1. Introduction
A Chinese solar greenhouse (CSG) is a horticultural facility that uses solar energy to promote a growth environment for crops and provides high-efficiency thermal storage performance to meet the demand of vegetables’ growth in winter [
1]. By absorbing short wave radiation, storing heat and reducing convective and longwave radiative heat loss, a CSG with a closed structure maintains a suitable environment for the intensive production of various crops [
2,
3]. The main structures of CSGs include a steel frame, a north wall (north side), two sidewalls (east side and west side), a single-layer plastic cover (south roof), and a thermal insulation blanket [
4]. Besides being an important load-bearing structure in CSGs, the north wall is a heat sink while storing during the day in order to act as a heat source during the night [
5,
6]. Six shapes of greenhouses in Iran were compared from the perspective of energy demand, showing that an east–west oriented single-span greenhouse with a north wall was the optimum option. Moreover, insulation of the north wall reduced the heat losses substantially. [
7]. Therefore, the structural innovation of the north wall and its effects on heat storage and release have been a hot spot on the issue of improving the thermal performance of CSGs.
The thermal preservation and insulation performance of CSGs sustains a significant higher temperature inside the facility than outside in the winter, with a temperature difference of 15 to 40 °C [
8,
9]. However, cold damage will still be caused by low temperature at night in winter [
10]. Especially in cold northern China, additional heating is necessary for crop growth, but this increases the fossil fuel consumption and causes environmental pollution. In addition, the temperature of CSGs in sunny winter afternoons tends to rise to 40 °C, which necessitates ventilation and cooling [
11,
12]. Ventilation causes heat loss during the day, which is even more detrimental for greenhouses equipped with CO
2 enrichment devices. Therefore, collecting the surplus solar energy received by CSGs during the day and using it for indoor space low temperature heating at night is an effective way to improve the crop growth comfort and reduce fossil fuel consumption [
13].
Greenhouses need separate passive [
14] or active [
15] heating systems to maintain a suitable microenvironment during cold winters [
16]. Passive heating in CSGs relies heavily on the north wall [
17]. The structure and materials of the north wall were studied based on passive heating theory [
18]. Many studies focus on the combination of building materials for walls [
19,
20], the application of specialty materials [
21,
22] and the development of phase change materials [
23,
24,
25,
26,
27]. These studies provide a variety of options for the north wall of CSGs, and contribute to the issue of the environment balance of CSGs. However, the stability of the material combination, the range of applications, and the safety and cost of phase change materials require further analysis and evaluation. More importantly, merely changing the material and composition of the wall cannot increase the heat storage area of the north wall [
28], and expansion of the wall area involved in heat storage is beneficial to improve the efficiency of solar energy utilization. On the other hand, some studies have revealed that the depth of the north wall involved in heat storage is limited, and blindly increasing the wall thickness cannot improve the heat storage and preservation performance [
29,
30]. A way to improve the heat storage of the north wall is to mobilize the parts of the north wall that could not participate in heat storage. Therefore, it is crucial to expand the surface area involved in the heat transfer between the indoor air of CSGs and the north wall so that more volume of the north wall could be used for heat storage and release.
In order to expand the heat transfer surface area and increase the heat storage volume of the north wall in CSGs for optimizing the utilization of solar energy and improve the thermal comfort of the microenvironment in winter, we proposed a design of heat storage north wall with a hollow layer on the basis of air convection [
31]. There is a hollow structure within the north wall, and the internal space of the hollow layer exchanges heat with the cultivation space. By the air convection effects, the hollow layer collects and stores surplus solar energy in the air during the day and transfers it to the cultivation space for heating at night. Additionally, circulating fans are added to the wall structure to generate a stable airflow to enhance the heat exchange effect. Field experiments were conducted during the winter in the Tongzhou District, Beijing (39.9° N 116.6° E, elevation 8.2–27.6 m) to analyze the thermal performance under two air convection strategies.
3. Results and Discussion
3.1. Indoor Temperature Changes under Different Conditions
The most important effects of the wall structure on the cultivation environment of the CSG is the change in indoor day and night temperature.
Figure 5a shows the changes of indoor and outdoor temperatures of the NCW and NW. The average daily temperature from January 15 to 22 was −2.5 °C, with the lowest nighttime temperature low below −10 °C on January 18 to 21. The differences between indoor and outdoor air temperatures of NCW and NW at night were 17.3 °C and 16.0 °C. Compared to NW, both of the average and minimum nighttime temperatures of NCW were 1.3 °C higher, indicating that natural convection elevated the indoor temperature levels in CSG. The role of the north wall with hollow structure in improving the thermal environment of CSGs was demonstrated.
