Comparison of Application Effects of Capillary Radiation Heat Pump and Electric Heating Wire in Greenhouse Seedling Cultivation

Abstract

Heat pumps with water as heat exchange medium applied in greenhouse heating have not been used for vegetable seedling cultivation. In this work, a multi-connected direct expansion capillary radiation heat pump (MDCRHP) was designed for vegetable seedling cultivation, and a closed local temperature control (CLTC) method was proposed to accurately control air and substrate temperatures in seedling beds, enhance the heating effect and reduce the equipment cost. The results show that the CLTC method can achieve good heating effect and quickly raise air/substrate temperature at a speed of 1 °C/min from 8 °C to about 20 °C within 12 min. The air and substrate temperature fields in the seedling beds were evenly distributed. The temperature differences at different horizontal positions and different heights were less than 1 °C, and the relative humidity was 100%, which is conducive to vegetable grafting seedling. It can be concluded from test results that the MDCRHP had strong adaptability for greenhouse heating and was obviously superior to the electric heating wire (EWH). The output capacity of the compressor can be controlled by adjusting the operation frequency to meet the temperature requirements at different stages of seedling cultivation. Under the conditions of similar external ambient temperature and heating effect in the seedling beds, the energy saving rate of the MDCRHP was 47–50.7% compared with the EWH. The cost of the MDCRHP is about 7.2% lower than that of the conventional heat pump. It takes 3–3.5 years to recover the equipment investment.

1. Introduction

Seedling cultivation is the key link of vegetable production, whose main affecting factors in the greenhouse include light, temperature, humidity, water, nutrients and carbon dioxide [1]. Among them, temperature and humidity are the most important control parameters, especially for grafting seedling cultivation. During the grafting healing period, vegetable seedlings are more sensitive to the changes of temperature and humidity. Thus, it is particularly necessary to control the temperature and humidity within a suitable range for seedling cultivation. At present, the relevant research mainly focuses on the heating and anti-freezing measures for the growth of vegetable in winter. The main technical measures include nylon film covering for heat preservation [2], electric wire heating [2], coal-fired boiler heating [3] and heat pump heating.

Heat pump is expected to become a promising heating technology for greenhouse seedling cultivation due to its high efficiency, energy saving, none-pollution and automatic control. Currently, it has been widely used in commercial buildings and households. However, heat pump in greenhouse heating has not been widely promoted and applied, especially in vegetable seedling cultivation due to low coefficient of performance (COP) [4] and high cost of installation [5]. In addition, humidity control during vegetable growth is very important in the greenhouse. At present, the heat pump cannot be used to adjust the humidity of greenhouse. It was not until the 1970s that heat pumps began to be used for greenhouse heating due to the oil shock [6,7]. However, so far, there is no report on the commercial application of heat pump in greenhouse vegetable seedling cultivation. Therefore, it is necessary to study the effects of heat pump heating on the temperature, humidity and energy conservation of vegetable seedling cultivation.

At present, the research of heat pump used in greenhouse heating for vegetable growth mainly focuses on the heat pump equipment development, heating effect, energy saving and thermo-economic assessment.

Heat pump equipment used for greenhouse heating mainly includes air source heat pump (ASHP), water source heat pump (WSHP), ground source heat pump (GSHP), solar-assisted air source heat pump (SAASHP) and solar-assisted ground source heat pump (SAGSHP). According to the consulted literatures, except for the test results about the heating effect and the COP [8], there is very little research on the application of a ASHP that only absorbs air energy for greenhouse heating because the design is simple and the technology is mature. The WSHP heating could basically meet the planting requirements in the multi-span greenhouse under short-term extreme weather conditions [9]. However, it has rarely been used for greenhouse heating due to the limitation of water resources. At present, most of the studies have been focused on GSHP, SAASHP and SAGSHP.

At the early stage of the heat pump due to the limit of technical conditions, the energy efficiency was generally not high, and the COP was generally lower than 3.0 [9,10,11,12,13]. In detail, the measurement results of COP were 2.13–2.84 of the SAGSHP [10], 2.64 of GSHP and 2.38 of the overall system [11], 2.84 of GSHP unit and 2.27 of the overall system [12], 2.62 of the GSHP [13], and 2.12 of WSHP [9] respectively. The energy efficiency has been improved since some corresponding design technologies and control strategies were applied in heat pumps. For example, a latent heat thermal storage tank was used in GSHP heating system, and the average values of the maximum COPs of the heat pump and the global system were reached 3.8 and 4.3, respectively [14,15]. Similarly, heat storage tanks and fan coil units were matched with the heat pump in the greenhouse, and an electric heater was used for supplemental heating; the energy conservations were determined to be maximum of daily 76.3% and monthly 25.7% [1]. An optimized flat plate solar collector was added in the ASHP, and the software TRNSYS was used to obtain the appropriate boundary conditions for the real system with the best efficiency; the COPs of the SAHP and heat pump system were 3.2 and 3.88, respectively [16]. The Ground Loop Design (GLD) software was used to calculate the minimum allowable depth of boreholes. An average field COP of 3.5 was obtained [17]. The heat recovery method was designed in the GSHP. The COPs of the GSHP and complete system were 3.93 and 3.56, respectively [18]. However, high drilling costs become an obstacle to widespread use of the GSHP and SAGSHP.

