You are on page 1of 6

Applied Thermal Engineering 41 (2012) 30e35

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

journal homepage:

Comparison of a R744 cascade refrigeration system with R404A and R22 conventional systems for supermarkets
Alessandro da Silva a, Enio Pedone Bandarra Filho b, *, Arthur Heleno Pontes Antunes b
a b

Bitzer Compressores Ltda, Av. Joao Paulo Ablas 777, Sao Paulo 06711-250, Brazil Faculty of Mechanical Engineering, Federal University of Uberlandia, Av. Joao Naves de Avila, 2121, Uberlandia-MG, Brazil

a r t i c l e i n f o
Article history: Received 29 April 2011 Accepted 8 December 2011 Available online 30 December 2011 Keywords: Cascade system CO2 Carbon dioxide R744 Supermarket R404A R22

a b s t r a c t
The present article focuses on the energy efciency and climate performance of three different systems used in supermarket applications. The refrigeration systems consist of a cascade cycle (CO2/HFC-404A) e provide nominal refrigerating capacity e with carbon dioxide for subcritical operation and HFC-404A in the high stage temperature stage (pump circuit for normal refrigeration and direct expansion for deepfreezing), and also HFC-404A and HCFC-22 with direct expansion systems. The cascade system presented a lower refrigerant charge, 47 kg of both uids, which represents less than a half of the refrigerant charge of the other systems. An important factor is the total GWP in case of leakage, where the impact in the atmosphere of the cascade system operating with CO2 was much less than the two direct expansion systems. 2011 Elsevier Ltd. All rights reserved.

1. Introduction Carbon dioxide is a climate-friendly refrigerant because it has a low direct global warming potential with the reference value of 1. Due to its specic thermodynamic properties, including high operating pressure, low critical temperature and low viscosity, CO2 offers a great potential as a new energy-efcient product. According to Parise and Marques [9], industry is now challenged to produce modern systems with zero leak and minimum refrigerant charge, leading to more compact and efcient heat exchangers. However, Bansal [1] suggests that there is very little information available in the open literature on the fundamental boiling and condensation heat transfer characteristics of CO2 at low temperatures below 30  C and 15  C, respectively. The appropriate optimal design of new heat exchangers may be impeded due to the lack of this information. Furthermore, it will encourage the development of modern systems that will put the refrigeration industry on a more sustainable footing. Two-stage cascade refrigeration systems are suitable for the supermarket refrigeration industry, where the evaporating temperature of frozen-food cabinets ranges from 30  C to 50  C [5].
* Corresponding author. Tel.: 55 34 32394022; fax: 55 34 32394206. E-mail addresses: (A. da Silva), bandarra@ (E.P. Bandarra Filho), (A.H.P. Antunes). 1359-4311/$ e see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2011.12.019

Refrigerant emissions from the commercial sector are relatively high [10] performed a study in 220 supermarkets in Norway and observed an annual leakage rate of 14% (not including stand-alone equipment). In these conditions, a considerable emission of greenhouse gases to the atmosphere is observed and it reinforces the need to reduce the leakage rate and also search for alternatives uids. Considering centralized systems, there are, basically, three possibilities to use the carbon dioxide. It can be used as a secondary refrigerant or it can be employed as a primary refrigerant in the low temperature stage of a cascade system. In all-CO2 centralized systems with the low temperature stage in cascade and, nally, this refrigerant can be used with separated circuits for LT and MT service, both rejecting heat directly to the environment. Casson [2] evaluated the COP for medium temperature (MT) and low temperature (LT) for CO2 systems, with evaporation temperatures between 10  C and 35  C, respectively. This author observed that the COP decreases with the increase of the external air temperature, obtaining a COP of 5.2 for MT and 2.5 for LT, when the ambient temperature was 0  C, and 1.5 and 0.8, respectively, for higher ambient temperature, 30  C. Girotto et al. [6] evaluated the efciency of a centralized all-CO2 system with 120 kW of capacity and compared it with a direct expansion R404A system. The authors concluded that, for medium temperature applications, the efciency was still somewhat lower if compared to the R404A system. For medium temperature

