1. Introduction
Solar energy boasts rich reserves, wide distribution, and a high degree of environmental friendliness. In China, over two-thirds of the country’s regions receive annual radiation exceeding 5020 MJ/m
2, while numerous winter heating areas possess considerable solar energy resources, providing an optimal basis for promoting the utilization of solar energy [
1,
2]. However, despite the potential for replacing conventional energy sources with solar energy to reduce building heating energy consumption and carbon emissions, solar energy utilization still encounters stability issues. Furthermore, the supply of solar energy utilization systems can conflict with the construction loads of building environments. This conflict is mainly due to the mismatch between the peak periods of solar energy supply and the trough periods of building heat loads and vice versa. Therefore, the development of high-performance and reliable heat storage equipment that can efficiently store and allocate solar energy resources is a crucial solution to address this problem.
The phase-change thermal energy storage (TES) device mainly consists of three parts: the phase-change material, the heat exchange component, and the heat exchange fluid. Therefore, the heat transfer characteristics of the phase-change material, the heat exchange effect of the heat exchange component, and the flow characteristics of the heat exchange fluid all affect the thermal performance of the TES device. Improving the thermal performance of the TES device has great research significance and application value. In recent years, many scholars have explored the enhancement of heat exchange and the improvement of TES performance. Among them, changing the structural form, adding fins, and adjusting the heat exchange fluid conditions are common methods used by scholars to enhance heat exchange.
According to whether the heat exchange fluid is in direct contact with the phase-change material, TES devices can be divided into direct contact and indirect contact types [
3]. The direct-contact-type TES device refers to the mixing of the phase-change material and the heat exchange fluid in the device, and heat exchange is carried out through direct contact. In the later stage of heat exchange, the two need to be separated. The application process of this type of device is more complex, and some phase-change materials cannot be in direct contact with the heat exchange fluid, so it is less common in practical applications. The indirect-contact-type TES device encapsulates the phase-change material through the heat exchange component, which is more convenient to operate than the direct contact type and avoids the loss of phase-change materials. Therefore, it is widely used in engineering [
4]. The phase-change materials are mainly encapsulated in rectangular containers, slender coils, and cylindrical containers. According to their encapsulation form, the indirect-contact-type TES device is mainly divided into three types of structures: flat plate; stacked; and tube and shell.
The structure of a flat-plate phase-change TES system is relatively simple, and the plate size is a crucial factor affecting its performance. Many researchers have explored this issue. Campos-Celador et al. [
5] conducted a comparative study between a pure water TES system and a plate heat storage system using paraffin as a PCM. The results showed that the volume of the phase-change heat storage system was half that of the pure water system when storing the same amount of heat, and it had better heat storage and release capacity. Prieto et al. [
6] investigated the relationship between the thickness of the phase-change plate and heat transfer efficiency, finding that reducing the thickness can enhance the heat transfer effect, increasing the heat transfer rate up to five times.
Stacked TES systems are filled with encapsulated small-volume PCMs, and the heat transfer fluid flows through the gaps in the PCM stack, providing a larger heat transfer area than other methods. Reda studied the effects of different diameters on the heat storage process of a stacked TES system with spherical encapsulation, finding that smaller diameter phase-change spheres have a relatively high heat transfer efficiency. Shamsi et al. simulated the heat storage process of a stacked bed TES system and found that increasing the temperature or flow rate of the heat transfer fluid can improve the heat storage rate and amount.
The shell-and-tube type TES system has a higher heat storage density per unit volume, has a simpler encapsulation of the PCM, and is easier to combine with various forms of heat exchangers, making it a preferred choice for over 70% of latent heat thermal energy storage systems (LHTESs) [
7]. Pipe-and-shell heat exchangers are common forms of heat exchange components in phase-change thermal energy storage systems. Coil-shaped heat exchangers can achieve longer heat exchange paths in small sizes, effectively increasing the heat exchange time between the fluid and the phase-change material and improving the heat exchange efficiency [
8]. Cheng et al. [
9] conducted numerical simulations on a phase-change thermal energy storage system with an embedded spiral coil and analyzed the effects of natural convection, the heat transfer fluid flow rate, and the initial parameters of the system on the system’s heat release characteristics. The results showed that increasing the inlet flow rate of the heat transfer fluid is beneficial for improving the heat release efficiency, and natural convection has a significant impact on the heat release performance. Duan et al. [
10] established an experimental platform for the thermal performance of a phase-change thermal energy storage system and studied the effects of the flow rate of the heat transfer fluid and the number of series-connected coil-shaped heat exchangers on the heat release performance of the system. The results showed that reducing the flow rate and increasing the number of series-connected coils can both extend the effective heat release time.
