The project was carried out in collaboration between Winns, Sintef and NTNU with support from the Research Council. Synthetic refrigerants that are used in almost all heat pumps today are very potent greenhouse gases with a climate footprint that is 1,500 – 4,000 times higher than CO2. According to the UN, better leakage control and the phasing out of synthetic refrigerants as well as a transition to natural refrigerants can reduce future global warming by up to 0.5°C up to 2050. The EU has introduced strict requirements for the phasing out of synthetic refrigerants and implementation has already come a long way. Hence a transition to the use of natural refrigerants is being forced. CO2 is an excellent naturally occurring refrigerant with 750 to 4,000 times lower climate impact than the synthetic media in use today.
When the water or air to be heated has a low incoming temperature, a heat pump that uses CO2 as a refrigerant achieves very good energy efficiency (low power consumption in relation to the heat produced). At an input temperature above 30°C, the energy efficiency falls relatively quickly. In many heating systems, the return temperature is 35°C - 40°C. It is therefore very interesting to improve the performance of CO2 heat pumps in this return temperature range.
With an ejector implemented in a heat pump, it is possible to increase the compressor suction pressure and reduce power consumption. One of the main goals in the project was the development of a dedicated ejector for CO2 heat pumps where the return temperature from the heating system is 35°C - 40°C. Also, the ejector should be as simple and inexpensive as possible. To achieve the lowest possible cost of the ejector, it was built up from standard components as far as possible. In collaboration with NTNU and Winns, SINTEF has further developed its existing numerical models for dimensioning ejectors, as well as investigated the flow conditions in an ejector with variable geometry and developed an associated controller.
Testing of the ejector showed that the suction pressure on the compressor could be increased by up to 6 bar, which results in a reduction in power consumption of around 14%. The greatest increase in suction pressure was achieved at 100% system performance. As many heat pumps run at partial load for large parts of the year, it was interesting to test the partial load performance of the adjustable ejector. The ejector was easily regulated down to 50%, but when regulated down, a marked drop in performance was observed. To maximize the effect of the ejector in heat pumps with adjustable performance, several ejectors in parallel should therefore be considered, where each ejector is dimensioned for different performance adapted to the operating conditions.
Another main activity in the project was experimental testing of evaporators to extract heat from greywater and to test different evaporator geometries in combination with ejectors. Gray water for drainage often maintains 20 - 30°C and is, thermally speaking, a very good source of heat for a heat pump. The biggest disadvantage of greywater is various pollution. A filled evaporator (shell and coil) and a plate heat exchanger were tested. The main findings from here were that a shell & coil/ filled evaporator outperforms a plate heat exchanger in performance, but at the expense of a high pressure drop in the coil and increased power consumption on the pump, furthermore the tank must be sized for a high pressure and the CO2 filling in the system will be high. However, the high pressure drop and turbulent flow in the coil mean that this solution is less susceptible to build-up from pollution in the gray water than a plate heat exchanger.
A combined solution where the water was first cooled down in the shell & coil evaporator and then further cooled down in the plate evaporator was also tested. The ejector was used to raise the pressure on the gas from the plate exchanger. The ejector then ended up in a poor performance area and, not surprisingly, no better performance was achieved with a two-stage solution.
Oil return from a shell & coil/filled evaporator can be a problem as there is no driving pressure differential to return the oil. A primitive ejector that was supposed to ensure oil return was assembled from tubing parts. It proved to do the job well and no problems were observed with oil return to the compressor.
The implementation of the project has given the participants an increased understanding of the process and learning from the project has already been implemented in heat pumps built by Winns.
Mad basis i læring fra prosjektet har Winns designet, bygget og idriftsatt en CO2 varmepumpe med gråvann som varmekilde. Selskapet har de seinere årene også utviklet kjøleuniter, chillere og varmepumper for offshore. Det tette samarbeidet med Sintef/ NTNU forskningsmiljø har gitt viktige bidrag til å etablere Winns som en ledende leverandør til offshoremarkedet.
The underlying idea of the innovation is realization of a CO2 ejector heat pump cycle where the isenthalpic throttling in expansion devices would be entirely replaced with expansion work recovery in ejectors. The innovative CO2 heat pump ejector cycle would contain 2 main ejector modules:
+ High-pressure refrigerant ejector, aimed at maintaining the requested compressor discharge pressure in the system,
+ Low-pressure refrigerant ejector, aimed at circulating the evaporating refrigerant through evaporator,
+ Use the said low pressure ejector to intermittently return compressor oil to MT receiver where it can be easily separated and returned to the compressor in a conventional way.
The innovation is intended to minimize the unnecessary loss of expansion work available in the cycle and thus to elevate the overall energy performance of transcritical CO2 heat pumps, especially for higher return water temperatures. Moreover, by applying ejectors to enforce the circulation of both refrigerant and oil, several other issues will be addressed, namely:
+ Reduction of compression ratio of compressors.
+ Flooded evaporators.
+ Effective and reliable oil management system.
The anticipated results of the project will comprise (i) upgraded simulation tools for refrigerant-oil mixture flows in heat pump and refrigeration installations, (ii) a line of newly developed and validated components (ejectors, separation/accumulation tanks, heat exchangers) for throttling-free CO2 transcritical cycles, (iii) a demo CO2 heat pump unit developed, manufactured and tested in both laboratory and field conditions.
The project owner is Winns AS. In addition to administration duties Winns will contribute throughout the project with knowledge from years of designing, building and operating CO2 heat-pumps as well as experimental testing and field testing. SINTEF ER together with NTNU will lead and perform the research and development activities in the project in close cooperation with Winns.