Figure 5b shows the indoor air temperature changes of the FCW, the NCW, and the outdoor air from January 27 to February 3. The average outdoor temperature during the experiment was −4.2 °C (the lowest temperature reached −16.2 °C). The indoor temperature was maintained at a relatively high temperature, and the average indoor and outdoor temperature differences of the NCW and FCW at night were 17.2 °C and 18.2 °C, respectively. Even on a cloudy day, the indoor temperature at night was kept above 5 °C with both strategies, which effectively avoided cold damage caused by low temperatures at night. The average night temperature (17: 30 to 8: 00 the next day) of the FCW was 1.0 °C higher than that of the NCW, and the average lowest temperature at night of the FCW and the NCW were 9.1 °C and 8.1 °C, respectively. The results show that the forced convection can further enhance the contribution of the north wall to raise the indoor temperature of CSG compared to a system solely based on the natural convection. The higher nighttime indoor temperature of FCW means that forced convection mobilizes more available heat from the north wall than natural convection, resulting in better efficiency of heat release.
3.2. Temperature Distribution along the Depth of the Wall
A typical sunny day (16 January 2018) was chosen as the representative day for analysis to compare the effect of the presence or absence of air convection on the temperature distribution of the wall (
Figure 6a,b). The average nighttime temperature was −4.3 °C and the average daytime temperature was 6.0 °C. The highest outdoor temperature was 11.2 °C, and the lowest outdoor temperature was −7.3 °C. From the temperature distribution along the depth direction, the NCW shows several ‘peaks’ and ‘valleys’, and there are temperature fluctuations in the depth range of 600–1000 mm. It implies that there is a temperature difference inside the walls of NCW, which facilitates the transfer of heat. In contrast, the temperature fluctuations in the NW are limited to 0–600 mm, with little change in the deeper parts of the wall throughout the day. Besides, the temperature level of the NCW was significantly higher than that of the NW at different depths.
The temperature distribution of the NCW demonstrates that the air convection catalyzed an increase in the heat exchange between the cultivation space and the hollow layer and caused the deeper wall materials to engage in heat storage and release. The temperature distribution of the NW indicates that the heat storage depth of the wall was limited. The NW only delayed the temperature drop and had a thermal insulation effect [
37].
Additionally, another day (31 January 2018) was chosen as the representative day for comparing the FCW and NCW (
Figure 6c,d). The average nighttime temperature was −5.0 °C and the average daytime temperature was 3.4 °C. The highest outdoor temperature was 6.2 °C, and the lowest outdoor temperature was −10.3 °C. Both the NCW and FCW demonstrate temperature fluctuation in deep part, indicating that both natural convection and forced convection promote heat exchange between the hollow layer and cultivation space. However, the FCW had a better temperature level in contrast to the temperature distribution of the NCW. This is attributed to steady circulating airflow of forced convection without relying on thermal pressure.
After specifying the variation throughout the day, the focus is on comparing the temperature distribution at a typical time. First, 14:00 was chosen as the typical heat storage time, and 06:00 was chosen as the typical heat release time.
At the typical heat storage time, both the NCW and NW had the highest temperature at the wall surface in the inner layer (0–400 mm) (
Figure 7a,b). Then, the NCW demonstrated a slower temperature decrease than NW, and it was 2.3 °C warmer than the NW at the depth with the lowest temperature (340 mm). However, the temperature of NCW had a rapid rebound and the temperature difference expanded to 4.3 °C at 400 mm. The circulating airflow as a low temperature heat source effectively improved the temperature distribution in the inner layer of the wall. In the outer layer, thanks to the hot air being introduced into the hollow layer, the temperature difference between NCW and NW at 1000 mm reached 4.8 °C, which is a good proof that the circulating airflow can drive the deep wall material to participate in the heat storage process. Additionally, judging from the temperature difference of up to 8.1 °C between the two hollow layers, the heat that can be brought into the interior of the wall with the circulating airflow is significant enough to have an impact on the temperature distribution of the deep material.
At the typical heat release time, the temperature level of the NCW was significantly higher than that of the NW, and the average temperature of the NCW was higher than that of the NW by 1.7 °C. Thus, natural convection enhanced the heat release capacity of the wall at night. In the hollow layer (400–1000 mm), the air temperatures of the NCW and NW were lower than those of the walls on both sides, indicating that there was a heat release from the wall surfaces to the air of hollow layer at night.
