Applied Energy 91 (2012) 222234

Contents lists available at SciVerse ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Buildings dynamic simulation: Water loop heat pump systems analysis

for European climates
Annamaria Buonomano, Francesco Calise, Adolfo Palombo
DETEC, University of Naples Federico II, P.le Tecchio, 80, 80125 Naples, Italy

a r t i c l e

i n f o

Article history:
Received 13 June 2011
Received in revised form 21 September
2011
Accepted 21 September 2011
Available online 19 October 2011
Keywords:
WLHP system
Building dynamic simulation
HVAC performance analysis
Energy and economic saving

a b s t r a c t
In this paper, a purposely designed code for the performance analysis of the Water Loop Heat Pump
(WLHP) systems is presented. Hourly, daily and seasonal energy system consumptions, operating economic costs and environmental impact assessments are dealt with. For the scope of comparison, the performances of two reference HVAC system are investigated too. For the computation of the building
heating and cooling requirements, a suitable dynamic performance simulation model is being developed.
All the relevant algorithms are implemented in MATLAB. A case study of an ofce building undergoing
simulation in different European climatic areas is being presented. Here, different building thermal features are considered. In order to maximize the system performance an additional optimization procedure
to the operating devices temperatures is carried out. Results show that primary energy savings and
avoided CO2 emissions of the WLHP system vary in relation to the compared reference systems and
can be obtained only in several European weather zones. The feasibility of the WLHP system strongly
depends on electricity and natural gas national costs.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
In buildings where space heating and cooling loads simultaneously occur, a Water Loop Heat Pump (WLHP) system can be
conveniently adopted [1,2]. Basically, it consists of a set of heat
pumps that reject to a water loop the excess heat from cooled
space. Such heat is recovered by other heat pumps and transferred
to spaces in need of heating. In the water loop, the occurring heating or cooling decits are balanced by additional heaters and/or
cooling towers. WLHP systems are typically installed in edices
with distinguished core and perimeter zones or commercial building with deep freeze or cold stores. A basic scheme of a WLHP system is reported in Fig. 1.
Such systems were developed in the 1960s in USA, they became
widely popular and applicable since 1990s mostly in USA and
Japan. In recent years, several studies were carried out aiming at
evaluating the system component features and relative operating
parameters. An investigation concerning the WLHP systems environmental contribution to a green building environmental control
is reported in [3]. In this study, alternative options to increase the
buildings energy performance were considered. A comparison between conventional air-conditioning systems and a WLHP system
for a number of Chinese climatic zones is carried out in [4]. An
interesting analysis of the WLHP performances on four different
Corresponding author.
E-mail address: palombo@unina.it (A. Palombo).
0306-2619/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2011.09.031

kinds of buildings was presented, where the devices efciencies

of the simulation model are kept constant [4]. In order to increase
the system energy saving, other authors studied the combination
of WLHP systems with gas-engine-driven heat pump (GHP) [5]
and coupled with low-temperature geothermal sources [6,7]. In
particular in [6] the evaluation of system performance and energy
saving for commercial and public buildings is carried out. The efciencies of the water source heat pumps are considered dependent
on the loop water temperature that ranges between 16 and 32 C. A
constant cooling load prole is adopted for the core building zone.
For the perimeter zone, the heating load is assumed linear to outdoor temperatures. In [7] WLHP system is applied to three towershaped apartment building in Beijing (China) where well water is
used as the low-temperature heat-source. The system performances are analyzed using a eld-test data obtained by running
the system over two winters and a summer. The system controlling
conditions are also investigated. In [8] a given test building load
prole and a single type of WLHPs equipped with a variable speed
compressor and a cooling tower with a variable speed fan are considered in order to nd out the optimal loop water temperature
minimizing the WLHP overall energy consumption.
In this paper, a detailed, purposely-designed performance simulation model for the building-WLHP system is presented. Its computer implementation, obtained by MATLAB, allows assessing
hourly, daily and seasonal building-HVAC system performances,
from an energy, economic and environmental points of view. This
tool allows the variation of system running parameters in contrast

A. Buonomano et al. / Applied Energy 91 (2012) 222234

223

Nomenclature
C
b
c
ce
cg
E
G
H
h
K
I
L_
_
m
Q_
Q
r
R
S
t
T
THVAC
U
Z

Subscripts
A
air
AW
air to water electric chiller and traditional natural gas
boiler
B
boiler
bal
balance
C
compressor
Co
condenser
Cool
cooling mode
CT
cooling tower
db
dry bulb
Ev
evaporator
e
electric
ext
external surface
g
natural gas
Heat
heating mode
HVAC
referred to the HVAC system
in
indoor
int
internal surface
o
outdoor
TB
traditional system boiler
TOT
total
V
ventilation
w
water
WLHP
water loop heat pump
WW
water to water electric chiller and traditional natural
gas boiler
wb
wet bulb
y
year

economic cost ()
capacitance (J/K)
specic heat (kJ/kg K)
electricity unitary cost (/kW h)
natural gas unitary cost (/N m3)
primary energy consumption (MW h/y)
Gebhart matrix for long wave radiation ()
hour
specic enthalpy of the moist air (kJ/kg)
solar radiation (W/m2)
solar ux intensity (W/m2)
electricity rating (kW)
mass ow rate (kg/s)
heat load (kW)
heat (kW h)
resistance (K/W)
saving
surface (m2)
time (s)
temperature (C)
traditional HVAC
thermal transmittance (W/m2 K)
building thermal zones

Greek letters
a
absorptance ()
C
long wave internal radiation (W)
U
external radiation (W)
W
short wave internal radiation (W)
e
long wave radiation emissivity ()
g
efciency ()
h
running hourly ratio ()
k
conductivity (W/m K)
q
density (kg/m3)
s
optical transmissivity ()
r
StefanBoltzmann constant (W/m2 K)

Superscripts
entering the WLHPs

exiting the boiler
e
economic
H
heating mode only
S
load simultaneity
C
cooling mode only
0

to other available commercial software which do not allow several

system conguration to be stimulated. A comparison of the WLHP
system performance vs. the Traditional HVAC (THVAC) systems is
also carried out. The devices-efciencies are variable in relation
to the systems operating conditions and an optimization procedure
on the water loop temperatures to maximize the systems performance is also implemented. A case study relative to large ofce
buildings is nally presented. Simulations correlate to a number

of European climatic zones. Both existing building and new construction components features are selected according to the outdoor climate. The performance analysis of the WLHP system for
different European climates and kinds of buildings is novel with respect to what is published in the recent literature. A primary and
basic simulation model in addition to some partial results about
the system performance are presented in [9,10].

2. Modelling
Pump

Boiler

Cooling tower
Pump

Tw

Heat exchanger

Tw

WLHP

WLHP

WLHP

WLHP

Fig. 1. Basic scheme of a WLHP system.

WLHP

In recent years, several numerical simulation models for the

building-HVAC system performance evaluation were developed
[1114]. In general, such simulation tools can be categorized with
respect to the tasks they are meant to full (equipment sizing and
selections, energy performance analysis, system optimization, control analysis, etc.). Although the current generation tools tend to be
fully integrated with respect to different building performance aspects, the integration between building and HVAC system models
is accomplished at different levels. However, for comparing HVAC
system alternatives and evaluating different control strategies,
detailed HVAC system models are required [15]. In this paper, in
order to well quantify the effects of the input parameters to the
output results of a numerical simulation study, a purposely designed building-HVAC dynamic simulation model was developed.