Vyshak et al. [
11] studied the total time required for PCM to melt in containers with three different geometric configurations (rectangular, cylindrical, and cylindrical shell) with the same volume and heat transfer surface area, finding that cylindrical shells require the shortest time for equal energy storage under the same PCM mass and heat transfer surface area, and this geometric effect becomes more pronounced as the PCM mass increases. Liang et al. [
12] simulated the effect of different pipe length-to-diameter ratios and PCM volume ratios on the performance of a shell-and-tube type phase-change TES system, finding that the effective energy storage ratio increases with an increase in the pipe length-to-diameter ratio and that there is an optimal PCM volume ratio. Only when the PCM volume ratio is greater than a certain value can an increase in the effective thermal conductivity of the PCM effectively improve the effective energy storage ratio.
Increasing the heat transfer area between the heat transfer fluid and PCM is an effective means of enhancing heat transfer efficiency, and attaching fins is a common way to increase the heat transfer area. Numerous studies have shown that the thermal performance of phase-change thermal storage devices can be significantly improved by adding fins. Tay et al. [
13] attached fins and pin fins to the heat transfer tube to enhance heat transfer efficiency in the intensification device and compared the improvement effects of the two through numerical simulation and experimental research. The results showed that the heat storage effect of fins was better, and the heat storage time was shortened by 25% compared with pin fins. Rathod et al. [
14] conducted a comparative study on the heat storage performance of the phase-change thermal storage device before and after the addition of fins. The results showed that the heat storage efficiency was significantly improved after adding three longitudinal fins. When the inlet temperatures of the heat transfer fluid were 80 °C and 85 °C, the heat storage time was shortened by 12.5% and 24.52%, respectively. Mat et al. [
15] studied the inner fins, outer fins, and inner and outer fin enhancement technologies in the triplex-tube heat exchanger (TTHX) as well as the influence of fin length on the enhancement technology. The results showed that the three enhancement technologies did not have a significant difference on the PCM melting rate. Compared with the three-tube heat exchanger without fins, the time for the complete melting of inner and outer fins with a length of 42 mm was reduced by 43.3%.
The number, thickness, shape, and arrangement of fins have an impact on the performance of phase-change TES systems. Al-Abidi et al. [
16] numerically simulated the melting heat transfer process of PCM with fins inside and outside of a triple-tube heat exchanger (TTHX) and found that the melting time was influenced by the number of fins, the length of fins, and the thickness of fins. The complete melting times for four, six, and eight fins were 69.5%, 56.5%, and 43.4%, respectively, compared to no fins. The melting times for fins with lengths of 10 mm, 20 mm, 30 mm, and 42 mm were 73.9%, 60.8%, 47.8%, and 43.4%, respectively, compared to no fins. The melting times for fins with thicknesses of 1 mm, 2 mm, 3 mm, and 4 mm were 43.4%, 39%, 39%, and 35%, respectively, compared to no fins. Yuan et al. [
17] simulated a cylindrical phase-change TES unit using paraffin RT82 as the PCM and established different numerical models by changing the number and width of fins and the radius of the cylindrical TES tube to compare their TES performance and obtain optimal model parameters. Castell et al. [
18] conducted experimental research on a phase-change TES system with fins and pointed out that the melting rate increased with decrease in fin spacing and increases in the temperature difference between the heat transfer fluid and the PCM. Shatikian et al. studied the effect of different fin thicknesses on the TES process of a phase-change TES system and found that increasing the fin thickness accelerated the melting of the PCM but shortened the fin spacing and reduced the amount of PCM. A reasonable configuration of fin thickness and spacing can improve the heat transfer performance of the TES system.
Prieto et al. [
19] investigated the effect of different PCMs on the heat storage and release performance of the same structure, and the results showed that paraffin with higher enthalpy had stronger heat storage capacity and longer heat release duration, while stearic acid with higher heat transfer coefficient had better heat transfer effect and could better meet the load demand during peak hours. Bansal et al. [
20] used organic acid as the phase-change heat storage material and verified the applicability of the heat transfer model combining solar collectors and phase-change heat storage under various operating conditions using numerical simulation. The results showed that the system combining PCMs and collectors was superior to the system separating PCMs and collectors. Rathod et al. [
21] used the analytic hierarchy process (AHP) method combined with the technique for order preference by similarity to ideal solution (TOPSIS) and fuzzy TOPSIS methods to select the optimal PCM for a solar hot water system. Xu et al. [
22] used the AHP and TOPSIS methods to select the optimal PCM for a solar air conditioning system and validated the selected results through Ashby’s method.
Improving the structure of the device and changing the heat flux density can effectively enhance the heat transfer performance of the device. Some scholars have changed the parameters of the heat transfer fluid and the device during the heat storage and release process to achieve optimal operating conditions. Hosseini et al. [
23] conducted experimental and simulation studies on a horizontally placed shell-and-tube latent heat storage device, which showed that raising the temperature of the heat transfer fluid can shorten the melting time of the device. When the inlet temperature was 80 °C, the melting time of the phase-change material inside the device was reduced by 37%. Rathod et al. [
24] conducted experimental studies on a vertically placed shell-and-tube device during the heat storage process and observed the melting differences of the device by changing the inlet flow rate and heat transfer fluid temperature. The results showed that reducing the mass flow rate or inlet temperature of the heat transfer fluid would increase the total melting time of the phase-change material. During the phase-change process of the material, the inlet temperature of the fluid had a more significant impact on the melting of the device than the mass flow rate. Wang et al. [
25] simulated the heat storage performance of a shell-and-tube phase-change heat storage device and found that increasing the inlet flow velocity of the heat transfer fluid would decrease the total energy storage efficiency of the device while increasing the heat storage rate; raising the inlet temperature of the heat transfer fluid could simultaneously improve the heat storage rate and energy efficiency ratio of the device. Hou et al. [
26] conducted experimental studies on a phase-change heat storage device with annular phase-change units and found that increasing the flow rate of the heat transfer fluid could enhance the heat storage and release rate of the device, but the flow rate should not be too large; otherwise, the heat storage and release effect could not be guaranteed.
