Sie sind auf Seite 1von 8

Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 101 (2016) 280 287

71st Conference of the Italian Thermal Machines Engineering Association, ATI2016, 14-16
September 2016, Turin, Italy

Refrigerant alternatives for high speed train A/C systems: energy


savings and environmental emissions evaluation under variable
ambient conditions
R. Mastrulloa, A. W. Mauroa,*, C. Velluccia
a
Dipartimento di Ingegneria Industriale, Universit degli Studi di Napoli Federico II, p.le Tecchio, 80 80125, Napoli

Abstract

The environmental concerns due to global warming are pushing A/C industries to new eco-friendly refrigerants in several fields.
In this paper a model to simulate the dynamic evolution of the temperature inside an air-conditioned high-speed train
compartment is presented. The dynamic modeling of both the reversible heat pump unit and the thermal loads of the cabin are
presented, including the possibility of adapting the frequency of the compressor and the return air fraction for maintaining the
internal comfort conditions. Under different dynamic load conditions (in terms of ambient temperature, solar radiation, train
speed, number of passengers) the energy consumptions and the TEWI related to the use of new refrigerants, (like R1234yf and
R1234ze), are calculated being the R134a a baseline for comparison.


2016
2016TheTheAuthors. Published
Authors. by by
Published Elsevier Ltd.Ltd.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of ATI 2016.
Peer-review under responsibility of the Scientific Committee of ATI 2016.
Keywords: train, energy saving, environmental effect, heat pump;

* Corresponding author. Tel.: +39 081 7682198


E-mail address: alfonsowilliam.mauro@unina.it

Nomenclature
A surface [m2] Subscripts
C capacitance [Wh m-2 K-1] a air inside cabin
COP Coefficient Of Performance amb ambient
GWP Global Warning Potential aux auxiliaries
Hz hertz [s-1] ceiling ceiling
Q power [kW] ext external

1876-6102 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of ATI 2016.
doi:10.1016/j.egypro.2016.11.036
R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287 281

T temperature [C] furniture furniture


TEWI Total Equivalent Warming Impact [kgCO 2 ] g solar
U heat transfer coefficient [W m-2 K-1] glass glazing surfaces
Vol volume [m3] hp heat pump
cp specific heat at constant pressure [kJ kg-1 K-1] int internal
k thermal conductivity [W m-1 K-1] p passengers
s thickness [mm] plant floor
Greek symbols rd radiative
absorptivity coefficient roof roof
curtains coefficient side left/right
long wave emissivity vertical vertical
time [s] wall wall of cabin
density [kg m-3]
Stephan-Boltzmann constant
transmissivity coefficient

1 Introduction

In order to decrease the concentration of greenhouse gases in the atmosphere F-gas regulation (EU Reg.
517/2014) is pushing A/C industries to phase at refrigerants which do not conform with threshold values of the
global warming potential (GWP) [1], among which the most widely used refrigerants for heat pumps. In this paper a
dynamic model to simulate the air conditioning of high speed train is presented. Up to now several thermal cabin
models have been presented. For instance, other authors such as Li and Sun [3] have modelled a cabin coupled with
an AC unit. Liu et al. [4] built a mathematic model to simulate dynamic cooling load of an air-conditioned train
compartment. Maidment and Missenden [5] presented the evaluation of an underground railway carriage operating
with a sustainable groundwater cooling system. Torregrosa-Jaime at al. presented a lumped-parameter model of a
vehicle's cabin with thermal zones [6]. Luger et al. [7] introduced methodologies and tools to identify representative
operating conditions of the heating, ventilation and air conditioning (HVAC) systems in passengers rail vehicles.
Dullinger et al. [8] built a modular thermal simulation tool for computing energy consumption of HVAC units in rail
vehicles. The dynamic modelling here presented combines the reversible heat pump unit and the thermal loads of the
cabin and calculates, under different dynamic load conditions (in terms of ambient temperature, solar radiation, train
speed, number of passengers), the energy consumptions and the TEWI related to the use of two HFOs refrigerants
(R1234yf and R1234ze), being the R134a, one of the most used HFCs, a baseline for comparison. HFOs are low
GWP fluids and they have been verified to be valid alternatives for HFCs [2]. First the cabin thermal model will be
presented. Then the characteristics of the reversible heat pump and its modelling results in IMST-ART [8] will be
shown. Finally a comparison among the three refrigerants considered in terms of energy consumption and TEWI
will be shown.