For the comparison of FCW and NCW (
Figure 7c,d), the air temperature in the hollow layer of the FCW was 4.2 °C higher than that of the NCW, indicating that the temperature of the airflow introduced via forced convection was higher. The reason for this is that the generation of natural convection requires a temperature difference, leading to temperature lag. Forced convection produces a greater heat storage capacity effect. Besides, the wall temperature level of the FCW was higher at the typical heat release time than that of the NCW. Additionally, the air temperature of the FCW hollow layer was clearly lower than the wall temperature on both sides. This was due to a more thorough heat exchange due to forced convection.
3.3. Heat Flux on Different Surfaces of the North Wall
Negative heat flux indicates that heat flows from the north wall to the indoor environment (heat release) and positive heat flux indicates that heat flows from the indoor environment to the north wall (heat storage). The results of the north wall temperature distribution (
Figure 6) indicated that there were three surfaces involved in heat storage and release in the north wall: the indoor surface, the hollow layer’s south surface, and the hollow layer’s north surface. Thus, the continuous heat flux changes of the three surfaces were monitored (
Figure 8). The heat storage and release patterns of the hollow layer and the indoor surface were similar over seven consecutive days of observation, and there were evident heat absorption and release effects. The heat flux of the indoor surface was significantly greater than that of the other surfaces. During the day, the maximum heat flux of the NCW indoor surface reached 220.7 W m
−2, and at night, the heat flux reached 72.1 W m
−2. These results demonstrate that the north wall has an essential role in heat storage and release in CSGs. The two surfaces of the hollow layer also assisted in heat storage and release. During the day, the heat storage of the north surface was slightly higher than that of the south surface. The FCW demonstrated a more stable and efficient ability for heat storage, and the maximum heat flux was 67.3 W m
−2 on the north surface. At night, the south surface heat flux was better than that of the north surface. The heat flux of the FCW and NCW reached 19.8 and 13.8 W m
−2, respectively. The outer wall was exposed to the outside environment, resulting in a large heat loss. The temperature difference between the air and the south surface was lower than the difference between the air and the north surface, therefore, the north surface absorbed more heat during the day and released less heat at night.
The ratios of heat storage and release of the three surfaces were calculated (
Figure 9). The key heat storage and release portion was the indoor surface, contributing 77–84% (NCW) and 62–66% (FCW) of the heat storage as well as 73–76% (NCW) and 66–72% (FCW) of the heat release. For the NCW, the heat storage generated via natural convection accounted for 16–23%, and the heat release accounted for 25–28%. The results show that the heat storage of the north surface is higher and that the heat release of the south surface is higher. For the FCW, the heat storage and release generated via forced convection accounted for 34–38% and 28–34%, respectively. From the overall ratio, the heat storage and release effect of forced convection is better than that of natural convection.
The heat storage and release ratios of the north and south surfaces in both of the FCW and NCW were unequal. The north surface (belonging to the outer layer of the hollow wall) absorbed more heat during the day, but released less during the night. The reason is that the outer layer of the hollow wall is closer to the external environment in terms of location, and the continuous heat dissipation into the outdoors reduces the amount of heat that can be used for heating, resulting in a lower contribution to the indoor temperature. In comparison, the south surface of the hollow layer loses less heat, allowing more heat to be available for CSG’s air heating. Therefore, the south surface of the hollow layer provided a lower ratio of heat storage and a larger ratio of heat release than the north surface. Increasing the thickness of the insulation layer on the basis of the current north wall design would improve the thermal performance of the outer layer of the hollow wall in extremely cold conditions.
Figure 10 shows the thermal images of the indoor surfaces of the FCW and the NCW at 06:00 when the outdoor temperature was −16.6 °C (24 January 2018). At 06:00, the temperature ranges of the indoor surfaces of the FCW and NCW were 7.6 °C–9.8 °C and 8.1 °C –9.7 °C, respectively. At this time, the wall remained in an exothermic heat state and the upper and lower vents of the FCW had a large temperature difference. The temperature of the upper vents was lower than that of the lower vents because of the constant operation of the fans. Additionally, the temperature of the upper and lower vents was closer in the NCW, therefore, the natural convection effect at this time was weak.
3.4. Air Convection under Different Strategies
The basic condition for the formation of natural convection is a certain difference in temperature between the hollow layer and the space of cultivation, therefore, the velocity of air circulation and the temperature difference characterize the airflow pattern. The air velocity of the upper and lower vents of the NCW and the FCW revealed different trends during observation over several days (
Figure 11).
