6% 2.8%
Southern Europe country visibly represent most of the cooling energy needs in Europe. We can reduce this information to a set of 6
coefficients that will be used to weight the SEER obtained from London, Milan and Seville load curves. Final weighting coefficients for the
available load curves types are presented Table 8.17.
Table 8.17. Cooling energy needs hourly load curves weighting coefficients for the air cooled chillers, for CAV
and FCU, free cooling and VAV specific applications (WC : weighting coefficient)
Climate London Milan Seville
CAV 5.3% 16.1% 18.4%
FCU 8.2% 22.2% 29.8%
WC
Free Cool. 6.5% 41.6% 51.9%
VAV 10.6% 40.3% 49.1%
CAV (air distribution) and FCU (water distribution) systems enough to represent the present stock. Figures
of the two first lines of Table 8.17 for CAV and FCU are to be understood as a complete set of coefficients
for 6 load curves. The sum of the 6 weighting coefficients equals 1. For the values for CAV+Free Cooling
and VAV applications, only the air distribution systems are concerned . The 3 weighting coefficients issued
from CAV market shares are here to generate scenarios of improved efficiency.
162
At last, these coefficients enable to weight the results of Table 8.15 and to present final ESEER average
European conditions in Table 8.18. The values of the two first arrays are our proposal for EU chillers rating.
The other results show how free cooling and VAV application could change those recommended values.
Table 8.18. Application of the method to derive a ESEER for air cooled chillers used with the free cooling or
VAV options
ESEER Free-cooling VAV
Weighting Weighting Weighting
Part load ratio Temperatures Temperatures Temperatures
coefficients coefficients 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).
163
It clearly appears Table 8.19 that the temperature dependency is not linked first to climatic conditions but to the sizing and to the cooling
tower choice. It has to be recalled that the cooling towers have been sized for maximum wet bulb temperature with 1% of yearly occurrence.
The lower temperature results in Seville as compared to Milan show that on average wet bulb temperatures are higher in Milan. The main
differences amongst climates and systems are found for weighting coefficients differences.
Table 8.19. Applying the reduction methodology to the set of available load curves
Condenser inlet water temperatures (°C) Weighting coefficients
Load 100% 75% 50% 25% 100% 75% 50% 25%
Climate System
London CAV 28.3 24.8 21.7 18.5 0.38% 10.96% 41.53% 47.13%
CAV-FC 27.2 26.1 22.3 20.3 2.23% 22.34% 44.54% 30.89%
FC4P 28.3 24.5 21.5 18.3 0.32% 10.38% 46.35% 42.96%
VAV 26.8 26.2 22.4 18.3 0.07% 9.42% 27.23% 63.28%
Milan CAV 28.8 25.9 21.9 18.4 2.96% 35.53% 39.66% 21.85%
CAV-FC 30.4 26.5 23.2 19.8 0.04% 43.26% 42.65% 14.05%
FC4P 27.5 25.3 21.6 18.0 1.87% 32.39% 43.29% 22.45%
VAV 28.7 26.1 22.5 18.5 2.90% 32.51% 34.75% 29.84%
Seville CAV 29.0 26.2 22.6 18.6 3.31% 38.69% 38.16% 19.85%
CAV-FC 29.0 26.2 22.8 19.5 4.15% 45.90% 40.06% 9.89%
FC4P 28.9 26.2 22.8 18.6 2.30% 36.20% 38.97% 22.54%
VAV 28.9 26.5 23.2 18.8 2.03% 30.28% 42.67% 25.02%
The same market shares can be used for water cooled chillers than for air cooled chillers. because a constant
share of water cooled and air cooled systems has been used, the same for all countries. However, since Table
8.19 exhibits different load weighting coefficients, the final weighting for the ESEER for water cooled
chillers slightly differs from the air cooled chillers coefficients, Table 8.20.
Table 8.20. Application of the method to derive a ESEER for air cooled chillers used with the free cooling or
VAV options
ESEER (Water) Free-cooling VAV
Condenser inlet Weighting Condenser inlet Weighting Condenser inlet Weighting
Part load ratio
temperatures (°C) coefficients temperatures (°C) coefficients temperatures (°C) coefficients
100 28.6 2% 29.4 2% 28.7 2%
75 25.8 34% 26.2 44% 26.2 30%
50 22.3 40% 22.9 41% 22.8 39%
25 18.5 24% 19.6 12% 18.6 29%
8.5 Is there a method good enough for classification of products by order of merit?
We are comparing now the numerical results and the way each of the existing methods would sort chillers by
order of merit.
164
Table 8.21. Comparison of the ESEER conditions with the EMPE and IPLV for air cooled chillers
ESEER ARI EMPE
Weighting Weighting Weighting
Part load ratio Temperatures Temperatures Temperatures
coefficients coefficients coefficients
100 35 3% 35 1% 35 10 %
75 30 33% 26.7 42 % 31.3 30 %
50 25 41% 18.3 45 % 27.5 40 %
25 19 23% 12.8 12 % 23.8 20 %
Temperatures of the ESEER are 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
Weighting Weighting Weighting
Part load ratio Temperatures (°C) Temperatures Temperatures
coefficients coefficients coefficients
100 30 3% 29,4 1% 29.4 10%
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.
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?
165
HSEER versus EER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5
EER
IPLV and EMPE are more accurate than EER for classification but do not give enough
accuracy for comparison of chillers
Figure 8.16. comparison of US-IPLV with HSEER for the tested chillers
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
IPLV
Figure 8.17. comparison of EMPE with HSEER for the tested chillers
166
HSEER versus EMPE
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
EMPE
Based on similar assumptions, the two methods, IPLV and EMPE have the same advantages and disadvantages.
The proposed ESEER method allows grading and ranking of chillers by order of merit
Figure 8.18 . comparison of ESEER with HSEER for the tested chillers
HSEER versus ESEER
3,5
2,5
HSEER
1,5
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4
ESEER
Conclusion : the differences are relatively large between existing methods and reality, and not always in the
same direction. The newly proposed ESEER method is more accurate in a noticeable manner and satisfies
the needs of Eurovent certification process as well as the expectations of the DGTREN in a market
transformation effort.
First way to realise the testing needed for the ESEER proposed certification method
At this point there is still a choice to be made between an experimental approach based on ARI-IPLV
(knowing that it will be completely changed in a few years due to the arrival of an ISO standard) or based on
the draft CEN standard close to publication and more consistent with the upcoming ISO standard. Since there
is no European specific standard to perform part load testing, the analysis is based on :
• the full load testing definition [CEN, 1998],
• the IPLV standard [ARI, 1998], that contains some remarks about testing,
• the experience gained during the “Joint project” at EDF R&D facility, DMT and manufacturer
laboratories visited.
167
The testing must fulfil the following associated constraints :
• it must enable to start the part load certification next year,
• it must not require the manufacturer presence for the testing,
• it must respect the chiller ESEER sequence,
• it must minimize the number of testing points, it is to say, the time needed to perform all tests, the
real cost factor.
