Alternative Refrigerant Evaluation for Mini-Split Air Conditioners
Finding R-22 and R-410A alternatives for high-ambient-temperature environments
The Oak Ridge National Laboratory (ORNL) High-Ambient-Temperature Evaluation Program for low-global warming potential (GWP) refrigerants aims to develop an understanding of the performance of low-GWP alternative refrigerants to hydrochlorofluorocarbon (HCFC) and hydrofluorocarbon (HFC) refrigerants in mini-split air conditioners under high-ambient-temperature conditions. This article highlights the parties involved, the alternative refrigerant selection process, the test procedures, and the final results.
ORNL designed a test matrix of 84 tests. Table 1 shows the refrigerants identified for testing by ORNL with guidance from an expert panel. The expert panel is composed of members from various nations, as well as United Nations Environment Programme and United Nations Industrial Development Organization personnel. Guided by input from the expert panel, ORNL selected the refrigerants based on their GWP, commercial availability, and physical properties, also considering whether information about the characteristics of the refrigerants is readily available. ORNL conducted tests using two “soft- optimized” ductless mini-split air conditioners provided by Carrier. Both units have a cooling capacity of 5.25 kWth (1.5 TR). One unit is designed to operate with R-22 refrigerant (2.78 coefficient of performance [COP], equivalent to a 9.5 energy efficiency ratio [EER]). The other is designed to use R-410A refrigerant (3.37 COP, equivalent to an 11.5 EER).
Testing was conducted in ORNL’s Multi-Zone Environmental Chambers. This facility contains several controlled chambers, which allows for the setup of an outdoor unit and one or more indoor units, each in separate spaces. The test procedure involved soft-optimization (moderate modifications to the units to enable them to run with each refrigerant) and testing at six different environmental testing conditions.
For all refrigerants, including R-22 and R-410A, efficiency degraded with increased ambient temperature. Further, in evaluating the results, it is important to keep in mind that the test units were not designed specifically for the alternative refrigerants, but rather were soft-optimized. As a result, the alternative refrigerants should not be expected to perform as well as they would if the system designs were fully optimized for them.
R-22 UNIT RESULTS
Table 2 lists the characteristics of the alternative refrigerants evaluated in the R-22 unit.
Table 3 summarizes the test results at moderate ambient temperatures (AHRI Standard 210/240 A and B). At each of these two ambient conditions, the results from the R-22 unit showed that all the alternative refrigerants except R-290 (propane) resulted in lower performance than R-22 (in terms of both COP and cooling capacity). R-290 led to lower cooling capacity but higher COP than the baseline.
Table 4 summarizes the results at high ambient temperatures (hot and extreme). At the highest-ambient-temperature test condition (“extreme” 55°C outdoor temperature), the system with R-290 achieved an 8 percent higher COP than the baseline with a 9 percent drop in cooling capacity. R-444B resulted in only modest performance degradation relative to R-22 at the extreme condition: a 7 percent lower COP and 4 percent lower cooling capacity than the baseline refrigerant. Also at the extreme condition, ARM-20B had a cooling capacity only 3 percent lower than the baseline refrigerant, with an 11 percent drop in COP.
Based on the uncertainty analysis, the air-side capacity had an uncertainty of ±2.3 percent and the air-side COP had an uncertainty of ±2.4 percent. Considering these uncertainties and the potential for further performance enhancements, refrigerants with performance values within 5 percent of the baseline may be expected to match the performance of R-22 with further soft-optimization. Furthermore, values within 10 percent of the baseline indicate an acceptable match that requires additional engineering design to reach parity with R-22 performance. For performance losses greater than 10 percent, significant redesign of the unit would likely be necessary to match the performance of the baseline. This suggests that, at high ambient temperatures, at least a few alternative refrigerants could be expected to perform at least as well as R-22, if not better, with additional soft-optimization, while others might require additional engineering redesign.
R-410A UNIT RESULTS
Table 5 lists the alternative refrigerants evaluated in the R-410A unit and their characteristics.
Table 6 summarizes the results at moderate ambient temperatures (AHRI Standard 210/240 A and B). Three of the alternatives to R-410A resulted in lower COPs and cooling capacity, partly due to the fact that the unit was only soft-optimized and didn’t undergo engineering design for zeotropic refrigerant mixtures. Nevertheless, the difference between R-410A and these three alternatives was not as significant as the difference between R-22 and its alternatives. One of the other alternatives, R-32, led to both higher COP and higher cooling capacity than the baseline refrigerant at both test conditions. In addition, the system with DR-55 had slightly lower cooling capacity but slightly higher COP than the baseline R-410A at both test conditions.