In terms of energy saving rate, monthly mean energy conservations of a surplus air heat pump were 0.63%, 10.36% and 25.72% in January, February and March, respectively [1]. Compared with the traditional air/water heat pump system, the ground source heat pump saved energy of 43% [19]. The GSHP with basket geothermal heat exchangers obtained the yearly primary energy savings from 20% to 41% for the different examined years [20].

Increasing greenhouse air temperature can enhance the accumulation and transportation of photosynthesis products of crops, and improve the yield and quality of crops [21]. From the research situation, the increase range of the air temperature in the greenhouse was generally about 2–7 °C [8,15,18,22]. However, there are relatively few studies about greenhouse temperature control, especially the distribution of temperature field in greenhouse. In fact, the temperature field in the greenhouse was more uniform, and its fluctuation with time was smaller than in the household [8,23].

Increasing crop rhizosphere temperature can improve the activity of crop roots, increase the absorption of water and nutrients by roots, enhance the synthesis of substances in roots and the activity of rhizosphere microorganisms [24,25,26]. From the test results, the heat pump has limited effect on the substrate temperature increase in a large space greenhouse, generally about 2–10 °C [21,27,28].

Humidity is another important factor affecting the growth of vegetables. High humidity is easy to cause diseases of vegetables. Under high temperature and high humidity environment, vegetable seedlings are very easy to grow in vain and form tall and slender shapes. However, for grafting seedling cultivation, the relative humidity during grafting healing period is required to be more than 95%. Therefore, humidity regulation is very critical to seedling cultivation, but there is relatively little research at present. The test results of Wang Qiang [29], Sun Weituo [30], Sun Xianpeng [31,32] and Hassanien [22] show that the evaporation of water vapor in the greenhouse will reduce the relative humidity of the air during the operation of the heat pump system, which is conducive to the prevention and control of vegetable diseases during the growth period.

The above kinds of heat pumps can stably operate with high energy efficiency and can save energy consumption. However, the boiler has both lower capital costs and lower operating costs [3].Therefore, the use of heat pumps for greenhouse heating requires a certain payback time to recover the cost. For conventional air-to-water heat pump, it was found to have a payback period of approximately six years and reduce liquefied petroleum gas consumption by 16% [33]. The payback time of a GSHP system with solar collector instead of gas heaters and fuel oil furnace would be around 14 and 12 years, respectively [34]. The GSHP system with basket geothermal heat exchangers could reach a reduction from 10% to 30% of the average operating costs for heating [21]. Compared to a conventional hot air generator using liquefied petroleum gas, when a photovoltaic-geothermal heat pump integrated system was used in greenhouse heating, pay back time for energy and for carbon emissions were 1 year and 2.25 years, respectively [35].

The application of the above heat pump technologies in the greenhouse heating has promoted the development of industrialized vegetable planting, but there are still some problems. For example, the heat pump with water as the heat exchange medium has the risk of water leakage and freezing damage of the pipeline. The heat pump is used to heat the whole greenhouse, instead of directly adjusting the temperatures of soil/substrate and vegetable rhizosphere, resulting in limited increase of temperature. If the whole space of the greenhouse is heated to a set target temperature, the heat load of the greenhouse is some number of times of the heat needed for vegetable seedling growth. In fact, most of the heat in the upper part of the greenhouse cannot be used for the seedling heating, which is wasted in vain. As a result, the equipment investment cost is several times higher than that required for local temperature control. At present, the heat pump technology is difficult to accurately control the temperature and humidity to meet the required conditions of vegetable seedlings cultivation at different stages, especially grafting seedlings. The application of heat pump in greenhouse seedling cultivation has not been reported yet. Therefore, a safe, efficient and precise temperature control system with strong adaptability is urgently needed in the greenhouse seedling cultivation in winter.

The multi-connected direct expansion capillary radiation heat pump (MDCRHP) with the refrigerant as the heat exchange medium based on the control principle of the variable refrigerant flow air conditioning system(VRF system) was proposed for room radiation heating and cooling [36]. On the basis of the previous investigation, the MDCRHP with the local temperature and humidity control method was first applied in the vegetable seedling cultivation. It filled the gap of precise control of temperature and humidity in the process of seedling cultivation.