First, if we follow the present ARI approach the manufacturers have to give to Eurovent and to the testing
laboratory the expected results close to the ESEER conditions, so as to minimise interpolation and iterations;
the computation of ESEER will be only a weighting/interpolation of testing results. In case we want a
“blind” checking, more time is needed to guess out the proper conditions for testing in such an ARI
approach. We shall present afterwards another way of doing, which seems to us more in the “spirit” of
Eurovent certification : once the manufacturer gives the full table of performance, three points are taken
“randomly” and checked; it is enough to allow the use of the full table and the calculation of the ESEER.
In a certification approach, the manufacturers must give to Eurovent the cooling capacities, the electric
powers and the efficiencies of each one of the point that will be tested and the inlet fluid temperature at the
condenser according to the ESEER temperature load “curve”. Table 28 gives an example, for chiller number
7, that will be explained just after. Under this format, the table allows to answer directly to the question :
what happens to performance when the load (resp. the temperature) decreases.
Table 8.23. For chiller N°7, stage capacities, part load ratio (% of full load, OAT = 35°C), electric power, and
efficiencies.
Decreasing capacity
Toe = 7 (°C) Stage 4 Stage 3 Stage 2 Stage 1
Decreasing % of FL % of FL % of FL % of FL
OAT (°C) at 35°C at 35C at 35°C at 35°C
35 CC 100% 153.7 116.8 81.5 38.4
EP P1 60.0 43.8 26.9 14.2
EER 2.6 2.7 3.0 2.7
30 CC 166.3 82% 126.4 57% 88.1 41.6
EP 53.9 P2a 39.4 P2’ / P2b 24.2 12.7
EER 3.1 3.2 3.6 3.3
25 CC 176.3 134.0 61% 93.4 28% 42.9
EP 48.9 35.8 P3a 22.0 P3’ (P3b) 11.9
EER 3.6 3.7 4.3 Fan cycling 3.6
19 CC 185.0 137.0 93.0 29% 43.9
EP 43.8 32.8 20.7 P4 10.9
EER Fan cycling 4.2 Fan cycling 4.2 Fan cycling 4.5 Fan cycling 4.0
For continuous control chillers : the chiller can supply the adapted capacity within the allowed tolerance.
Point 2, (75%,T2) noted after P2.
168
For capacity staged chillers, we guess in general, the chiller cannot supply the required capacity within the
allowed tolerance ; thus, either the step is above, either the step is beneath. If a step has been triggered above,
it is not counted, and only the step just beneath will be. To reach the step, the inlet evaporator water
temperature is decreased until a capacity stage under 10.75°C has been reached. Correlated inlet at the
condenser according to the ESEER load versus temperature curve is then imposed, for instance, (70%, T2’)
P2’.
The procedure enables not to be forced to check the performances of the step above 75%, thus economizing
testing time, but at the cost of a further interpolation procedure.
Then, for both chillers, with continuous or discontinuous capacities, the following acquisition is made
respectively at P2 and P2’ :
[ARI, 1998]
“C3.1.2 To confirm that steady-state conditions have been established at the specific set of conditions and
within the tolerances set forth in C6.2.1, three sets of data shall be taken, at a minimum of five-minute
intervals. To minimize the effects of transient conditions, test readings should be taken as nearly
simultaneously as possible.”
The procedure is then repeated for the 3rd and 4th testing points.
For the chiller N°7, the points needed to calculate the ESEER are noted P1, P2’, P3’ and P4. From these
points, the interpolation scheme and cycling correction are then applied.
The testing time and precision are gathered Table 8.24. Evaluation of the time needed for the complete
testing procedure is given with and without the interpolation procedure. In this latter case, two points are
needed for the 75% and 50% points. Depending whether two stages embed the 25% load point or not, it will
respectively lead to 7 and 6 testing points. The higher option is kept. The points needed at each % load are
noted P2a, P2b. It would lead in that case to 6 points.
Table 8.24. Evaluation of the first ESEER testing methodology (the +1 testing point corresponds to the nowadays
non nominal testing point defined in the Eurovent testing procedure)
4 points with interpolation 4 points without interpolation
Testing Time Precision Testing Time Precision
CTS (-) (-)
DS (-) (-)
ST [CEN, 1998] 1 hour (+) 1 hour (+)
2 hours 2 hours
ST, PID
(P2’,P3’,P4’) (P2ab,P3ab,P4ab)
1 hour 1 hour
FC, PID
(P3’,P4’) (P3ab,P4ab)
IP WITH : not satisfying WITH : satisfying
Testing points 4 (+1) 7 (3 at 1 hour) (+1)
Set up 1 hour 1 hour
Disassembling 1 hour 1 hour
TOTAL 2 days 3 days
Consequently, the interpolation scheme leads to a non precise enough index for classifying the chillers. But
the results can be obtained in 2 days only.
When applying the method without interpolation, the ESEER sequence is exact but the testing time increases
to 3 days.
For both methods, the PID, sensor temperatures and fan cycling problems are likely to create insolvable
testing problems.
For fan cycling there is no guarantee that the behaviour of the chiller be the same for the tested unit and for a
sold unit, since the parameter could be modified to get favourable behaviour for testing.
The same rationale applies to chillers that have several possible staging to output the same capacity.
169
These two formats of the same scenario are not satisfying ; as a consequence, the second scenario is
proposed.
Second way to realise the testing needed for the ESEER proposed certification method
The manufacturer must give the Table 8.25 to Eurovent for the chiller that will be tested. All the stages
programmed in the soft of the chiller must be supplied. For continuous control chillers, 6 capacity steps at
least must be supplied.
Table 32 is the Eurovent tested chiller N°7 : a 2 circuits, 4 scroll compressors (1 tandem by circuit), 3 fans on
each circuit.
• C1 and C2 are the 2 distinct circuit.
• The percentages refer to the full load point at 35°C OAT for the cooling capacity, the electric power
and the EER.
• The percentages for fans and compressors refers to the circuit full load and not to the chiller full
load. Here they are electric power ratios.
For chiller specific configurations, supplementary information should be gathered for testing :
• For fans, their position should be given to the experimenter if they supply the air for different part of
the air coil and that consequently stopping 1 fan is not equivalent to stopping another one. For
variable speed chillers, the manufacturer should also explain to the experimenter how to reach the
published points.
• For reciprocating chillers, supplementary information should be supplied to the experimenter to
enable to make the difference between compressor unloading or compressor ON-OFF.
• For screw chillers with slide valve, the manufacturer should explain to the experimenter how to
activate the slide valve (access and postion).