Figure 11 shows the variations of air velocity for natural convection and forced convection during the observation period. The difference in air velocity between daytime and nighttime natural convection was large, with average air velocity ranging from 0.47 to 0.53 m·s
−1 during the day on sunny days, but falling below 0.15 m·s
−1 at night. The strongest natural convection occurred in the midday, and the air velocity was comparable to that of forced convection. On January 28 and 29, both of the maximum air velocities of the natural convection were 0.90 m·s
−1, which were higher than the air velocity of the forced convection. The results show that sufficient circulating airflow can be created for heat storage during the daytime by the air density difference.
The diurnal air velocity variation of forced convection was relatively small, remaining between the range of 0.55–0.95 m·s−1. The daytime air velocity of natural convection was weakened (0.39 m·s−1) by the reduced temperature difference between the hollow layer and the cultivated space on a cloudy day, but forced convection provided stronger airflow (0.70 m·s−1) which promoted heat storage on a cloudy day. Under unfavorable weather conditions, the stronger convective heat transfer effect of forced convection mobilized more accumulated heat in the wall which is of great significance for maintaining the thermal environment in CSGs.
Besides, the direction of airflow circulation can be optimized by fans. From the perspective of benefiting the crop growth, it is necessary to configure double-direction fans for reversing the direction of natural convection that forcing the cooled air into the CSG from the upper vents during the day and forcing the warm air into the CSG from the lower vents at night.
Figure 12 shows the temperature differences between the upper and lower vents of the NCW and the FCW from 27 to 31 January 2018. 27 January was a typical cloudy day. The temperature differences between the upper and lower vents of the NCW and the FCW were higher during the day and lower at night, and the peak in temperature differences occurred at noon whether it was a sunny or a cloudy day. This trend is closely connected to the solar radiation obtained by the CSG.
For the NCW, the temperature difference changed dramatically as the solar radiation increased during the day, and the mean temperature difference at noon exceeded 12 °C. Thus, the wall structure had the requisite conditions for air convection during operation. The temperature difference at night was lower, resulting in weak air velocity, but the hollow layer still played the role of heating the air according to the night temperature distribution and heat flux of the wall (
Figure 6 and
Figure 8).
The change pattern of the FCW was similar to that of the NCW, but there was a slightly smaller average temperature difference. During the day, the maximum difference in temperature was only 6.2 °C. The lower temperature difference demonstrates that air exchange in the hollow layer is faster because of the fans, allowing a more efficient heat transfer between the hollow layer and the cultivation space.
The temperature difference between the upper and lower vents varies with the environmental temperature. When the temperature difference is zero, it means that the circulating airflow stagnates and enters the flipping period, following which the circulating airflow of natural convection is gradually enhanced in the opposite direction as the temperature difference increases again. The air circulation directions of natural convection and forced convection at night are the same, and the role of forced airflow is to promote the exchange of heat at night. However, the temperature difference between the upper and lower vents of the NCW was positive, whereas the difference of the FCW was negative. The high air velocity generated by the fans creates a "local low temperature zone" near the upper vents, resulting in a negative temperature difference between upper and lower vents.
3.5. Heat Transfer Capacity of the Air Convection in the Hollow Layer
Natural convection heat transfer is a typical thermal pressure ventilation process, and the airflow can be calculated on the basis of the temperature of the upper and lower vents. Forced convection is constant because of the strong airflow, and the measured air velocity is used to calculate the airflow. The convective heat exchange process of the circulating airflow includes changes in sensible heat and latent heat. The change in latent heat produced by the condensation of water vapor in the hollow layer is not negligible. Thus, the heat exchange of the air convection in the hollow layer is calculated using the air state of the upper and lower vents. In this section, the airflow and heat transfer through the upper vents are calculated based on a 60 m north wall in a CSG and the results are shown in
Figure 13.
During the day, both natural and forced convections generated good air circulation, and the convection heat transfer became stronger. To calculate the airflow and heat transfer of air convection, 14:00 was chosen as the typical heat storage time. At 14:00, the airflow of natural convection and forced convection were 2.62 and 4.29 m³·s−1, respectively. At 14:00, the temperature difference between the hollow layer and the cultivation space was the greatest and the convection heat exchange was the highest, both of which resulted in the best heat exchange performance. The efficient airflow brought extensive heat exchange. The maximum heat transfers of natural convection and forced convection during the day were 31.5 and 52.2 kW, respectively. Forced convection maintained an uninterrupted and strong air circulation because of the fans, and the airflow and temperature levels of the hollow layer were better than those in natural convection.