This work comparatively studies the temperature and humidity control effect and electricity energy consumption of the MDCRHP and electric heating wire (EHW) in order to provide theoretical and technical support for the application of MDCRHP in greenhouse seedling cultivation in the winter.

2. Heating Methods and Test Methodology

2.1. Heating Methods

The multi-span glass greenhouse is located in Ningbo High Technology Agricultural Experimental Park. There are two adjacent independent small glass greenhouses of about 48 m², which are in the shape of a single roof, with a height of 2.8 m and a maximum height of about 3.5 m in the middle of the glass greenhouse. The greenhouses are equipped with agronomic sodium lamps and an openable skylight. Two tidal seedling beds are installed in every glass greenhouse. Each seedling bed is 1.3 m wide and 10 m long, and the spacing between the two seedling beds is 1 m. Two heating devices, MDCRHP and EHW, were used to test in a greenhouse on the north side of the experimental park.

The EHW with a length of 120 m of each heating wire is composed of metal wire and wrapped with silica gel for insulation protection. The heating temperature can be controlled by the thermostat, which is connected with the heating wire, and the target temperature can be set by the dial of the thermostat. The temperature control range of the thermostat is 0–50 °C, and the voltage is 220V±10%.

The system principle diagram of the MDCRHP is shown in Figure 1.

The outdoor unit is matched with multiple sets of capillary groups, which can be considered as indoor units. The electronic expansion valve (EEVi) assembled in each capillary group can be used to adjust the refrigerant flow rate of corresponding capillary group to realize independent control of indoor unit.

In Figure 1, HPS is the discharge pressure sensor, and the T_d, T_s, T_ao and T_def are the discharge temperature, suction temperature, outdoor ambient temperature and evaporator coil temperature, which are measured by the corresponding sensors, respectively. The T_W(E),1 and T_W(E),2 are the temperatures at the centers of west and east sides of two seedling beds, respectively, which are measured by the corresponding temperature sensors. The average value of two temperatures is used as the judgment condition for thermo-on and thermo-off state of the corresponding capillary radiation heating of the seedling bed. The main electronic expansion valve (MEEV) and EEVi are the EEV in the outdoor unit and in the capillary group, respectively, and the SV is the solenoid valve.

The basic parameters of the heat pump include the operation basic parameters, such as condensation temperature (T_con) and operation frequency, and performance parameters, such as heating effect, coefficient of performance (COP) and electric power. The heating capacity and COP cannot be measured in the greenhouse because they must be tested in the enthalpy difference lab. In this work, the air and substrate temperatures in the seedling beds can reflect the heating effect. The electric energy consumption (EP) can be easily measured by the electricity meter and can be used to evaluate the energy conservation of the heat pump.

The arrow indicates the refrigerant flow direction during heating operation. In detail, the variable-frequency compressor discharges a high-temperature vapor refrigerant, enters the capillary group in each seedling bed through the four-way valve, radiates heat to the air in the seedling bed and the air is heated. Then the refrigerant enters the main liquid pipe and flows into the evaporator after throttling of the MEEV. The refrigerant absorbs air energy from outside air. Finally, it returns to the compressor to complete the heating cycle. During this cycle, the EEVi and the SV of thermo-on indoor units are opened, but those of the thermo-off indoor units are closed, so as to achieve zero energy consumption. The compressor frequency is adjusted according to a certain target condensation temperature or target air/substrate temperature in the seedling bed. The condensation temperature is the saturation temperature corresponding to the discharge pressure of the heat pump unit. The MEEV opening is regulated according to a certain target super-heating degree, which is defined as the temperature different between suction temperature (T_s) and evaporator coil temperature (T_def).

The control logic of the system is described as follows. When the average air temperature or substrate temperature in the seedling bed is 2 ℃ lower than the set target temperature, the unit starts heating operation. When the average temperature in the seedling bed is 2 ℃ higher than the set target temperature, the EEVi in the corresponding capillary group is closed. If all seedling beds are shut down or in standby state, the system will stop operation. For the thermo-off indoor unit, the EEVi and SV in the seedling beds are closed, so the capillary group has no radiation heating effect and realize zero energy consumption.

Based on the above principle of MDCRHP, a set of 3HP unit with two capillary groups was designed, with a rated heating capacity of 8 kW. The system was charged with 2.2 kg R410 refrigerant. The capillaries are made of copper tubes with an outer diameter of 4.2 mm and an inner diameter of 3.0 mm. The distance between the copper tubes laid on the seedling beds was 10 cm.

