The Development of a Novel Direct-Expansion Ground Source Heat Pump (DE-GSHP) for Embankment Heating in Cold Regions

Abstract

This study aimed to investigate the practical applications of a ground source heat pump (GSHP) system for heating applications in embankment engineering. A special direct-expansion GSHP (DE-GSHP) was designed and manufactured, and its performance was experimentally assessed. Subsequently, the newly developed DE-GSHP structure demonstrated an excellent heating performance with respect to positive heat-supplying temperatures in cold regions. The temperature could be automatically controlled at 30, 45, 60, and 75; additionally, the heat-absorbing temperatures were maintained below 0 °C, which was extremely lower than that of the deep frost-free stratum. Further, the experimental data were used to evaluate the coefficient of performance of the DE-GSHP, which was more than 3.5. The findings indicated that the GSHP has potential applications for embankment heating in cold regions.

1. Introduction

Seasonally frozen regions constitute 53.5% of China’s total land area [1]. Ground freezing in winter damages engineering constructions, such as embankments, tunnels, pipelines, canals, and line towers. Currently, several major engineering projects, including the Harbin–Dalian high-speed railway, Harbin–Qiqihaer high-speed railway, Lanzhou–Wulumuqi high-speed railway, and Beijing–Zhangjiakou high-speed railway, pass through seasonally frozen regions where the roads and railways are prone to damage from freeze–thaw cycles. Frost heaving of the embankment is a critical obstacle to the normal operation of transportation [2,3]. The easy-to-frost characteristic indicates that anti-freezing measures should be undertaken to protect the embankment layers from frost heaving in winter to ensure the stability of the engineering constructions.

An effective method for reducing frost-heaving-induced damage is to modify the frost-heaving susceptibility of the subgrade fills within the freezing depth, by altering essential factors such as soil property, temperature, and moisture content [4,5]. According to the frost heave monitoring of the coarse-grained fillings of the Harbin–Dalian high-speed railway subgrade [6], an evident frost heave deformation was predicted in the first protective layer, with asphalt concrete and well-graded gravel, and the second protective layer, with A/B group fillings, both of which were insensitive to frost heave because of their high porosity and weak hydrophilic performance. The thermal resistance technique applies the low thermal conduction characteristics of the insulation material to reduce or prevent thermal propagation and lower the freezing rate [7,8]. Inadequate drainage also contributes to frost heave. However, ditches present beyond the tracks do not support drainage, as cleaning and maintaining the ditches only within the track area is insufficient [9]. The anti-freezing effect of the above-mentioned measures is based on the effects of soil property improvement, moisture content control, and heat insulation; therefore, these measures cannot effectively adjust the embankment temperature during extremely cold winters. Consequently, considerable unfavorable frost heaving cases still exist in cold regions, especially in deep-frozen soil regions. For example, monitoring data from various units along newly constructed high-speed railways in China revealed that frost heaving occurs under some anti-freezing embankments [10,11]. Therefore, these conventional measures are not applicable to extremely cold regions.

Recent studies have proposed several methods for increasing the efficiency of anti-freezing technologies. Among these, artificial heating technologies have received considerable attention and also been practically applied. In ref. [12], electric heat tracing was used and a method for preventing frost heave damage in tunnels was proposed. In refs. [13,14], an active heating technique using a heat pump for applications in tunnel engineering was developed, resulting in an efficient freeze-proof effect. In ref. [15], a geothermal heating technology for deicing and snow removal on bridges was created. Their technique was based on passive anti-freezing that differed from active ground heating, which can increase heat input more effectively using external heat sources. However, there has been no research on the application of artificial heating methods in embankments located in seasonally frozen regions.

In this study, heating methods using a heating boiler, electric heating, and renewable-energy-based heating were compared. Furthermore, a novel heating method based on geothermal energy was introduced, and heat pump technology was adopted to collect and transfer geothermal heat. A direct-expansion ground source heat pump (DE-GSHP) was designed and manufactured for embankment heating, and experiments were carried out to evaluate its feasibility. Furthermore, the distribution of soil temperature and heat supply temperatures nearby, as well as the coefficient of performance (COP), were investigated.

2. Key Theories and Technical Points of Ground Source Heat Pump

2.1. Heat Resources

According to the viewpoints of civil engineers, artificial heating methods are expensive for application in most civil engineering projects. In other words, heating technologies have advanced significantly. The required heat can be supplied from different sources such as electrical heaters, boilers, cogeneration plants, heat pumps, solar thermal collectors, and thermal storage equipment [16].