224

A. Buonomano et al. / Applied Energy 91 (2012) 222234

The assessment of the building heating and cooling requirements is

the rst step to be carried out in a detailed simulation code. In the
rst part of this paragraph such model is presented in a short form.
In the second part, the model created for the performance analysis
of both the WLHP and THVAC systems is reported.
2.1. Dynamic model for the building heating and cooling energy
assessment
A completely new simulation model was suitably written. Here,
the thermal conduction through the opaque and transparent elements is assumed as one-dimensional (thermal bridges are separately taken into account). The internal heat gain, the building
ventilation and the solar radiation loads are included in the calculation process. The radiation heat transfer of the external opaque
surfaces takes into account the solar and the long wave effects
[16,17]. In particular, the sun incidence angles are calculated as a
function of time and surface building orientation [18].
The modelled windows are supposed to be multi-glazed and
lled by different gases to meet the desired thermal transmittance.
Solar radiation entering through the windows is absorbed, reected and distributed within the internal space by selected
absorption, reection and view factors, respectively. The internal
long-wave radiation is handled by the Gebhart method [16,17].
Depending on the buildings envelope thermal inertia, such absorbed heat contributes to the heat gain during the winter season
and to the space cooling load in summer. In order to reduce the
summer solar heat gains, the windows external solar shadings
are considered. The control model of their tilt angle, and thus of
the windows shading coefcient, is here based on the standard
requirement of the average indoor horizontal lighting level [19].
In order to simulate the dynamic thermal response of each building element (wall, ceiling, oor and window), a thermal network of
n nodes is implemented, [20,21]. Here, for each m building element,
n depends on the modelled isothermal layers and it is selected by a
trade-off analysis taking into account the computational time and
the accuracy of the calculation procedure, simultaneously.
In order to calculate the heating or cooling energy requirement
(QHVAC) for each building space, the following system of differential
equations is used:

8 bm;n dT m;n T 1;n1 T 1;n T 1;n1 T 1;n

>
< ds rm;n1 rm;n Cm Wm Um
M
P
T m;N T in
>
_ V ho  hin  Q_ HVAC
: binddTs in
Q_ GAIN m
r m;N
m1

Cm Sm em;int r

M
X

Gm;j T 4j;N  T 4m;N

j1

Cm 0 only for internal surface n N

Wm Sm am;int Im Wm 0 only for internal surface n N
h

Um Sm em;ext rT 4sky  T 4m;1 am;ext K m

Um 0 only for external surface n 1

m 1M
n 1N
where T, r and b are the thermal network temperatures, resistances
and capacitances, respectively. Q_ GAIN is the total internal gain due
to human activity and electrical devices. I is the total solar radiation
ux received by an internal surface and is calculated as the sum of
radiation directly received without considering multiple reections,
plus the sum of the solar irradiance reected by others surfaces [16].
_ V and h represent air ventilation rate and moist air enthalpy,
m
respectively. Q_ HVAC is calculated according to a Proportional Integral

(PI) control on the indoor air conditions [18]. Temperature and

humidity can range in xed intervals, avoiding the HVAC system
activation.
Note that the results obtained by the mentioned new dynamic
model were compared with those achieved by a standard simulation code (TRNSYS 17, type 56 for buildings [17]). For different
building geometries and European weather data, the detected differences ranged from 4% to 8%. All the details concerning this specic proposed dynamic simulation code will be presented in a
future paper.
2.2. WLHP system model
Starting from calculation results of space design loads, two different logics can be followed by such code. The rst one includes
specic purposely sized WLHP for each space while in the second
logic different WLHPs of the same size are taken into account. In
order to balance the eventual water loop thermal energy decits
or excesses due to the WLHPs running, a natural gas boiler or a
cooling tower are implemented accordingly. Steady state conditions are assumed in each systems running hour (H-th) [22].
The eventual heat provided by the boiler in the H-th hour is calculated by:

_ w H  c  T 0w H  T w H  H
Q B H m

_ w H  c is the product of the water loop mass ow rate in

where m
the H-th hour by the liquid water specic heat while Tw(H) and
T 0w H are respectively the entering and exiting water temperatures
of the boiler in the H-th hour (Fig. 1). T 0w H is modelled taking into
account the running constraints of the WLHP system devices.
The primary energy consumption and operating economic cost
of the boiler are calculated as a function of the boiler efciency, the
natural gas lower heating value and the natural gas unitary cost.
While boilers allow controlling the outgoing water temperature, this is in general not possible when using cooling towers. In
this model, the steeper the water loop temperature increasing
trend is, the stronger and faster the cooling towers reaction is.
Here, different operational steps are possible: when the outdoor
air temperature is lower than the temperature in the water loop
a free cooling of the loop water is activated using the cooling tower
as a dry heat exchanger. In the case that such process is not sufcient to decrease the water temperature (i.e. for still high exiting
temperature) the cooling tower water pumps are switched on.
For medium-difcult working conditions, a rst fan set is activated
increasing the cooling tower efciency. For heavy working conditions, a second fan set in the cooling tower is switched on with a
further increase of its efciency. This algorithm is implemented
in a new simulation code. Subsequently, the hourly water exiting
temperature is iteratively calculated taking into account the design
and operating conditions of closed circuit cooling towers [23].
The primary energy consumption and operating economic cost of
the cooling tower in the H-th hour are calculated according to the
conventional average electricity production efciency in the power
plant and the electricity unitary cost (which vary between countries). The cooling tower water consumption is here disregarded.
The loop water at T 0w (Fig. 1) reaches each WLHP where the following calculation is done. Depending on the supply water loop
temperatures, variable cooling and heating capacities Q_ Ev H and
Q_ Co H and compressor ratings L_ C H are accounted. Since a WLHP
onoff regulation is considered, the hourly rate hC(H) in which each
WLHP compressor when switched on is calculate as:

Cooling : hC H Q CHVAC H=Q_ Ev H

Heating : hC H Q HHVAC H=Q_ Co H

225

A. Buonomano et al. / Applied Energy 91 (2012) 222234

where Q CHVAC H and Q HHVAC H correspond for each i-th indoor space
to term: QHVAC obtained by equations system (1).
In the occurring H hour the Tw that has to be considered with
respect to the following simulation time step (H + 1) is calculated
by the following energy balance on the water loop:
Z
Z
X
X
Q_ Co H  hC Hi;Cool 
Q_ Ev H  hC Hi;Heat
i1

i1

_ w H  cT w H  T 0w H
m

where Z is the buildings space. If on the water loop the sum of the
heat rejected by the WLHPs running in cooling mode is higher than
the sum of the cooling load of the heating WLHPs, then: T w > T 0w
and vice versa.
For each building space and relative to hC(H) the systems primary energy consumption and electricity cost in the H-th hour
are calculated. For the whole building and in the same simulation
hour the sum of these primary energy consumptions (EWLHP(H))
and costs (Ce,WLHP(H)) are added to the eventual ones due to the
boiler (EB(H), Cg,B(H)) or to the cooling tower (ECT(H), Ce,CT(H)). Thus,
the yearly WLHP system overall primary energy consumption and
operating cost are respectively computed by:

year
X
EWLHP H EB H ECT H

year
X
C e;WLHP H C g;B H C e;CT H

A parametric optimization procedure is carried out on the boiler

and cooling tower temperatures activation in order to get the systems minimum primary energy consumption, according to the
WLHP system constraints. In general, the higher the water loop
temperature is, the lower the heating mode WLHPs energy consumption is. This is due to the higher relative COPs. On the other
hand, the consumptions of boiler and cooling mode WLHPs are
higher. Conversely, the lower the loop water temperature is, the
lower the cooling mode WLHPs energy costs are. In this case the
consumptions of cooling tower and heating mode WLHPs are higher. In order to reach the optimal water loop temperatures, the following cases have to be distinguished because of the different
system constraints:
 case 1. Only heating is supplied to the building spaces by the
WLHPs (heating only mode);
 case 2. Both heating and cooling are simultaneously supplied to
the spaces by the WLHPs (simultaneous heating and cooling
mode);
 case 3. Only cooling is supplied to the building spaces by the
WLHPs (cooling only mode).
The optimization procedure in case 2 is obtained by simulating
all the possible combinations of boiler and cooling tower temperatures activation by an interval of 1 C.