A. da Silva et al. / Applied Thermal Engineering 41 (2012) 30e35


applications, indirect cooling and low temperature CO2 cascade system showed to be the best option available to reduce HFC refrigerant charges. Sawalha [8] used a numerical simulation model to investigate the performance of two main CO2 transcritical system solutions: centralized with accumulation tank at the medium temperature level and parallel, that consists in two separate direct expansion (DX) circuits; one serves the medium temperature level cabinets and the other serves the freezers for low temperature level. For the ambient temperature range from 10  C to 40  C (different climate conditions: cold, moderate and hot) the two-stage centralized system solution presented the highest COP. In addition, the author concluded that the CO2 systems, except parallel, presented better performance for cold climates, whereas NH3eCO2 cascade system is better for hot climates. In this climate case, the CO2 modied centralized system presented only 1% higher annual energy consumption in comparison with the conventional R404A and the modications on the centralized system solution proved to be more important for high ambient temperature operating conditions. Both systems proved to be better alternatives in relation to the R404A (DX) system for supermarket refrigeration. Ge and Tassou [4] used the supermarket model called SuperSim to compare the performance of a conventional R404A refrigeration system and a CO2 booster system. Both systems were found to lead to very similar energy consumption. Floating head pressure control was implemented for both systems when they were in subcritical cycles; this strategy reduced signicantly heat recovery opportunities from the R404A system. The booster CO2 system, on the other hand, due to the higher cycle pressures and temperatures lends itself for heat recovery even during operation at subcritical conditions. It was found that heat recovery can satisfy 40% of the space heating demand of the supermarket. With the focus primarily on supermarket applications, this paper will analyze energy efciency comparisons carried out between the CO2 cascade system and the direct expansion conventional system using R404A and R22, and discusses their advantages and disadvantages, along with a comparison of a cost analysis with carbon dioxide. Relevant issues for the application of CO2 will also be highlighted. These energy efciency comparisons were conducted in the CO2 Technology Center that has been

operating in the Laboratory of the Bitzer Brazil since 2008. In this center three refrigerating systems with similar cooling capacities were installed. 2. Experimental facilities The experimental facilities consist of a cascade system with carbon dioxide for subcritical operation in the high temperature stage and HFC-404A (CO2 is pumped in the liquid phase to the evaporator with liquid recirculation and direct expansion for deepfreezing). Two others direct expansion systems were used for comparison, one with the refrigerant HFC-404A and other with the HCFC-22. Figs. 1 and 2 show the three refrigeration racks. They cool down storage rooms from 2  C and a deep-freeze room, to minus 25  C. There are also two deep-freeze islands working at minus 25  C that are only connected to the carbon dioxide circuit. The cooling capacity for normal refrigeration is about 20 kW, and about 10 kW in the deep-freeze range. The evaporators of the three refrigerating systems are designed as air-coolers and tted under the ceiling of each cold room. The condensers operate with either air-cooling or water-cooling. All machines and cold rooms are equipped with infrared sensors and a carbon dioxide extraction system, since in case of leakage a dangerous situation can occur and also to maintain the CO2 concentration at lower level. Only one system was in use at any one time to allow for energy comparisons. They have been running, week on week off, so an accurate comparison can be drawn among them. Table 1 presents the major technical data for each refrigeration rack of both the Medium Temperature (MT) and Low Temperature (LT) systems. The cooling capacity of the MT and LT multicompressor refrigeration systems is higher than the required thermal load from cold rooms and islands, as shown in the Table 2. 2.1. Compressors The compressors of each rack also have the operating option of both a frequency inverter and a head pressure capacity control unit, except for compressor model 4TCS-8.2 which is used in connection

Fig. 1. Refrigeration racks used in the present research.


A. da Silva et al. / Applied Thermal Engineering 41 (2012) 30e35

Fig. 2. (a) Thermostatic and electronic expansion valves used in air-coolers with R404A and R22; (b) CO2 manual expansion valve used in evaporators with liquid recirculation.