The design and enhancement of heat exchange in phase-change thermal storage devices can be summarized as follows: (1) Compared to direct- and indirect-contact-type phase-change thermal storage devices, the tube-and-shell structure has advantages, such as versatility in combination with various heat exchangers, convenient packaging, and a high heat storage density, making it more favored by researchers. (2) The performance of phase-change thermal storage devices is affected by factors such as the number, thickness, spacing, shape, and arrangement of fins. (3) The heat storage effect of phase-change thermal storage devices is influenced by factors such as the inlet temperature and flow rate of the heat exchange fluid and the flow direction.
There are still several issues in the current research on phase-change thermal storage devices: (1) In terms of numerical simulation, the heat exchange tube is often simplified as a partial tube section, and the heat storage and release process is simulated and analyzed. However, the numerical model is significantly different from the actual structure. For example, for longer coiled pipes, the temperature of the heat exchange fluid changes during flow, resulting in discrepancies between the results obtained from analyzing partial pipe sections and the actual performance. (2) In terms of operating conditions, most studies focus on the effects of heat exchange fluid temperature and flow rate on the heat storage performance of the device and the effects of the heat exchange fluid flow rate and initial device temperature on the heat release performance of the device. Research on the effects of heat exchange fluid flow direction on the heat storage performance of the device and the effects of heat exchange fluid temperature on the heat release performance of the device is limited. (3) In terms of performance analysis, some studies only evaluate the performance of the device based on the melting or solidification effect of the PCM, without considering the heat storage capacity, instantaneous power, and effective heat release efficiency of the device under different operating conditions.
The objectives of this study are as follows: (1) establish a physical model and a mathematical model to use numerical simulation software to simulate the heat storage process of the phase-change thermal storage device and to summarize the dynamic melting law of the phase-change thermal storage device; (2) set up a test bench to test the thermal performance of the phase-change thermal storage device, clarify the operation mode of the system and the specific experimental scheme, and investigate the influence of different factors on the thermal storage and release performance of the device; (3) based on the experimental data, analyze the variations in the dynamic characteristics of the thermal storage device with the inlet water temperature; inlet water flow rate; and thermal storage method and the variations in the dynamic characteristics of the thermal release device with the initial temperature; inlet water temperature; inlet water flow rate; and thermal release method of the device, then use the analysis to select the optimal operating conditions and compare the numerical simulation with the experimental results to verify the accuracy of the numerical simulation.
The structure of this paper is as follows:
Section 2 introduces the physical model, control equations, and numerical simulation methods.
Section 3 describes the phase-change experiments.
Section 4 summarizes and analyzes the experimental results.
Section 5 uses numerical simulation methods to study the performance of the phase-change thermal storage device.
2. Materials and Method
2.1. Physical Model
This device is primarily designed for small-scale solar water heating systems in buildings. A small wooden building model in Inner Mongolia, China is shown in
Figure A2 in
Appendix A. Energyplus software was used to simulate and calculate the unit area heat load of the building under the coldest design condition, which was determined to be 22.56 W/m
2. The heating area of the designed device is 15 m
2, and the device provides heat for 12 h. The calculated heating capacity of the device is 4.06 kWh [
27].
The physical model diagram of the phase-change TES device studied in this paper is shown in
Figure 1. The physical model of the device mainly consists of a metal coil with fins and a square box outside the device. The internal space of the box is 820 mm long, 290 mm wide, and 800 mm high. There is one group of serpentines with fins inside the box, and each group of coils is composed of 10 straight pipes with a length of 700 mm, 9 semicircular pipes with a radius of 38 mm, and inlet and outlet pipes. The fin thickness is 1 mm, and the fin spacing is 50 mm. The heat transfer fluid enters from the top of the coil and exits from the bottom.
Due to the extremely complex phase-change process, the heat storage calculation process of the device is simplified by the following assumptions to reduce the computation time:
- (1)
The outer wall of the device box has good insulation properties, and the heat loss of the phase-change device is ignored.
- (2)
The PCM is isotropic, and the physical parameters of the solid and liquid phases are constant and do not change with temperature.
- (3)
Natural convection heat transfer between liquid PCMs is considered by introducing the Boussinesq assumption.
- (4)
The heat transfer fluid is an incompressible Newtonian fluid.
- (5)
Axial heat conduction and viscous dissipation in the heat transfer fluid are neglected.
