ASHRAE Journal - February 2009 - 19

and decreased system maintenance. Concerns about increased energy consumption have largely been overcome through proper design practice.1 Low-temperature secondary systems were introduced in the late 1990s using various potassium-based salts. Dozens of systems were installed by multiple manufacturers, but some difficulties were experienced with leakage of the secondary fluid and resulting corrosion of surrounding materials. Although the potassium-based fluids exhibit superior performance to other single-phase fluids,2 material compatibility continues to be a concern. Use of the potassium salts has slowed dramatically in the last several years with only a handful of new installations. However, recent introduction of cost-effective plastic piping materials and components may allow these systems to become a viable alternative for some future applications. In searching for alternative fluids suitable for low-temperature application, it became clear that CO2 as a two-phase secondary coolant showed several advantages compared to the singlephase salts. Primarily, these were lower pumping power, smaller pipe sizes, excellent heat transfer properties, and good material compatibility with the additional benefit of the low cost of the fluid. The main disadvantage of CO2 appeared to be the higher operating pressures and availability of components. CO2 Supermarket Installations in North America In 2001, laboratory testing of low-temperature CO2 secondary systems was initiated. After extensive investigation of the system operation, display-case and unit-cooler performance, and various piping configurations and control methods, the first U.S. system was installed in the field in mid-2006. By the end of 2008, nine low-temperature carbon dioxide systems were operational in the U.S. and Canada. All systems used CO2 as a low-temperature two-phase secondary fluid with stores ranging in size from small markets to large supercenters and warehouse-style stores, and with system loads ranging from 22 to 160 kW (75 to 550 kBtu/h). The systems also universally included a primary refrigeration system using an HFC (R-404A or R-507A). Most installations included a medium-temperature secondary coolant system using propylene glycol, although one site was installed with a centralized direct-expansion HFC system that served as a comparison site for energy and performance monitoring. Higher pressures and availability of components has not proven to be problematic. Concurrent to the introduction of these systems, significant introduction of components suitable for application with R-410A, increasingly becoming the domestic A/C refrigerant of choice for replacement of R-22, made the majority of these components readily available. Successful installation of the systems has relied heavily on comprehensive contractor training programs developed specifically for these secondary coolant applications and will continue to be critical to the implementation of these and other types of CO2 systems in the future. The selection of CO2 grade or purity-level has been carefully considered. Initial systems used CO2 gas of 99.99% purity February 2009 (Coleman- or instrument-grade). However, some systems have started using 99.5% industrial-grade materials when better grades are not readily available. Charging the CO2 through liquid filter-driers and a purge of non-condensable gases during start-up seems to make specification of a higher purity unnecessary. CO2 costs and purity-level availability appear to vary widely throughout the country, however most installations have been able to obtain the materials around 1.10 $US/ kg (0.50 $US/lb). CO2 systems have shown to be susceptible to the same types of mistakes that can plague any field-installed refrigeration system. Problems have included contractors not evacuating 100% of the piping network and charging of impure refrigerant. Similar to a direct expansion system, non-condensable gases not evacuated from the system or charged into the system from impure refrigerant tend to accumulate in the condenserevaporator heat exchangers, resulting in what appears to be reduced condensing capacity. In extreme cases, flow of the CO2 thermosiphon effect to these heat exchangers can be blocked completely, requiring system reevacuation and recharging. Good-quality filter-driers installed in the liquid lines and changed shortly after start-up have eliminated any problems associated with moisture in the systems. Since a secondary-loop CO2 system is essentially a recirculated liquid system with wet returns (circulation rate greater than one), balancing the flow between loads had not been a problem. Proper application of the piping network combined with careful coil design has shown that balancing can be done during the design-phase of the project and removes any complicated field-balancing procedures from consideration. Since CO2 systems are typically designed for maximum working pressures well below the saturation pressure at ambient temperatures (60 bar at 23°C, or 900 psig at 75°F), back-up or auxiliary refrigeration units have been installed on some stores to provide a source of cooling for the CO2 during extended power outages or maintenance procedures. Experience has shown that it takes several hours for the pressure in the system to reach levels where relief to ambient would be required, and opinion remains divided on future application of this cooling depending on customer experiences, reliability of the power supply, and availability of a back-up power-supply. Energy consumption comparisons have been made between one CO2 field installation and three similar HFC DX systems nearby. Results have so far indicated a 2% to 3% average reduction in energy required by the low-temperature CO2 secondary coolant store during several summer months in 2007 compared with the surrounding stores. This was better than anticipated as a dual-suction group HFC DX system was compared with a single-suction group CO2 secondary system, and at best, energy consumption was not expected to be favorable during the warmer months, however a more detailed analysis is required to get an idea of annual performance expectations. Distribution Piping Effects In the first implementations of low-temperature systems as ASHRAE Journal 19

ASHRAE Journal - February 2009

Table of Contents for the Digital Edition of ASHRAE Journal - February 2009