International Journal of Refrigeration 25 (2002) 421427 www.elsevier.

com/locate/ijrefrig

CO2 heat pump systems

Petter Neksa *,1
SINTEF Energy Research, Refrigeration and Air Conditioning, N-7034 Trondheim, Norway Received 23 November 1999; received in revised form 29 February 2000; accepted 8 March 2000

Abstract After the CFCs and the HCFCs were deemed unt as working uids in refrigeration, air conditioning, and heat pump applications, there has been a renaissance for carbon dioxide technology. Heat pumps is one of the application areas where theoretical and experimental investigations are now performed by an increasing number of research institutions and manufacturers. This paper gives an overview of some of the current activities in the CO2 heat pump eld. Discussed are the important characteristics of the transcritical CO2 process applied to heat pumps, and also discussed are theoretical and experimental results from several heat pump applications. Provided that calculations and system designs are performed on the premises of the working uid, and that test plants are constructed and operated to fully exploit the specic characteristics of both the uid and the transcritical process, the results show that CO2 is an attractive alternative to the synthetic uids. Competitive products may be launched in the near future. # 2002 Elsevier Science Ltd and IIR. All rights reserved.
Keywords: Heat pump; Refrigerant; CO2; Transcritical cycle; Performance

` Pompes a chaleur utilisant le dioxyde de carbone

Resume Depuis que les CFC et les HCFC ont ete interdites en tant que frigorige`nes dans les applications frigoriques, de conditionnement dair et de pompes a` chaleur, la technologie utilisant du dioxyde de carbone (CO2) connat un renouveau. Des etudes theoriques et experimentales sur les pompes a` chaleur sont desormais eectuees par un nombre croissant dorganismes de recherche et de fabricants. Cette communication donne une vue densemble de certaines activites actuelles ` dans les domaine des pompes a chaleur utilisant le CO2. Les caracteristiques cles du cycle transcritique applique aux pompes a` chaleur sont decrites, ainsi que des resultats theoriques et experimentaux obtenus dans plusieurs applications des pompes a` chaleur. Les resultats montrent que le CO2 est un frigorige`ne attrayant qui permet de remplacer des frigorige`nes synthetiques, a` condition dadapter les calculs et la conception du syste`me pour le CO2, et de construire des bancs dessai permettant dexploiter au mieux les proprietes du uide et du cycle transcritique. Les produits competitifs seront lances dans un avenir proche. # 2002 Elsevier Science Ltd and IIR. All rights reserved.
` ` Mots cles : Pompe a chaleur ; Frigorigene ; CO2 ; Cycle transcritique ; Performance

* Tel.: +477-359-3923; fax: +477-359-3950. E-mail address: petter.neksa@energy.sintef.no (P. Neksa). 1 Visiting Researcher at Consiglio Nazionale delle Ricerche, Istituto per la Tecnica del Freddo,z Corso Stati Uniti 4, 35020 Padova, ` Italy and Visiting Researcher at Universita di Padova, Dipartimento di Fisica Tecnica. Vice president of IIR Commission E2. 0140-7007/02/$22.00 # 2002 Elsevier Science Ltd and IIR. All rights reserved. PII: S0140-7007(01)00033-0

422

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

1. Introduction The Montreal Protocol dictates the abolishment on the CFC (chlorouorocarbon) and HCFC (hydrochlorouorocarbon) refrigerants (or working uids), known to be harmful to the environment. These substances have been, and still are, commonly applied in refrigeration, air conditioning, and heat pump systems. At present, two main groups of refrigerants are potential replacement candidates for the CFCs and the HCFCs, namely, the HFCs (hydrouorocarbons) and the natural refrigerants. The global warming potential (GWP) of the HFC gases are high. Typically, a release of one kilogram of an HFC gas contributes 10003000 times more to global warming, than release of one kilogram CO2. This is the reason why these gases are included in the Kyoto-agreement as compounds to be regulated. CO2 is one of the few natural refrigerants, which is neither ammable nor toxic. It is inexpensive, widely available and does not aect the global environment as many other refrigerants. CO2 has a GWP=1, but the net global warming impact when used as a technical gas is zero, since the gas is a waste product from industrial production. There is a net surplus of CO2 in the world today which can be used for dierent applications, for example for carbonising Coca Cola and other beverages or as a refrigerant in refrigeration and heat pump systems. CO2 is an excellent alternative among the natural refrigerants, especially in applications where the toxicity and ammability of ammonia and hydrocarbons may be a problem. The remaining issues are to prove that the energy eciency and the cost of the systems are at an acceptable level. Dierent kinds of heat pumps applying CO2 as the working uid are currently being investigated both in theory and in the laboratory. The intention with this paper is to give an overview of some of these research activities, and also to show that both types of activities yield very competitive performance compared to the halocarbons provided they are done correctly. Additional advantages are also discussed.