Table 7 summarizes the results at high ambient temperatures (hot and extreme). At the highest- ambient-temperature test condition (“extreme” 55°C outdoor temperature), R-32 resulted in a 6 percent higher COP and 13 percent higher cooling capacity than the baseline. All of the other alternatives delivered higher COPs than the baseline, with nearly the same cooling capacity at the extreme test condition. For instance, HPR-2A led to a 6 percent higher COP and approximately the same cooling capacity as the baseline. At high ambient temperatures, the results demonstrate that these alternative refrigerants allow for equivalent or better performance than the baseline, in terms of both COP and cooling capacity, for soft-optimized conditions.
Based on the uncertainty analysis, the air-side capacity had an uncertainty of ±1.5 percent and the air-side COP had an uncertainty of ±1.6 percent. Considering these uncertainties and the potential for further performance enhancements, refrigerants with performance values within 5 percent of the baseline may be expected to match the performance of R-410A with further soft-optimization. Furthermore, values within 10 percent of the baseline indicate an acceptable match that requires additional engineering design to reach parity with R-410A performance. This suggests that, at high ambient temperatures, all of the alternatives to R-410A tested could deliver performance equivalent to, or better than, the baseline. In many cases, achieving such performance would not require further soft-optimization or minor redesign; however, further engineering may still be required to ensure safe and reliable operation.
The test results from this evaluation program demonstrate that there are several viable alternatives to both R-22 and R-410A at high ambient temperatures. In some cases, there was a significant improvement in the performance of the alternatives over that of the baseline, in terms of both COP and cooling capacity. In other cases, the performance of the alternatives fell within 10 percent of the baseline, which suggests that parity with baseline performance would likely be possible through additional engineering design.
The R-22 alternative refrigerants showed promising results at high ambient temperatures: although both of the A1 alternative refrigerants lagged in performance, some of the A2L refrigerants showed capacity within 5 percent and efficiency within approximately 10 percent of the baseline system. The A3 refrigerant (R-290) exhibited higher efficiency consistently; however, it did not match the cooling capacity of the baseline system. The most promising A2L refrigerants exhibited slightly higher compressor discharge temperatures, while the A3 refrigerant exhibited lower compressor discharge temperatures.
The R-410A alternative refrigerants are all in the A2L safety category. Most of them showed significant potential as replacements. R-32 was the only refrigerant that showed consistently better capacity and efficiency; however, it resulted in compressor discharge temperatures that were 12-21°C higher than those observed for the baseline refrigerant. These higher temperatures may negatively impact compressor reliability. DR-55 and HPR-2A had higher COPs than the baseline and matched the capacity of the baseline at both the hot and extreme test conditions. R-447A and ARM-71a had lower cooling capacity than the baseline at all ambient conditions. The system efficiency of R-447A showed improvement over the baseline at high ambient temperatures; for ARM-71a, the efficiency was similar to the baseline at all test conditions.
The efficiency and capacity of the alternative refrigerants could be expected to improve through design modifications that manufacturers would conduct before introducing a new product to market. However, given that the scope of this study covered only soft-optimized testing, no detailed assessment can be made of the extent of potential improvements through design changes. Within the bounds of what is possible in optimizing the units for soft-optimized tests, the ORNL test plan included only minor optimizations, including refrigerant charge, capillary tube length, and lubricant change. Therefore, these are conservative results that probably could be improved through further optimization. Additional optimization, including heat transfer circuiting and proper compressor sizing and selection, would likely yield better performance results for all of the alternative refrigerants.
Losses in cooling capacity are typically easier to recover through engineering optimization than are losses in COP. The primary practical limit to improvements in capacity is the physical size of the unit; but that is not expected to be a significant concern in this case, based on the magnitude of the capacity losses exhibited in this evaluation program. Thus, the COP losses and the increases in compressor discharge temperature are particularly important results of this testing program, in that these variables will be the primary focus of future optimization efforts.
This performance evaluation shows that viable replacements exist for both R-22 and R-410A at high ambient temperatures. Multiple alternatives for R-22 performed well. Many R-410A alternatives matched or exceeded the performance of R-410A. These low-GWP alternative refrigerants may be considered as prime candidate refrigerants for high ambient temperature applications. Before commercialization, engineering optimization carried out by manufacturers can address performance loss, the increase in compressor discharge temperature that many alternatives exhibited (particularly the R-410A alternatives), and any safety concerns associated with flammable alternatives.
Publication date: 3/7/2016