- (6)
Since the flow and heat transfer of water in the finned tube are axisymmetric, the phase-change device can be simplified to a two-dimensional model.
2.2. Control Equation
Based on the above assumptions, a transient two-dimensional control equation is established to describe the phase change region. The continuity equation is used to describe the law of density change of the fluid, which states that the net mass inflow into a certain control volume per unit of time is equal to the increase in mass of the control volume during that time [
9]. In a Cartesian coordinate system, its differential form is as follows:
Here,
represents fluid density,
denotes time in seconds (s),
represents the
x-component of the velocity vector in meters per second (m/s), and
represents the
y-component of the velocity vector in meters per second (m/s).
The momentum conservation equation is used to describe the momentum equation of viscous incompressible fluids, which states that the rate of change of the total momentum of a control volume microelement in a unit of time is equal to the sum of all external forces acting on the microelement control volume during that time [
16]. In the Cartesian coordinate system, its differential form is expressed as follows:
Here, and represent the components of the volume force in the and directions, respectively, in units of Pa; represents the dynamic viscosity coefficient in units of N·s/m2; represents the pressure, in units of Pa.
The energy equation is the differential form of the first law of thermodynamics, which is a fundamental equation for analyzing flow systems with heat exchange. It describes the rate of change of the total energy of a control volume element over a unit of time, equal to the sum of the net rates of energy transfer in and out of the element, and the impact of surface and body forces on the element [
16]. In rectangular coordinates, its differential form is expressed as follows:
Here, denotes the temperature in Kelvin (K), represents the thermal conductivity of the fluid in units of W/(m·K), is the specific heat capacity at constant pressure in units of J/(kg·K), and stands for the source term.
The governing equations for the enthalpy-based model in this study are given as follows [
9]:
Here, represents the enthalpy of PCM in kJ/kg, represents the density of PCM in kg/m3, and represents the thermal conductivity of PCM in W/(m·K).
When the specific heat of the solid–liquid phase of the PCM is constant, the relationship between temperature and enthalpy is as follows [
15]:
Here, represents the latent heat of the phase change material, measured in kJ/kg, represents the reference temperature, measured in K, and represents the enthalpy of the phase change material at the reference temperature, measured in kJ/kg.
2.3. Numerical Simulation
A 2D model of a phase-change TES device was established using DesignModeler in ANSYS Workbench software for simulation. The model was then imported into the Meshing module for grid generation. Mesh selection is a prerequisite for numerical calculations and one of the decisive factors affecting the computational efficiency and accuracy of the results [
28]. Mesh selection should consider factors such as the type of physical problem and computational efficiency and should follow the following two principles [
29]: (1) save computation time while ensuring accuracy; (2) select several sets of meshes with different densities for comparison to obtaining mesh-independent solutions.
The results of the grid independence test can be found in
Figure A2 in
Appendix A. It can be seen that, under the same operating conditions, the simulation results of the three grid models with different numbers of grids almost overlap. To balance computational accuracy and speed, this study chose the grid model with 32118 grids, as shown in
Figure 2. The model was then imported into the FLUENT software for simulation, which is based on enthalpy–porosity and finite volume methods [
9].
2.4. Indicators
The liquid fraction, denoted as
β, is defined as the percentage of liquid substance in the entire control volume and is calculated using Equation (9) [
30]. In the liquid mixing region, the liquid fraction ranges from 0 to 1. If the phase-change material completely solidifies, the liquid fraction is 0, while if it completely melts into the liquid state, the liquid fraction is 1.
represents the solid-phase line temperature of the PCM, expressed in Kelvin (K). represents the liquid phase line temperature of the PCM, expressed in Kelvin (K).
The energy storage performance of the phase-change TES system determines the amount of usable energy during the night. The evaluation indicators for the energy storage process mainly include the inlet and outlet water temperature difference, the total amount of stored heat, and the average stored heat power.
Here, represents the stored heat capacity of the device in kJ, represents the average stored heat power in kW, is the start time of the energy storage process in seconds, is the end time of the energy storage process in seconds, ρ is the density of water in kg/m³ (assumed as 1000 kg/m³), is the specific heat capacity of water at constant pressure in kJ/(kg·K) (taken as 4.2 kJ/(kg·K)), V is the flow rate of the inlet water in m³/h, is the inlet temperature of the device in °C, and is the outlet temperature of the device in °C.
The heat release performance of the phase-change TES system determines the energy utilization efficiency on the demand side. The evaluation indicators for the heat release process mainly include the inlet and outlet water temperature difference, effective heat release time, effective heat release efficiency, and heat release power. Among them, the effective heat release efficiency is defined as the ratio of the energy content in the water discharged from the TES during the effective heat release time to the stored heat energy in the TES at the beginning of the heat release process.
Here, represents the heat released by the device, measured in kJ; denotes the effective energy release efficiency; and and represent the start and end times of the heat release process, measured in seconds.