2 Thermal cabin model


The dynamic model here presented has the aim of calculating the thermal load of a cabin of a high speed train
during its travel and to couple its loads dynamically with the heat pump performance in order to estimate the energy
consumption, while guaranteeing the comfort for passengers, both in summer and winter.
The cabin considered is 27 m long, 3 m high and 3.5 m wide. Besides 30 m2 of glazing for each wall are
assumed. The construction of vertical walls and ceiling is assumed to be 50 mm thick with polyurethane insulation
(thermal conductivity of 0.026 W m-1 K-1). A cavity separates the ceiling of the cabin by the external roof. The
global thermal balance related to the cabin which provides the evolution of the internal air temperature is
represented by Eq. (1):

, + = () + , () + () + + () + () + () +

() + () (1)
282 R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287

In Eq. (1), the capacitance of furniture is set to 11 Wh m-2 K-1, with reference to the surface of the floor, as indicated
in reference [10]. In a similar way, the variation with time of the temperature of the wall T is determined with
Eq. (2):
,
= , () , () , () (2)

In Eq. (2) the thermal capacitance of the wall is fixed to 30 Wh m-2 K-1. is calculated in the middle of the
thickness of the cabin and is considered constant along the thickness of the wall. The pedix side indicates left,
right and ceiling depending on case considered. The energy balance on roof is represented in Eq. (3), in which
thermal inertia is neglected.
, () , () = , () (3)
The treatment of these thermal loads is shown in next section, with the exception of , intrinsically positive in the
case of air conditioning in winter and intrinsically negative in the case of air conditioning in summer, as it represents
the interaction with the heat pump. Its absolute value depends on the current operating conditions of the system in
terms of volumetric flow rates and internal and external temperatures, as it is shown in section 3 and Appendix A.

2.1 Solar radiation

Sun radiation absorbed by the walls of the cabin and transmitted through the glasses is considered using Eqs. (4 - 5)
and (6) respectively. The radiative heat flow between the external roof and the ceiling is considered with Eq. (7).

, () = () , (4)
, () = () (5)
,_ () = () , (6)

() ()

, () = (7)
( )

where is the solar radiation incident on the roof [W m-2], the solar radiation incident on vertical walls
[W m-2], and the absorptivity coefficient and transmissivity coefficient, respectively, assumed equal to 0.7 and
0.6; is the long wave emissivity, equal to on the assumption of gray body, is Stephan-Boltzmann constant.
Besides, a curtains coefficient is considered and assumed as 0.3.

Fig. 1: Global thermal balance related to the single cabin


R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287 283

2.2 Conduction and convection loads through the cabin.

The heat conduction is calculated as in Eq. (8):


, () = , () , , () () (8)

where [W m-2 K-1] is the total internal heat transfer coefficient and is calculated as in Eq. (9):

= ( + ) (9)

where [W m-2 K-1] is the convective heat transfer coefficient calculated for each vertical wall and for the ceiling
(see Appendix A). The thermal storage of the windows is neglected and the heat load due to window conduction is
calculated as in Eq. (10):
, () = , () () (10)
-2 -1
where U glass is the total heat transfer coefficient for the windows, assumed as 3.67 W m K , as indicated in [5]
The external heat load due to interaction with the outside is calculated as in Eqs. (11 - 12):
, () = , () , (, () ()) (11)
, () = , () ( () ()) (12)
-2 -1 -2 -1
where , [W m K ] and , [W m K ] are the total external heat transfer coefficients referred to
vertical walls (right and left) and to roof (horizontal) of cabin respectively, calculated as in Eqs. (13 - 14):

, = ( + ) (13)

, = ( ) (14)
where [W m-2 K-1] is the convective heat transfer coefficient, which is a function of the train velocity,

calculated for each vertical wall and for the roof (see Appendix A). , is independent of because no

insulating material is adopted on the roof (a cavity separates the roof from the ceiling).

2.3 Equipment load

The heat load due to the presence of auxiliaries is calculated as in Eq. (15):
() =
(15)
where i is the i-th auxiliary e M is the number of auxiliaries.
This parameter is assumed to be constant and equal to 3 kW, as indicated in [4].