Table 8.25. Scenario 2, part load and reduced temperature performance table
Toe =
7 (°C)
Stage 1 Stage 2 Stage 3 Stage 4
% (full % (full
OAT load load
% (full load % (full
(°C) 35°C) load 35°C)
35°C) 35°C)
19 C1 C2 CC 29% 43.9 C1 C2 CC 61% 93.0 C1 C2 CC 89% 137.0 C1 C2 CC 120% 185.0
Fan 66% 0% EP 18% 10.9 66% 66% EP 35% 20.7 66% 66% EP 55% 32.8 66% 66% EP 73% 43.8
Comp 50% 0% EER 157% 4.0 50% 25% EER 175% 4.5 50% 100% EER 163% 4.2 100% 100% EER 165% 4.2
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.
170
Using this scenario, most of the testing problems are solved .It still remains to manufacturers that build
chillers with variable speed fan control to make sure a precise fan speed reduction can be set manually.
As for the precedent scenario, it is not guaranteed the sold chillers will have exactly the same characteristics.
But this problem is part of the Eurovent certification scheme.
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.
The tool is not gifted of any prediction power of the yearly consumption for any real installation. It is just an
indicator of the seasonal performance, whose only aim is to classify the chillers a fairer way the simple EER
does.
It has been shown that the reduction methodology enabled to successfully extract four weighted temperature
conditions : the bias introduced was lower than the experimental uncertainty. As a consequence, it could be
applied to any other stock of load curves to build other indexes for other domains. This method also enables
to increase the number of points ; however, for some of the load curves considered, more than 5 points would
give a useless 100% load point.
Given the differences within the weighting coefficients for the 3 selected climates, different values should be
published at least as a function of the country. To do so, a simple method based on cooling degree days or
more simulations could be developed, if manufacturers require it, to adapt the coefficients and temperatures
by country. Similar spreading could also be done by type of building. Moreover, the present study enabled to
171
show that the free cooling and VAV options should be differentiated from typical CAV installations, mainly
because of the differences in the weighting coefficients.
As given, the coefficients are nearer from Southern Europe operating temperatures and weightings (Italy,
Spain and Greece represent 85% of the installed chiller based systems according to the EECCAC stock and
market study).
The load weighting coefficients are the main seasonal efficiency drivers. Temperatures can be shifted easily
of 1°C if needed (as practised to round the operating temperatures).
This index could be used for single circuit units. However, a method must be adapted to determine the
cycling degradation versus the load. The default cycling law could then be revised for all chillers.
The reduction methodology for a dedicated load curve associated to the presentation of individual testing
results or more generally of part load performances and reduced temperature efficiencies for chillers would
be a highly efficient simple selecting tool for buyers when following the choice method steps hereafter :
• Hourly simulation of the project gives the building or chiller hourly load curve for the specific
project.
• Then, the reduction methodology enables to characterize 4 capacity step points and operating
temperatures.
• The presentation of part load results by the manufacturers enables to select the chiller on 4 efficiency
points.
Certifying seasonal performances for chillers means indicating an average efficiency generally higher than
nominal efficiency as has been largely figured. But it also means to avoid the efficiency competition may be
based on non-representative indicators, as nominal full load EER is. Thus, the seasonal performance index is
thought to be a huge and necessary progress to strengthen energy efficiency of chillers.
Future versions should consider the extension to single stage units that generally operate at different
conditions. It has been shown however that the methodology could be applied to these units for the load
curves treated.
Similar work has to be performed in the heating mode since air to water reversible chillers is an increasing
end use in Europe. The same philosophy could also be applied to ground water condensing chillers and heat
pump, also a developing market in France and Germany.
The applicability to each country and building type should also be studied in order to give a full range of
testing conditions and weighting coefficients nearer from specific and climatic applications.
In order to approach field reality, the integration of dynamics into the part load testing index should also be
considered. Nevertheless, supplementary work has to be performed to reach a such far away goal from actual
chillers characterization.
This work could also serve as a basis for developing a seasonal index for room air conditioners, the more
developing end-use in Europe nowadays.
172
9. ENERGY AND ENVIRONMENTAL BENEFITS: HIGHER EFFICIENCY CAC
SCENARIOS
In this section, efficiency policy and technology scenarios which project energy consumption, peak power demand and
CO2 emissions will be produced and the results compared with the base case for period of 1990 to 2020. The
scenarios corresponding to some policies and measures are defined by impact time and can be
translated at that time in new specific consumptions for the market after that time.
9.1-Scenarios
Scenario 1 MOVING ALL COOL GENERATORS TO AVERAGE PERFORMANCE
All packages RAC and chillers presently under average reach by 2005 the EER level corresponding to the
average of present market but part load is not taken into account in Eurovent grading and so the
corresponding improvement is not obtained. The policy measure associated is banning some classes of
equipment either directly (Directive ) or by voluntary agreement. We can also expect that a certain number of
years of labelling and communication by energy agencies reaches the same point, nobody wanting to buy a
« poor » image equipment.
Chillers AC : the average being 2.50, the classes E F and G should be banned, the weighted gain is 0.23 on
full market average and the factor is 96.2% to be applied to compressor consumption. Chillers WC : the
average being 3.85, the classes E F and G should be banned, the weighted gain is 0.175 on full market
average and the factor is 93.0% to be applied to compressor consumption. Packages and large splits the
average being 2.46, the classes E F and G should be banned, the weighted gain is 0.05 on full market average
and the factor is 98.0% to be applied to compressor consumption. RAC classes E F and G should be banned
and the average gain corresponding is 0.10 on an EER around 2.50 so the factor is around 96%
Scenario 2 THE BEST CHOICE AMONG EXISTING COOL GENERATORS BASED ON FULL
LOAD INFO
On average packages and chillers reach in 2005 the EER level corresponding to the minimum LCC (BAT
with present information) but part load is not taken into account in Eurovent grading and so the
corresponding improvement is not obtained. The policy measure associated is banning many classes of
equipment or a negotiated agreement on average full load performance like ACEA agreement for cars.
Chillers either the average moves from 2.50 (present market) to 2.80 or the classes D E F and G are banned,
the factor is 89.3%.Packages and large splits either the average moves from 2.46 (present market) to 3.22 by
a voluntary agreement or the classes B to G are banned, the gain is 0.78 on full market average so the factor
is around 76.4% There will be a consequence for reversible winter heating, that will be provisionally taken as
the same factor RAC either the average moves from 2.50 (present market) to 3.20 by a voluntary agreement
or the classes B to G are banned, the gain is 0.7 on full market average so the factor is around 78.1% There
will be a consequence for reversible winter heating, that will be provisionally taken as the same factor.
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)
173
Scenario 4 FREE COOLING
Obligation of introducing free cooling on air side of air based distribution systems at a certain value of air
flow (Portuguese regulation and Ashrae) even for primary air (which is the case of our simulations, at
comfort level TC)
There is a reduction in cooling demand which is climate dependant but has been expressed here in relative
terms on table 8.5
Table 8.5 Consumption of each system, relative to present, after introduction of scenario 4
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.