fully utilise the advantages introduced by the high pressures, for instance in reducing component sizes due to high volumetric capacity. The critical temperature of CO2 is 31.1 C. This implies that sub-critical operation, as known from common refrigeration technology, is only possible when the average heat sink temperature is rather low. However, when subcritical operation is possible it may be stated that the CO2 systems compete very well with respect to energy eciency with systems using other refrigerants. At higher average heat sink temperatures transcritical operation is necessary. A typical one-stage CO2 process with internal heat exchange is shown in the T-s diagram of Fig. 1. A simplied sketch of a heat pump, with its main components and corresponding process points, is included. Counter-ow heat exchange with a heat source and a heat sink, are indicated by the broken lines. Characteristic for the process is heat rejection at a supercritical pressure, introducing a gliding temperature instead of condensation at constant temperature. Operation of the transcritical process introduces the need to control the high-side pressure in order to obtain optimal eciency. Fig. 2 shows how the heating capacity, heating-COP and compressor shaft power vary with the compressor discharge pressure. The heating capacity increases steeply in the beginning, then, for higher discharge pressures it attens out. The compressor shaft power is increasing almost linearly with the discharge pressure. Since the eciency of the system (heating-COP) is given as the ratio between the heating capacity and the power input, a maximum of the COPcurve can be found at a certain optimum discharge pressure. As illustrated in the gure, the variation in COP around the optimum may be rather at. How at depends, among other factors, on the boundary conditions, the shape of the compressor isentropic eciency curve and the length of the internal heat exchanger. For example, a small capacity internal heat exchanger results

2. Fundamentals The application of CO2 as a working uid introduces challenges, but certainly also opportunities and possibilities, compared to o-the-shelf technology. First of all, components for much higher design pressures are needed, typically 150 bar maximum operating pressure. When CO2 started loosing market shares around 1940, the lack of such components was one of the reasons for the decline. However, with the manufacturing technology and the knowledge base existing today, it is possible to

Fig. 1. T-s diagram showing the transcritical CO2 cycle used for water heating. Fig. 1. Diagramme T-s montrant le cycle transcritique du CO2 utilise pour chauer de leau sanitaire.

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

423

Fig. 2. Variation of heating capacity, heating-COP and compressor shaft power with the discharge pressure for a CO2 heat pump. Fig. 2. Variation de la puissance calorique, du COP et de la e puissance absorbe sur larbre du compresseur en fonction de la ` pression de refoulement dune pompe a chaleur au CO2. Fig. 3. Coecient of performance for the heat pump when producing hot water at a temperature of 60 C, at varying evaporation temperature. ` e Fig. 3. Coecient de performance dune pompe a chaleur utilise ` ` rentes tempe ratures pour chauer de leau chaude a 60 C, a die vaporation. de

in a more pronounced optimum. For further explana tion, see Neksa et al. [1]. In subcritical operation, the COP is in principal limited by the highest heat sink temperature. In transcritical operation, on the contrary, the COP is most often limited by how low refrigerant temperature it is possible to achieve after heat rejection. Consequently, the lowest heat sink temperature gives the limitation. This must be reected in the design of the total heat pump system, including the heat distribution system. For instance, at a given average heat rejection temperature, a heat distribution system with a high temperature glide would benet the CO2 heat pump. A common assumption when comparing the energy eciency of CO2 with other working uids is to assume the same temperature approach in the cold end of the condenser/gas cooler of the two systems. Most often this assumption disfavours CO2. A fair basis for comparison is to assume equal mean temperature dierences in the heat exchangers, and thus also for assessing the trends in the temperature approach. The exact values will be design specic, but it is essential to take the dierence into consideration in order to make a fair comparison to CO2. Operation near the critical point oers a good potential for ecient compression [2]. This is due to low pressure ratios and that relatively high valve pressure drops can be accepted for a given loss in temperature [3]. The volumetric capacity of CO2 is in the order of 5 10 times higher than for the common alternatives. This opens the possibility for compact compressors, even though the pressures are high. Dorin [4] also indicates the possibility to run semi-hermetic CO2 compressors at 3000 rpm, which of course would give a very compact compressor for a given capacity. The heat transfer characteristics of CO2 is very good, even in the supercritical region. Combined with the high volumetric capacity, this makes it possible to develop compact and ecient heat exchangers [5,6].