2.4 Occupancy load

The heat load due to the presence of passengers is calculated as in Eq. (16):
= (16)
where n is the number of passengers and is the heat load from a single passenger.
The heat emission per person, sensible and latent, is assumed to be constant and equal to 116.3 W at the indoor air
temperature of 26 C [11].

3 Description and performance of the heat pump

The key element of the A/C system is the reversible heat pump, which is an air-to-air model. Its nominal cooling
capacity is 40 kW at 50 Hz when cooling air from 29C to 22 C at the evaporator, with a superheating of 5 K, a
refrigerant charge of 11 kg and working with 12000 m3 h-1 of air leaving the condenser at 45C. The heat pump is
284 R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287

modelled with the vapour compression software package IMST-ART [9] (which includes the main elements of the
vapour compression circuit: evaporator, condenser, compressor, expansion valve and connecting pipe work)
considering the catalogue data provided by the component manufacturers. IMST-ART software has been
experimentally validated in several studies, such as [12], where it was demonstrated that the model can predict the
heat transferred in finned tube exchangers with an accuracy of 5% for a wide range of operating conditions. The
heat pump can work with variable air temperatures at the inlet of the condenser and evaporator and has the objective
of cooling air to 22 C in summer (a summer comfort temperature of 26 C for the passengers is considered). The
capacity of the heat pump (of cooling for summer and heating for winter) and COP are correlated as shown in Eq.
(17). Coefficients correlating the main performance indicators in heating are obtained reversing the vapour
compressor cycle modeled for cooling. In Eq. (17) x 1 is the air inlet flowrate at the evaporator [m3 h-1], x 2 is the air
inlet flow-rate at the condenser [m3 h-1] and x 3 the external temperature [C].
= + + + (17)
These correlations are inserted in the dynamic model of the cabin train, written in MatLab, in which they receive
as inputs the working conditions (flow rates of the secondary fluids entering the condenser and the evaporator and
temperature at the inlet of the internal side heat exchanger), and provide as outputs the performance parameters of
the heat pump. The correlations are obtained for three different refrigerants (R134a, R1234yf and R1234ze)
considering two different frequencies of the inverter-driven compressor: 30 Hz and 60 Hz. The values of the
coefficients obtained for the main performance indicators in the case of cooling and heating are shown in Appendix
B.

4 Assessment of environmental impact during realistic operating conditions of an high speed train

An Italian railway line is supposed for dynamic simulation. A return trip of 1600 km from Napoli to Milano
(being 800 km the distance between these two cities), passing through Roma, Firenze and Bologna is considered.
While not traveling, it is assumed that the train is inside a covered station. Details of the railway (departures, arrivals
and durations) are shown in Table 1. A high speed train is considered, with a cruise speed of 350 km h-1. Number of
passengers is assumed to be 60 per cabin and constant. Meteorological data (solar radiation incident on horizontal
surface, solar radiation incident on vertical surface and ambient temperature) are taken from the Meteonorm
database [13] and they are interpolated in space knowing the position as function of time.

Table 1. Italian railway line details


Departure Arrival Duration

Napoli - Roma 8:00 am 9:16 am 1 h 16 min


Roma - Firenze 9.27 am 10.39 am 1 h 12 min
Firenze - Bologna 10.49 am 11.29 am 40 min
Bologna - Milano 11.40 am 12.41 am 1 h 1 min
Milano - Bologna 1.00 pm 2.01 pm 1 h 1 min
Bologna - Firenze 2.12 pm 2.52 pm 40 min
Firenze - Roma 3.02 pm 4.12 pm 1 h 12 min
Roma - Napoli 4.23 pm 5.39 pm 1 h 16 min