 Water to water electric chiller (supported by a cooling tower)

and Traditional natural gas Boiler (TB) for cooling and heating
respectively. This plant is called WW traditional system.
For evaluating the performance of these systems a new suitable
model was written. The systems devices are sized by the building
design heating and cooling loads previously calculated. The heat
provided to the building by TB and the cooling energy extracted
by the AW and WW chillers are calculated by algorithms similar
to those considered for the WLHP system boiler and heat pumps.
Also in this case variable COPs are taken into account. Thus, the
yearly AW and WW traditional systems total operating cost is computed by:

C AWor WW

year
X
C g;TB H C e;AW Chilleror WW H

Included here are electricity consumptions from the fans of the

AW chiller and cooling tower and the pump of the WW chiller.
For both the WLHP and THVAC systems, the performance
parameters of all the devices are extracted from the manufacturers
input regarding a wide range of HVAC operating conditions. Note
that for outdoor air temperature below 20 C the AW chiller COP
increases while it remains almost constant for the WW chiller. This
is due to the following reasons: (a) for the AW chiller, the condenser fans between 20 and 0 C are proportionally and gradually
turning off, below 0 C the system is in free-cooling mode (fans and
compressor are switched off); (b) the COP curve value is constant
for the WW chiller when cooling tower water temperatures correspond to outdoor air temperatures and are between 20 and 0 C.
An optimization procedure for minimizing the primary energy
consumptions for the WW chiller cooling tower is also carried out.
For the economic analysis, the capital cost of WLHP and THVAC
systems and the operating cost due to the space air diffusion and
the water pumping are assumed to be almost the same [4].

3. Simulation
The simulated performance of a WLHP system is compared with
those obtained by the above described traditional systems. The
considered case study refers to a multi-oor ofce building. Length
and width are 40 and 20 m, respectively, the height of each oor is
3.5 m. The buildings longitudinal axis is EastWest oriented
(Fig. 2). The reported results were generated from a simulated 7oor building block. For each oor, a core zone (interior zone without direct transmission or radiation external loads) is divided by
partition walls with a surrounding perimeter zone (Fig. 2). For
the present case study, the ratio between the building core zone
volume and the total (core plus perimeter) volume is 0.54. For such
a building block, a self-sufcient HVAC system is considered. Since
adiabatic partition oors are hypothesized, only nine different
thermal zones are modelled (Fig. 2). The ratio between the windows surface areas and the building lateral areas is 0.50. Uniform
window distribution over the whole lateral surface is considered.

2.3. THVAC system model

Each building thermal zone is equipped by several 4-pipe fan
coils, supplied by chilled or hot water depending on the temporal
cooling or heating load. In the present simulation model, it is assumed that such processes are obtained by two different reference
systems [24]:

perimeter zone

N
core zone

perimeter zone

 Air to water electric chiller and Traditional natural gas Boiler

(TB) for cooling and heating, respectively. This plant is called
AW traditional system;

40.0 m
Fig. 2. Simulated building: plan view.

20.0 m

3.50 m

226

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Simulated indoor ventilation rates and internal load assumptions are reported in Table 1, here the considered values are selected according to the typical standards for ofce buildings.
The absorbed solar radiation of the external surfaces and the
sky temperature are calculated following the hourly TRY weather
data proles [25,26]. For all the exterior opaque surfaces on the
building, the absorption and convective heat transfer coefcients
are 0.5 and 18 W/m2 K, respectively.
The wall and window thermal properties (specic heat, density
and conductivity), are chosen starting from suitable U-values selected according to the building type and the outdoor climate.
From this point of view, the Heating Degree Days (HDD) index
[27] is considered and calculated with the TRY weather data for
all the considered climatic zones:

HDD

year
h X
t bal  tdb;o H
24 H

By the HDD indexes, six climatic zones are dened (Table 2).
The new ofce buildings (Building I) were simulated according to
the relevant HDD zone, with the referential thermal transmittances
equated to the threshold values reported in an Italian Decree concerning the building energy efciency (D.Lgs. n. 192/05 receiving
the EPB Directive 2002/91/EC) (Table 2). For existing building sim-

ulations (Building II), higher U-values are considered (Table 2). The
thickness of building walls ranges between 30 and 40 cm, depending on the weather zone: the harsher the winter climate, the higher
the considered wall thickness. The walls are composed by concrete
bricks (k = 0.429 W/mK, q = 1200 kg/m3, c = 921 J/kg K) and insulation (k = 0.039 W/mK, q = 25.0 kg/m3, c = 1250 J/kg K). For interior
partitions, here considered as adiabatic, the same conguration
and materials of the above mentioned walls are taken into account;
the thickness in this case is 23 cm. Ceiling and oor material properties are similar to those previously reported; the thickness in this
case is 30 cm. For windows, a double-glazed system is considered:
from 464 (4 mm glass, 6 mm internal gas, 4 mm glass) to 416
4. Standard or low-e glasses and different gasses (air or argon) are
considered in accordance with the desired relative thermal transmittance. The building time constant calculated according to [28]
ranges from 27.3 to 25.7 h for Buildings I and II, respectively.
Concerning the solar heat gain through the windows, it is assumed that the direct solar radiation is absorbed only by the oor
with an absorption factor of 0.3. Regarding long wave radiation, the
absorption and emission factors of interior surfaces are assumed
equal to 0.2 and 0.8, respectively. The simulated shading coefcient varies according to standard lighting requirements for ofces
[19]. The solar and visible window transmissivities for different climate zones are reported in Table 3 [29].

Table 1
Simulation assumptions.
Ventilation rate

Internal gain

Crowding index

Sensible to total heat ratio

1.4 Vol/h

Core zone lights: 10 W/m2 from 8.00 to 18.00

Perimeter zone lights: 10 W/m2 from 13.00 to 18.00
Core machinery: 15.0 W/m2
Perimeter machinery: 10.0 W/m2
People: 115 W/person

Core zone: 0.12 person /m2

Perimeter zone: 0.06 person/m2

0.870.91

Table 2
Reference U-values for Building I and Building II.
U (W/m2 K)

Climate zone

HDD < 600

601 < HDD < 900
901 < HDD < 1400
1401 < HDD < 2100
2101 < HDD < 3000
HDD > 3001

A
B
C
D
E
F

Building I ultimate limits (Italian D.Lgs. n. 192)

Building II

Opaque vertical wall

Window glass

Opaque vertical wall

Window glass

0.62
0.48
0.40
0.36
0.34
0.33

3.7
2.7
2.1
1.9
1.7
1.3

0.96

4.5
3.8
3.4
3.0
2.7
2.5

Table 3
Glazing system transmissivity.
Climate zone

Solar and visible transmissivity

Slat angle (deg)