3. Heat pump water heaters Neksa et al. [1,7] describe experimental results for a CO2 heat pump water heater (HPWH) application. The prototype system results show that CO2 is very well suited as a working uid for tap water heat pumps. The energy consumption can be reduced by 75% compared to electrical or gas red systems, when hot tap water is supplied at 60 C, with the ambient air as the heat source. Fig. 3 shows measured hp-COP as a function of the evaporation temperature. The tap water is heated from 10 to 60 C, typical Norwegian conditions. HpCOP is dened as the heat output divided by the power input to the compressor motor. The shaft power, rather than the input power to the motor, has been measured. The power input to the motor was obtained assuming a motor eciency of 0.9. The high process eciency is partly due to good adaptation of the process to the application, but also ecient compression and good heat transfer characteristics for CO2 are contributing to the high COP. Fig. 4 depicts isentropic and volumetric eciency for the prototype compressor at dierent pressure ratios. The isentropic and volumetric eciency at design conditions (pressure ratio 2.6, heat rejection 90 bar and heat absorption 35 bar), are 0.84 and 0.86, respectively. The corresponding pressure ratio for a halocarbon reference process is about 4, illustrating the favourable pressure ratio for the CO2 process. The performance of the compressor is quite satisfactory considering that this is a prototype under development. A CO2 heat pump water heater may produce hot water with temperatures up to 90 C without operational problems and with only a small loss in eciency.

424

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

Fig. 4. Isentropic and volumetric eciency for the CO2 compressor at varying pressure ratios. trique du compresseur Fig. 4. Rendement isentropique et volume ` rents taux de compression. de CO2 a die

Fig. 5. System design for a combined space and water heating system. The process is also illustrated in the T-s diagram. `me mixte utilise pour chauer et Fig. 5. Conception dun syste pour produire de leau chaude sanitaire. Le diagramme T-s montre ce processus.

Increasing the required hot water temperature from 60 to 80 C reduces the hp-COP from 4.3 to 3.6 at an evaporation temperature of 0 C. The area of application is thus much larger than for the traditional heat pump systems, often restricted to hot water temperatures lower than 55 C. Due to the high volumetric eciency, leading to small ow areas and the good heat transfer characteristics of CO2, it should be possible to manufacture the systems compact and cost ecient. The market potential for heat pump water heaters is large. Roughly 20% of the energy use in residential and commercial buildings goes to water heating, [8]. In addition, there is a substantial need for water heating in industry, where often the tap water heating may be combined with refrigeration and /or freezing, by utilising the cold side of the system. Several research organisations are now developing CO2 HPWH [912].

covered when space heating was not required, in order to simplify the system design. A comparison to using R-134a as a working uid showed favourable seasonal performance for CO2 when more than 30% of the power demand for space heating were covered by the air heating system. The rest is then covered by the radiator system. A 70/50 C radiator system and heat recovery eciency of the balanced ventilation system of 60% was assumed. In larger buildings in Norway, typically more than 50% of the heating demand is air heating and this percentage is increasing due to better insulation and increased air quality requirements. This indicates that CO2 may be a promising candidate for this application.