For the glass greenhouse with a maximum height of 3.5 m, it is a tall space relative to the height of the seedling bed. If the whole greenhouse is heated to achieve the ideal seedling cultivation temperature, more MDCRHP systems will be needed, resulting in high investment cost. However, the actual height of the seedling bed is generally 0.5–0.8 m. If only the vegetable seedling space on the seedling bed is heated, the energy consumption, equipment cost and operation cost will be greatly reduced. So, the heating style was designed as a local temperature control of seedling bed, that is, copper pipes were laid on seedling bed to heat local seedling space. It was implemented with two modes: closed and open local temperature control mode. In the closed mode, the seedling bed frame around the seedling bed was covered with drip-free nylons for thermal insulation. The layout diagram of seedling beds in the glass greenhouse is shown in Figure 2. In the open mode, drip-free nylons are uncovered from the seedling bed frame.

A set of 12 capillary copper tubes were arranged on each seedling bed to form a capillary group. In order to realize the uniform distribution of refrigerant in the capillary and the consistency of temperature in the seedling bed, the capillary arrangement was designed as a symmetrical type; that is, the refrigerant distributors connected with the vapor tubes and liquid tubes were arranged in the middle of the seedling beds, and these copper tubes were arranged to both ends of the seedling beds. Aluminum foils with a thickness of 0.1 mm were laid on the copper tubes to strengthen the radiation of heat from the refrigerant in the copper tubes into the seedling beds, shown in Figure 3.

2.2. Test Device

The test was carried out in the glass greenhouse of Ningbo High Technology Agricultural Experimental Park. The heating effects and electric energy consumptions of the two heating methods were comparatively tested. The layout of the two heating methods is shown in Figure 4.

Aluminum foils were laid on the capillary copper pipe, and nylons were placed on the aluminum foils, which is convenient for tidal irrigation without water leakage. After the MDCRHP was tested, the heat insulation layer and electric heating wires were laid on the seedling beds. The seedling bed frames were covered with drip-free nylons to prevent condensation water from falling on the seedling beds.

The test device includes a heat pump control subsystem, a monitoring subsystem, a temperature acquisition subsystem and a humidity acquisition subsystem, as shown in Figure 5.

The heat pump control subsystem included an outdoor unit PCB board, two indoor unit PCB boards, communication lines and a wire controller. The indoor unit PCB boards were connected in series through communication lines, and then connected to the outdoor unit PCB board. Two temperature sensors in a seedling bed were connected with the corresponding indoor unit PCB board. The wire controller was used to turn on and turn off the system and display the operation parameters of the indoor units, including the EEVi opening and temperatures of air or substrate in the seedling bed. Two temperature sensors of every seedling bed were respectively arranged at the center of the west side and east side of each seedling bed and connected to the corresponding PCB board of the indoor unit.

The monitoring subsystem was composed of monitoring software, USB-RS485 converter, computer and electricity meter (DDS1666). One indoor unit PCB board communication port is connected to the computer through a USB-485 converter. The monitoring software can monitor the system operation state and set the operation control target parameters, synchronously collect and record the unit operation parameters and generate data curves [37]. The electricity meter was used to record the electric energy consumption of the unit.

The temperature acquisition subsystem was composed of 35 temperature thermocouples, an Agilent 34970A temperature acquisition instrument with 40 channels and a computer. A thermocouple was installed in the adjacent comparable greenhouse, which is not shown in Figure 5. The humidity acquisition subsystem was composed of 7 relative humidity probes and a multi-channel humidity recorder. These thermocouples and humidity probes measured the temperature and humidity every 2 min.

For the EHW heating, the heating wires on each seedling bed were connected with an electric energy meter of the same model and a thermostat. The thermostat probe was placed on the seedling bed or into the substrate to independently control the on/off state of the electric heating wire of each seedling bed.

In order to test the distribution of the temperature field in the seedling bed and determine the locations of the temperature sensors of the capillary groups and thermostat probe of the electric heating wire, 8 groups of thermocouples were evenly arranged along the horizontal direction of the seedling bed racks, and each group had two thermocouples, which were located at the height of 0.1 m and 0.5 m, respectively.

At the same time, two relative humidity sensor probes were arranged on the seedling bed frames of the west and east sides of each seedling bed, which were located at the height of 0.3m above the seedling bed. In addition, three thermocouples and three relative humidity sensor probes were arranged in the aisle between the two seedling beds of the test greenhouse, in the adjacent comparable greenhouse and outside the greenhouse to timely detect the changes of temperature and relative humidity. The layouts of these thermocouples and humidity sensors are shown in Figure 5.

The uncertainties of the test parameters are shown in

Table 1. Uncertainties of the test parameters.