174
Figure 9.1 Demand evolution, depending on scenarios
140,00
120,00
100,00
BAU
80,00 Scenario 1
TWh for AC
Scenario 4
Scenario 5
Scenario 2
60,00 Scenario 6
Scenario 3
40,00
20,00
0,00
2000 2005 2010 2015 2020
Scenario 1
4000
3500
3000
2500
GWh saved in AC
2000
1500
1000
500
0
CAHORE Education Hospitals Households Offices Trade Total
Scenario 2 THE BEST CHOICE OF COOL GENERATORS FOR THE CUSTOMER BASED ON
FULL LOAD INFO
The policy measure associated is banning many classes of equipment or a negotiated agreement on average
full load performance like ACEA agreement for cars. Figure 9.3 shows that this potential is more on RAC
and packages and so that the savings induced benefit more to households and trade than with the previous
one.
Fig 9.3 savings in 2020 in scenario 2 2020
175
Scenario 2 in 2020
16000
14000
12000
10000
GWh
8000 Série1
6000
4000
2000
0
CAHORE Education Hospitals Households Offices Trade Total
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
176
Table 9.3
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
Such a regulation provides a gain not so high as a very strong prescription, but is safer because the designer
can put the effort on any segment of the project.
177
10. POLICY OPTIONS AND RECOMMENDATIONS TO IMPROVE CAC ENERGY
PERFORMANCE
10.1 Some fundamental considerations regarding policy measures
The results of the energy scenario analyses of the preceding chapter have illustrated that there are significant
potential energy savings to be attained by the optimisation of CAC systems. As CAC systems are diverse in
nature, are often designed on-site rather than simply being factory made packaged-systems and are installed
and operated in diverse circumstances, viable policy measures will need to take account of the diverse
circumstances which apply to them in order to realise the potential energy savings. CAC equipment, like
other tradable goods are subject to the terms of the European single market and therefore it is appropriate for
the European Commission to develop policy measures which will raise the average efficiency of new
equipment sold within the European Union. These type of measures include certification, energy labelling,
and minimum energy performance requirements (either mandatory or voluntary in nature). Proposals for
each of these are made in section 10.3. In the case of minimum energy performance requirements, these
could be developed within the mandate of the proposed Directive on Ecodesign of End-Use Equipment;
however, the application of energy labelling for central air conditioning equipment would either require an
amendment of the existing energy labelling Directive to include energy-using equipment destined for usage
sectors other than just households, or it would require the issue of a new primary Directive giving authority
to the Commission to develop energy labels for commercial and tertiary equipment.
Important as these measures are, they only address the efficiency of the end-use equipment used in CAC
systems as determined under standard test conditions and will not realise many of the potential energy
savings which are related to the design, installation and operation of specific CAC systems. Policy measures
which can realise these savings at the design and installation stage are typically provided through building
thermal regulations. The new Energy Performance in Buildings Directive places some obligations on
Member States to develop policy measures which will address some part only of these savings; however,
there are many more areas for action than are specified within it. The most advanced national building
thermal regulations addressing the energy consumption of central air conditioning systems are in the UK and
Portugal; yet even these are not as mature or as encompassing as the US ASHRAE 90.1-1999 standard.
Many Member States are in the process of revising their building regulations to take account of the
requirements of the Energy Performance in Buildings Directive and thus this offers an ideal opportunity for
them to co-operate at least regarding measures applying to the energy performance of building cooling
systems. Specific proposals regarding this are made in sections 10.2, 4 and 5.
10.2 Policies and measures to encourage the selection of more efficient equipment
The analysis presented in this study has shown that there is a significant variation in energy efficiency for all
types of CAC equipment that have been investigated when tested under standard test conditions. The
measures which can be considered to encourage a higher energy efficiency levels for new CAC equipment
are:
• 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)
178
these values may lack credibility therefore Eurovent currently operates a certification scheme of their
members products in order to ensure the reliability of claimed performance levels. While Eurovent
Certification reports the energy efficiency levels of most CAC equipment as tested under standard test
conditions it does not yet compare the efficiency of equipment in a simplified form, thus the user of the
information is required to have a high degree of expertise to be able to interpret the reported efficiency level.
An effective way of indicating the relative energy efficiency of products in a simplified manner is to use a
categorical energy label, i.e. to apply a label which rates the energy performance of the equipment into one
of a limited set of efficiency ranges which are graded using a simple scale. The current EU energy label
applied to household appliances, which grades appliance energy efficiency using an A to G rating scale, is an
effective and widely known example of this.
CAC equipment are generally large and are not generally ‘on display’ in a shop window at the moment of
purchase. It is therefore debatable whether there is much advantage in applying a removable energy label to
the product itself; however, ensuring that the purchaser and subsequent end-user be able to see the relative
efficiency of the equipment is likely to be an effective market transformation tool. The information provided
in a comparative energy label provides the basic language of energy efficiency that enables many other
market transformation efforts to be realised. Even though central AC equipment are subject to the classic
split incentives situation where the purchaser or procurer is unlikely to be the entity responsible for paying
the energy bills the provision of relative efficiency information is still a fundamental component supporting
more complicated policy measures that may aim to bridge the split incentives problem. It is therefore
strongly recommended that mechanisms be put in place which will allow such information to be passed
through the equipment procurement and usage chain.
The current EU energy labelling Directive is restricted to household appliances hence would require
amendment to address this issue. It certainly makes sense to exploit the high brand recognition and public
understanding of the current A to G energy-label efficiency scale for the comparative rating of CAC
components. However, a key question is how that information should be presented to the public? In the case
of products having split-incentives such as CAC components do, it is desirable that not only the procurer
should be aware of the comparative energy efficiency of the equipment they are purchasing but also that
subsequent potential users of the piece of equipment should be able to see this easily. As mentioned a
removable paper label as currently applied to household appliances is not likely to satisfy these requirements.
Alternatively it seems essential that the information on the comparative energy rating of the equipment (A to
G) should be presented in all product catalogues and literature, including on-line sales, and that the label
information should be indicated on the fixed metal rating plate that is applied to the equipment before it
leaves the factory.
For the time being there is no such scheme in place and it may be some time before one is formally
developed; however, industry associations, such as Eurovent, have expressed an interest in adopting such a
comparative grading approach, which could be applied by them on a voluntary basis (e.g.. it could be made
mandatory within Eurovent for all manufacturers who wish to place their products in the Eurovent catalogue
to report the A to G grading of their equipment). By so doing it would allow manufacturers to express the
relative energy performance of their products and, as has been seen for household appliances, would allow
greater differentiation of products within the market place. The European Commission could take advantage
of this informal adoption of a grading system to prepare the ground for a formal efficiency rating scheme in
the years ahead. In the spirit of aiding the rapid adoption of an efficiency grading scheme for central air
conditioning equipment, the remainder of this section contains explicit recommendations regarding the
thresholds which could be applied to denote the A to G efficiency grades for each equipment type examined
in this study.
179
Structural issues
There are certain technical structural issues that need be addressed prior to the formulation of proposals for
an A to G efficiency grading.