with the R22 in low temperature with controlled injection cooling (CIC), and also for the CO2 compressor model 2KC-3.2K which has only one head (two cylinder). As a result it is not possible for both compressors to operate with a head capacity control unit. The application range varied from 30 Hz to 70 Hz for the compressors with frequency inverters. 2.2. Condensers Each rack has the operating option of both air-cooled and watercooled condensers (mainly the high stage of the subcritical CO2 rack). Air-cooled and water-cooled condensers can be used to compare the energy efciency of the system. The air-cooled condenser fans also have the option of operating with frequency inverters as well as On/Off control pressure switches to control the condensing temperature. The water-cooled condenser types are of the shell-and-tube type and they operate with a cooling tower. 2.3. Evaporators The air-coolers that use R404A and R22, which are tted in the cold rooms, are direct expansion (DX) type and use either thermostatic expansion valves (TEV) or electronic expansion valves (EEV). The CO2 air-coolers are used for both MT cold rooms and LT deep-freeze room. The air-cooler used in the LT deep-freeze room is DX and uses only an electronic expansion valve. The other two CO2 air-coolers, for medium temperature, run with liquid recirculation and only use manual expansion valves to control the refrigerant ow.
Table 1 Technical data of multicompressor refrigeration systems. Subcritical CO2 rack (CO2/R404A) MT design condition TCond 5  C (CO2) TEvap 10  C (high stage) TCond 40  C (high stage) TEvap 30  C (CO2 e DX) TCond 5  C (CO2) 01 2KC-3.2K (CO2) 01 4CC-9.2.Y (R404A) MT cooling capacity LT cooling capacity 21.0 kW 9.8 kW R404A rack TEvap 10  C R22 rack TEvap 10  C

All relevant information comes together in a central monitoring unit that can also be controlled via LAN and the Internet. Medium temperature (MT) evaporator defrosting is achieved by off cycle (air) while low temperature (LT) evaporator defrosting is achieved with an electric heater, mainly for the LT CO2 evaporators (Islands and LT deep-freeze room). 3. System architecture 3.1. R404A and R22 refrigeration racks Both refrigeration racks work with two semi hermetic piston compressors (Octagon model 4CC-9.2 for MT and 4TCS-8.2 for LT) in parallel applications. Each rack has a common discharge line, but the suction header is split for MT and LT suction lines, while the discharge is collected in a common header and directed into a single oil separator. The oil return pipe enters into an oil receiver, which pushes the oil since it is equipped with a combination of an oil separator-reservoir. No oat valve was installed into oil separator-reservoir. Refrigerant ow from the discharge line to the condenser and then goes to a vertical liquid receiver, where a header liquid line distributes the liquid to the evaporators. The operating conditions are 30  C for low temperature (LT), 10  C for medium temperature (MT) and 40  C for condensation temperature. 3.2. CO2/R404a subcritical rack Fig. 3 shows a schematic diagram of the cascade CO2/R404A system. According to this gure, the MT CO2 evaporators run with liquid recirculation at 5  C, while the LT CO2 evaporators run with direct expansion at 30  C of evaporating temperature, through the vapor compressor cycle using a semi hermetic reciprocating compressor. As shown in this gure, in the R404A/CO2 cascade system, the CO2 and the R404A are in two separate circuits. These two circuits come into thermal contact in the interstage heat exchanger (also called cascade condenser) where they exchange heat with each other without mixing the two refrigerants. The interstage heat exchanger serves as a condenser for the CO2 system and as an evaporator for the R404A system. CO2 is used as pumped liquid for normal refrigeration and direct expansion for deep-freezing. The design of the CO2 rack has some unusual features, which are required to maintain the compressor temperatures at the recommended level. It was found that the performance of the CO2 compressor decreases in very low operating temperatures, which if left unchecked, would result in a high concentration of refrigerant

TCond 40  C TEvap 30  C TCond 40  C 01 4CC-9.2Y (MT) 01 4TCS-8.2Y (LT) 21.0 kW 10.7 kW

TCond 40  C TEvap 30  C TCond 40  C 01 4CC-9.2 (MT) 01 4TCS-8.2 (LT) 19.8 kW 9.9 kW

LT design condition Compressor models

A. da Silva et al. / Applied Thermal Engineering 41 (2012) 30e35 Table 2 Overview of refrigeration points. MT cold room Dimensions Thermal load Internal temperature


MT walk-in cooler 3.5 m 4 m 3.5 m 7.5 kW 2  C

LT Deep-freeze room 3.5 m 4 m 3.5 m 7.5 kW 25  C

LT islands

3.5 m 4 m 3.5 m 7.5 kW 0 C

5 m total length 2.5 kW 25  C

The two LT Islands only run with the CO2 refrigeration rack.