Parameters	Instrument	Uncertainty	Full Scale
Temperature in seedling beds (a-h,A-H)	T-type thermocouple	±0.5 °C	−50–200 °C
Relative humidity	Humidity sensor	±0.5%	0–100%
Electricity energy consumption	Electricity meter	±0.2%	—

.

2.3. Test Methodology

The purpose of the experiment was to comparatively test the effects of the MDCRHP and EHW heating on the air temperatures, substrate temperatures and relative humidity in the seedling beds, as well as to measure the electricity energy consumptions when the closed and open local temperature control were used.

For the MDCRHP, under the condition of different condensation temperatures, variations of air temperature fields, relative humidity in seedling beds and electricity energy consumptions with time were tested. When the MDCRHP operated, the compressor operation frequency was controlled according to a certain condensation temperature. The condensation temperature can be set by the monitoring software [38]. For the EWH, under the condition of different electric heating wire spacings, variations of these parameters with time were tested.

After the air temperature distribution measurements of two heating methods in the seedling bed were completed, the seedling trays were placed on the seedling beds for seedling cultivation with tray substrate. The variations of the substrate temperature fields, relative humidity in seedling beds and electricity energy consumptions with time were tested.

During the test process, the average temperature detected by the temperature sensors was used to control the thermo-on/thermo-off state of each seedling bed.

3. Test Results and Analysis

3.1. Effects of Closed Local Temperature Control on Air Temperature Distribution

The target air temperature in every seedling bed was set as 24 °C. The air temperature distributions in the closed seedling beds under the condition of different condensation temperatures of the MDCRHP were tested for 24 h. Figure 6 shows the average air temperatures at different heights on the west and east sides of seedling beds and the temperature differences between the total average air temperatures and minimum air temperatures on the west and east sides of the seedling beds. In Figure 6, in the subscript of average temperature is “T”; “1 “and “2” represent seedling bed 1 and 2, respectively; “W” and “E” mean the west and east sides of the seedling beds, respectively; and “0.1” and “0.5” denote the height of 0.1 m and 0.5 m in the seedling beds, respectively. The average temperatures at the height of 0.1 m and 0.5 m are defined as the average values of the temperatures measured by A-H and a-h thermocouples, respectively.

The heating effects of the EHW with the spacing of 5 cm and 10 cm are shown in Figure 7.

Comparing the heating effects of the two heating methods, the following conclusions can be drawn.

(1) The MDCRHP had strong adaptability for greenhouse heating. It was obviously superior to the EHW. For the EHW, when the spacing between electric heating lines was 5 cm, the heat was excessive, and the frequent ON-OFF phenomenon occurred because the air temperature easily reached the set temperature. When the spacing was 10 cm, the heat generated was not enough to heat the cold air in the seedling bed to the set target temperature of 24 °C. For the MDCRHP, when the target condensation temperature was 38, 42 and 46 °C, the air temperatures in the seedling bed were stable at about 20, 22 and 24 °C, respectively. It can be concluded that the output capacity of the compressor and the air temperature in the seedling bed could be controlled by adjusting the frequency of the compressor, so as to meet the different temperatures required for different growth periods of seedlings, and there was no frequent thermo-on and thermo-off phenomenon when the air temperature reached the set target temperature.

(2) The closed local temperature control method can achieve uniform temperature distribution in the seedling bed. For the two heating methods, in the horizontal direction at the same height, the corresponding average temperatures of the east and west sides in the seedling beds were almost the same, and the temperature differences were less than 1.0 °C. At the height direction, the temperature differences at the height of 0.1 m and 0.5 m is also less than 1.0 °C. The temperature on the easternmost and the westernmost sides were the lowest, and the maximum temperature differences from the total average temperature of all temperatures were less than 1.0 °C.

(3) The closed local temperature control method achieved good heating effect and fast heating speed. With two heating methods of MDCRHP and EHW, the air temperatures reached about 20 °C within 12 min. For the MDCRHP, under the condensation temperature of 38, 42 and 46 °C, the air temperatures in the greenhouse reached 21.1, 22.0 and 20.4 °C from 8.5, 9.2 and 7.6 °C in 12 min, respectively; 22.2, 23.6 and 23.3 °C in 20 min, respectively, and increase amplitudes compared with the indoor temperature of greenhouse were about 11, 12 and 14 °C, respectively. In addition, the increase amplitude can be further lifted by improving the condensation temperature, i.e., the operation frequency of the compressor. For the EHW with spacing of 10 cm, the average air temperature was increased from 11.3 to 19.7 °C in 12 min, and to 21.0 °C in 20 min. The increase amplitudes compared with the indoor temperature of greenhouse were about 12.5 °C. Obviously, the heating effect of the closed local controlling temperature method is superior to that in previous work.