The dynamic model allows the possibility of adapting the frequency of the compressor and the return air fraction
for maintaining the internal comfort conditions (26 C in summer and 20 C in winter, with a dead band of 2 K in
both cases). A frequency control, proportional to the absolute value of the temperature difference between the actual
cabin temperature and the comfort temperature, is imposed: for frequencies between 30 to 60 Hz, performance
resulting from the correlations of the heat pump at 30 and 60 Hz are interpolated linearly. Two simulations have
been carried out: for summer, the hottest week of the year (in July); for winter, the coldest week of the year (in
February), with the purpose of comparing energy performance, in terms of mean COP, and the environmental
impact, in terms of TEWI. Fig. 2 (a) and Fig. 2 (b) show the trends of the temperatures of interest and thermal loads
for a day (29th of July) considering R134a: T ext is the external temperature, T wall,right is the temperature of right
vertical wall, T wall,left is the temperature of left vertical wall, T roof is the temperature of roof, T ceiling is the
temperature of the ceiling, T a is the temperature of air inside the cabin. From Fig. 2 (a) it can be seen that T a remains
R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287 285

inside the comfort band. Highest temperature is T roof ; in summer this temperature shows an increasing trend when
the train slows down due to solar radiation and a decreasing trend when the train runs at maximum speed due to
forced convection which cools down the roof. In Fig. 2 (b), the blue line represents the cooling capacity of the heat
pump.
For the other two fluids, the temperature trends are very similar. In order to indicate the overall environmental
impact of the A/C system during its operation, two contributions are summed up: direct effect of refrigerant released
during the lifetime of the equipment and indirect impact of CO 2 emissions (from fossil fuels used to generate energy
to operate the equipment). For this reason TEWI is calculated as in Eq. (18):
= + ) ( = + () (18)
where L is the annual leakage rate in the system in kg (3 % of refrigerant charge annually), N is the life of the
system in years, n is the system running time in years, Ea is the energy consumption in kWh per year and is carbon
dioxide emission factor in CO 2-eq . per electrical kWh (for Italy 0.483 kg CO 2 kWhe-1 as indicated in [14]).
In this case TEWI is calculated considering a week of functioning (both for cooling and for heating) and it is
used to compare the environmental impact of the three refrigerants. For comparison in terms of consumption and
performance, mean COP and heating/cooling energy per week are considered. In Table 2. the summer and winter
weekly results are summarized; the term Q useful,heat represents the cooling capacity in summer and the heating
capacity in winter. As shown in Table 2., in summer the most environmentally friendly refrigerant is the R1234ze,
with total CO 2 equivalent emissions of 178.7 kg per week. Part of this result is due to the 16 % higher mean COP
than R134a, as evidenced by the lower indirect emissions. In winter the mean COP of three fluids are very similar
(with R1234ze the mean COP is 1% higher than R134a). It can be noticed that in winter the energy supplied by the
heat pump is lower than what is subtracted in summer, and so TEWI is lower than in summer, given that the
contribution of solar radiation, passengers and auxiliaries allows to maintain a comfortable temperature inside the
cabin with a lesser need of heat pump operation during winter.
5 Conclusions
In this work a model to simulate the dynamic evolution of the temperature inside an air-conditioned high-speed
train compartment is presented. The dynamic model combines the reversible heat pump unit and the thermal loads of
the cabin (in terms of ambient temperature, solar radiation, train speed, number of passengers and auxiliaries). An
Italian railway line was supposed (return trip Napoli - Milano). Two simulations have been carried out: for summer,
the hottest week of the year (in July); for winter, the coldest week of the year (in February). Three refrigerants were
considered (R134a, R1234yf and R1234ze) and they were compared in terms of energy consumptions and TEWI.
The results evidence that, in terms of temperatures inside the cabin, comfort temperature is maintained regardless
of the refrigerant employed. Energy performance is better in the case of R1234ze: in particular, in summer R1234yf
has a 2% higher mean COP than R134a, while R1234ze has a 16% higher mean COP than R134a. In term of
environmental impact R1234yf has a 5% lower TEWI than R134a, while R1234ze has a 18% lower TEWI than
R134a. The winter consumption is less than 10% than in summer, thanks to the contribution of solar radiation,
passengers and auxiliaries. In terms of TEWI, R1234ze has the lowest environmental impact due to the best energy
performance and low direct GWP (10% lower TEWI than R134a, while R1234yf has a 1% lower TEWI than
R134a). In conclusion, the most performant and environmentally friendly refrigerant in this application is R1234ze.
The results on TEWI are related to the value used for the specific CO 2 emissions. This value is set to decrease in
Europe given the commitment to introduce more renewable sources and therefore in perspective the relative weight
of the direct contribution will increase making low GWP fluids a necessity. While there is no specific regulation for
the industry of trains, even in this area there are already economically and technically possible solutions that allow a
significant reduction in environmental impact.
Table 2. Cooling and heating weekly results (GWP R134a =1300 GWP R1234yf = 4, GWP R1234ze = 6)
COOLING HEATING
R134a R1234yf R1234ze R134a R1234yf R1234ze
mean COP 3.41 3.48 4.09 4.57 4.64 4.62
Q useful,heat [kWh/sett] 1495.7 1499.4 1513.0 117.9 118.4 106.6
TEWI [kgCO 2 /sett] 220.2 208.1 178.7 20.83 12.34 11.18
Dir.Emis [kgCO 2 /sett] 8.376 0.025 0.038 8.376 0.025 0.038
Ind.Emis [kgCO 2 /sett] 211.8 208.1 178.7 12.46 12.32 11.14
286 R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287