0

30

45

60

90

svis
ssol

0.781
0.606

0.419
0.322

0.184
0.140

0.117
0.088

0.002

svis
ssol

0.781
0.606

0.419
0.322

0.184
0.140

0.117
0.088

svis
ssol

0.695
0.511

0.376
0.275

0.169
0.122

0.109
0.078

svis
ssol

0.695
0.511

0.376
0.275

0.169
0.122

0.109
0.078

svis
ssol

0.695
0.511

0.376
0.275

0.169
0.122

0.109
0.078

svis
ssol

0.753
0.510

0.400
0.273

0.174
0.120

0.109
0.076

A. Buonomano et al. / Applied Energy 91 (2012) 222234

The simulation starts at 00:00 of January 1st and ends at 24:00

of December 31st. The considered daily HVAC running time is from
8:00 to 18:00 (10 h per day). The calculations are carried out for 28
different European TRY weather zones (Table 4).
The threshold temperature, above which the HVAC cooling system is activated, is adjusted between summer and winter months
to assure satisfactory thermal comfort between zones during the
whole year. Activation temperatures were set at 26 C and 23 C
for summer and winter, respectively. A heating mode was considered for winter months for indoor temperatures lower than 20 C.
The adiabatic behavior hypothesis is supported by these temperature conditions, along with the assumption of thermally well-insulated interior partitions (3 cm, k = 0.039 W/mK) [18]. The relative
humidity of the indoor air is maintained between 50% and 60%.
The considered WLHPs temperatures constraints and the HVAC

systems simulation assumptions are reported in Table 5. The performance of all the considered devices for both WLHP and THVAC
systems are calculated for a wide range of HVAC operating conditions, as dened by manufacturers.
The operating algorithm of the centralized boiler and cooling
tower is summarized in Table 6, and varies according to the WLHPs
operating mode. The logic of the algorithm is as follows: if only
heating is required by the WLHPs, the loop water temperature is
maintained at the desired boiler exiting temperature T w . When
both heating and cooling are simultaneously required by the
WLHPs, the loop water temperature is maintained between T w
and the optimal cooling tower activation temperature T 0w;CT . For
the latter scenario, temperature control is achieved through several strategies: (1) boiler activation; (2) both boiler and cooling
tower standby; (3) cooling tower standby; (4) cooling tower free

Table 4
Climatic zones, HDD index and HDD zones.
Country

Climatic area

Latit. N

HDD (Kd)

HDD zone

UK

Lerwick
Eskdalemuir
Aberporth
Kew (London)
Copenhagen
Dublin
Valentia
Eelde
De Bilt
Vlissingen
Oostende
Uccle (Bruxelles)
Saint Hubert
Trappes
Nancy
Macon
Limoges
Carpentras
Nice
Bolzano
Venezia
Milano
Genova
Roma
Foggia
Cagliari
Crotone
Trapani

60080
55190
52080
51280
55460
53260
51560
53080
52060
51270
51120
50480
50020
48460
48410
46180
45490
44050
43390
46280
45300
45260
44250
41480
41310
39150
39040
37550

4024
3970
3178
2900
3696
3133
2741
3427
3194
2877
3147
3020
4188
3069
3245
2980
2899
2266
1650
3087
2317
2551
1560
1663
1759
1349
1409
976

F
F
F
E
F
F
E
F
F
E
F
F
F
F
F
E
E
E
D
F
E
E
D
D
D
C
C
C

Denmark
Eire
Netherlands

Belgium

France

Italy

Table 5
HVAC systems assumptions.
WLHP system
Heating
Minimum WLHP inlet water:
Tw = 8 C (DT = 6 C)
Maximum inlet water Tw = 23 C
Cooling
Minimum WLHP inlet water
Tw = 13 C
Maximum inlet water: Tw = 44 C
(DT = 6 C)
Boiler efciency, gB = 1.06
Natural gas lower heating value,
LHV = 9.59 kW h/N m3

AW chiller and boiler

WW chiller and boiler

Design heating water temperatures = 4540 C

Design cooling water temperatures = 712 C

Design AW chiller inlet air, T = 44 C (DT = 6 C),

minimum inlet air T = 16 C

227

Design WW chiller inlet water, Tw = 50 C (DT = 6 C), minimum inlet water

Tw = 20 C, design cooling tower water DT = 6 C

228

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Table 6
Boiler and cooling tower control logic.

cooling; (5) cooling tower activations determined by different temperature trends. The boiler activation depends on the water loop
temperature and the desired boiler exiting temperature. When
the water loop temperature ranges between the desired boiler
exiting temperature and the cooling tower activation temperature,
both boiler and cooling tower are on standby. For water loop temperatures higher than the cooling tower temperature, activation or
non-activation are both possible according to exertion strategy options. When the water loop temperature is lower or higher than the
outdoor air wet bulb temperature, the cooling tower is on standby
or in free cooling mode, respectively. The free cooling exertion on
the cooling tower is also determined by outdoor wet bulb temperature. Finally, the fth case is when the cooling tower activation
takes place for a water loop temperature higher than the cooling
tower activation temperature, T 0w;CT .
Here, the devices exertion is regulated proportionally to the
temperature increasing trend. The exertions are regulated by three
conditions. The rst one requires that the water loop temperature
is higher than the outdoor air wet bulb temperature while the latter is higher than zero; the second condition is based on the difference (k) between the water loop temperature and the cooling
tower activation temperature; the third condition is based on the
water loop increasing trend (p). Depending on the values of k
and p three exertions on the cooling tower pump and fans are
hypothesized. The very same cooling tower control logic is applied
when only cooling mode is required. Note that the optimization
procedure for minimizing the primary energy consumption led to
nding the ultimate activation temperatures for the boiler and
cooling tower devices in reference to the different European climatic zones.

4. Results and discussion

Table 4 reports the considered TRY weather areas, the HDD indexes calculated by (9) and the corresponding climate zones. Note
that the investigated locations are included between C and F HDD
zones and thus the corresponding buildings elements are modelled
starting from the corresponding features reported in Tables 2 and 3.
For Building I and Building II, the yearly heating and cooling energy requirements (QHVAC) vs. the HDD index for all the investigated
climatic regions are reported in Fig. 3. Although other weather indexes are often considered in literature [27], the HDD is preferred
here as an independent variable, not only for its popularity and availability with regard to many locations, but mostly for its satisfactory
correlation obtained with the calculated results. Regarding the potential saving of the WLHP system vs. the traditional ones, it is possible to observe in Fig. 3 that for both extremely low and high HDDs,
very low simultaneity is obtained because of very low heating and
cooling requirements. Thus, for these weather zones low or no savings can be achieved. Conversely, for HDDs around 3000 Kd for
Building I and 2000 Kd for Building II the heating or cooling demands
are quite similar and a potential saving is achievable. It is possible
however, that such occurrence is not realized as it is better highlighted in the following.
WLHPs operating modes are selected in relation to building
spaces heating and cooling energy requirements, subsequently.
The boiler and cooling tower switch-on temperatures are evidently a function of such modes. Note that the performance of
the WLHP system strongly depends on these activation temperatures. For the heating only mode and for all the considered zones,
a boiler minimum activation temperature of 8 C (Table 5) in light

229

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Building I
Building II

QHVAC [kWh/m2y]

QTOT

110

Heating Builiding I
Cooling Building I
Heating Building II
Cooling Building II
100

70
30
500

2500

4500

HDD

50

0
500

1500

2500

3500

4500

HDD [Kd]
Fig. 3. Building I and II: heating and cooling energy requirements.