in the lubricating oil within the compressor sump, causing premature compressor failure. Superheating degrees between 20 K up to 30 K at the CO2 compressor suction were required to maintain acceptable sump temperatures in the CO2 rack. To prevent this, an additional heat exchanger was added between the CO2 suction line and the R404A high temperature stage liquid line, which maintained the CO2 suction gas temperature at the compressor between 10  C and 0  C. It is interesting to observe that the low temperature of the vapor at the inlet of the compressor is sometimes a problem, since the high density of CO2 in the vapor phase, the CO2 has a much larger capacity to absorb heat of the compressor castings than other gases. This can result in the compressor being chilled to a point where the compressor discharge line, and the compressor crankcase are covered in frost and ice and this will almost certainly means that the oil is being diluted by the refrigerant. Any refrigerant dilution will have an adverse effect on the life expectance of the compressors running gear. It is best to keep the compressor sump temperature at least at body temperature and the discharge should always be with high temperatures. The control of the CO2 superheating degree had an important hole to be provided by some means such as liquid-suction heat exchangers, using the R404A in the liquid phase thus providing the subcooling of the high stage liquid. Some types of control must be installed to limit the compressor return vapor temperature, either a bypass system or multiple heat exchangers staged to provide accurate control of the vapor inlet temperature. Either discharge vapor temperature or suction return temperature can be used to control the heat exchanger operation. Low return vapor superheating will give rise to oil and lubrication problems, while high superheat levels will cause motor overheating and subsequent failures, as well as high discharge temperatures.

4. Results The cascade system design can also take advantage of a high degree of liquid subcooling in the high stage circuit with R404A, which results in substantial reductions in pipe line diameters and, as consequence, a reduced refrigeration charge based on the tube diameter, compared to conventional refrigerants such as R404A and R22. Table 3 shows this comparison, including some data from Table 4, using the softwares [7] and [3]. The Table 5 presents the comparison among the tubes in terms of mass divided by length (kg/m), only relating the suction and liquid lines used in each evaporator. As a general guide, pipeline sizes, mainly the suction pipe work lines, can be reduced to approx 1/5 of the line sizes currently used with R404A and R22 for the same system capacity. Table 6 shows the total refrigerant charge used each rack. Due to the purchase price of CO2 being considerably less expensive than that of refrigerants currently in commercial use, such as R404A and R22, the total cost of the refrigerant charge can be signicantly reduced.

4.1. Energy analysis These three refrigeration systems were monitored and a supervisory system was used to acquire and integrate the variables such as temperature, pressure, refrigerant mass ow and also, it was able to capture the total power consumption of the entire system. Power was recorded at 15min intervals for the systems in operation, and included all aspects, such as the compressor motors and sump heaters, fan motors defrost heaters, evaporator fans, and so on. The energy efciency comparisons were an average over one year

Fig. 3. Schematic diagram of the cascade CO2/R404A system.


A. da Silva et al. / Applied Thermal Engineering 41 (2012) 30e35 Table 5 Total pipe work used in two cold rooms for MT as well as in the deep-freeze room for LT. Pipe work real length (m) e only suction line (SL) and liquid line (LL) each evaporator Diameter (mm) R22 Cold room 01 R22 Cold room 02 R22 Deep-freeze room R404A Cold room 01 R404A Cold room 02 R404A Deep-freeze room CO2 Cold room 01 CO2 Cold room 02 CO2 Deep-freeze room ASTM B-280 e kg/m R22 R404A CO2 9.52 12.7 15 LL 11 LL 15 LL 15.88 28.58 15 SL 11 SL 15 SL 15 SL 11 SL 34.93

Table 3 Suction and liquid line size comparison used in the two cold rooms for MT using CO2, R404A and R22. Refrigerant Suction line (wet return line for CO2, dry return line for R404A and R22) Liquid line (5  C for CO2) Cooling capacity (kW) DT (K) Velocity (m/s) Diameter (mm) Velocity (m/s) Diameter (mm) CO2 10 0.67 6.64 12.7 1.36 9.52 R404A 10 0.47 10.36 28.58 0.84 15.88 R22 10 0.55 9.02 28.58 0.57 12.7

Cold room # 1 and # 2 for MT (TEvap 10  C; TCond 40  C); Leqv 20 (m).