a b

Fig. 2
Fi 2. (a)
( )TTrain
i velocity
l i andd temperatures trends
d on 29th
29 h July
J l (R134a);
(R (b) Main thermal loads trends on 29th July (R134a).

Acknowledgements
The present work was financially supported by Universit degli Studi di Napoli Federico II through the
Italian project Studio e sperimentazione di sistemi iNnovativi EMBedded caratterizzati da elevata
efficienza per applicazioni ferrOviarie' (NEMBO - PON03PE_00159_1), which is gratefully acknowledged.
Appendix A. Convective heat transfer coefficient calculation
The convective heat transfer coefficient [W m-2 K-1] is calculated with the correlations described in Table A.1,
for natural and forced convection.
Table A.1 Natural and forced convection correlations
Natural convection
Geometry Flow regime Correlations Number Formula

Laminar flow
= 0.54 (10 < < 10 ) (A.1)
Upper surface of hot plate or

Horizontal flat lower surface of cold plate = 0.15

(10 < < 10 )
Turbulent flow (A.2)
plate:

Upper surface of cold plate Laminar flow =
0.27 (10 < < 10 ) (A.3)
or lower surface of hot plate

0.837
= 0.825 +

Vertical flat plate Laminar/Turbulent flow
(A.4)

1 + (0.492)
Forced convection

Laminar flow = 0.664
(A.5)

Turbulent flow = (0.037
871) (A.6)

where L is the characteristic length, is the thermal conductivity of the fluid [W m-1 K-1], is the kinematic
viscosity [m2 s-1], g is gravitational acceleration [m s-2], and are the temperatures of the surface and air,
respectively, and is the volumetric thermal expansion coefficient [K-1].
Appendix B. Coefficients correlating the main performance indicators in cooling and heating
Table B1. Coefficients correlating the main performance indicators during the mode cooling.
R134a b1 b 2 (104) b 3 (104) b4 b1 b 2 (105) b 3 (106) b4
Q ev30Hz 20.910 2.4000 1.3000 0.5415 COP 30Hz 1.6963 1.1520 1.8186 -0.2342
Q ev60Hz 29.533 5.5000 3.7000 0.7935 COP 60Hz 7.1040 1.5060 0.5812 -0.1660
R1234yf b1 b 2 (104) b 3 (104) b4 b1 b 2 (105) b 3 (106) b4
Q ev30Hz 23.946 2.4000 1.5000 0.3661 COP 30Hz 11.735 1.1566 -4.0799 -0.2303
Q ev60Hz 34.911 4.9000 4.5000 0.4624 COP 60Hz 7.1679 1.4464 0.5514 -0.1639
R1234ze b1 b 2 (104) b 3 (104) b4 b1 b 2 (105) b 3 (106) b4
R. Mastrullo et al. / Energy Procedia 101 (2016) 280 287 287