The formula for calculating the instantaneous heat release power is as follows:
The heat release conditions considered in this paper are classified into two types: effective heat release occurs when the device releases heat for heating and its heat release power is higher than the building heat load (P ≥ 0.338 kW). According to the Standard, the recommended water temperature for shower use is 37–40 °C, while for washbasin spout use it is 30 °C. Additionally, the maximum daily water consumption quota per person per day is 4080 L. Therefore, when the device releases heat for a domestic hot water supply, an outlet water temperature higher than 37 °C ( ≥ 37 °C) is considered effective for showering; for a domestic hot water outlet, a water temperature higher than 30 °C ( ≥ 30 °C) is considered effective. This paper particularly emphasizes the heat release of the device when the outlet water temperature exceeds 37 °C.
5. Optimizing Phase-change TES System Performance
In this section, the impact of various factors on the TES process in the system was investigated using numerical simulation, providing a theoretical basis for the structural design and practical application of phase-change TES systems.
5.1. Fin Arrangements
To investigate the effect of fin arrangement on the phase-change process, simulation models with aligned fins and staggered fins were selected. As shown in
Figure A3 in
Appendix A, the liquid fraction and temperature distribution characteristics were analyzed at different times, and the time differences required for completing the phase-change heat storage were compared.
Figure 26 shows the simulation results of the two different fin arrangements. As shown in
Figure 26a, the distribution of the melting region in the device with aligned fins is different from that in the device with staggered fins. The unfused area between the coils in the device with aligned fins is circular, while in the device with staggered fins it is semicircular. However, the liquid-phase fraction of the two devices is almost the same at the same time, and both devices have a liquid-phase fraction of 0.95 at the end of the heat storage process. As shown in
Figure 26b, there are several small areas with relatively low temperatures between the coils in the device with aligned fins, while the temperature distribution inside the device with staggered fins is more uniform at the same time. After 4 h of heat storage, the device with aligned fins has the highest average temperature, and the enclosed area formed by the fins and coil bends has a better heat transfer efficiency and therefore heats up faster. When the heat storage time reaches 6 h, the high-temperature area between the coils in the device with aligned fins is rectangular, while the high-temperature area between the coils in the device with staggered fins is trapezoidal with a wider upper part and a narrower lower part. The average temperature of both finned devices is around 59.7 °C. When the heat storage time reaches 8 h, the average temperatures of the two devices are 61.55 °C and 61.59 °C, respectively, and the device with staggered fins exhibits the best melting effect. Based on the entire phase-change heat storage process, it can be concluded that the melting of the PCM in the heat storage device is mainly influenced by the distance to the pipe wall and fins, and there is no significant difference in melting efficiency between devices with aligned or staggered fins. The device with staggered fins has a slightly better heat storage effect than the device with aligned fins, but the difference is not significant.
Figure 27 shows the variation in the liquid fraction and average temperature of a phase-change TES unit with different fin arrangements over time. The gray line represents the liquid fraction curve for the straight fins, while the red line corresponds to the temperature curve of the TES unit. The time required for 50% and 90% of the PCM to melt inside the TES unit with the two fin arrangements is 10,170 s and 20,920 s, and 10,050 s and 21,190 s, respectively. During the entire charging process, the liquid fraction curves of the TES unit with the two fin arrangements almost overlap, indicating that the difference in melting performance of the aligned or staggered fin arrangement is negligible. From the graph, it can be seen that the time required for the average temperature of the TES unit to reach 50 °C with the two fin arrangements is 6020 s and 5280 s, respectively. During the entire charging process, the average temperature curves of the TES unit with the two fin arrangements are almost the same, and the average temperature of the TES unit with the staggered fins is slightly higher than that with the aligned fins most of the time.
Therefore, it can be concluded that the TES unit with the staggered fin arrangement has a better heat storage performance than the TES unit with the aligned fin arrangement, but the difference is not significant. Considering that the length of the fins is limited to less than half the tube spacing for aligned fin arrangements, while the length of the fins and tube spacing can be adjusted over a wider range for staggered fin arrangements, the fins in the TES unit should preferably be arranged in a staggered manner. However, if the conditions are restricted, the aligned fin arrangement can also be used, and the heat storage performance of the TES unit with the two fin arrangements will not differ significantly.
5.2. Fin Spacing
To investigate the effect of fin spacing on the thermal storage process, device models with fin spacings of 30 mm, 50 mm, and 70 mm were selected for simulation, as shown in
Figure A4 in
Appendix A. The liquid fraction and temperature distribution characteristics within the models were analyzed at different time points, and the differences in the time required to complete phase-change thermal storage were compared.
Figure 28 shows the simulated cloud maps of the devices with different fin spacings.