where the condensing temperature was maintained of the order of 38  C. It is most likely that with a CO2 system, a good proportion of the energy savings can be attributed to the subcooling of the high stage liquid, by the low stage suction gas. According to the Table 7, CO2 presented a higher efcient, around 22.3% in comparison with the R404A system, and 13.7% with the R22 system (both systems operates with frequency inverter and electronic expansion valves). However, when both the R404A and R22 systems used thermostatic expansion valves, CO2 became even more efcient, in which it represented 24.7% in relation to the R404A system and 15.5% to the R22 system. Electronic expansion valves save more energy costs since it is more reliable and precise in its way to control the refrigerant mass ow through the evaporator, as it receives all the information regarding the temperature and pressure in the evaporator outlet in order to control the opening and closing of the valve, according to the superheating. 4.2. Environmental analysis The direct global warming potential (GWP) of the three systems, due to direct emissions in the event of a total loss of the entire refrigerant charge, was also evaluated and represented great importance. CO2 is used as the base unit for measuring GWP, in this case is equal to 1. One kg of R404A has a GWP of 3260; one kg of R22 has a GWP of 1500. Therefore, the CO2/R404A cascade system has a total value of GWP of 48932, the R404A system has 407500 and the R22 system has 172500, according to each refrigerant charge in the system. As can be noted, the difference between the CO2/R404A cascade system and R404A system is of the order of 358468 and for R22 is 123568. 4.3. Cost analysis While the medium temperature system through the liquid recirculation system does not generally provide signicant reductions in energy costs, substantial savings can be achieved through a reduced refrigerant charge and a real reduction in the actual cost of the refrigerant. The cost of the three racks, the six air-cooled evaporators and the condensers, were all tracked and a complete comparison could
Table 4 Suction and liquid line size comparison used in the two deep-freeze room for LT using CO2, R404A and R22. Refrigerant Suction line (dry return line for CO2, R404A and R22) Liquid line (5  C for CO2) Cooling capacity (kW) DT (K) Velocity (m/s) Diameter (mm) Velocity (m/s) Diameter (mm)

15 LL 11 LL 15 LL 0.186 0.294 41 LL

15 11 15 15 11 15


15 SL

0.424 41 LL 41 SL

0.971 41 SL 26 SL

1.314 15 SL

41 LL

TOTAL (kg) 51.86 62.34 25.01

be drawn, since it is important to highlight that the CO2 rack and CO2 evaporators were built and had to be air freighted from abroad. The rack system costs were calculated separately. This has been done because the contractor supplies the interconnecting pipe work, as well as the pipe insulation, between the various items. The two racks that make up the cascade system using CO2 on low temperature and R404A on high temperature stage were found to be 18.5% (based on 2008 values) more expensive than single stage racks using R22 and R404A based on the same cooling capacity. This higher cost was largely due to the additional safety equipment that the CO2 system required under the Brazilian occupational health and safety codes, and the fact that a reasonable amount of the components were specially built and had to be air freighted from Australia. As CO2 gains in popularity and more CO2 equipment becomes available this additional cost will be reduced. The main factors at stake here are the large reduction in the size of the pipe work and insulation respectively. In addition, the CO2 evaporators were physically smaller and less expensive due to the increased specic cooling capacity of the refrigerant. It was found that the both R404A and R22 evaporators need approximately 20% more surface area to achieve the same thermal performance as the CO2 evaporators (based on the same temperature difference between evaporating and room temperature). Refrigerant charge in each of the three systems also has an inuence on the total cost. According to the Table 6 the cascade system has 32 kg of CO2 as well as an additional 15 kg of R404A, (32 15 47 kg). The other two racks using R404A and R22, they have 125 kg and 115 kg, respectively. In Brazil, the refrigerant HFC-404A has an average cost in 2011 of $35 (thirty ve dollars) per kg, the HCFC-22 costs $13 (thirteen dollars) per kg, while CO2 has a cost of $ $2.0 (two dollars) per kg. The CO2/R404A cascade system has an advantage over the R404A system of the $3786 and for the R22 system of $906. It is important to observe that the price of R22 in Brazil will increase since according to the Montreal Protocol the production and consumption will be frozen in 2013. Fig. 4 shows the comparison of each tested system. This analysis shows the total charge of refrigerant as well as the cost of pipes and
Table 6 Total of refrigerant charge used in each rack. Refrigeration rack Total refrigerant charge CO2/R404A (subcritical rack) CO2 e32 kg R404A e 15 kg R404A rack R404A e 125 kg R22 rack R22 e 115 kg

CO2 10 0.35 8.35 15.88 0.85 9.52

R404A 10 0.53 11.42 34.93 0.97 15.88

R22 10 0.39 10.28 28.58 0.68 12.7

Deep-freeze room LT (TEvap 30 C; TCond 40 C); Leqv 20 (m).