3.2. Effects of Closed Local Temperature Control on Substrate Temperature

The above experiments show that the air temperature fields in the seedling beds were evenly distributed. When the substrate temperatures were tested, the thermocouples at the height of 0.1 m were arranged in the substrate, and the temperature sensors of the capillary group were respectively inserted in the substrates on the east and west sides of the seedling beds. The test device is shown in Figure 8.

The experimental test was made under the conditions of different target condensation temperatures and a certain target substrate temperature, respectively. The test results are shown in Figure 9, where the subscript of “subs” means the substrate. T_{1(2),W(E),subs}, T_{1(2),W(E),air} denotes the total average substrate and air temperatures of the west and east sides of the seedling beds.

The test results in Figure 9a show that when the MDCRHP was used to heat the seedling beds, the substrate temperature differences between the east and west sides of the two seedling beds and the corresponding temperature differences between the two seedling beds are less than 1 °C, which indicates that the substrate temperature was uniform under the condition of the closed local temperature control. The average substrate temperature was about 5–5.5 °C higher than the average air temperature in the seedling bed. When there was no sunlight on rainy days, the daytime and night temperatures were basically stable at the target temperature, and the system operated stably without frequent thermo-on and thermo-off phenomenon.

In order to further verify the stability of the unit operation, the test was carried out on a sunny day. The ambient temperature in the daytime was significantly higher than that at night, and the target substrate temperature was set at 20 °C. The test results are shown in Figure 9b. The test results show that the average substrate temperature measured by the temperature sensors was 2 °C higher than the target temperature after the unit operated for about 5 h, and the unit stopped operation for about 2.5 h. When the average substrate temperature was 2 °C lower than the target temperature, the unit started to operate. When the sunlight appeared at noon, the glass greenhouse absorbed the sunlight, the indoor temperature rose significantly and the unit stopped operation. During the whole heating period, there were only two times of thermo-on and thermo-off occurrence because the set temperature was reached. It can be concluded that the system operated stably, and the temperature control was accurate.

For the EHW heating, the thermostat probes were placed in the substrates of the seedling beds to control the substrate temperature. The test results are shown in Figure 10.

Comparing Figure 9b and Figure 10, it can be seen that, under the condition of similar ambient temperature, when electric heating wire was used for seedling bed heating, a frequent ON–OFF phenomenon occurred. This phenomenon was determined by the thermostat characteristic of none-temperature return difference control. The target temperature was set as 20 °C by the thermostat. When the substrate temperature was lower than 20 °C or higher than 20 °C, the thermostat responded, and the electric heating wire was frequently turned on and turned off.

For the MDCRHP, the output capacity of the compressor could be adjusted to make the average temperature in the seedling bed approach the set target temperature. In detail, the compressor frequency was adjusted according to the target condensation temperature. As the temperature in the seedling bed increased, the saturation temperature corresponding to the discharge pressure of the system lifted. When it was higher than the target value, the operation frequency of the compressor was reduced until it reached the lowest value of 20 rps. Finally, when the average temperature in the seedling bed was more than 2 °C higher than the set target value, the EEVi was closed. Furthermore, the temperature return difference control was designed in the control program. When the temperature return difference was ±2 °C, the shutdown and startup could be delayed to avoid frequent thermo-on and thermo-off occurrences.

The above results show that the air temperatures and substrate temperatures in the seedling beds were evenly distributed, and the air temperatures and substrate temperatures of the two seedling beds were almost equal. Therefore, the average temperature detected by the temperature sensors in the substrates of the two seedling beds can be used to control the operation frequency of the compressor.

3.3. Effects of Open Local Temperature Control on Substrate Temperature and Air Temperature

In order to compare the heating effects of open and closed local temperature control methods, the drip-free nylons were uncovered to test the temperature distributions in the seedling beds. Considering that heat would be emitted to the greenhouse when the seedling beds were opened in order to test the influence of heat dissipation on the greenhouse ambient temperature, the temperature and humidity of the test greenhouse and the adjacent chamber as a comparable greenhouse were compared. A thermocouple and a relative humidity probe were arranged in the comparable greenhouse.

Under the condition of open local temperature control, the MDCRHP was used to heat the substrate and the air around the seedling bed. The test results are shown in Figure 11. The subscript “compa” denotes the comparable greenhouse.

The test results show that the substrate temperatures were stable around the target temperature of 20 °C, and the system did not stop frequently when it reached the set target temperature. Moreover, the substrate temperatures and air temperatures were evenly distributed in the two seedling beds and the temperature control accuracy was high. From the unit operation monitoring curve, the condensation temperature was stable between 33–33.5 °C, the compressor operated in the low frequency range, 20–21 rps, the substrate temperature reached the target temperature, but it was only about 5.6 °C higher than the greenhouse ambient temperature (T_ai), and the average air temperature above the seedling bed was only about 1.5 °C higher than the greenhouse ambient temperature.