Figure 29 shows the variation curves of liquid fraction and average temperature with time for the phase-change TES device under different fin spacings. The gray line represents the liquid fraction curve, and the red line represents the device temperature curve. It can be observed that there are differences in the rate of change of the liquid fraction under different fin spacings, but the overall trend is consistent: the melting rate of the device is fast in the early stage of TES, and a clear inflection point appears in the later stage, with the liquid fraction curve relatively flat and the melting rate decreasing. This is because the solid PCM dominates the device in the early stage, and the presence of fins increases the heat exchange area, resulting in a faster melting rate. In the later stage, most of the PCM around the fins and pipelines has melted, and the heat transfer coefficient of the PCM is small, resulting in a significant slowdown in the melting rate. The time required for 50% of the PCM in the device to melt under three different fin spacings is 7980 s, 9940 s, and 11,170 s, respectively. When the fin spacing is reduced from 70 mm to 50 mm or 30 mm, the time required for 90% of the PCM to melt in the device is reduced by 11.01% and 28.56%, respectively. The time required for 90% of the PCM to melt in the device is 19,140 s, 21,190 s, and 22,900 s, respectively. When the fin spacing is reduced from 70 mm to 50 mm or 30 mm, the time required for 90% of the PCM to melt in the device is reduced by 7.47% and 16.42%, respectively. It can be seen from the figure that there are differences in the average rate of change of the device under different fin spacings, but the overall trend is consistent. There are two clear inflection points during the TES stage: the first one appears when the solid–liquid phase change is completed, at which point the device changes from latent heat storage to sensible heat storage, and the temperature rise rate increases. The second one appears when the average temperature of the device is around 60 °C in the later stage of TES. This is because most of the PCM inside the device has melted, and the temperature distribution is more uniform than in the early stage, resulting in weaker natural convection. Heat transfer mainly occurs in the form of conduction, and the heat transfer coefficient of the PCM is small, resulting in a significant slowdown in the melting rate. The time required for the average temperature of the device under three different fin spacings to reach 60 °C is 21,510 s, 23,710 s, and 25,110 s, respectively. When the fin spacing is reduced from 70 mm to 50 mm or 30 mm, the time required for the average temperature of the device to reach 60 °C is reduced by 5.58% and 14.34%, respectively.
Given the aforementioned, it can be inferred that for a given heat storage condition, a smaller fin spacing leads to a faster melting rate of the PCM (PCM), resulting in a shorter completion time of the storage process, a more uniform melting degree of PCM within the device, a higher average temperature of the device, and a greater TES capacity. However, a reduction in the fin spacing will increase the number of fins, thereby reducing the PCM capacity within the device and increasing the manufacturing cost of the fins. Therefore, in practical applications, the choice of fin spacing should be based on a comprehensive consideration of the demand for use and the economic feasibility of the device. For solar-energy-rich regions with support for longer storage times, there is no need for an excessively small fin spacing, whereas for shorter storage times a reduction in the fin spacing can be appropriate to enhance the TES performance of the device.
5.3. Fin Thickness
To investigate the effect of fin thickness on the thermal storage process, device models with fin thicknesses of 1 mm, 2 mm, and 3 mm were selected for simulation, as shown in
Figure A5 in
Appendix A. The liquid-phase ratio and temperature distribution characteristics inside the models were analyzed at different times, and the differences in the time required to complete the phase-change thermal storage were compared.
Figure 30 shows the simulation cloud maps of the devices under different fin thicknesses.
Figure 31 shows the comparison of the liquid-phase and average temperature variation trend for a phase-change TES device with different fin thicknesses. The gray line represents the liquid-phase comparison curve, while the red line represents the device temperature curve. The melting times of 50% of the phase-change materials (PCMs) inside the device correspond to three different fin thicknesses of 1 mm, 2 mm, and 3 mm, which are 9940 s, 8580 s, and 7290 s, respectively. When the fin thickness increases from 1 mm to 2 mm or 3 mm, the melting times of 50% of the PCM inside the device are reduced by 13.68% and 26.66%, respectively. The melting times of 90% of the PCM inside the device correspond to the three fin thicknesses of 1 mm, 2 mm, and 3 mm, which are 20,710 s, 18,620 s, and 16,760 s, respectively. When the fin thickness increases from 1 mm to 2 mm or 3 mm, the melting times of 90% of the PCM inside the device are reduced by 10.09% and 19.07%, respectively. In conclusion, increasing the fin thickness can significantly enhance the melting efficiency of the device.
It can be inferred from the above that, under the same heat storage conditions, an increase in fin thickness leads to a faster melting rate of PCM (PCM), a shorter time required for completing the heat storage process, a more uniform degree of PCM melting at the end of the process, a higher average temperature of the heat storage device, and a greater heat storage capacity. However, increasing the fin thickness reduces the PCM capacity of the device and raises the manufacturing cost of the finned tubes. Therefore, in practical applications, the selection of fin thickness should be based on a comprehensive consideration of the requirements for use and the economic feasibility of the device. For solar-rich regions with relatively long heat storage times, a fin thickness of no more than 2 mm is recommended, whereas for shorter heat storage times an increase in fin thickness can be considered to enhance the heat storage performance of the device.
5.4. Inlet Flow Direction and Installation Method
To investigate the effects of the device’s water inlet direction and installation method on the heat storage process, we set the installation method with the longer edge of the device aligned with the direction of gravity as the vertical installation. We selected models of the device with the fluid entering from the bottom and exiting from the top for horizontal installation and with the fluid entering from the top and exiting from the top for vertical installation and analyzed the liquid-phase fraction and temperature distribution characteristics inside the models at different time points, as shown in
Figure A6 in
Appendix A. We compared the differences in the time required to complete the phase-change heat storage.