Table 7
Building I: optimal boiler and cooling tower activation temperatures.
Climatic area

Lerwick
Eskdalemuir
Aberporth
Kew (London)
Copenhagen
Dublin
Valentia
Eelde
De Bilt
Vlissingen
Oostende
Uccle (Brux.)
Saint Hubert
Trappes
Nancy
Macon
Limoges
Carpentras
Nice
Bolzano
Venezia
Milano
Genova
Roma
Foggia
Cagliari
Crotone
Trapani

T 0w;B (C)

T 0w;CT (C)

Heating only

Heating and cooling

8 (6 6)

16
15
16
15
15
15
13
15
15
15
15
13
15
15
15
15
14
13
13
13
13
13
13
13
13
13
13
13

(11)
(9)
(12)
(11)
(9)
(9)
(10)
(9)
(9)
(9)

Cooling only
23 (25)
23 (25)
22
23
22
23 (24)
23 (25)
23 (24)
22
23 (24)
22
23 (25)
23 (25)
23
22
22
23 (23)
22
23
22
22
23 (25)
23 (24)
23
23
22
22
23 (26)

31
29
27
29
29
28
29
31
30
30
30
31
33
28
32
33
29
33
32
32
32
32
31
31
32
31
31
31

ture must not surpass 23 C (Table 5). The corresponding cooling

tower activation temperatures T 0w;CT , detected for Building I are reported in Table 7. Here, it is possible to observe that the theoretical
optimal temperatures often surpasses the constraint temperature.
If cooling only mode occurs, the maximum water loop temperature cannot exceed 44 C (Table 5). In this case the obtained optimal cooling tower activation temperatures T 0w;CT for Building I
range from 27 to 33 C (Table 7). The optimal switch-on temperatures of the traditional system cooling tower for Building I range
between 21 and 24 C. For sake of brevity, the results of the optimization procedure for Building II, however slightly higher than
Building I, are not reported.
An example of loads simultaneity on the water loop is reported
in Fig. 4. Here, the heating, cooling and total heat load proles from
a sample of 3 days in October are depicted for Building I which is
located in the weather zone of Copenhagen. In the high side of
the gure the whole simulated season is reported. In these days
a heating Q_ WL;Heat;TOT and cooling Q_ WL;Cool;TOT load simultaneity
on the water loop occurs (Fig. 4) and a potential WLHP system energy saving is detected. Note that the space cooling load becomes a
water loop heating load (lines in positive graph zone) and vice versa for the space heating load. The net result of thermal load on the
water loop Q_ WL;TOT is obtained by the simultaneous Q_ WL;Heat;TOT and
Q_ WL;Cool;TOT . In general, such net thermal load is usually balanced by
the centralized system devices. For the rst sample day of Fig. 4, to
balance the detected Q_ WL;TOT (>0) the cooling tower is activated
only when the water loop temperature reaches the optimal activation threshold T 0w;CT (22 C, Table 7). A similar phenomenon occurs
in the second sample day reported in Fig. 4. In the third day to balance the detected Q_ WL;TOT (<0) a boiler activation is necessary as a

80

QWL,Cool ,TOT
40

Q [kW]

of WLHP system constraints must be respected. The simulation

however shows more energy saving when temperature is lower
or equal to 6 C (see Table 7, where for Building I the theoretical
optimal temperatures T 0w;B are reported closed by round brackets;
such temperatures could be useful when future generation machines can over lap system constraints). Although a higher water
loop temperature causes higher WLHPs COPs, the overall energy
efciency of system decreases.
It was found that for the simultaneous heating and cooling
mode, the optimal boiler activation temperatures T 0w;B range from
9 to 16 C, all that in relation to the investigated zones, for both
Building I (Table 7) and Building II. To maximize the WLHPs COP
in the northern climate zones, the boiler optimal activation temperatures is likely higher than 13 C (which is a system constraint,
Table 5), unfortunately for others zones the calculated theoretical
optimal temperatures are not allowed (Table 7). When heating
and cooling are simultaneously required, the loop water tempera-

8760

QWL,TOT

-40

QWL, Heat ,TOT

7032

7056

7080

7104

hours [h]
Fig. 4. Building I: water loop heating and cooling loads proles (October at
Copenhagen).

230

A. Buonomano et al. / Applied Energy 91 (2012) 222234

30

8760

Tw

Tw [C]

25
20

Tdb

15
10

Twb

5
7032

7056

7080

7104

hours [h]
Fig. 5. Building I: temperature proles (October at Copenhagen).

result of the water loop temperature reaching the optimal activation threshold T 0w;B (15 C, Table 7). All such operating conditions
are clearly visible in Fig. 5, where for Copenhagen the temperature
prole of the water loop T 0w , Fig. 1), for the same above mentioned
sample days, are reported. Here, the proles of the TRY outdoor air
dry bulb and wet bulb temperatures are also shown. Both dry and
wet bulb temperature proles are provided for the whole simulated season in the high side of the same gure.
For both Buildings I and II the yearly primary energy WLHP system consumptions E are reported in Table 8. Lower yearly primary
energy consumptions of Building I vs. Building II are detected only
for the northern Europe zones. The opposite is true for the southern zones where lower yearly primary heating consumptions do
not offset the higher cooling ones. In the same table, the percentage share of energy consumptions due to the WLHPs (EWLHP), the
boiler (EB), and the cooling tower (ECT) are also shown. The colder

the zone the higher the boiler consumption, and vice versa for
the cooling tower. When comparing Building II to Building I the
boiler usage decreases; this is caused by two different effects.
The rst one is ascribed to the lower building heating demand
thanks to the better insulated envelope in Building I, the second effect is due to the increase of simultaneous heating and cooling
requirements. A higher utilization of the cooling tower is observed
in this comparison, however the corresponding primary energy increase is lighter than the boiler energy consumption decrease. Table 8 reports the relative primary energy saving of WLHP system
compared to both traditional systems (AW(RAW) and WW(RWW)).
The present study shows that for the northern European zones
RAW often surpasses 20% reaching in several cases 3035%. Higher
relative primary energy savings are achieved in Building II mostly
because energy requirements are higher when compared to Building I. RAW values are always found higher than RWW ones in both
Building I and II for all the investigated zones while they are somewhat similar for very high HDDs.
In Fig. 6 the primary energy requirements for all the considered
HVAC systems are reported for two sample locations. This graph is
to highlight the systems performances relative to different weather zones. A higher energy consumption is always detected in Building II when compared to Building I for northern Europe zones (e.g.
Copenhagen). This is mainly due to a better envelope response to
the typical high heating, low and very low cooling loads. Conversely, a lower energy consumption is always detected in Building
II when compared to Building I for southern Europe zones (e.g. Trapani). This is mostly a result of moderate heating loads in addition
to a difculty of the well insulated building to dissipate the high
internal and solar cooling loads. Note that, for locations typically
located at intermediate latitudes, where remarkable heating and
cooling loads are obtained (e.g. Bolzano and Venezia) the energy
consumption trends depend on the ratio of cooling and heating

Table 8
Primary energy consumptions and savings.
Country

Climatic area

Building I


h
E kW
y m2

EWLHP (%)

EB (%)

ECT (%)

RAW (%)