When there was sunlight at noon, the glass greenhouse absorbed the sunlight and the greenhouse temperature rose significantly. The system stopped operation when the substrate temperatures in two seedling beds reached the set target temperature of 20 °C.

On the other hand, it can be seen from Figure 11 that the ambient temperature of the comparable greenhouse (T_ai,compa) was slightly, 1–2 °C, higher than that of the test greenhouse. The west wall (4 m wide) and north wall (12 m long) of the test greenhouse were in contact with the outdoor air. The heated air in the greenhouse dissipated heat through these two walls. While only the west wall (4 m wide) of the comparable greenhouse is in contact with the outdoor air. Therefore, the test greenhouse air emitted more heat than the comparable greenhouse, so the indoor ambient temperature of the test greenhouse was lower than that of the comparable greenhouse.

3.4. Effects of Closed and Open Local Temperature Control Methods on Relative Humidity

Environmental humidity has a great impact on seedling cultivation. For conventional seedling cultivation, if the environmental humidity is high, it is easy to produce diseases. However, for grafting seedlings, the relative humidity is required to be more than 95% during the grafting healing period, which is conducive to the healing of the grafting interface and improving the survival rate of scions. Obviously, the current heat pump with open temperature controlling is not suitable for grafting seedling cultivation.

The test results show that under the condition of open local temperature control, the two heating devices of the MDCRHP and EWH made the same effect on humidity in the seedling beds, but the influence of closed and open local temperature control methods on humidity was different. Taking the MDCRHP heating as an example, the test results are shown in Figure 12. These relative humidity data were taken from the corresponding test results in the experiments in Figure 9b and Figure 11, respectively.

The above test results show that for the closed local temperature control, the relative humidity in the seedling beds could be maintained at 100%. The reason is that water vapor accumulated in the closed seedling bed due to the transpiration of substrates and seedlings, resulting in a rapid increase of humidity, which is very beneficial to grafting seedling cultivation.

For the open local temperature control, the humidity of the air above the seedling bed gradually decreased with the continuous heating, and finally approached the ambient humidity of the greenhouse, which was only about 2% higher than the greenhouse relative humidity. When the sun shined at noon, the relative humidity above the seedling bed and in the greenhouse decreased sharply, falling to a minimum of about 60%. As the sunlight weakened, the relative humidity began to rise. In addition, the humidity of the comparable greenhouse was 2–3% lower than that of the test greenhouse because there was no transpiration of substrates and seedlings in the comparable greenhouse. The test results are different from those measured by Wang Qiang [30], Sun Weituo [31] and Sun Xianpeng [32,33]. The reasons are analyzed as follows. For conventional heat pump heating, when the large space of greenhouse is heated, the greenhouse temperature rose and the water vapor in the greenhouse air evaporated, resulting in the reduction of air humidity in the greenhouse. For local temperature control, the air around the substrates and seedling beds was heated. The water vapor in the substrates and seedlings was constantly evaporated, and the evaporated water vapor rose. For closed local temperature control, the water vapor produced due to substrate and seedling transpiration was confined in a narrow space and could not be distributed. The wet air measured by the humidity sensor on the upper part of the substrate reached saturation, and the measured relative humidity was 100%. For the open local temperature control, the water, substrates and seedlings on the seedling bed diffused due to transpiration, and the relative humidity was higher than that of the comparable greenhouse. When there was sunlight during the day, the temperature of the whole greenhouse rose, and the water vapor in the air evaporated rapidly and rose to the top of the greenhouse. The water vapor in the upper part of the seedling beds evaporated and diffused, resulting in decrease of the relative humidity.

4. Economic Analysis

4.1. Electricity Energy Consumption Comparison

In order to objectively and reasonably evaluate the energy conservation of the MDCRHP, electricity energy consumptions of the MDCRHP and EHW closed heating for 24 h were compared under the same outdoor ambient temperature and the same greenhouse temperature. Considering that the power of the MDCRHP is different at different target condensation temperatures, the comparison of electric energy consumption must be carried out under the condition of the same heating effect, that is, the same temperature in the seedling beds. The electric energy consumption was measured by the electricity meter and was used to evaluate energy conservation of the heat pump.

Under the almost the same heating effect, the comparison results of the two heating methods are shown in

Table 2. Comparative results of heating effect and electricity energy consumption.

