Enhancing Heating Performance of Low-Temperature Air Source Heat Pumps Using Compressor Casing Thermal Storage

Zhongbao Liu 1,*, Fengfei Lou 1, Xin Qi 2 and Yiyao Shen 3 1 Department of Refrigeration and Cryogenic Engineering, College of Environmental and Energy Engineering, Beijing University of Technology, 100 Pingleyuan Road, Chaoyang, Beijing 100124, China; loufengfei@emails.bjut.edu.cn 2 China Household Electric Appliance Research Institute, 6 Yuetan beixiao Str, Xicheng, Beijing 100037, China; qix@cheari.com 3 Key Laboratory of Urban Security and Disaster Engineering of Ministry of Education, Beijing University of Technology, Beijing 100124, China; sshenyiyao@163.com * Correspondence: liuzhongbao@bjut.edu.cn; Tel./Fax: +86-010-67391613


Introduction
With the development of the global economy, energy shortage and environmental pollution have become major global problems. Air source heat pumps (ASHPs), as clean and efficient heating devices, have broad applications. However, when ASHPs are used in extreme temperatures, they suffer from problems such as low heating efficiency and excessively high compressor discharge temperature [1][2][3]. Researchers have proposed many solutions to these problems, such as the addition of auxiliary heating systems, the use of two-stage or multistage compression cycles, and the use of enhanced vapor injection (EVI). Existing studies have helped broaden the use of ASHPs, especially those with EVI, which has become a research hotspot in this field with good comprehensive performance [4,5].
Wang [6] reported that rotary compressors encounter severe performance degradation at low ambient temperatures, and thereby introduced a rotary compressor prototype with a novel end-plate injection structure. Compared with a single-stage rotary compressor, the proposed rotary compressor with a novel end-plate injection structure enhanced the heating capacity and coefficient of performance (COP) by 16.2-31.6% and 5.1-12.0%, respectively. Chen [7] proposed a new direct-expansion being throttled to point e by the capillary. The vapor-injected supply circuit (Branch 3) opens on its own at a low temperature. Hence, the refrigerant vapor discharged from the flashtank enters the vapor-injected supply opening of the compressor, and the refrigerant liquid at point f discharged from the flashtank is throttled to point h by the electronic expansion valve (EEV). The gas-liquid mixture from the EEV enters Branches 1 and 2, which is called Refrigerant A. Refrigerant A starts to split at point Y and enters Branches 1 and 2, which is called Refrigerant A 1 and Refrigerant A 2, respectively. By opening and adjusting the opening degree of the manual control valve, Refrigerant A 1 enters Branch 1, and the secondary evaporator is located on this branch. The thermal storage heat exchanger wrapped outside the compressor casing is used as a secondary evaporator to heat Refrigerant A 1 . Meanwhile, Refrigerant A 2 enters Branch 2, and the main evaporator (outdoor heat exchanger) is located on this branch. In the end, Refrigerants A 1 and A 2 are mixed into the compressor suction port at point Z as superheated steam.  The heat of compressor casing is used as a new auxiliary heat source, and its temperature is much higher than the ambient temperature, which is equivalent to the heat source of main evaporator. Theoretically, the suction and evaporation temperatures of the DE-CCTS are increased, so the suction and evaporation pressure are increased. In this manner, discharge temperature and pressure are decreased, and the heating capacity is increased. From the pressure-enthalpy diagram in Figure 2, it can be seen that h′-a′ and d′-c′ represent the suction and discharge pressure of the DE-CCTS, respectively. The suction pressure Po is higher than that of the original system, and the discharge pressure Pk is lower than that of the original system. At the same time, it can be seen that the compression ratio of the DE-CCTS is also reduced compared with that of the original system.