5. Residential heat pumps A detailed simulation study on the performance of HCFC-22 and CO2 in reversible residential heat pumps gave quite promising results [14]. Evaporator temperatures were higher in the CO2 system, and very small approach temperatures were estimated for the CO2 gas cooler. The mechanically expanded round-tube heat exchangers were designed within the same core dimensions and air-side pressure drop in both systems. The eects of pressure drop, particularly in the evaporator and suction line of the HCFC-22 system and the superheat characteristics of the expansion valve, gave cooling COPs (summer operation) that were similar in both systems, even at high ambient temperatures. Dierences in heating capacity characteristics between CO2 and HCFC-22 are important for the

4. Space heating The market for CO2 heat pumps would of course be extended signicantly if the demand for space heating could be covered eciently in addition to the demand for water heating. Schieoe and Neksa [13] investigated a system design as shown in Fig. 5. In order to achieve a lowest possible return temperature from the heating system, radiator and air heating are connected in series. Tap water heating is indicated pre-heated in parallel with the space heating and heat exchange against hot discharge gas is used to achieve the required hot water temperature. The tap water heating part could also be implemented as a separate system or

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

425

Fig. 6. Measured COP, cooling mode. Ambient temperatures are 28, 35 and 46 C for AC1, AC2 and AC3, respectively. ratures Fig. 6. COP mesure en mode refroidissement. Les tempe ambiantes sont de 28, 35 et 46 C en mode conditionnement dair (AC1, AC2 et AC3 respectivement).

Fig. 7. Measured COP, heating mode. Ambient temperatures are 2, 7 and 14 C for HP1, HP2 and HP3, respectively. ratures Fig. 7. COP mesure en mode chauage. Les tempe ambiantes sont de 2, 7 et 14 C en mode chauage (HP1, HP2 et HP3 respectivement).

seasonal heating performance (winter operation), since dierences in the need for supplementary heating aect the total system energy eciency. The CO2 system was able to maintain a higher heating capacity than the HCFC-22 system at low ambient temperatures, thereby saving supplementary heat of some kind (electric resistance heating was assumed). Even though the heating COP of the two heat pump circuits are similar, the overall result was a 20% increase in system energy eciency (HSPF) for the CO2 system, due to a lower need for supplementary heat. To verify the theoretical results, an HCFC-22 unit and a prototype CO2 system was set up in the SINTEFs laboratory. The capacity of the units is a little larger than the systems analysed in the theoretical study, and the heat exchangers applied in the CO2 lab-unit are of the multichannel (parallel ow) concept. Still, the results should give a fairly good indication of the eciency level between the two systems. Important is that the heat exchangers was designed with the same core dimensions. The main results from the experiments are shown in Figs. 6 and 7. In AC mode, the COP values (cooling COP) of the CO2 unit are slightly lower than those of

the HCFC-22 unit, whereas in HP mode, the COP (heating COP) of the CO2 unit is slightly higher than those of the HCFC-22 unit. In arriving at the compressor power consumption for the CO2 system, the shaft power is measured, and a motor eciency of 0.9 is assumed. The motor eciency is included in the HCFC22 power consumption measurements. The CO2 evaporation temperatures in AC mode were 34 K lower than for the HCFC-22 unit. One reason for this is water retention. With the at heat exchanger pipes running horizontally, and the unit only slightly tilted forward, water is eectively prevented from draining. The result is that water builds up until it reaches a critical mass, when it is suddenly released in a batch. The water condensation rates were more or less the same for both systems. The CO2 evaporation temperatures in HP mode were 3.7 K higher to 0.4 K lower than those of HCFC-22. The measured temperature approach was between 1.5 and 2.0 K in AC mode and between 7.2 and 15.0 K in HP mode. The relatively large temperature approaches in HP mode are too high for ecient operation. The reason for this ineciency is that the number of passes through the indoor heat exchanger is too low. There was only one pass through both the front and the rear section. More passes are easily introduced by installing baes in the headers. In conclusion, it can be said that the overall system eciency of the CO2 system in terms of COP is already competitive to that of the HCFC-22 system in both modes of operation. The heat exchanger design concept applied is neither optimal from a gas cooler point of view, nor from an evaporator point of view. Water retention has been a problem, and this has also caused uctuating system pressures. In some of the experiments, the CO2 compressor has not performed optimally, owing to worn out bearings, and leakage through the stung box. Based on the above, there should be room for further improvements in the performance of the CO2 system. Details about the work done so far are given in Aarlien and Frisik [15]. More work will be done to improve this system in the future.