Heat boilers use fuels and rapidly release a large amount of heat energy. However, their smooth operation requires large amounts of heat and expensive heat exchangers and radiators. Consequently, boilers have high capital and operating expenses. Electrical heaters run on commercially purchased electricity; hence, they increase the demand for electricity. The construction and operation of boilers and electric heaters is costly during embankment heating, thereby placing undue pressure on environmental conservation and sustainable development. Hence, using renewable energy for embankment heating is the most effective approach.

Renewable thermal energy includes solar and geothermal energy. Solar energy is inexhaustible and free; however, its intensity and temperature are low in winter. The outputs of solar thermal collectors are affected by numerous factors, including the intensity of the solar irradiation and the flow rate of hot water. Conversely, the ground temperature is primarily influenced by the ambient temperature up to a few meters below the surface. Typically, the temperature of the ground rises with depth at a rate of 0.03 °C /m. However, the ground temperature may rise to 90 °C, even at a depth of 100 [17]. High underground heat levels may result from the absorption of solar radiation on the surface. However, during winter, the temperature decreases while the thermal inertia increases with depth. Hence, compared to the almost constant low underground temperature, the rapid changes in the Earth’s surface temperature are reduced.

A heat pump can be used to extract heat from the relatively warm ground and subsequently transfer it. Qian [18] conducted a survey on the feasibility of a geothermal heat pump for residential areas in China. According to the findings illustrated in Figure 1, the utilization of geothermal energy for heating purposes in cold regions is an appropriate and cost-effective approach. Recently, GSHP technology has been extensively investigated as a highly effective HVAC system, exhibiting superior characteristics compared to other technologies.

The heat capacities, efficiencies, and reliabilities of the four techniques (boilers, electric heaters, solar thermal collectors, and heat pumps) were compared. Furthermore, the drawbacks and benefits of these techniques were outlined for a clear comparison. Subsequently, a GSHP was the most beneficial for embankment heating, since it uses freely available energy, which is abundant even in winter and is inexhaustible.

2.2. Physical Configuration and Working Mechanism

A schematic illustration of the refrigerant circuit of the GSHP is presented in Figure 2a. The system comprises four primary components, including a compressor, evaporator, throttling valve, and condenser [19]. The compressor plays a crucial role in facilitating the efficient transfer of latent heat and ensuring that the compressor shell temperature is maintained at the same level as the condensing temperature. The refrigerant leaves the compressor and enters the condenser, where the surrounding medium extracts heat from the refrigerant, thereby condensing it; at the outlet of the throttle valve, a two-phase flow exists and the refrigerant flows to the evaporator, which is typically a wrapped tube. In the evaporator, the refrigerant extracts heat from the surrounding medium in the form of latent heat. The superheated vapors of the refrigerant then return to the condenser, completing a one-stage refrigeration cycle.

2.3. Coefficient of Performance (COP)

The COP is an index reflecting the heating efficiency, which is defined as the ratio of the useful heating provided to the work required [19]. In order to increase the operating efficiency, GSHP heaters transfer additional heat acquired from external sources. The pressure–enthalpy graph of the designed loop is shown in Figure 2b. The heat absorption per unit mass of the refrigerant in the evaporator can be calculated using the following expression [19]:

Q₀ = m(h₁ − h₄)

(1)

where m is the refrigerant mass (kg) and h₁ corresponds to the enthalpy of the refrigerant at Point 1 in Figure 1, (J/kg). The power consumption of the compressor is given by [19]:

W = m(h₂ − h₁)

(2)

The heat released in the condenser is calculated as [19]:

Q_heat = m(h₂ − h₃)

(3)

Ignoring losses and assuming h₃ = h₄, the theoretical COP of the refrigeration system can be expressed as follows [19]:

COP = Q_heat/W = (h₂ − h₃)/(h₂ − h₁)

(4)

2.4. Types of Heat Exchangers

Heat exchangers are of two types: DE-GSHP and a secondary loop ground-coupled source heat pump (SL-GSHP) system. An SL-GSHP comprises a heat pump and a subsystem for ground heat exchangers, a water pump, and a water–refrigerant exchanger. It exhibits a high heating capacity and is generally used for large-scale buildings. In a DE-GSHP, a buried refrigerant loop exchanges heat with the surrounding soil. Therefore, the heating capacity of DE-GSHP systems is relatively low, but a high thermal extraction and/or rejection from or to the subsurface or the aquatic environment can be achieved. Additionally, a DE-GSHP can reduce the cost of application by eliminating the secondary loop from the ground. In ref. [20], the DE-GSHP and SL-GSHP systems were compared based on their economics and technologies, and it was found that the DE-GSHP system exhibits a higher efficiency than the SL-GSHP system by 23.8%, so it is more economic than the SL-GSHP system.