DE-CCTS Device
As shown in Figure 1, DE-CCTS is based on a China Gree low-temperature ASHP water heater, whose model and operating refrigerant are KFRS-3.5JPd/NaA-1 and R410A, respectively. The main components of the proposed DE-CCTS are the compressor, water tank (with a condenser coil inside), flashtank, outdoor heat exchanger (main evaporator), and thermal storage heat exchanger (secondary evaporator). The rated power and maximum power of the compressor are 0.833 and 2 kW, respectively. The water tank is self-made with a total capacity of 50 L. It has an inner coiled pipe heat exchanger with a total length of 10 m that heats the water in the water tank. The main evaporator is a finned heat exchanger and the secondary evaporator is self-made.
The secondary evaporator is the core part of this experiment, which is used to fill the PCM and complete the thermal storage and release operation between the PCM and the refrigerant in the heating cycle. Its size is determined in accordance with the mass of the thermal storage PCM and the shape of the compressor casing. As shown in Figure 3, the secondary evaporator is made into a hollow cylinder. The secondary evaporator is filled with 1.74 kg paraffin with phase change temperature of 45 °C, and the inner disc of paraffin has a diameter of 6 mm and a length of 2.8 m serpentine copper tube. The secondary evaporator is surrounded by a thin iron sheet to form a circular tank. Because the secondary evaporator needs to closely fit the compressor casing to absorb the waste heat, a layer of silica gel between the casing and the iron sheet is applied to enhance heat conduction. The experiment prototype is a 1.5 P compressor with height of 200 mm and diameter of 100 mm. According to the shape and size of the compressor, the inner diameter and outer diameter of the The heat of compressor casing is used as a new auxiliary heat source, and its temperature is much higher than the ambient temperature, which is equivalent to the heat source of main evaporator. Theoretically, the suction and evaporation temperatures of the DE-CCTS are increased, so the suction and evaporation pressure are increased. In this manner, discharge temperature and pressure are decreased, and the heating capacity is increased. From the pressure-enthalpy diagram in Figure 2, it can be seen that h -a and d -c represent the suction and discharge pressure of the DE-CCTS, respectively. The suction pressure P o is higher than that of the original system, and the discharge pressure P k is lower than that of the original system. At the same time, it can be seen that the compression ratio of the DE-CCTS is also reduced compared with that of the original system.

DE-CCTS Device
As shown in Figure 1, DE-CCTS is based on a China Gree low-temperature ASHP water heater, whose model and operating refrigerant are KFRS-3.5JPd/NaA-1 and R410A, respectively. The main components of the proposed DE-CCTS are the compressor, water tank (with a condenser coil inside), flashtank, outdoor heat exchanger (main evaporator), and thermal storage heat exchanger (secondary evaporator). The rated power and maximum power of the compressor are 0.833 and 2 kW, respectively. The water tank is self-made with a total capacity of 50 L. It has an inner coiled pipe heat exchanger with a total length of 10 m that heats the water in the water tank. The main evaporator is a finned heat exchanger and the secondary evaporator is self-made.
The secondary evaporator is the core part of this experiment, which is used to fill the PCM and complete the thermal storage and release operation between the PCM and the refrigerant in the heating cycle. Its size is determined in accordance with the mass of the thermal storage PCM and the shape of the compressor casing. As shown in Figure 3, the secondary evaporator is made into a hollow cylinder. The secondary evaporator is filled with 1.74 kg paraffin with phase change temperature of 45 • C, and the inner disc of paraffin has a diameter of 6 mm and a length of 2.8 m serpentine copper tube. The secondary evaporator is surrounded by a thin iron sheet to form a circular tank. Because the secondary evaporator needs to closely fit the compressor casing to absorb the waste heat, a layer of silica gel between the casing and the iron sheet is applied to enhance heat conduction. The experiment prototype is a 1.5 P compressor with height of 200 mm and diameter of 100 mm. According to the shape and size of the compressor, the inner diameter and outer diameter of the secondary evaporator are 100 mm and 180 mm, respectively, that is, the wall thickness of the secondary evaporator is 40 mm and the height is 150 mm.
Energies 2020, 13, x FOR PEER REVIEW 6 of 21 secondary evaporator are 100 mm and 180 mm, respectively, that is, the wall thickness of the secondary evaporator is 40 mm and the height is 150 mm.