Full or part-load? The first issue to consider is whether the scale should be based on full or partial load
ratings. It is clear from the results reported in Chapter 8 that a part-load efficiency scale is more
representative of the true energy efficiency of CAC components when in real usage conditions and therefore
the ultimate goal should be to develop ratings based on part-load performance; however, at the present time
there is no accepted part-load test and only full load efficiency ratings are generally available. The analysis
of Chapter 8 has led to a proposal for an EU IPLV for chillers based upon specific rating conditions;
however, even if this is adopted without controversy and in the shortest imaginable time scale it is still likely
to be years before there are a large number of EU IPLV ratings available from which to develop an IPLV
rating scale. Thus a more pragmatic solution would be to adopt an efficiency grading scale based on an
analysis of the currently available full-load efficiency ratings and to use this in the near term. The results of
Chapter 8 have indicated that commonly, but not always, the relative efficiency of equipment determined at
full load is indicative of its relatively efficiency at part-load, therefore there is little risk of misleading the
public by adopting an initial grading system based on full load performance. Furthermore, the part-load
ratings are usually equivalent to or slightly higher than the full load ratings, which suggests that the same
efficiency range may be applicable to both the full and part-load efficiency grades. Accordingly this report
presents proposals for A to G efficiency grades based on the analysis of full-load performance data, which
are intended for use in the near term. In the future, at such a time when EU part-load ratings are widely
available, it would be appropriate to review the appropriateness of these grades for translation into part-load
grades.
Heating and cooling-modes. Two separate efficiency scales letters (one for the cooling function and one for
the heating function) are already used for RAC in the EU RAC energy labelling Directive. Therefore the
same approach is followed here.
Product categories. A key question is whether it is appropriate to mix all the chillers into a single product
group for efficiency grading or to adjust the efficiency scales depending on the product sub-category? Were
the systems using water and air completely comparable (i.e. were the energy consumed by the cooling tower
to be included) it would be possible to use the same scale for both; however, the uncertainties regarding the
tower control and the origin of the natural water being used are such that it is impossible to make a
meaningful comparison.
The basic chiller types to be separated correspond at least to the different testing conditions applied in the
test standard, which inherently generate incomparable figures as follows:
• condenser-less units.
The efficiency ratings are not directly comparable between these 7 distinct sets of test conditions.
The chillers could also be separated into two categories as a function of cooling capacity say, those with a
capacity less than 750kW and those with a capacity greater than 750kW. The separation for units currently in
the catalogue is not needed since the screw units (mostly) follow the same design as units with less than 750
kW of cooling capacity and the product ranges overlap across both capacity ranges. Some centrifugal units
are already integrated in the catalogue; although it seems they can be graded on the same scale.
Some markets require ducted condensers, which degrade the EER. It is proposed to introduce a specific
classification for them. Admittedly the additional consumption of the necessary fan for the condenser is not
included according to the test standard but this omission is arbitrary (efficiency of 0.3) and in fact heat
180
exchange drops in this case, and the correction of the standards cannot translate the reality. Establishing a
classification based on specific statistics for ducted units will eliminate the two problems. However, when
testing ducted units, the available static pressure is set by the design of the manufacturer. For this reason, the
testing does not give comparable figures. Therefore it is recommended that manufacturers should supply a
common static pressure for testing ducted units.
Proposed grading scale
In order to require the same effort for each type of chiller, it is proposed to make the average efficiency level
of the products on the market correspond to the threshold between the D and E grades for each chiller type
whenever possible. The average EER and COP values are of course based on data which include units using
HFCs. The following classes and values regarding R22 are given as an information only since the European
market will no longer have R-22 chillers.
The construction of the scale for the different categories, intends to respect the following rules, classified by
order of importance:
EER boarders Air Cooled Air Cooled, Water Cooled Water Cooled, Remote condenser
Floor heating and
Floor heating and cooling
cooling
A/B 3.10 3.65 5.05 4.75 3.55
B/C 2.90 3.50 4.65 4.60 3.40
C/D 2.70 3.35 4.25 4.45 3.25
D/E 2.50 3.20 3.85 4.30 3.10
E/F 2.30 3.05 3.45 4.15 2.95
F/G 2.10 2.90 3.05 4.00 2.80
Note: for borders, A, for air-cooled units, is strictly superior to 3.10.
COP boarders Air Cooled Air Cooled, Water Cooled Water Cooled,
Floor
Floor
A/B 3.25 4.20 4.45 4.50
B/C 3.05 4.05 4.15 4.25
C/D 2.85 3.90 3.85 4.00
D/E 2.65 3.75 3.55 3.75
E/F 2.45 3.60 3.25 3.50
F/G 2.25 3.45 2.95 3.25
Note: for borders, A, for air-cooled units, is strictly superior to 3.25.
Impact of the proposed grading on the chiller cooling market
181
Figure 10.1. Air cooled chillers, cooling-mode, < 750kW
40.0%
35.0%
30.0%
25.0%
R407C
R134a
20.0%
R22
HFC
15.0%
10.0%
5.0%
0.0%
A (>3,1) B (>2,9) C (>2,7) D (>2,5) E (>2,3) F (>2,1) G (<2,1)
182
Figure 10.2. Water cooled chillers, cooling-mode, < 750kW
45%
40%
35%
30%
R407C
25%
R134a
R22
20%
HFC
15%
10%
5%
0%
A (>5,05) B (>4,65) C (>4,25) D (>3,85) E (>3,45) F (>3,05) G (<3,05)
When we compare the model-based statistics from the directory with a sample of 2001 confidential market-
based figures we find the values given in Figures 10.3 and 10.4.
183
Figure 10.3. Sales and Market statistics
1,6 1,6
3,3 2000 1,7 3,3 250 1,7
3,2 1500 1,8 3,2 200 1,8
150
3,1 1000 1,9 3,1 1,9
100
500
3 2 3 50 2
0 0
2,9 2,1 2,9 2,1
Sales and Eurovent catalogue statistics for air condensing units are convergent: sales statistics show a better
efficiency.
2,7 2,7
5,5 500 2,9 20
5,5 2,9
400
5,3 3,1 5,3 15 3,1
300
10
200
5,1 3,3 5,1 3,3
100 5
0 0
4,9 3,5 4,9 3,5
Sales and Eurovent catalogue statistics for water condensing units are convergent: sales statistics show a
better efficiency..
184
Proposal of grading of packaged AC in Europe (extension of the RAC labelling scheme)
(1) Air-cooled air conditioners - cooling mode
Table 10.3 Proposed efficiency grades for large-split packaged AC in the cooling-mode
Table 10.4 Proposed efficiency grades for large unitary packaged AC in the cooling-mode
Table 10.6 Proposed efficiency grades for unitary packaged AC in the cooling-mode
185
(3) Air-cooled air conditioners – heating mode
Table 10.7 Proposed efficiency grades for large-split packaged AC in the heating-mode
Table 10.8 Proposed efficiency grades for large unitary packaged AC in the heating-mode
Table 10.10 Proposed efficiency grades for unitary packaged AC in the heating-mode
186
Market mixed statistics based on the scheme (splits and packages mixed)
Table 10.11 Air-cooled air conditioners – cooling-mode -mixed statistics (splits and packages mixed). Market
average efficiency = 2.46 W/W.