Frost damage to embankments mainly occurs in cold and extremely cold regions. Groundwater, the local microclimate, and other factors disperse frost damage. The transition section of a road bridge and road culvert is also highly vulnerable to frost damage. Hence, the active artificial heating apparatus should exhibit a decentralized heat source, large heat output depth, variable heating capacity, and compact size for quick deployment. The development and operation of SL-GSHP systems are complicated for embankment heating. In addition, for common single-track embankments in seasonally deep-frozen soil regions, the maximum instantaneous and average heating load along the longitudinal direction of the embankment are 200 W/m and 20 W/m. It is worth noting that these values are significantly lower than the typical heating loads in residential buildings [21]. Therefore, the DE-GSHP system is comparatively more suitable for embankment heating.

3. Active Heating Embankment and the DE-GSHP System

3.1. Active Heating Embankment

Figure 3 shows that the proposed system mainly consists of a heat pump with a buried evaporator under the frost-free stratum, while the condenser is buried at a lower depth and acts as a heat source. After the heat energy is extracted from the Earth, it is compressed, condensed, and expanded to the upper frozen soil layer, after which, the cycle re-starts. By repeating the above process, the heat energy can be stored in the frozen soil layer and the freezing depths of the embankments can be reduced by artificial heating. Thus, frost heave can be reduced or eliminated.

In order to extract geothermal energy from deep strata, a vertical helically coiled tube is used for the evaporator, while the condenser can either be a helically coiled tube or a multi U-tube, depending on the site conditions. Furthermore, the helically coiled tube can be installed through mechanical drilling. However, the multi-U-tube can be pre-buried when building an embankment. The condenser and evaporator can be installed either coaxially or vertically, allowing for rapid layout through a single hole (Figure 3a). Alternatively, they can be installed along different axes, which provides flexibility in changing the position and inclination of the heat exchangers to meet the heating needs.

3.2. Design of the DE-GSHP System for Embankment Heating

Figure 4 shows the structure of the proposed DE-GSHP. Except for the ground heat exchangers, all the components were installed in a stainless steel container above the ground surface. The condenser and evaporator were both helically coiled; thus, this geometry promoted heat exchange. In addition, there existed an adiabatic part between the evaporator and condenser to prevent thermal interference. It should be indicated that the condenser height H_cond and evaporator height h_evap could be adjusted by the spiral pipe pitches, b_cond and b_evap, respectively.

In practice, such mesoscale DE-GSHP systems can be widely used along both sides of an embankment, where the space to install ground heat exchangers is usually limited. Figure 5 presents a view plan of the DE-GSHP, in which the evaporator was installed vertically in case of undesirable thermal interference from the evaporator to the upper frozen layer of the embankment. Further, the condenser was installed at a variable inclined angle to reduce the freezing depth and damage induced by the frost heaving in the embankments, to ensure that embankment deformation did not hinder railway operations. In remote areas lacking a power grid, a small-scale photovoltaic system can be used for the “island” operations of DE-GSHP systems. A DE-GSHP system has the advantage of operating in a flexible management mode, allowing for adjusting the operating time and heat-supplying temperature output according to the heating requirements of the embankment. Consequently, the appropriate operational management, as well as an intermittent mode, is required for the sustainable and long-term operation of DE-GSHP systems.