Measuring Instrument and Measuring Point Layout
The experimental data required to be collected are the real-time changes in the values of parameters such as temperature, power, and humidity. The data are collected by automatic detection using Agilent 34,970, which is a multifunctional data acquisition instrument, and the data were collected, on average, once every 30 s. The measurement parameters and the instrument characteristics are shown in Table 1.

Measuring Instrument and Measuring Point Layout
The experimental data required to be collected are the real-time changes in the values of parameters such as temperature, power, and humidity. The data are collected by automatic detection using Agilent 34,970, which is a multifunctional data acquisition instrument, and the data were collected, on average, once every 30 s. The measurement parameters and the instrument characteristics are shown in Table 1. The DE-CCTS has nine temperature measurement points, namely, the temperatures of the compressor suction and discharge ports, compressor casing, thermal storage PCM and water, and inlets and outlets of the main evaporator and secondary evaporator. Multiple temperature sensors are arranged, and the average value is taken after performing multiple tests to ensure the accuracy of the experiment. The arrangement of temperature measuring points is shown in Figure 1.

Experimental Method
All tests were in accordance with low ambient temperature air source heat pump (water chilling) packages-Part 2: Heat pump (water chilling) packages for household and similar applications of the China National standard [18], and were completed in the Environmental Simulation Laboratory in the Department of Refrigeration and Cryogenic Engineering of Beijing University of Technology.
The length and width of the laboratory were 2 m, and the height was 3 m. The wall was mainly insulated by rock wool board. The temperature and humidity were regulated in the following three ways, respectively: The heating method was a fin heating tube, the cooling method was an air-cooled refrigeration unit, and the humidification method was the water vapor heated by heating tube. The laboratory could simulate nine different ambient temperature conditions so that the outdoor temperature met the required experimental requirements, and the humidity requirements could also be achieved, namely, −25 • C; −18 • C; −12 • C, 30% RH; −6 • C, 50% RH; 0 • C, 40% RH; 7 • C, 90% RH; 15 • C; 25 • C; 35 • C. When the outdoor temperature reached the condition, the indoor temperature was 25 • C.

On-Time Mode of DE-CCTS and Performance Comparison
The DE-CCTS was tested under nine different ambient temperatures of −25 The opening degree of the manual control valve (equivalent to adjusting the mass flow of Refrigerant A 1 ) and the opening time of Branch 1 at these temperatures served as the focuses of this study. Multiple tests on the proposed DE-CCTS were performed at each ambient temperature. By constantly changing the opening degree of the manual control valve and the opening time of Branch 1, the suction and discharge temperatures, the rate of water temperature rise, and the COP of the DE-CCTS were compared to determine the optimal on-time mode under different ambient temperatures.
In the experiment performed under the nine ambient temperatures, the water in the water tank was controlled to ensure the same initial water temperature, water temperature rise, and mass of water. Under the premise of controlling the same heating capacity and optimal on-time mode, the suction and discharge temperatures, the rates of water temperature rise, and the COP values of the original system and the proposed DE-CCTS were compared. Under nine temperature conditions, the original system and the DE-CCTS all heat a tank of water of 50 L from 20 • C to 52 • C, that is to say, the heating capacity of two systems is the same, and the COP is determined by the power consumption. The calculation formula of COP is shown in (1), where M refers to a tank of water with a mass of 50 kg; C refers to the specific heat capacity of water of 4.2 kJ/(kg· • C); ∆t refers to a water temperature rise of 32 • C; W refers to the integral of instantaneous power under each temperature condition, that is, the power consumption under each temperature condition.
M-Mass of the water in the water tank, kg, C-Specific heat capacity of water, kJ/(kg· • C), ∆t-Water temperature rise, • C, W-Power consumption, kJ.