Contents	Parameters	MDCRHP		EHW
Test Conditions	Tcon (°C)	38	38
	Spacing/cm			10
	Tai (°C)	8–11	11–13	10–13
	Tao (°C)	4–6	7–9	5–8
	Operation time(h)	24	24	24
Test results	Tavg,air (°C)	19–21	19–20	20–21
	Tavg,subs (°C)		23–25
	EP (kWh)	27.2	25.3	51.3
Comparison	Energy saving rate (%)	47.0	50.7

. The test results show that the energy-saving rate of the MDCRHP was 47–50.7% compared with the EHW. It can be concluded that the energy-saving effect of the MDCRHP was obvious, and the energy-saving rate is higher than that of previous work [1,20,21].

4.2. Cost and Economic Analysis

The electric heating wire was used for heating. The materials include electric heating wire and thermostat. The unit costs of the electric wire and the thermostat (Ningbo Electric Heating Equipment Factory, Ningbo, China) are ¥49/120 m and ¥144 respectively. Two seedling beds were equipped with two electric heating wires and two thermostats. The total cost of electric heating wires in seedling beds is ¥386 (excluding labor costs). The cost comparison between MDCRHP and conventional heat pump (excluding labor cost) is shown in

Table 3. Cost comparison between MDCRHP and conventional heat pump.

Unit	Material Cost of Equipment (¥)			Material Cost of Installation (¥)				Total Cost (¥)
3HP MDCRHP	Outdoor unit	Hydraulic module	Wire controller	Copper connecting-pipe	Branch pipe	Capillary group (Copper)	Auxiliary materials
	2806	0	100	800	80	2600	960 *A	7346
3 HP Conventional heat pump	Outdoor unit	Hydraulic module	Wire controller	Main pipe water pipe	Knockout trap+Actuator	Capillary pipe (PE)	Auxiliary materials
	2806	2620 *B	100	300	420	1050	600 *C	7896

*A: Auxiliary materials include: aluminum foil, thermal insulation material, karting, reflective film, etc. * B Hydraulic module includes: water pump, buffer tank, hydraulic module box, plate heat exchanger, water flow switch, auxiliary electric heating, AC contactor, control board, etc.* C Auxiliary materials include: insulation board, karting, reflective film, etc.

. The material cost of 3 HP heat pump was given by Zhongguang Electric Appliance Co. Ltd (Lishui, China).

The above comparison shows that the cost of the MDCRHP is much higher than that of EHW, but it is 7.2% lower than that of conventional heat pump due to the omission of hydraulic module. If the labor cost is included, the cost of a set of the MDCRHP is about ¥8000 higher than that of the EWH. However, the former can reduce the electricity charges by nearly 50%, saving about ¥25 per day. If the MDCRHP is used for greenhouse heating for 3 months every winter, ¥ 2250 of electricity energy consumption will be saved every year. It will take 3–3.5 years to recover the equipment cost. The service life of the MDCRHP is more than 12 years, which is economically feasible.

5. Conclusions

The heating effect and electricity energy consumption of MDCRHP and EHW in glass greenhouse heating were comparatively tested by using local temperature control method for seedling cultivation. The following conclusions can be drawn.

(1) The closed local temperature control method can obtain good heating effect and fast temperature increase. The air temperature in the seedling beds reached about 20 °C from 8 °C after 12 min of heating. The temperature fields in the seedling beds were evenly distributed, and the temperature differences in the horizontal direction and at different heights were less than 1.0 °C.

(2) The MDCRHP has strong adaptability to the greenhouse heating for seedling cultivation. It can operate stably at different ambient temperatures by adjusting the output capacity of the compressor and can achieve different air and substrate temperatures to meet the seedling cultivation demands at the different growth stages. The heating capacity of the EHW cannot be adjusted and controlled, and the on/off action was only controlled by the temperature thermostat, which is poor in adaptability.

(3) The relative humidity of the air in the seedling beds can reach 100% with the closed local temperature control, which is conducive to grafting seedling cultivation. With open local temperature control, the average substrate temperature and the upper temperature of the seedling beds were increased by about 5.6 °C and 1.5 °C, respectively. The relative humidity of the air in the seedling beds was close to that in the greenhouse, which is difficult to meet the temperature and humidity conditions required for seedling cultivation.

(4) Under the conditions of similar ambient temperature and heating effect, the electricity energy consumption of the EHW was 51.3 kw·h/d, and the energy-saving rate of the MDCRHP was 47–50.7%. In terms of cost and economy, the cost of this unit was about 7.2% lower than that of conventional heat pump, about ¥8000 higher than that of electric heating wire. Compared with the EHW, the payback time of the MDCRHP is 3–3.5 years.

(5) For greenhouse vegetable seedling cultivation, in order to improve the utilization rate per unit area of the greenhouse, the tidal seedling beds are designed to be movable, so the connecting pipe between the outdoor unit and the indoor units of the heat pump needs to be designed to be movable with the seedling beds. This is the problem to be solved in the next step.