% of market(density)
30%
26%
25%
22%
20%
15%
15%
11%
10%
7%
5%
5%
2%
0%
A B C D E F G
Table 10.12 Air-cooled air conditioners - Heating function -mixed statistics (splits and packages mixed).
Market average efficiency = 2.87 W/W.
187
Figure 10.6. Air-cooled air conditioners - Heating function -mixed statistics
% of market(density)
0,25
0,2
0,15
0,1
0,05
0
A B C D E F G
The last column gives a frequency independent of class width (so divided by two in the intermediate class for
heating. It's a pity that our statistics mix the various subtypes (packages and splits) because it seems that
there are really two populations in the data. Note also that we have here only full load COP and EER.
Removing less efficient equipment from the market (MEPS and voluntary agreements)
The draft Ecodesign of End-Use Equipment Directive proposes that policy measures should be enacted
which bring the market average efficiency of equipment up to the least-life cycle cost for the final user. This
implies the introduction of energy efficiency requirements for new equipment, which could be mandatory
(MEPS) or voluntary in nature but attaining the same goal. Table 10.13 lists the full-load efficiency levels for
CAC equipment associated with the least life cycle cost for the final user as determined in this study. The
adoption of policy measures which would move the average new equipment efficiency levels to those
indicated in Table 10.3 from 2005 onwards was simulated in Scenario 2 as reported in Chapter 9 and would
lead to energy savings of about 11% by 2020 compared with the Business As Usual scenario. Adopting
similar measures based upon an EU part-load performance indicator (EU IPLV and/or EU SEER) would
produce energy savings in 2020 at about 18% of the Business As Usual scenario total. By contrast simply
setting MEPS at the current average full-load efficiency levels for CAC equipment would only save energy
equivalent to 3% of all CAC energy consumption by 2020: see Scenario 1 in Chapter 9.
Table 10.13 Full-load efficiency levels associated with the least-life cycle cost by CAC component (W/W).
An alternative approach to MEPS setting is to harmonise levels with those applied internationally. From a
commercial perspective there is some logic to this approach, because most companies supplying the EU
188
market are multinationals with headquarters outside the EU and the CAC equipment they supply to the EU
market is based on the same product platforms that are used in other international markets. From a
programme management perspective this has the advantage that the MEPS efficiency thresholds are tried and
tested having been successfully applied elsewhere. Furthermore in the only case where the life cycle cost
analyses have been directly compared, for packaged AC units, the least life cycle cost efficiency level has
been found to be the same, which suggests that the US efficiency thresholds might be appropriate for
adoption in the EU.
Were the EU to adopt the same MEPS requirements as currently apply in the US through ASHRAE 90.1-
1999 (as either a mandatory or voluntary measure) it would imply the following efficiency thresholds:
Table 10.14 ASHRAE 90.1 MEPS levels for CAC equipment (W/W)
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.
• Establish public sector procurement guidelines (e.g. only class A equipment should be procured for
use in the public sector)
• Develop corporate procurement guidance documentation, analytical tools and training materials to
explain and quantify the advantages of procuring more energy efficient CAC equipment
• Develop and promote on-line directories of efficient equipment (e.g. as in the UK web-site
www.ukepic.org)
• Develop low cost credit lines for more efficient equipment (e.g. as in the UK Enhanced Capital
Allowance scheme)
• Arrange training programmes for associations of designers and installers to explain the cost-benefits
of more efficient AC equipment and to ensure lines of access to efficient equipment
• Provide rebates on efficient equipment (e.g. as in the Dutch rebate scheme for class A household
appliances)
189
• Create favourable tax differentials for efficient equipment (e.g. lower VAT levels, or corporate tax
breaks for manufacturers and/or corporate procurers)
10.3 Policies and measures to encourage the adoption of more efficient system
structures
Policy aims and potential measures targeting the adoption of more efficient system
structures
Measures that aim to encourage the adoption of more efficient components, as outlined in section 10.2, will
only realise some of the potential to save energy for CAC systems. Such measures are necessarily focused on
the individual components and not on the performance of the system as a whole, therefore they do not
encompass the freedoms and constraints applying to the system designer in trying to design an efficient yet
effective CAC system. Furthermore they do not address the activities of the installation engineer who is
responsible for executing the system design and commissioning the system. The comparison of the energy
performance of eighteen different CAC systems, each designed to provide total cooling in a typical EU
office, has shown that there can be a difference of a factor of 2 in the total energy consumption per m2 of
cooled space for typical configurations of CAC equipment using average efficiency components. The same
results have also indicated that while the proportion of energy required for heating and cooling may vary
appreciably from one climate and Member State to another, the absolute annual energy consumption per m2
shows a much smaller variation and follows a trend that can be related in a roughly proportional manner to
the annual cooling and heating degree days. These limited results imply that it might be feasible to develop
simple benchmarks of overall CAC system performance as a function of the level of cooling and air quality
required, the building type and of the cooling and heating degree days.
Policy measures would aim to encourage the adoption of more efficient CAC systems while maintaining the
freedom of the system designer to achieve a solution which meets the cooling, environmental and air quality
requirements of the client within acceptable cost boundaries.
As such building codes are the primary policy measure which promote the adoption of more efficient system
types; however, these can be supported by the following measures:
• The provision of analytical tools and technical guidance enabling the energy efficiency of CAC
systems to be optimised
• Training of system designers and installers on the options regarding energy efficient CAC systems
• The provision of financial and fiscal incentives to help overcome split incentives such as the
provision of cheap credit for efficient systems
Building code requirements are articulated in quite different ways among EU Member States. One difference
regards how the requirements for energy using systems in building codes should be expressed. In the UK the
primary policy goal is carbon abatement and therefore the requirements are expressed in terms of maximum
allowable emissions of carbon per m2 per year. In some Member States the building codes are articulated in
terms of limits regarding the maximum allowable energy consumption per m2. Once the fuel of the heating
and cooling system has been fixed these two approaches are effectively transposable; however, the carbon
approach provides an additional degree of freedom for designers to satisfy the requirements through
optimisation of the choice of fuel used by the system. The approach of setting limits for either energy
consumption or carbon emissions per m2 leaves the designer almost complete freedom to decide how they
are going to satisfy the requirements. In Portugal, however, there are no requirements on overall annual
energy or carbon but rather a set of simple prescriptions to follow, which are designed to save energy. Both
approaches have their merits and indeed can even be integrated as is the case in the US ASHRAE 90.1-1999
standard. In this standard the designer is obliged to follow some general requirements and mandatory
measures for each technical section but thereafter has a choice of two final compliance pathways: following a
further set of simple prescriptive requirements or demonstrating compliance by satisfying the “energy cost
budget”, method which requires the use of one of a number of designated detailed building thermal
simulation tools. Following the simple prescriptive measures is an easy way for a designer to demonstrate
190
their compliance with the standard; however, the prescriptions are relatively rigorous compared with the
requirements when a detailed simulation tool is used. Therefore the standard creates an incentive for
designers to use detailed building thermal simulation software. The combination of a prescriptive compliance
pathway and a pathway based on meeting overall energy limits demonstrated, through the use of detailed
simulation software, simultaneously meets the needs of designers “in a hurry” dealing with standard design
briefs and those who have specific and complicated design briefs; who may need more freedom to meet the
same energy goals than would otherwise be allowed through application of a set of simple prescriptions.