Figure 32 presents the simulation contour maps of the device under different water inlet directions and installation methods. According to the comprehensive analysis of the entire heat storage process, the melting effect of the vertically installed device is significantly better than that of the horizontally installed one, because the vertical installation provides a larger spatial scale in the direction of gravity, which is conducive to the formation of natural convection inside the device. When both are installed horizontally, the heat exchange method of downward inlet and upward outlet is better than that of upward inlet and downward outlet during the early stage of heat storage. After the heat storage time exceeds 6 h, there is not much difference in the heat storage effect of the two water inlet methods, because the heated fluid flows in the longer coil and exchanges heat with the device while flowing, and the fluid temperature inside the coil gradually decreases. Moreover, due to the density difference between the solid and liquid phase-change materials, there will be a temperature stratification phenomenon in the device during the heat storage process under the influence of gravity, with the upper part having a higher temperature than the lower part. When the device is horizontally placed, the downward inlet and upward outlet heat exchange method will increase the heat exchange temperature difference and improve the heat exchange effect.
Based on the overall heat storage process, it can be concluded that the melting effect of the vertically installed device is significantly better than that of the horizontally installed device. This is because the vertically installed device provides a larger spatial scale in the direction of gravity, which is conducive to the formation of natural convection inside the device. For devices installed horizontally, the heat transfer method of downward inlet and upward outlet performs better than the upward inlet and downward outlet in the early stage of heat storage. However, after a heat storage time exceeding 6 h the heat storage effect of the two inflow methods is not significantly different. This is because the hot fluid flows through the longer path of the coil, exchanging heat with the device as it flows, and the fluid temperature inside the pipe gradually decreases. Additionally, due to the density difference between the solid and liquid PCMs, the device experiences temperature stratification, where the upper part of the device has a higher temperature than the lower part during heat storage under the influence of gravity. When the device is horizontally installed, the downflow–upflow heat transfer method can increase the heat transfer temperature difference, thereby improving the heat transfer effect.
Figure 33 shows the variation curves of the liquid fraction and the average temperature of the phase-change TES device with different water inlet directions and installation methods over time. The gray line represents the liquid fraction curve, while the red line represents the device temperature curve. The time required for the liquid fraction to reach 50% under three operating conditions is 9940 s, 9530 s, and 8940 s, respectively. When the operating condition is changed from horizontal up-in and down-out to horizontal down-in and up-out or vertical up-in and up-out, the time required for 50% of the PCM in the device to melt is shortened by 4.12% and 10.06%, respectively. The time required for 90% of the PCM in the device to melt is 20,710 s, 20,680 s, and 17,770 s, respectively. When the operating condition is changed from horizontal up-in and down-out to horizontal down-in and up-out or vertical up-in and up-out, the time required for 90% of the PCM in the device to melt is shortened by 0.14% and 14.20%, respectively. In conclusion, the heat storage performance of the device is significantly better in the vertical installation mode than in the horizontal installation mode, while the effect of changing the water inlet direction on the heat storage performance of the device is small in the horizontal installation mode. Additionally, the time required for the device’s average temperature to reach 60 °C under the three operating conditions is 23,720 s, 23,250 s, and 19,420 s, respectively. When the operating condition is changed from horizontal up-in and down-out to horizontal down-in and up-out or vertical up-in and up-out, the time required for the device average temperature to reach 60 °C is shortened by 1.98% and 18.13%, respectively. From the graph, it can be seen that within the first 4 h of heat storage the average temperatures of the device under the three operating conditions are similar, while after 4 h the vertical installation device has a significantly improved temperature rise rate.
6. Conclusions
Based on the summary of the design and enhanced heat transfer research of phase-change thermal energy storage (TES) devices both domestically and internationally, this study presents a multifunctional phase-change TES and heat transfer device suitable for small wooden buildings in regions with abundant solar energy and other renewable energy sources. The device has a small footprint, high heat storage density, and good heat storage and release performance. To explore its performance characteristics, a TES device performance testing platform was built, and the variation in heat storage and release characteristics was analyzed based on experimental data. Furthermore, the effects of various factors on the heat storage process of the device were simulated and studied based on the validation of the experimental verification model.
The experimental results indicate that when the device is used for heating the effective heat release efficiency is higher than 78% at an initial temperature of 63 °C, an inlet temperature of 30 °C, and a flow rate of 0.36 m3/h. The device’s heat release power exceeds the building’s heat load for more than 14 h. When the device is used for providing domestic hot water, at an initial temperature of 58 °C, an inlet temperature of 25 °C, and a flow rate of 0.05 m3/h, the device can provide more than 130 L of hot water above 37 °C, meeting the shower water needs of at least two people, with an effective heat release of 9855 KJ.