33.2
30.1
20.1
20.6

43.1
50.8
63.5
75.4

55.6
46.9
31.0
17.8

0.6
1.1
2.5
3.1

1.2
10.2
24.6
33.3

UK

Lerwick
Eskdalemuir
Aberporth
Kew (London)

Denmark

Copenhagen

34.1

65.3

28.1

2.5

Eire

Dublin
Valentia

19.9
18.0

70.0
71.3

23.4
20.0

2.8
3.7

Netherlands

Eelde
De Bilt
Vlissingen

29.8
27.2
24.7

67.2
72.5
70.7

25.1
19.0
20.1

Belgium

Oostende
Uccle (Brux.)
Saint Hubert

23.8
27.3
40.6

73.1
76.5
50.0

France

Trappes
Nancy
Macon
Limoges
Carpentras
Nice

27.2
30.5
34.6
27.6
39.7
45.9

Italy

Bolzano
Venezia
Milano
Genova
Roma
Foggia
Cagliari
Crotone
Trapani

35.7
45.1
43.0
44.0
45.0
43.5
45.3
46.3
57.0

RWW (%)

Building II


EWLHP (%)
h
E kW
y m2

EB (%)

ECT (%)

RAW (%)

RWW (%)

42.3
52.7
64.2
61.2

57.7
47.1
35.5
36.9

0.0
0.1
0.2
0.8

4.3
24.1
37.6
31.4

4.2
23.9
36.9
30.3

2.6
8.8
14.3
19.6

58.4
44.0
26.9
29.4

16.8

5.8

46.5

60.4

38.3

0.5

19.3

17.1

28.2
22.2

20.9
18.9

31.7
25.4

58.4
59.6

40.8
39.5

0.3
0.4

27.2
28.2

26.5
27.2

3.2
3.5
3.8

20.2
25.7
22.7

13.4
12.6
15.5

35.4
33.6
29.1

68.8
69.3
71.8

29.2
28.7
26.7

0.8
0.8
0.6

34.5
33.9
35.9

32.1
31.3
35.2

18.6
12.2
47.1

3.4
4.7
1.2

26.5
27.7
1.7

16.1
12.2
2.6

31.2
33.4
53.5

67.7
71.4
55.8

30.9
25.4
43.6

0.6
1.3
0.2

34.1
33.7
20.3

33.4
30.0
19.9

76.8
77.4
78.9
81.6
80.7
80.6

13.6
9.7
7.9
6.4
1.6
0.5

3.4
4.7
4.8
4.3
6.4
6.8

23.8
23.3
22.0
25.8
18.8
15.2

9.1
5.4
0.5
4.7
24.2
30.7

32.5
34.1
34.5
32.5
31.8
36.0

75.3
78.9
81.3
73.6
79.2
76.3

22.5
17.7
13.3
22.6
11.4
10.7

0.8
1.2
1.9
1.4
3.4
4.7

32.1
35.1
34.5
29.9
29.3
17.8

30.2
32.6
28.7
26.5
13.2
5.8

84.0
79.3
80.8
78.8
80.7
81.3
79.8
80.4
80.2

2.2
5.3
3.8
2.6
0.0
0.8
0.0
0.0
0.0

5.4
6.0
6.0
7.3
7.5
7.0
7.9
7.7
7.7

29.7
19.7
24.2
16.2
19.8
21.2
15.7
19.0
15.2

5.2
14.3
4.4
27.6
30.0
31.1
40.1
34.7
44.2

33.3
38.0
39.5
34.5
35.0
34.8
32.4
35.1
39.7

85.6
78.0
79.4
76.0
78.3
75.7
79.4
78.0
79.4

6.9
9.8
11.9
10.9
8.2
8.3
4.2
6.3
1.2

2.9
4.8
3.4
5.1
5.3
6.2
6.4
6.1
7.6

40.7
29.8
31.6
21.1
23.2
25.1
21.8
22.2
18.8

33.2
16.0
20.1
0.8
6.2
3.1
9.0
11.9
24.6

231

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Building I -WLHP
Building I - AW
Building I - WW

E [kWh/m2y]
80

Building II - WLHP
Building II - AW
Building II -WW

60
40
20
0
Copenhagen

Trapani

Fig. 6. WLHP and THVAC systems primary energy consumption.

loads as well as the HVAC efciencies. These ndings are also displayed in above described Fig. 3 which additionally shows that
lower heating and higher cooling requirements are always obtained by the better thermally insulated Building I vs. the traditional Building II. Heating requirements surpass the cooling ones
for high HDDs, the opposite is true for low HDD indexes. In particular, passing from the Building I to the Building II, the higher the
HDDs the higher the increase accounted for the heating requirements, for the cooling ones the opposite is true. The sum of the
heating and the cooling energy requirements of Building I and II
is depicted in the right top graph in Fig. 3. Here it is possible to observe that for high HDDs (northern Europe) the yearly overall energy requirement for both heating and cooling (QTOT) of Building
I is remarkably lower than that of Building II. The contrary occurs
for low HDDs (southern Europe). In addition a quasi steady state
behavior of QTOT is exhibited for Building II for a large HDD interval.
Fig. 7 shows primary energy savings as a function of the HDD index of the investigated climatic areas for both Buildings I and II. The
WLHP system energy saving vs. the traditional systems increases
with HDD until about 3000 Kd. For higher HDDs a decrease is detected. For all the southern Europe climatic regions (low HDD)
the considered WW traditional systems perform better than WLHP
ones. In general, the WW traditional system has lower energy consumption than the AW system while for high HDDs they report
similar results. This is essentially due to the AW and WW traditional system running principle. In particular, for zones with very
high HDDs the system running is obtained, for more than 75% of
the occurrences, with outdoor air temperatures below 20 C. For
this reason, and for the above mentioned considerations concerning the AW and WW chillers efciencies, the two traditional systems show similar performances. The energy saving depends also
on the electricity to gas consumption ratio. An increase of such ratio is everywhere obtained when shifting from traditional systems

to the WLHP one. Remarkable energy savings are detected in the

cold winter zones (about 3000 Kd), where high gas consumptions
were expected. For the coldest winter zones (HDDs higher than
3500 Kd) energy saving are very low, this is because of the low ratio between cooling and heating load. When no signicant cooling
loads are detected (see Fig. 7 for Building II) the savings are due
only to the gas to electricity consumption shifting. The very low ratio between cooling and heating load implies that only few WLHPs
are in cooling mode. In light of this, a higher consumption of the
boiler caused by the water loop temperature constraints (see Table 7, Tw > 13 C) is not counterbalanced by a lower heating consumption of the WLHPs due to the increase of their COPs. Under
these conditions Building I shows a weak simultaneity while the
WLHP system performs similarly to the traditional systems.
Fig. 8 shows how a remarkable amount of primary energy saving can be achieved through a decreased natural gas usage in the
WLHP system compared to the traditional ones in Uccle (Bruxelles)
and Limoges.
Table 9 presents the environmental impact analysis in terms of
CO2 emission due to the national electricity production (elCO2 and
the natural gas combustion gasCO2 . Note that such parameters return the same results as emission factors achieved by LCA (Life Cycle Analysis) since CO2 emissions of the manufacturing processes
are almost the same for the considered HVAC systems. CO2 emis2
_ CO2 , the AW system m
_ CO
sion for the WLHP system m
AW and the
2
_ CO
WW system m
,
are
reported
for
both
Building
I
and
II. In the
WW
same table, CO2 emission relative savings of the WLHP system vs.
CO2
2
both traditional systems RCO
AW and RWW are reported as well.
Remarkable CO2 savings are reached for many European climatic
regions. Note that all such values are strongly dependent on the
adopted method for electricity production in the relevant countries
CO2
2
[30]. The highest RCO
AW and RWW reached in France are due to the
very low corresponding elCO2 .
In Table 10 system operating costs and relative economic savings are reported. Here, the total WLHP operating costs C (VAT excluded) and the relative economic savings of WLHP vs. AW and
WW traditional systems ReAW and ReWW ) are reported. These calculations were subject to the countrys unitary costs of electricity and
natural gas (ce and cg, Table 10 [30]). Results show remarkable
money savings are obtained only in the high HDD zones. Note that
the percentage economic saving of the WLHP system vs. the traditional ones ReAW and ReWW can increase (e.g.: Denmark, Netherlands and France) or decrease (e.g.: all the other considered
Countries) vs. the corresponding energy saving. A primary energy
saving may even become an economic loss (e.g.: Bolzano vs. WW
traditional system). Conversely, a primary energy loss can become
a monetary saving (e.g.: Nice vs. WW traditional system). These
behaviors are mainly due to: (i) the electricity to natural gas