Natural convection is weak at night because of the small temperature difference. However, the changes in heat exchange in the hollow layer of the two convection strategies at night must be confirmed, so the airflow and heat exchange throughout the night were calculated (
Figure 13). The airflow of natural convection at night was maintained at a low level of 0.5–0.6 m
3 s
−1 from 20:00 to 06:00 (the next day). Forced convection was maintained at a high flow level. After 20:00, airflow rose slightly and reached a maximum value of 4.7 m³ s
−1 at 00:00. Afterward, it showed a downward trend. The large airflow difference between the two strategies led to the difference in heat transfer capacity. The forced convection heat transfer change trend was similar to the change in the airflow trend. The heat transfer capacity of air convection rose continuously from 20:00 to 00:00 and achieved a maximum value of 10.1 kW. This result indicates that the hollow structure had a high heat exchange potential when using the forced convection strategy. The heat release slowed down after midnight and fluctuated in the range of 2.9–4.6 kW. The heat transfer of natural convection at night is significantly lower than that of forced convection, fluctuating between 0.7 and 2.4 kW. The results indicate that forced convection creates a relatively more fluctuating heat transfer and higher heat release than natural convection, which facilitates the full release of the stored heat in the hollow layer. Therefore, forced convection can compensate for natural convection, which is limited by the air density difference at night and made the heat release more sufficient.
3.6. Dehumidification Potential of Hollow Layer
The convective heat exchange process of the circulating airflow in the hollow layer contains both sensible heat and latent heat changes. The latent heat change is mainly the process of condensation of hot air after entering the hollow layer. Thus, the wall structure has a certain dehumidification potential. During the experiment, a large amount of condensed water was observed in the hollow layer, especially at noon when the convection heat exchange was strong. However, there was no water collection tank to collect the condensate in the current wall design scheme, even if it is possible to dehumidify during the day, this moisture will be brought back into the CSG at night. Therefore, the dehumidification potential is calculated to provide an evidence for improving the wall design. It should be noted that the experiment was in a CSG without any plants, which modify seriously the heat and water vapor balances. Theoretically, the dehumidification potential of an empty CSG can be estimated by the humidity change of the air flowing through the upper and lower vents and the airflow rate. When the natural convection was strongest at 14:00, the circulating air of the entire north wall removed 4.8 g of water vapor per second, meaning that the maximum instantaneous dehumidification capacity of the entire CSG (480 m2) reached 17.3 kg·h−1. The 4.8 g of water vapor per second means 11 kW of latent heat and compared with the 31.5 kw of natural convection heat transfer at noon, the ratio of latent heat to sensible heat is 0.53. Therefore, when the convection heat exchange was strong, the wall structure had a certain dehumidification potential, which had a positive effect on the regulation of the humid environment in the CSGs. The condition for exercising the dehumidification capacity of the wall is to configure condensate collection tank on the existing design to remove excess water vapor from the hollow layer and prevent condensed water to be re-evaporated from the wall during the night.
3.7. Accumulated Heat and COP
Figure 14 shows the comparison of the FCW and the NCW daily accumulated heat storage and release results for 5 days. The results indicate that both strategies produced strong heat storage and release results under sunny conditions, and the FCW was superior to the NCW in terms of heat storage and release. There was a large difference between the heat storage of the two strategies. The total heat storage of the FCW and the NCW was 3.8 and 3.0 MJ·m
−2·d
−1, respectively. The heat storage of the FCW was 25.7% higher. The FCW and the NCW had an average heat release of 2.7 and 2.4 MJ·m
−2·d
−1, respectively, and the heat release of the FCW was 10.7% more than the NCW. Natural convection clearly enhanced the heat storage and release potential of the wall, and this effect was further improved via forced convection. Hence, a mixed working strategy of natural and forced convections can be appropriately adopted to obtain economical and effective heat exchange effects.
For forced convection, the power of fans was 3.6 W·m
−2 and the energy consumption of fans was 0.3 MJ·m
−2·d
−1. Taking the hollow layer with the fans as an active solar heating system, the COP floats in the range of 1.9 to 3.4, which is a relatively efficient performance compared with some other types of heating systems, such as water source heat pumps (2.9–7.0) [
38,
39,
40], soil source heat pumps (3.8) [
41] and air source heat pumps (2.7) [
42]. However, it should be noted that the circulating airflow actually only provides no more than 20 °C of hot air at night, far less than other heat pump systems, while the benefit is the lower energy consumption.
In this study, the fans’ runtime has an impact on the COP. The results of air velocity and heat flux proved that natural convection has a strong heat storage capacity during the daytime. Therefore, a combination of natural convection and forced convection should be used to improve the COP by adjusting the fans’ runtime, which is beneficial for reducing energy consumption.