Generally, a hermetic piston-type compressor was selected for the DE-GSHP system due to its simple structure and robustness to leakages and vibrations. A passive capillary tube was applied in a simple small structure, and was thus highly favorable for the proposed mesoscale heating system. The entire system was controlled by a computer and operated automatically. The controller included integrated functions, including programmable control and regulation, constant temperature control, alarm control, and time programs. Furthermore, the system adopted the temperature control method of return difference. The simple structure, reliable control, and low cost served as advantages of the system. During practical applications, temperature sensors were deployed at the evaporator, condenser, and surrounding soil. When the target temperature was higher than a particular value, the DE-GSHP stopped, while it re-started automatically when the temperature was close to the preset limit value. Subsequently, by repeating the above process, the DE-GSHP achieved heating at a constant temperature, and this temperature could be adjusted along with the additional protection of the surrounding filling materials. Therefore, it could not only exhibit anti-freezing functions, but was also effective in energy saving. The need for high capital investment is a critical factor that limits the application of DE-GSHPs. More specifically, the average price of a conventional system is 1500 RMB. DE-GSHP systems are separated by 4.0 m along an embankment according to the layout of the ground heat exchangers in building-heating applications. Accordingly, the number of DE-GSHP sets required per kilometer would be 500, which corresponds to approximately 750,000 RMB, which is cheaper than constructing and replacing the gravel fillers of common thermal insulation structures and their long-term manual maintenance cost. Hence, the distribution of a DE-GSHP along the roadways and railways in extremely cold regions can be potentially developed.

In this study, a piston-type compressor with a rated power of 166 W was adopted to make a set of prototypes for model testing. The copper tubes in the condenser and evaporator were 45 m and 30 m long, respectively, with a diameter of 6 mm. To verify the results, the H_cond and h_evap values were 1.0 m and 2.0 m, respectively. All the components were carefully welded by connecting copper tubes to construct the DE-GSHP prototype (Figure 5.). The size of the stainless steel container above the ground surface was 30 cm × 45 cm× 45 cm (width × length × height). During the coiling process of manufacturing, to prevent the copper tubes from flattening, the tubes were filled with fine sand prior to bending to safeguard the smooth inward surface, which was washed with a high flow of hot water after the process.

4. Experiment Description and Results Analyses

4.1. Test Site and Operation Scheme of the DE-GSHP

A test was carried out to evaluate the performance of the DE-GSHP under real working conditions. The test site was located in a seasonally cold region of China. In the study area, winter lasts from November to February and the mean daily temperature varies between −8 °C and 4 °C. January is the coldest month, with the lowest temperature reaching −19.8 °C. The surface layer of the foundation bed at the test site is filled with 0.5 m thick non-frost-heave Group A filling material, and the lower part is a 0.1 m thick medium-coarse sand cushion layer. A test setup was created by filling a quadrangular frustum pyramid with a height of 3.2 m and a top width of 1.6 m. Figure 6 schematically illustrates the experimental facilities with the DE-GSHP at the center of the bench. The thermal conductivity, density, and specific heat of the soil were 1.74 W/(m·K), 1690 kg/m³, and 1800 J/(kg·K), respectively.

During the experiment, a thermocouple (PT100) with an accuracy of ±0.1 °C was utilized to measure the temperature. Three thermistor cables were installed to the right of the DE-GSHP, and their horizontal spacing was 0.25 m. In the present study, 12 temperature sensors were used to measure the temperatures at different points. Five sensors were installed from 0.0 m to 1.0 m, while seven sensors were installed from 1.2 m to 3.2 m. The experimental values were recorded every 10 min by a data acquisition system.

In this study, the operation modes of the DE-GSHP were selected to acquire a constant heat-supplying temperature. Specifically, the heat-supplying temperatures at the mid-position of the condenser for four groups were fixed at 30, 45, 60, and 75 °C, respectively, by using the temperature control method of return difference. The return difference was 5 °C. The experimental duration of each scheme was 24 h. The experiment was conducted in winter and the average daily temperature was below −5 °C. Observations commenced on 21 February 2018. At the end of one group of tests, the ground temperature was returned to the original level, after which, the next group of tests began.

4.2. Heat-Absorbing Temperatures and Heat-Supplying Temperatures

Figure 7 shows the temperature on the outer walls of the evaporator and condenser at a heat-supplying temperature of 30 °C. It was observed that, as the system activated, the supplying and absorbing temperatures increased and decreased, respectively. Due to the gradual heat transfer process of the helically coiled tubes and the distance between the compressor and condenser, the temperatures on the outer walls decreased gradually along the flow direction. When the temperature at the mid-position of the condenser was higher than 30 °C, the DE-GSHP stopped, while it re-started automatically when the temperature decreased to 5 °C. Through these start–stop cycling processes, the heat-supplying temperatures varied from 15 to 40 °C. Furthermore, the heat-absorbing temperature varied from −5 °C to −10 °C, indicating that the flow temperature was cooler than the surrounding soil. The DE-GSHP efficiently collected heat from the deep stable stratum through the cyclic evaporation and condensation of the refrigerant, driven by the temperature difference between the ground heat exchangers and the surrounding soil. Thus, the temperature remained constant, even when the air temperature fluctuated significantly.