Uncertainty Analysis COP Value
Based on the mathematical model of synthesis uncertainty, the uncertainty of COP value is analyzed from the perspective of Type-A uncertainty and Type-B uncertainty. Type-A uncertainty Energies 2020, 13, 3269 7 of 18 analysis refers to that under each temperature condition, the experiment is carried out many times, and finally multiple COP values are taken. The DE-CCTS has been tested many times (six times) under each temperature condition. Under the premise that the data size and trend are basically the same with time under each temperature condition, these COP values are recorded as COP 1 , COP 2 , COP 3 , COP 4 , COP 5 and COP 6 , respectively. In addition, the average of the six COP values is calculated according to formula (2) and recorded as COP'. Based on the statistical analysis method, the experimental standard deviation can be obtained, so the calculation formula of uncertainty of Type-A of the measured COP value are shown in (3).
Type-B uncertainty analysis means that the uncertainty is not obtained by repeated date acquisition of experiment, while the uncertainty is evaluated according to the uncertainty of measurement device. Temperatures are measured using Pt100 temperature transducers with an uncertainty of ±0.1 • C, and the power input to the compressor is measured using a power meter with an uncertainty of ±0.2%. During the test of heating COP, the C (specific heat capacity) and (density) can be regarded as constants, and the COP is determined by three independent variables, namely t in (inlet temperature of water), t out (outlet temperature of water) and W (power consumption). Type-B uncertainty for the heating COP can be calculated with the follow Equation (4). The composite standard uncertainty of the heating COP can be obtained from the combination of Type-A uncertainty and Type-B uncertainty, and the calculation formula is shown in (5).
Since COP values are varied under different conditions, the range of standard relative errors are calculated according to the following formulas. The range of standard relative error under nine temperature conditions is 1.01-1.45%, and the maximum standard relative error of COP is 1.45%, so experimental COP value reliable enough.

Thermal Storage PCM
Phase change temperature, the type and mass of thermal storage PCM are the key issues explored in this study. Under the premise that the compressor is maximally insulated and the remaining volume outside the compressor is all occupied by the insulation cotton, the average temperature of the compressor casing is 58 • C when the compressor is continuously operated at low temperatures for multiple cycles. Hence, the phase change temperature of the thermal storage PCM should be lower than 58 • C.
In the phase change temperature range, inorganic materials are primarily crystalline water and salt materials, which have high thermal conductivity. However, during repeated phase changes, inorganic materials are prone to dehydration, which results in changes in their physical and chemical properties and reduced reliability. In this phase change temperature range, organic materials are mostly multicarbon organic matter. Such materials have good stability but low thermal conductivity and long duration of phase change. In the current study, the paraffin is used as thermal storage PCM. Liquid paraffin with a 17 • C phase change temperature is included in the PCM to reduce the phase change Energies 2020, 13, 3269 8 of 18 temperature and thereby make it suitable for the compressor casing temperature. The solid-liquid ratio of the paraffin is 6:4. The thermophysical parameters of the PCM according to Differential Scanning Calorimetry (DSC) are shown in Table 2. Calculating Formulas (6) and (7)