45

45

Building II

30

30

15

15

R [%]

R [%]

Building I

0
-15

-15

WLHP vs. AW
-30
-45
500

WLHP vs. AW
-30

WLHP vs. WW

WLHP vs. WW

-45
1500

2500

3500

4500

500

1500

HDD [Kd]

2500

HDD [Kd]
Fig. 7. Primary energy savings vs. HDD index.

3500

4500

232

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Uccle (Bruxelles)

Limoges

WLHP

WLHP

AW

AW

WW

WW

Building I

50

100

50

WLHP

WLHP

AW

AW

WW

WW

Building II

100

[%]

[%]

50

100

50

[%]

100

[%]
Electricity

Natural Gas

Fig. 8. Building I and II: electricity to natural gas consumption ratio for WLHP and traditional systems.

Table 9
Systems environmental impact.
Country

Climatic area

elCO2

g

CO2

kW h

gasCO2

g

CO2

kW h

Building I
_
m

UK

Lerwick
Eskdalemuir
Aberporth
Kew (London)

514

198

CO2

7.4
6.8
4.6
4.7

Building II
2

(kg/y m )

2
RCO
AW

(%)

3.3
8.2
22.9
31.7

_ CO2 (kg/y m2)

m

2
RCO
AW (%)

4.7
6.7
11.9
17.0

13.1
9.9
6.1
6.7

2.0
21.9
35.4
29.3

2
RCO
wW

(%)

2
RCO
wW (%)

1.9
21.6
34.7
28.0

Denmark

Copenhagen

604

7.7

15.1

3.4

10.5

17.1

14.7

Eire

Dublin
Valentia

575

4.8
4.3

24.3
18.7

15.8
15.0

7.5
6.0

22.1
23.4

21.3
22.2

Netherlands

Eelde
De Bilt
Vlissingen

548

6.6
6.1
5.5

19.3
24.8
21.8

12.3
11.3
14.4

7.9
7.5
6.5

33.4
32.8
34.8

30.9
30.1
34.0

Belgium

Oostende
Uccle (Brux.)
Saint Hubert

250

3.0
3.2
6.4

49.4
50.7
22.4

45.6
44.9
20.7

4.3
4.4
8.2

56.8
57.1
43.3

56.6
55.9
43.2

France

Trappes
Nancy
Macon
Limoges
Carpentras
Nice

81

1.5
1.5
1.5
1.1
1.3
1.4

70.2
73.3
71.6
74.6
61.1
49.3

69.2
72.2
70.0
73.1
54.2
36.4

2.3
2.1
1.9
2.3
1.6
1.8

75.6
79.0
80.5
73.2
73.2
61.7

75.5
78.9
80.2
73.0
72.0
59.1

Italy

Bolzano
Venezia
Milano
Genova
Roma
Foggia
Cagliari
Crotone
Trapani

483

6.8
8.6
8.2
8.3
8.5
8.2
8.6
8.7
10.8

33.9
22.3
27.7
17.8
21.1
22.6
16.7
19.9
15.5

12.7
8.7
2.3
23.7
26.6
27.2
37.3
32.2
43.5

6.4
7.3
7.6
6.6
6.7
6.7
6.2
6.7
7.5

46.4
34.7
36.8
25.0
26.4
28.5
25.1
24.9
20.4

40.3
23.2
27.3
5.8
0.2
3.7
2.6
6.0
20.7

consumption ratio obtained for different zones by zone shifting

from traditional systems to WLHP one; (ii) the national electricity
and natural gas costs. From this point of view a sensitivity analysis
should be carried out for each investigated climatic area: Trappes
(Paris) and Roma were selected as an example, results are shown
in the graphs reported in Fig. 9. For such locations the economic
savings ReWW increase and decrease almost linearly vs. cg and ce,
respectively. Fig. 9 also shows that for Roma no economic protability of the WLHP system is detected for whatever combination
of ce and cg. Almost the contrary occurs in Trappes.

5. Conclusions
Although the considered simplied approach does not allow
accurate system feasibility or operating analyses measures, interesting operating guidelines can be found in the presented results.
Optimal setting of the system operating conditions aimed at
reaching the lowest energy and economic costs were obtained by
the implementation of a new simulation model. The latter allows
to take into account the system constraints and to avoid traditional
empirical eld attempts. Special attention was paid in order to

233

A. Buonomano et al. / Applied Energy 91 (2012) 222234

Table 10
System operating cost and relative economic savings.
Country

Climatic area

ce

c
kW h

20500 MW h/y

cgas

 

GJ

01000 GJ/y

Building I
2

UK

Lerwick
Eskdalemuir
Aberporth
Kew (London)

11.71

12.06

Building II

C (/y m )

ReAW

1.6
1.5
1.0
1.1

8.6
3.1
18.6
27.6

(%)

ReWW

(%)

10.3
1.3
5.6
10.4

C (/y m2)

ReAW (%)

2.8
2.1
1.3
1.4

4.2
15.8
29.5
23.5

ReWW (%)
4.4
15.5
28.6
21.9

Denmark

Copenhagen

11.00

21.26

1.8

35.6

30.5

2.6

40.3

39.4

Eire

Dublin
Valentia

15.49

15.15

1.3
1.2

21.7
16.5

12.5
12.5

2.0
1.6

18.6
20.1

17.7
18.8

Netherlands

Eelde
De Bilt
Vlissingen

13.6

16.79

1.7
1.5
1.4

23.0
28.3
25.4

16.8
16.5
18.9

2.0
1.9
1.7

37.7
37.1
39.2

35.6
34.8
38.6

Belgium

Oostende
Uccle (Brux.)
Saint Hubert

14.8

16.20

1.4
1.6
2.4

25.1
26.4
0.2

14.1
10.2
4.3

1.9
2.0
3.2

32.5
32.1
18.6

31.7
28.1
18.1

France

Trappes
Nancy
Macon
Limoges
Carpentras
Nice

7.81

12.62

0.8
0.9
1.0
0.8
1.1
1.3

39.4
39.8
37.0
40.5
27.0
20.4

31.2
30.0
24.5
28.3
5.5
18.5

1.0
1.1
1.0
1.0
1.0
1.1

50.3
53.2
52.7
47.3
43.5
29.8

49.4
52.0
50.0
45.6
34.5
14.6

Italy

Bolzano
Venezia
Milano
Genova
Roma
Foggia
Cagliari
Crotone
Trapani

16.27

14.26

2.3
2.8
2.7
2.8
2.9
2.8
2.9
2.9
3.6

23.8
16.2
19.4
14.2
18.2
19.5
14.5
17.9
14.9

6.2
22.1
14.1
32.8
34.4
36.2
43.5
37.9
45.2

2.1
2.4
2.5
2.2
2.2
2.2
2.0
2.2
2.5

31.8
22.4
23.7
15.7
18.9
20.3
17.4
18.6
16.7

21.4
4.8
8.6
10.6
15.5
13.0
18.3
20.2
29.7

60
22
18
14
10

e
RWW
[%]