Figure 8 shows the average temperature variations on the outer walls of the DE-GSHP. When the preset heat-supplying temperatures were 30, 45, 60, and 75 °C under the constant temperature operation mode, the actual average temperature ranges were 16.80–30.69 °C, 28.61–41.78 °C, 34.40–50.89 °C, and 46.29–63.45 °C, respectively. The temperatures at different positions of the GSHP were mainly influenced by the refrigerant flow direction, which could be adjusted by changing the coil direction or the temperature sensor location of the computer-based controller. A stable performance of the GSHP was achieved, despite significant fluctuations in air temperature during the test, as the ground temperature was only affected by these fluctuations up to a few meters below the ground surface.

Figure 9 depicts the daily average heat-supplying and -absorbing temperatures observed during the tests. The average heat-supplying temperature for the DE-GSHP reached above 50 °C, while the average heat-absorbing temperature was maintained at less than 0 °C. Despite the low ambient temperature, the system was able to maintain a desirable output temperature level after being activated. However, the heat-absorbing temperature unfavorably increased with an increase in the preset heat-supplying temperature. This was because the GSHP operated for a long duration in a high preset constant heat-supplying temperature mode, and the heat supplied to the soil surrounding the condenser was transferred to the distant ground. A high proportion of heat around the condenser can deteriorate the condensation effect and increase the circulating temperature of the refrigerant, which, in turn, affects liquefaction [22,23]. During practical applications, soil frost heave can be eliminated if the temperature is increased to above 0 °C. Hence, fluctuations in the heat-supplying temperature will not affect the anti-frost effect of the DE-GSHP. Conversely, the heat-supplying temperature levels should be determined according to the actual heating demand of the embankments.

4.3. Thermal Evolution of Soil

Figure 10 shows the geothermal regimes at the fixed mode of the heat-supplying temperature at 30 °C. The initial temperature isotherm of the platform was uniform, and the soil temperature was negative within a depth of 0.5 m (Figure 10a). After the activation of the DE-GSHP, the ground temperatures outside the evaporator and condenser decreased and increased, respectively. The platform formed an oval heating zone around the condenser of the DE-GSHP, and the heating effect gradually diffused from the center to the outside. Simultaneously, a trapezoidal cooling area developed around the evaporator, and the 0 °C isotherm gradually moved to the outside, resulting in an endothermic effect. This phenomenon indicated that the device could actively transfer heat from the stable stratum to the upper frost-heaving stratum. Therefore, the DE-GSHP system proved to be an effective technique for adjusting the thermal exchange between the embankment and ambient air.

4.4. COP

By neglecting the radial heat transfer conduction through the helically coiled tubes of the heat exchangers, the measured temperature reflected the refrigerant temperature at that point. Based on the average measured temperatures, the specific enthalpy of the refrigerant (R600a) at different points was obtained, as shown in

Table 1. Statistics of the specific enthalpy of the refrigerant under different operation modes of the DE-GSHP.

C1/°C	T1/°C	T2/°C	T4/°C	h1/(J/g)	h2/(J/g)	h4/(J/g)
30 °C	−4.78	31.96	14.74	549.12	598.45	234.67
45 °C	−4.33	44.13	25.72	549.72	614.61	261.36
60 °C	−3.17	55.56	31.51	551.28	629.57	275.76
75 °C	−1.63	71.80	38.59	553.34	650.30	293.68

[24,25]. The theoretical COP of the DE-GSHP was obtained using Equation (4). The COP results are illustrated in Figure 11. Thus, the daily COP value was high at the fixed mode of the heat-supplying temperature at 30 °C, and it decreased later with an increase in the heat-supplying temperature. This was because the ground temperature around the condenser increased rapidly under a high heat-supplying temperature after the operation of the system; consequently, the heat loss to the ambient surroundings increased. Thus, the high heat-supplying temperature of the DE-GSHP lowered the COP value.