A. da Silva et al. / Applied Thermal Engineering 41 (2012) 30e35 Table 7 Total power consumption of the usage data. Power consumption per year CO2 system [kWh] (compressor with frequency inverter; LT evaporator with EEV) Power consumption per year R404A system [kWh] (compressor with frequency inverter; evaporators with EEVs) Power consumption per year R22 system [kWh] (compressor with frequency inverter; evaporators with EEVs) Difference in percentage [%] e CO2 vs. R404A; CO2 vs. R22 103.234 126.295 117.435 22.33 (R404A); 13.75 (R22) 103.234


Power consumption per year CO2 system [kWh] (compressor with frequency inverter; LT evaporator with EEV) Power consumption per year R404A system [kWh] 128.701 (compressor with frequency inverter; evaporators with TEVs) Power consumption per year R22 system [kWh] 119.212 (compressor with frequency inverter; evaporators with TEVs) Difference in percentage [%] e CO2 vs. R404A; CO2 vs. R22 24.67 (R404A); 15.47 (R22)

systems. Many advantages of CO2 system in relation to R404A and R22 can be listed, such as: (i) Reduction of the electric energy consumption (in this case it can varies between 13 and 24%); (ii) Low compression ratio and increase of the useful life of the CO2 compressor; (iii) High CO2 density and high pressure in the low pressure stage; (iv) Reduction of CO2 piping diameter sizes; (v) Reduction of CO2 refrigerant charge; (vi) Low price of CO2 purchase; (vii) Higher enthalpy, degree of liquid subcooling and cooling capacity; (viii) Low GWP and less carbon taxes (CO2); (ix) Small volumetric displacement and smaller sized CO2 compressors; (x) Smaller refrigeration rack, compact installation and lower compressor numbers; (xi) Smaller and efcient evaporator coils; (xii) Reduced installation and lower maintenance costs. Due to the rapidly changing cost of refrigerants and the expected reduction in the cost of CO2 compatible components and taking in account the high volatility of the energy cost around the world, it will be possible to verify the increase of the new installations using CO2 as refrigerant.

Acknowledgements The authors gratefully acknowledge all support, technical assistance and research materials, which attributed to this investigation by BITZER Brazil, BITZER Australia, Carel Brazil, Carel Italy and FAPEMIG.

[1] P.K. Bansal, 2011, In-tube boiling heat transfer of CO2-lubricant mixture at low temperatures: Preliminary results, ASHRAE conference, ASHRAE-86018, Las Vegas (USA), p. 9. [2] V. Casson, Theoretical and experimental analysis of CO2 as a refrigerant in retail refrigeration (in Italian). PhD Thesis. Universit di Padova, Italy; 2002. [3] Coolpack software, (2006) [4] Y.T. Ge, S.A. Tassou, Performance evaluation and optimal design of supermarket refrigeration systems with supermarket model SuperSim. Part II: model applications, International Journal of Refrigeration 34 (2011) 540e549. [5] H.M. Getu, P.K. Bansal, Thermodynamic analysis of an R717-R744 cascade refrigeration system, International Journal of Refrigeration 31 (2008) 45e54. [6] S. Girotto, S. Minetto, P. Neksa, Commercial refrigeration system with CO2 as refrigerant experimental results, International Journal of Refrigeration 27 (2004) 717e723. [7] Micropipe software, (2006) [8] S. Sawalha, Theoretical evaluation of trans-critical CO2 systems in supermarket refrigeration. Part I: modeling, simulation and optimization of two system solutions, International Journal of Refrigeration 31 (2008) 516e524. [9] J.A. Parise, R. Marques, Editorial in The role of heat transfer in refrigeration, Heat Transfer Engineering 26 (9) (2005) 1e4. [10] O.J. Veiby, Internal Records, Documentation in the ICA Supermarket Chain in Norway 2003. Oslo, Norway (2003).

Fig. 4. Comparison between the three racks designs.

refrigerant. It is important to observe that the direct impact of the GWP (Global Warming Potential) obtained for the R404A rack is almost 10 times higher in comparison with the R744/R404A rack. 5. Conclusion This comparison showed high performance and environmentally friendly process can be applied to reduce the effects of direct and indirect of the global warming, achieving long term cost reduction of the equipment. Clearly, there are numerous advantages, which will ensure that carbon dioxide cascade systems have a place in refrigeration