Legal basis for policy measures targeting more efficient system structures
As mentioned, building codes are the primary policy measure available to encourage the adoption of more
efficient system types. The recent Energy Performance in Buildings Directive obliges Member States to
develop mandatory minimum energy performance requirements for buildings and specifies that these should
encompass the energy used by mechanical cooling systems. In the case of the UK this Directive may simply
require minor revisions of the existing regulations for building cooling energy performance, but for most
other Member States it will require completely new regulations to be developed addressing the cooling
system. The requirement that a simple calculation method should be constructed against which the
compliance of the MEPS is to be judged may also require modification of the Portuguese regulations.
The need to develop minimum energy performance requirements within building codes for so many EU
states raises the question of whether it would be appropriate for Member States to co-operate with each
other.
The ENPER-TEBUC study within SAVE deals with the issue of harmonisation in European Building Codes
and has set up a platform for information exchange on Energy Performance (EP) standardisation and
legislation among the prominent national players. The intention has been to systematically collect and
summarise the different approaches and to develop suggestions for a European 'model code'. Ultimately such
a code could be the EU equivalent to the US ASHRAE 90.1 standard, which is non-binding in itself but can
be brought either wholly or in-part into national regulations as deemed appropriate by the authorities in each
Member State. Sharing development and analytical effort makes considerable sense for such a major
undertaking and is a key recommendation from this study. In the longer term the EU model code could be
designed to enable a energy efficiency labelling or grading system for CAC systems of say an A to G type.
The difficulty in obtaining the best grades (closer to A) should be increasing and involve not only the
manufacturer but also other elements of the chain. Moving from G to F or E might be based only on full-load
ratings i.e. readily available EER values. Higher grades such as E or D could be reached on the basis of a
certain value of SEER, taking part load optimisation efforts into account. Part load is not only a phenomenon
to be computed, it's by itself one of the most important energy saving features.
Following these measures the importance of system design cannot be neglected. A designer could refer to a
design procedure proving he has considered all cost effective options of the project. and reached a certain
performance level like C (- 25 %) or B (-50 %) or even A (-75%) compared with the average European
performance level, just as is currently the case in the US Energy Star for buildings scheme.
In absence of CEN standardisation on many subjects, existing methods in some countries could be
provisionally approved as ways to reach an A or a B. This and others can make it possible to go quickly for
the third generation suggested here. As any other standard this one would be applicable only voluntarily,
here by the will of the households, companies, local authorities wishing to have buildings consuming ¾ or ½
of European average of the new buildings in 2000 and to have this environmental effort certified from
outside.
The satisfaction of the requirements would need to be determined at the design/installation stage through
verification using simulation tools. This implies that public domain simulation tools would need to be
developed to support the model building code. There are many such efforts under way in individual Member
States and there would be considerable value in co-ordinating national efforts within an EU umbrella project.
191
Specific recommendations
The 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)
For HVAC control systems introduce building code measures which encourage:
• Restrictions on dead-bands
192
1. Shutoff damper controls that automatically close when the systems or spaces served are not in use
(these dampers should also satisfy a maximum allowable leakage rate.)
2. Zone isolation capabilities that permit areas of the building to continue operating while others are
shut down
3. Automatic shutdown
4. Setback controls
5. Optimum start controls after the system airflow exceeds a minimum level
For refrigeration plant systems introduce building code measures which encourage:
• 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)
• The use of variable flow hydronic systems and wherein pumps draw substantially less power at part-
load than full-load
• Limits on the Joule heating (e.g. electric heating power provided by the Joule effect should not
exceed 5% of the total heating power installed, nor 25kW by independent zone).
• Limits on terminal re-heating. (e.g. terminal re-heating is allowed for cooling-only systems but can
not exceed 10% of the installed cooling power).
For the installation and commissioning of CAC systems introduce building code measures which
encourage:
• commissioning tests to be conducted for boilers, chillers (power and efficiency), cooling towers,
pumps, hydraulic tests, heat exchangers, controllers, noise levels and overall functionality
In addition the following specific measures which impact of CAC system design and installation are required
under the Energy Performance in Buildings Directive:
193
• Energy certification of new and existing buildings to verify their compliance with minimum energy
performance requirements
• Certificates are to be made available to the owner or prospective tenant when the building is
constructed and when it is sold or rented out. Certificates shall not be valid for more than 10 years.
• The certificate shall include reference values such as legal standards and benchmarks. It shall be
accompanied by recommendations for the cost-effective improvement of the energy performance .
Recommendations about these requirements are that Member States should consider:
• Ensuring that a company or entity independent from the designer and installer should conduct the
building certification
• That certifiers are trained and clear certification procedures have been developed and adopted
• That the certification would be automatically triggered for all new installations and would verify that
both the installed systems and their individual components meet the energy performance
requirements and attain their pre-declared performance ratings (this would require be able to
measure the electrical energy, flow rates and temperatures of installed systems and their
components)
• That individual items of equipment which are performing at lower than rated efficiency levels are
clearly reported in the certification procedure.
194
This requirement for “independent” inspectors presumably precludes that the inspection should be conducted
by the entity holding the operation and maintenance contract and therefore should provide some measure of
independent verification of the proper conduct of those contracts. However, it is important that Member
States implement this measure in such a way that a review of operation and maintenance contracts are
encompassed within the inspection. In parallel it would be very useful that efforts be made to define best
practice in operation and maintenance so that best practice guidelines can be issued against which ever
existing contracts would be compared. Some aspects of these guidelines would necessarily be specific to the
situation applying in individual Member States, but some would be common to all Member States. Therefore
it would be very beneficial were the European Commission to take a lead in organising an EU working party
charged with defining best practice in the operation and maintenance of AC systems, so that the findings
could be fed into the national provisions being drawn-up by Member States. If the Commission is unable to
initiate this process actors at the Member State level charged with implementing the Energy Performance of
Buildings Directive could take the initiative to establish a working party of willing Member States.