This study explains the performance changes of the system under different operating conditions, including heat storage, space heating, and domestic hot water heating, through experiments: (1) Under heat storage conditions, increasing the inlet water temperature, inlet water flow rate, and using an electric heating plate to assist in heating can all contribute to improving the heat storage performance of the system. Increasing the inlet water temperature can significantly enhance the heat storage effect, and when the heat storage temperature is low increasing the inlet water temperature has a more obvious effect on improving the heat storage rate. However, the effect of increasing the flow rate on heat storage performance is limited, and after a certain flow rate increasing the flow rate will actually cause the heat storage power to decrease. When the flow rate is increased from 0.18 m³/h to 0.27 m³/h or 0.36 m³/h, the heat storage capacity of the system only increases by 9.26% or 4.66%, respectively. Simultaneously heating hot water using an electric heating plate can significantly shorten the time required for heat storage. Compared to not using an electric heating plate, the time required for heat storage is shortened by 4.1 h when the temperature difference between the inlet and outlet water drops to 0.5 °C. (2) When the system releases heat for winter heating, reducing the inlet water temperature, increasing the inlet water flow rate, and raising the initial average temperature of the system can all contribute to improving the heat release performance of the system. The inlet water temperature has the most significant effect on the heat release effect of the system. When the inlet water temperature is decreased from 35 °C to 30 °C or 25 °C, the effective heat release of the system increases by 43.23% or 57.17%, respectively. Increasing the flow rate has a relatively weak effect on improving the heat release performance of the system. When the heat release flow rate is increased from 0.18 m³/h to 0.27 m³/h, the effective heat release of the system only increases by 8.56%. Increasing the initial average temperature of the system can significantly improve the sensible heat release power of the system. When the initial temperature of the system is increased from 53 °C to 58 °C or 63 °C, the effective heat release of the system increases by 11.96% or 28.22%, respectively. However, the improvement of heat release capacity is not significant in the later stages of heat release. (3) When the system releases heat to provide domestic hot water, increasing the make-up water temperature or reducing the inlet water flow rate can significantly increase the effective hot water output and heat release of the system. When the make-up water temperature is increased from 25 °C to 30 °C, the effective heat release of the system increases by 35.23%. When the inlet water flow rate is reduced from 0.11 m³/h to 0.08 m³/h or 0.05 m³/h, the effective heat release of the system increases by 16.46% or 65.49%, respectively.
This study investigates the effects of fin arrangement, fin spacing, fin thickness, inlet flow direction, and installation orientation on the thermal storage process through simulations. The results show that the staggered fin arrangement performs better than aligned fins, but the difference is not significant. Reducing the fin spacing and increasing the fin thickness can significantly improve the melting efficiency of the phase-change material. When the fin spacing is reduced from 70 mm to 50 mm or 30 mm, the time required for 90% of the phase-change material to melt is reduced by 7.47% and 16.42%, respectively. Increasing the fin thickness from 1 mm to 2 mm or 3 mm reduces the time required for 90% of the phase-change material to melt by 10.09% and 19.07%, respectively. The vertical installation orientation exhibits superior thermal storage characteristics compared to horizontal installation, with the difference mainly manifested in the later stage of thermal storage. When the device is horizontally installed, the thermal storage effect of the bottom-inlet and top-outlet flow direction is superior to that of the top-inlet and bottom-outlet flow direction, with the difference mainly manifested in the early stage of thermal storage. Changing the operating conditions from the horizontal installation with the top-inlet flow to horizontal installation with the bottom-inlet flow or vertical installation reduces the time required for 90% of the phase-change material to melt by 0.14% and 14.20%, respectively.
Compared with other studies [
30,
31,
32,
33,
34], the phase-change thermal storage device developed in this study has a strong specificity. It is designed to match the energy utilization characteristics of small wooden structures in areas with abundant solar energy, with a target of ultra-low emissions and serving the specific needs of solar hot water systems with a dedicated phase-change thermal storage device. The device has a small footprint and high energy storage density, making it easy to install and move. The effective energy release efficiency of phase-change thermal storage devices is generally between 50% and 80%. In this study, the effective heat release efficiency under conventional heating conditions can reach more than 75%, with a relatively superior heat release performance. When used to provide domestic hot water, the device has a lower make-up water flow rate than similar devices but can provide more domestic hot water. Under outdoor temperatures of about −27 °C, it can provide more than 130 L of domestic hot water at temperatures above 37 °C. The device can effectively coordinate the energy supply–demand contradiction in small wooden solar buildings, with a conventional heat storage efficiency of more than 1.6 kW, and the phase-change material can be melted almost entirely in 5 h. The heating power of the device is stable for more than 12 h when it is above 0.4 kW. The design of the electric heating plate in the device ensures both the heat storage effect and the comprehensive utilization of solar and other renewable energy sources.
The phase-change thermal storage device designed in this paper combines the functions of extreme and non-extreme climate conditions, and the optimization of the internal structure integrates the load performance characteristics required for building heating and domestic hot water. The hot water output external interface can switch between two functions by adjusting the valve, making the use scenarios of the device more flexible and extensive, which is conducive to the deep exploration of solar energy and other renewable energy utilization.