6. Air heating systems Rieberer and Halozan [16] and Rieberer et al. [11] made detailed theoretical studies of controlled ventilation air heating systems with an integrated CO2 heat pump. The results look very promising. Fig. 8 shows overall and hp-COP as a function of ambient temperature. Ambient air, preheated in a ground-to-air heat exchange system, is used as the heat source. At an ambient air temperature of 20 C, the air is preheated to 5 C before entering the heat pump evaporator. The overall seasonal performance factor for a Graz, Austria climate is calculated to be in the range 6.156.5. This

426

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

9. Other heat pump applications Late professor Gustav Lorentzen has published several papers describing the possibilities of using CO2 as a working uid in heat pumps and refrigeration systems. In [21,22] he describes a system design of large heat pumps for district heating. This is a high-capacity application where turbo expanders may be possible to realise a cost ecient manner. Described is also the possibility to combine refrigeration/freezing and tap water heating, which will give a very high overall system eciency. The global warming issue has led to increased focus on energy eciency and heat recovery in various appli cations. Neksa et al. [23] describe theoretical and experimental results for an all-CO2 system for commercial refrigeration. Less excess heat in cars due to more ecient motors and introduction of electrical cars has led to the interest of using the mobile air-conditioning system as a heat pump in the cold season. Hafner et al. [24] report favourable features of using CO2 in these kinds of systems.

Fig. 8. System eciency and compressor speed. (qtran,-12 C=15 W/m2, ua=0.8 1/h, pcond=70 bar) From Rieberer and Halozan [16]. `me et vitesse du compresseur (qtran, Fig. 8. Ecacite du syste 12 C = 15 W/m2, ua = 0,8 1/h, pcond = 70 bars) (fonde sur les travaux de Rieberer et Halozan [16]).

corresponds to a seasonal performance factor of the heat pump of above 4 (authors remark).

7. High temperature hydronic heating systems Enkemann et al. [17] made a theoretical study on CO2 heat pumps for retrot in typical hydronic heating systems in Western Europe. A system originally designed for temperatures 70/50 C was proposed modied by reducing the mass ow rate of water to obtain a 93/40 C system, which should give a corresponding heat output using the existing radiator system. The seasonal performance was then increased from 2.8 to 3.2. In addition, this system will be able to supply hot tap water without any loss in energetic eciency. Experimental results from two prototype systems for this application is reported in Brandes [18]. Eciency gures in the same range as those calculated were reported.

10. Conclusions Experimental results show that CO2 may be successfully used as a working uid in heat pumps with very competitive performance, if the system is properly designed. There is a good accordance between theory and experimental results when calculations are done properly, i.e. when the properties of CO2 and process characteristics of the transcritical CO2 process are fully exploited. Very promising simulation results exist for application areas where experimental results are not yet available. There are good reasons to continue with experimental work in these areas. With all the benets of a natural working uid and the very competitive performance demonstrated, CO2 is and will continue to be a viable alternative to the synthetic working uids.

8. Heat pump dryers Another interesting application is heat pump dryers. Steimle [19] reports, based on theoretical considerations, that energy saving is possible due to better temperature adaptation in the heat exchangers, compared to subcritical processes. It is also possible to achieve higher air temperatures without loss in eciency, thus increasing the moisture extraction rate. Experimental results from Schmidt et al. [20] reports hp-COPs in the range 5.5 and 55% reduction in the energy consumption, including fan power, compared to a traditional electrically heated dryer. The results were achieved after a rst optimising of the prototype system and it is hoped that further essential improvements still can be reached.

Acknowledgements For the projects reported here that are accomplished at SINTEF Energy Research, I would like to thank the companies Hydro Aluminium, FINSAM Int. Inc, Frostmann, OSO Hotwater, STATOIL, Tipro Nova, Sperre Thermo, Costan SpA and Ocine Mario Dorin SpA, in addition to the Norwegian Research Council and the EU commission.