The objective of the proposed Directive on Energy Demand Management (also known as Energy services) is
to complete the internal market for energy by developing and encouraging energy efficiency on the demand
side, especially as it is provided by utilities and service companies in the form of energy services. It is
envisaged that Member States will set targets to promote and support energy efficiency services, (e.g., third
party financing) and programmes, especially for smaller energy consumers such as SMEs. This could
certainly be used as an opportunity to improve the O&M of CAC if appropriate rules can be defined.
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 performance contracting
With an aim of informing national building thermal regulations and the implementation of the Energy
Performance in Buildings Directive.
Member States should consider enacting measures to ensure that:
• Building owners and occupiers are provided with a logbook and an adequate operation guide for the
CAC system deployed
195
• A regular maintenance and monitoring system be adopted for all CAC systems (e.g. impose a
requirement for regular maintenance and independent metering of CAC systems above a minimum
size)
• That a competent inspectorate be developed capable of carrying the provisions of the Energy
Performance in Buildings Directive applying to AC systems
• One of the roles of the inspectorate required under the terms of the Energy Performance in Building
Directive should be the independent review and evaluation of CAC system operation and
maintenance contracts
Member States could also consider the development of low cost finance mechanisms to encourage the
adoption of good practice for CAC operation and maintenance.
196
Definitions and general terms used in the study
Appliance category A group of appliances or equipment that have similar technical characteristics from the
perspective of their user utility.
Categorical energy label An energy label that classifies product efficiency into one of several classes. Examples
include the EU’s energy labels, which rank efficiency from A to G, and Australia’s energy
label, which ranks efficiency from 1 to 6 stars. Korea, Thailand, Iran, Brazil, Mexico and
India have all developed categorical energy labels.
Control cycle The period between two successive starts or two successive stops of the compressor in a
refrigerating system.
Defrost cycle The period between two successive starts or two successive stops of a defrost heater in a
appliance with an automatic defrost system.
Design temperature The temperature within a conditioned space that needs to be achieved during a test for the
energy-consumption measurement.
MEPS Minimum energy performance standards (sometimes known as ‘minimum energy
efficiency standards’).
Montreal protocol The internationally binding agreement to phase out ozone-depleting substances such as
CFCs.
Net present value (NPV) The monetised value of future costs expressed in terms of their discounted value at the
present time.
Payback period (PBP) The period of time it takes for a consumer to recover the extra investment made in a
higher-efficiency appliance through savings in operating costs. The payback period can be
‘simple’ in that no discounting of future savings is applied, or it can be the converse in
which the future savings are discounted using a real discount rate.
Thermal bridge A high thermal conductivity pathway.
Top Runner The term applied to the Japanese appliance energy-efficiency policy, wherein MEPS have
been set at efficiency levels equivalent to those of the highest efficiency appliance on the
market.
List of abbreviations
AC Air Conditioning
AS/NZS Joint test standards issued by Standards Australia and Standards New Zealand
197
CAC Central Air Conditioners
CEC Central and Eastern European Countries or Commission of the European Communities
CFC Cloroflurocarbons
DD Degree Days
EU European Union
HFC Hydrofluorocarbons
MEPS Minimum Energy Performance Standards (same as MEES, Minimum Energy Efficiency Standards)
198
RSECE The Portuguese building thermal regulations which include requirements for AC systems
SPF Specific Fan Power, in Watts (of motor rating) per litre/second of airflow.
VA Voluntary Agreements
REFERENCES
AFCE, 2002, Adnot J., Becirspahic S., Marchio D., Colomines F., Rivière P., Seasonal efficiency of primary
air conditioning systems, Procedings of the AFCE conference, Ecole des Mines de Paris, Sept 2002.
AICARR,2001 Average weighed efficiency of compression chillers: AICARR’s proposal for a calculation
method
AICARR, 2001, E. Bacigalupo, C Vecchio, M. Vio, M. Vizzotto., 2001, Average weighed efficiency of
compression chillers: a proposal to AICARR for a calculation method, Permanent Technical Committee for
"Refrigeration" in AICARR's Technical Activity Commission.
ARI, Standard 550/590, Water Chilling Packages using the vapor compression cycle, 1998.
ASHRAE Handbooks Fundamentals and Systems. J.F Kreider, A.Rabl, 1994, Heating and Cooling of
Buildings, Design for Efficiency, Mc Graw-Hill Book Company.
J.Bouteloup, M.Le Guay, J.Liguen, 4 vol, 1996, 1997, 1998, 1999, Air-conditioning, Air Handling, “Les
Editions Parisiennes”, France.
CEN, 1997, « Air conditioners and heat pumps with electrically driven compressors – Heating mode”,
standardCenelec EN 255, 1997.
CEN, 1998, « Air conditioners and heat pumps with electrically driven compressors – Cooling mode”,
standardCenelec EN 12055, 1998.
199
CEN, 2002, prEN 14511 – 1 à 5, proposed revision of [CEN98] and [CEN97], 2002.
CIBSE, 2004, "Guidance for the use of the carbon emissions calculation method: CIBSE TM32:2003" (ISBN
1 903287 41 3)
Ph Davy de Virville, 1994, Control Guide for Ventilation and Air-conditioning, thermic school, Paris,
France.
EERAC [1999] Energy efficiency of room air-conditioners. Paris: Ecole des Mines de Paris et al, for DG-
TREN, the Commission of the European Communities , SAVE contract DGXVII4.103/D/97.026, May.
Eurovent, 1998 Eurovent-Certification. 1998, Annuaire des Produits Certifiés. www.eurovent-
certification.com.
LBNL (2002) ‘Commercial Unitary Air Conditioner & Heat Pump: Life-Cycle Cost Analysis -
Inputs and Results’, Lawrence Berkeley National Laboratory for the US Department of Energy, December
2002.
RHEVA, 1993, The international dictionary of heating, ventilating and air-conditioning, E&FN Spon,
England.
REHVA, 1997 , Preliminary European Guide on HVAC design, Dominique Marchio, Eric Auzenet, ECEEE
1997 summer study proceedings
Rivière, 2001 general presentation and use of a method of calculation of consumption of the RAC (Room Air
Conditioners) by Ph. RIVIERE, J. ADNOT, M. ORPHELIN
Tiax (2002) ‘Engineering-analysis cost-efficiency curves. Commercial unitary air-cooled air-conditioners and air-source
heat pumps’, TIAX LLC, Cambridge Massachusetts for the US Department of Energy, January 2002.
TRIBU, (1994). "Etude Comparative des Réglementations Thermiques Européennes des Bâtiments non
résidentiels", TRIBU, 1994 for Ademe
UNI, (2002) UNI Standard 10963 : Air conditioners , chillers and heat pumps- part load tests.
Westphalen, 1999) "Energy Consumption Characteristics of Commercial Building HVAC Systems Volume
II: Thermal Distribution, Auxiliary Equipment, and Ventilation" Prepared by Detlef Westphalen and Scott
Koszalinski, Arthur D. Little, Inc. For U.S. Department of Energy, October 1999
200