Q ev30Hz 16.348 2.4000 0.8000 0.4472 COP 30Hz 13.379 1.6475 -4.0314 -0.2528
Q ev60Hz 26.210 3.8000 2.6000 0.6454 COP 60Hz 8.4918 1.2411 0.4009 -0.1840
Table B2. Coefficients correlating the main performance indicators during the mode heating.
R134a b1 b 2 (104) b 3 (106) b4 b1 b 2 (105) b 3 (106) b4
Q co30Hz 20.998 2.4000 9.8628 0.8186 COP 30Hz 6.7819 -6.4714 -0.2768 0.0782
Q co60Hz 37.083 4.2000 2.6641 1.2149 COP 60Hz 4.3801 -1.9994 0.3263 0.0258
R1234yf b1 b 2 (104) b 3 (105) b4 b1 b 2 (105) b 3 (106) b4
Q co30Hz 21.869 2.2000 2.7830 0.7136 COP 30Hz 7.0203 -7.0221 -0.2822 0.0650
Q co60Hz 36.175 4.0000 5.8070 1.1083 COP 60Hz 4.4315 -2.0775 0.3169 0.0207
R1234ze b1 b 2 (104) b 3 (106) b4 b1 b 2 (105) b 3 (106) b4
Q co30Hz 15.828 1.9000 6.4673 0.7067 COP 30Hz 7.5548 -8.8500 -0.6303 0.1212
Q co60Hz 28.676 3.3000 0.2043 0.9963 COP 60Hz 5.3026 -3.6111 8.9679 0.0407

Fig. B.1. reports a comparison among the three refrigerants (R134a, R1234yf and R1234ze) in terms of COP and
Q ev (in the cooling mode):

a b

Fig. B.1 Cooling performance comparison in terms of COP 60Hz ad Qev 60Hz among R134a, R1234yf and R1234ze, for fixed volumetric rate at
condenser (12000 m3h-1), varying boundary conditions (air inlet flowrate at the evaporator [m3 h-1] and external temperature [C] ). For
comparison, frequency of the inverter-driven compressor is fixed at 60Hz.

References
[1] Regulation (EU) No 517/2014 of the European Parliament and the Council of 16 April 2014 fluorinated greenhouse gases and repealing
Regulation (EC) No 842/2006. Off. J. Union, 2014.
[2]Invernizzi C. M, Iora P., Preinger M., Manzolini G. HFOs as substitute for R-134a as working fluids in ORC power plants: A thermodynamic
assessment and thermal stability analysis, Appl. Therm. Eng 2016, 103, p. 790797.
[3] Li W., Sun J., Numerical simulation and analysis of transport air conditioning system integrated with passenger compartment, Appl. Therm.
Eng. 2013, 50, p. 37-45.
[4] Liu W., Deng Q., Huang W., Liu R., Variation in cooling load of a moving air-conditioned train compartment under the effects of ambient
conditions and body thermal storage, Appl. Therm. Eng., 2011, 31, p. 1150 - 1162.
[5]Maidment G.G.,. Missenden J.F, Evaluation of an underground railway carriage operating with a sustainable groundwater cooling system, Int.
J. Refrig.,2002, 25, 569 - 574.
[6] Torregrosa-Jaime B, Bjurling F., Corberan J. M., Di Sciullo F., Paya J., Transient thermal model of a vehicle's cabin validated under variable
ambient conditions, Appl. Therm. Eng., 2015, 75, p. 45 - 53
[7] Luger. C., Kallinovsky J., Rieberer R., Identification of representative operating conditions of HVAC systems in passenger rail vehicle based
on sampling virtual train trips, Adv. Eng. Inf., 2016, 30, p. 157 - 167.
[8] Dullinger C., Struckl W., Kozek M., A modular thermal simulation tool for computing energy consumption of HVAC units in rail vehicles,
Appl. Therm. Eng., 2015, 31, p. 619 - 629.
[9] Corbern J.M., Gonzalvez J., 2002, "ART", a computer code to assist the design of refrigeration and A/C equipment, Proc. of 2002 Purdue
Refrigeration Conference, West Lafayette, Indiana, USA., p. 94 95
[10] Goulart, S, Thermal inertia and natural ventilation Optimisation of thermal storage as a cooling technique for residential buildings in
Southern Brazil. PhD thesis, Universidade Federal de Santa Catarina, Florianopolis, Brazil, 2004.
[11] China Ministry of Railways, China Ministry of Railways Standard TB 1951-87: Design Parameters for Air-conditioning Units of Passenger
Trains. China Ministry of Railways, Beijing, China, 1988.
[12] Corbern J. M., Garca M., Modelling of plate finned tube evaporators and condensers with R134A, Int. J. Refrig., 1998, 21, p. 273-284
[13] www.meteonorm.com - checked on 20th 2016
[14] International Energy Agency, CO 2 Emissions from Fuel Combustion (IEA 2015).