On-Time Mode of DE-CCTS under Different Ambient Temperatures
Figures 4a, 5a, 6a, 7a, 8a, 9a, 10a, 11a and 12a show the inlet and outlet temperatures of the main evaporator and secondary evaporator under nine ambient temperatures; here, the mass of Refrigerant A 1 and the opening time of Branch 1 are different. The mass flow of Refrigerant A 1 is small. As a result, the mass flow meter does not work so that Refrigerant A 1 cannot be measured. After several tests, the opening degree of the manual control valve is based on the inlet and outlet temperatures of the secondary evaporator. According to the trend of the inlet and outlet temperatures of the secondary evaporator, the following nine temperature conditions are divided into two parts for analysis: the first opening mode is at −25 • C, −18 • C, −12 • C, −6 • C and 0 • C, the second opening mode is at 7 • C, 15 • C, 25 • C and 35 • C. In Figures 4-12, the green vertical line represents the opening of Branch 1, which means that the secondary evaporator starts to work.
Firstly, the inlet and outlet temperatures of the two evaporators at a low temperature were introduced. At −25 • C, −18 • C, −12 • C, −6 • C and 0 • C, with the increase of temperature, the proportion of the opening time of Branch 1 to the total time is 33%, 42.9%, 54.4%, 58.3% and 60.3%, respectively. When the Branch 1 is not opened, the inlet and outlet temperatures of the secondary evaporator are significantly higher than those of the main evaporator, and the inlet and outlet temperatures of the secondary evaporator keep increasing. This stage is defined as the first stage of DE-CCTS at a low temperature. At the end of the first stage, the manual control valve is opened and the opening degree of the two valves is adjusted. The opening degree is adjusted according to the inlet and outlet temperatures of the Refrigerant A 1 . When the valve is adjusted to the optimal opening degree, the inlet temperature of the secondary evaporator drops rapidly but is still higher than that of the main evaporator, and the outlet temperature of the secondary evaporator suddenly rises to a maximum value and then decreases. This stage is defined as the second stage of DE-CCTS at low temperature.
Secondly, inlet and outlet temperatures of the two evaporators at medium-high temperatures were introduced. At 7 • C, 15 • C, 25 • C and 35 • C, with the increase of temperature, the proportion of the opening time of Branch 1 to the total time is 63.6%, 65.9%, 74.4% and 100%, respectively. When the Branch 1 is not opened, temperature trend is basically the same as that at low temperature. This stage is defined as the first stage of DE-CCTS at medium-high temperatures. At the end of the first stage, when the valve is adjusted to the optimal opening degree, the inlet temperature of the secondary evaporator drops rapidly but is still higher than that of the main evaporator, and the outlet temperature         Firstly, the inlet and outlet temperatures of the two evaporators at a low temperature were introduced. At −25 °C, −18 °C, −12 °C, −6 °C and 0 °C, with the increase of temperature, the proportion of the opening time of Branch 1 to the total time is 33%, 42.9%, 54.4%, 58.3% and 60.3%, respectively. When the Branch 1 is not opened, the inlet and outlet temperatures of the secondary evaporator are significantly higher than those of the main evaporator, and the inlet and outlet temperatures of the secondary evaporator keep increasing. This stage is defined as the first stage of DE-CCTS at a low temperature. At the end of the first stage, the manual control valve is opened and the opening degree of the two valves is adjusted. The opening degree is adjusted according to the inlet and outlet temperatures of the Refrigerant A1. When the valve is adjusted to the optimal opening degree, the inlet temperature of the secondary evaporator drops rapidly but is still higher than that of the main evaporator, and the outlet temperature of the secondary evaporator suddenly rises to a maximum value and then decreases. This stage is defined as the second stage of DE-CCTS at low temperature.
Secondly, inlet and outlet temperatures of the two evaporators at medium-high temperatures were introduced. At 7 °C, 15 °C, 25 °C and 35 °C, with the increase of temperature, the proportion of the opening time of Branch 1 to the total time is 63.6%, 65.9%, 74.4% and 100%, respectively. When the Branch 1 is not opened, temperature trend is basically the same as that at low temperature. This stage is defined as the first stage of DE-CCTS at medium-high temperatures. At the end of the first stage, when the valve is adjusted to the optimal opening degree, the inlet temperature of the secondary evaporator drops rapidly but is still higher than that of the main evaporator, and the outlet temperature of the secondary evaporator slowly rises. This stage is defined as the second stage of DE-CCTS at medium-high temperatures.