P. Neksa / International Journal of Refrigeration 25 (2002) 421427

427

References
[1] Neksa P, Rekstad H, Zakeri GR, Schieoe PA. CO2-heat pump water heater: characteristics, system design and experimental results. Int J Refrigeration 1998;21:1729. [2] Neksa P, Dorin F, Rekstad H, Bredesen AM, Serbisse A. Development of semi-hermetic CO2-compressors. 20th International Congress of Refrigeration, IIR/IIF, Sydney, 1999. [3] Pettersen J, Neksa P, Nesje OM, Schieoe PA, Rekstad H. Recent advances in CO2 refrigeration. 19th IIR Congress of Refrigeration, IIR/IIF, Haag, The Netherlands, 1995. [4] Dorin F. Semi-hermetic CO2 compressors for commercial applications. In: CO2 technology in refrigeration, heat pump and air conditioning systems. Mainz, Germany, 1999. [5] Bredesen AM, Aekt K, Pettersen J, Hafner A, Neksa P, Skaugen G. Studies on CO2 heat exchangers and heat transfer. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May, 1997. [6] Bredesen AM, Hafner A, Pettersen J, Neksa K. Heat transfer and pressure drop for in-tube evaporation of CO2. In: Heat transfer in natural refrigerants. R. Radermacher, College Park (MD) 1997. p. 115. [7] Neksa P, Rekstad H, Zakeri GR, Schieoe PA, Svensson MC. Commercial heat pumps for water heating and heat recovery. In: CO2 technology in refrigeration, heat pump and air conditioning systems. Mainz. Germany, 1999. [8] IEA-HPC. Domestic hot water heat pumps in residential and commercial buildings. IEA Heat Pump Center, 1993. [9] Saikawa M, Hashimoto K, Hasegawa H, Iwatsubo T, Straub J. A basic study on CO2 heat pumps especially for hot tap water supply. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997. [10] Hwang Y, Radermacher R. Carbon dioxide refrigeration systems. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997. [11] Rieberer R, Kasper G, Halozan H. CO2 a chance for once-through heat pump heaters. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997. [12] Heyl P, Preusser S, Kraus WE. The CO2 heat pump project at the TU Dresden. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997.

[13] Schieoe PA, Neksa P. CO2 varmepumpe for bygningsoppvarming, forprosjekt (in Norwegian). SINTEF Report. Trondheim, SINTEF Energy Research, 1999. [14] Pettersen J, Aarlien R, Neksa P, Skaugen G. A comparative evaluation of CO2 and HCFC-22 room air conditioners. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997. [15] Aarlien R, Frivik P-E. Comparison of practical performance between CO2 and R-22 reversible heat pumps for residential use. In: Natural working uids 98, IIR Gustav Lorentzen conference. Oslo, Norway, IIR, 25 June 1998. [16] Rieberer R, Halozan H. CO2 heat pumps in controlled ventilation systems. In: Natural working uids 98. IIR Gustav Lorentzen conference. Oslo, Norway, IIR, 25 June 1998. [17] Enkemann T, Kruse H, Oostendorp P. CO2 as heat pump working uid for retrotting hydronic heating systems in western Europe. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway, IEA Heat Pump Centre, 1314 May 1997. [18] Brandes H. Energy ecient and environmentally friendly heat pumping system using CO2 as working uid in existing buildings. In: CO2 technology in refrigeration, heat pump and air conditioning systems. Mainz, Germany, 1999. [19] Steimle F. CO2-drying heat pumps. In: CO2 technology in refrigeration, heat pumps and air conditioning systems. Trondheim, Norway. IEA Heat Pump Centre, 1314 May 1997. [20] Schmidt EL, Klocker K, Flacke N. Heat pumps for dehumidication and drying processes in residential and commercial applications. Hot air drying heat pump using a transcritical CO2 process. In: CO2 technology in refrigeration, heat pump and air conditioning systems. Mainz, Germany, 1999. [21] Lorentzen G. Large heat pumps using CO2 as refrigerant. In: Energy eciency in refrigeration and global warming impact. Gent, Belgium, IIR, 1214 May 1993. [22] Lorentzen G. Revival of carbon dioxide as a refrigerant. Int J Refrigeration 1994;17(5):292301. [23] Neksa P, Girotto S, Schieoe PA. Commercial refrigeration using CO2 as refrigerant system design and experimental results. In: Natural working uids 98. IIR Gustav Lorentzen conference, Oslo, Norway, IIR, 25 June 1998. [24] Hafner A, Pettersen J, Skaugen G, Neksa P. An automobile HVAC system with CO2 as the refrigerant. In: Natural working uids. 98, IIR Gustav Lorentzen conference, Oslo, Norway, IIR, 25 June 1998.