40
20
0

Trappes

-20
Roma

-40
-60
0.06

0.08

0.1

0.12

0.14

0.16

0.18

ce [ /kWh]
Fig. 9. Building I: economic saving vs. electricity and gas unitary costs.

demonstrate the difference between the WLHP systems performance and both traditional systems equipped by air to water
and water to water chiller and natural gas boiler.
A case study with two types of ofce buildings was developed.
As a function of the European TRY climatic areas and space loads
proles, the following simulation results were obtained.

- in cooling only mode when the cooling tower activation temperature is the optimal one obtained for the selected climatic
area.
 The WLHP system relative primary energy saving vs. traditional
systems depends on building heating and cooling simultaneity.
Concurrently, it depends on the increase of the electricity to gas
consumption ratio obtained by shifting from traditional to
WLHP systems. Such criteria are summarized in the HDD index:
the highest primary energy savings are detected in climatic
areas around 3000 Kd. In weather zones with extreme low
and high HDDs, moderate (or even null in water to water chiller)
energy savings were achieved.
 The CO2 emission of the investigated systems depends on the
obtained primary energy saving and in particular on the power
resource of the analyzed European Countries. The relative
avoided CO2 of the WLHP system reaches in some cases remarkable rates (around 40%).
 The eventual operating economic saving vs. the considered traditional systems strongly dependents on national electricity
and natural gas costs. The best results are obtained in Denmark
and northern France (between 35% and 40%).

References
 The minimum total energy consumption of the WLHP system
are obtained:
- in heating only mode when the ultimate boiler activation
temperatures are set equal the current WLHPs constraint;
- in heating and cooling simultaneous mode, for largest boilercooling tower temperature standby intervals. The WLHPs COP
increase for smaller intervals resulted not sufcient to counterbalance the subsequent higher energy consumption of the
system;

[1] ASHRAE 2005. ASHRAE handbook, Fundamentals. Atlanta: American Society of

Heating, Refrigerating and Air-Conditioning Engineers, Inc.; 2005 [Chapter 30].
[2] Wang SK. Handbook of air conditioning and refrigeration. 2nd ed. New York
(NY): McGraw-Hill; 2001.
[3] Gottfried DA, Schoichet EA, Hart M. Green building environmental control: a
case study. HPAC 1997;2:718.
[4] Lian Z, Park SR, Qi H. Analysis on energy consumption of water loop heat pump
system in China. Appl Therm Eng 2005;25:7385.
[5] Lian Z, Park SR, Huang W, Baik YJ, Yao Y. Conception of combination of gas
engine driven heat pump and water loop heat pump system. Int J Refrig
2005;28:8109.

234

A. Buonomano et al. / Applied Energy 91 (2012) 222234

[6] Xinguo Li. Thermal performance and energy saving effect of water-loop heat
pump system with geothermal. Energy Convers Manage 1998;39(3/
4):295301.
[7] Chen Chao, Sun Feng-ling, Feng Lei, Liu Ming. Underground water-source loop
heat-pump air-conditioning system applied in a residential building in Beijing.
Appl Energy 2005;82:33144.
[8] Yuan S, Grabon M. Optimizing energy consumption of a water-loop variablespeed heat pump system. Appl Therm Eng 2011;31:894901.
[9] Buonomano A, Calise F, Palombo A. Thermodynamic and economic simulations
of water loop heat pump systems: a computer based approach. In: Proceedings
of ECOS 2007 XX intl conference on efciency, costs, optimization,
simulation and environmental impact of energy systems, Padova, Italy, vol.
1; 2007. p. 61927.
[10] Buonomano A, Calise F, Palombo A. Water loop heat pump system
performances in European climates. In: Proceedings of CLIMAMED 2007
energy, climate and indoor comfort in mediterranean countries, Genova, Italy,
vol. 1; 2007. p. 47190.
[11] Zanetti Freire R, Mazuroski W, Abadie MO, Mendes N. Capacitive effect on the
heat transfer through building glazing systems. Appl Energy 2011;88:43109.
[12] Wong SL, Wan Kevin KW, Lam Tony NT. Articial neural networks for energy
analysis of ofce buildings with daylighting. Appl Energy 2010;87:5517.
[13] iroky J, Oldewurtel F, Cigler J, Prvara S. Experimental analysis of model
predictive control for an energy efcient building heating system. Appl Energy
2011;88:307987.
[14] Bichiou Y, Krarti M. Optimization of envelope and HVAC systems selection for
residential buildings. Energy Build 2010. doi:10.1016/j.enbuild.2011.08.031.
[15] Trcka M, Hensen JLM. Overview of HVAC system simulation. Automat
Construct 2010;19:939.
[16] Fissore A, Mottard JM. Thermal simulation of an attached sunspace and its
experimental validation. Solar Energy 2006;81:30515.
[17] TRNSYS 17. Manual of TRNSYS. University of Wisconsin-Madison, USA; 2010.

[18] Kreider JF, Rabl A. Heating and cooling of buildings. New York (NY): McGrawHill; 1994.
[19] EN 12464-1:2011. Light and lighting lighting of work places part 1: indoor
work places.
[20] Wang S, Xu X. A simplied dynamic model for existing buildings using CTF and
thermal network models. Int J Therm Sci 2008;47:124962.
[21] Shengwei S, Xinhua X. Parameter estimation of internal thermal mass of
building dynamic models using genetic algorithm. Energy Convers Manage
2006;47:192741.
[22] Bourdouxhe J, Grodent M, Lebrun J. Reference guide for dynamic models of
HVAC equipment. Atlanta: American Society of Heating, Refrigerating and AirConditioning Engineers, Inc.; 1998.
[23] Zweifel G, Dorer V, Koschenz M, Weber A. Building energy and system
simulation programs: model development, coupling and integration, EMPA.
[24] Sparber W, Thuer A, Besana F, Streicher W, Henning HM. Unied monitoring
procedure and performance assessment for solar assisted heating and cooling
systems, Eurosun 2008, Lisbon, Portugal; 2008.
[25] Adelard L, Pignolet-Tardant F, Mara T, Lauret P, Garde F, Hoyer H. Sky
temperature modelisation and applications in building simulation. Renew
Energy 1998;15:41830.
[26] Test Reference Year TRY. Weather data sets for computer simulations of solar
energy systems and energy consumption in buildings Commission of the
European Communities; 1985.
[27] Bellia L, Mazzei P, Palombo A. Weather data for building energy cost-benet
analysis. Int J Energy Res 1998;22:120515.
[28] EN ISO 13789:2007. Thermal performance of building components dynamic
thermal characteristics calculation methods.
[29] Windows and glazing system software. LBNL. <http://windows.lbl.gov/
software>.
[30] Eurostat 2009. Environment an energy, electricity and gas prices. <http://
epp.eurostat.ec.europa.eu>.