Summary of the Law of on-Time Mode
The total time required for heating 50 L water in the water tank from 20 • C to 52 • C and the opening time of Branch 1 under nine temperature conditions are shown in Table 3. It can be seen from Table 3 that as the temperature increases, the proportion of the opening time of Branch 1 to the total time increases gradually. As the ambient temperature gradually increases, the heat stored in the secondary evaporator is more and more, so the opening time of the Branch 1 is longer and longer, but at the same time, the opening degree of the manual control valve is another important factor. The opening degree of manual control valve at nine ambient temperatures is different, and the mass flow of Refrigerant A 1 is too small to be measured, so the abstract method of the inlet and outlet temperatures of the main evaporator and the secondary evaporator is used to determine the optimal opening degree of the DE-CCTS. With the increase of temperature, the opening degree of the manual control valve is smaller and smaller. Since the opening time is extended, in order to avoid Refrigerant A 1 quickly taking away the heat stored in the secondary evaporator, the mass flow of Refrigerant A 1 should not be too large, and the opening degree of valve is relatively small.
In the first stage of the nine temperature conditions, the reason why the inlet and outlet temperatures of the secondary evaporator of the DE-CCTS is relatively high is that the Branch 1 is not opened, at this time, the secondary evaporator is in the thermal storage stage, and the heat absorbed by the PCM is more and more, so the temperature of the copper tube immersed in the PCM is higher and higher. The reason why the outlet temperature trend of the secondary evaporator of DE-CCTS is different at low and medium-high temperature is that the opening degree of valve at medium-high temperature is smaller than that at low temperature, so the mass flow of Refrigerant A 1 is relatively small, and Refrigerant A 1 slowly absorbs the heat stored by PCM, so the outlet temperature of the secondary evaporator rises slowly. In the case of low temperature, the opening degree of manual control valve is larger than that of medium-high temperature, so when the Branch 1 is opened, the mass flow of refrigerant A 1 is larger, and Refrigerant A 1 quickly absorbs the heat stored by PCM, and then the outlet temperature drops slowly.  Table 4. Theoretically, as the temperature decreases, the increase rate of the COP values will increase. The proposed DE-CCTS is not effective at −18 • C and −25 • C because minimal heat is accumulated at low temperatures. By comparing the temperature values of the original system and the proposed DE-CCTS, the cause of this phenomenon is theoretically analyze using the lgP-h diagram. Figure 2 presents the lgP-h diagrams of the original system and DE-CCTS when the vapor-injected supply circuit (Branch 3) is opened at a low temperature. Relative to the original system, the proposed DE-CCTS shows increased suction temperature and pressure and reduced compression ratio. Therefore, the discharge temperature, discharge pressure, and power consumption are all reduced. The ambient environment and the heat of the compressor casing are considered as the ambient heat sources of the main and secondary evaporators, respectively. The temperature of the compressor casing is relatively high. Thus, the temperature of the mixed refrigerant from both evaporators rises, that is, the suction temperature increases. Therefore, the effects of the secondary evaporator are evident when the ambient temperature is low. Theoretically, the rate of increase of the COP rises as the ambient temperature decreases. When the proposed DE-CCTS is at −18 • C and −25 • C, the COP drops rapidly because the ambient temperature is excessively low, thereby causing minimal heat accumulation in the paraffin.

Performance Comparison between Original System and DE-CCTS
By testing the performance of the system under nine constant temperature conditions, we can analyze the performance changes of the system based on the daily temperature changes in Beijing, the capital of China in winter. In the winter of Beijing, the temperature can reach −12 • C in the morning, and then slowly rises to −6 • C. The maximum temperature can reach 0 • C at about two o'clock in the afternoon and drops to −12 • C at night. We can see that the increase rate of COP value is the largest at −12 • C and the smallest at 0 • C, which is fully consistent with the increase rate of COP values of the system under nine temperature conditions.
In addition, it should be mentioned that the secondary evaporator had the effect of reducing compressor noise. Under the premise that the indoor and outdoor temperatures were set at 25 • C, the experiments of measuring the compressor noise were measured in the quiet environment. First, the background noise was measured in the laboratory, and the average noise value of 37.2 dB was taken. Secondly, the experiments of original system were carried out, that is, there was no package of the second evaporator outside the compressor. The experiment measured a group of data every 5 minutes for 1 h. The average compressor noise value was 58.9 dB, which was the working noise of exposed compressor. Finally, under the same conditions, the experiment of DE-CCTS was carried out, taking the average noise value as 49.7 dB, which is the working noise of the compressor wrapped by the secondary evaporator. Compared with the two systems, the DE-CCTS could reduce noise by 15.6%, which showed that the secondary evaporator had a significant effect on noise reduction.

Conclusions
Experimental research using a China Gree low-temperature ASHP water heater with EVI was conducted. The conclusions are as follows: (1) The proposed DE-CCTS uses thermal storage PCM, which is filled in the secondary evaporator, to recover the waste heat of the compressor casing. The waste heat of the compressor casing is considered a high-temperature ambient environment of the secondary evaporator. It increases the suction temperature and decreases the discharge temperature, especially at low temperatures.