Generally, the COP value of the GSHP could exceed 3.5, that is, users could receive more than 3.5 kWh of heat by consuming only 1 kWh of energy. In contrast, boilers can convert only 70–90% of the internal energy of the fuel into heat for practical applications. Therefore, the proposed DE-GSHP systems are an efficient method for embankment heating.

5. Numerical Analysis of the Roadbed Heating and Frost Heaving Prevention

5.1. Computational Model

The thermal computation module of the OpenFOAM permafrost calculation platform was used to compute the effect of the active heating in terms of preventing frost heaving in a roadbed case study, as shown in Figure 12. Equations for heat transfer and water flow were linearized, mathematically coupled, and spatially discretized using the finite volume method. This model considered the effects of water–ice phase changes on the hydraulic and thermal properties of the soil and the effect of latent heat during phase transition [26]. The model adopted unstructured tetrahedron mesh with a side length of 0.5 m. Hydrothermal processes upon freezing and thawing were simulated. As for the thermal boundary conditions, the bottom boundary was set to a constant temperature of 7 °C, both sides of the boundary were adiabatic, and the upper boundary was set to the Dirichlet thermal boundary condition, as shown in Equation (5). The lengths of the heat pump’s heating section and heat collection section were 3.5 and 7.0 m, respectively, and the heating capacity was proportionally scaled, with reference to the prototype used in Section 3.1.

$$T\left(t\right)=T_0+A_0\sin \left(\frac{2\pi }{3.1104\times {10}^7}\times t+\frac{\pi }{2}\right)$$

(5)

In the above equation, T₀ is the annual average temperature, A₀ is the annual amplitude, and t is the time in s. The values of T₀ and A₀ are 7.6 and 14.2 °C on the roadbed surface, 3.7 and 15.8 °C on the natural soil surface, and 4.9 and 15.5 °C on the sides of the slope, respectively.

5.2. Analysis of the Effect of Frost Heaving Prevention

Figure 13a shows the cross-sectional profile of the temperature field in the natural roadbed on February 1 in winter. The top and slope surfaces of the roadbed and the surrounding ground surface formed a continuous region of negative temperature that was affected by frost heaving. The freezing depth was 0.83 m at the center of the roadbed and up to 1.24 m at the road shoulders on both sides, indicating serious frost-heaving damage. Figure 13b–d show the characteristics of the roadbed temperature field distribution on 1 February of the second year for heating temperature target points of 40, 50, and 60 °C, respectively. It can be seen that, under the effect of the heat supplied by the heat pumps, two symmetrical warming regions were created beneath the roadbed surface, and only a substantially shallow sub-zero temperature region remained at the surface layer, with a freezing depth of less than 0.2 m. This indicates that the frost damage was effectively controlled. The higher the heating temperature target point was, the larger the warming region of the roadbed and the higher the resultant temperature rise were. At the same time, a visible cooling region emerged around the heat pump evaporator. Hence, it was evident that the heat pump effectively collected, transformed, and transferred the heat energy from the stable stratum surrounding the roadbed to the frost-heaving affected layer of the roadbed.

Since the evaporator in the computational model was 2.0 m away from the foot of the roadbed slope, the heat absorbed by the evaporator would not have a secondary effect on the roadbed. In practical applications, the size and placement of the heat pump heat exchangers should be reasonably designed to suit the roadbed freezing depth and extent, as well as the geothermal energy distribution in the underlying foundation.

5.3. Freezing Depth

Figure 14 shows the variation in the freezing depth at the center of the roadbed under natural conditions and artificial heating conditions. It is evident that, under natural conditions, the roadbed started to freeze on 24 November, and the freezing depth continued to increase until 17 March of the following year, when it reached a maximum value of 0.89 m, owing to the climatic environment; subsequently, the frozen layer started to thaw from both directions as the temperature rose, and the freezing range was gradually reduced and completely thawed by 4 April, staying frozen for approximately 100 days. Under artificial heating conditions, the onset of freezing occurred on the same day as that of the natural roadbed, since no insulation measures were taken on the roadbed surface. However, the freezing depth was greatly reduced: the maximum freezing depths were 0.15, 0.13, and 0.11 m with heating temperatures of 40, 50, and 60 °C, respectively. Meanwhile, the freezing depth began to gradually decrease in mid-February without undergoing a bidirectional thawing process, and the layer of the roadbed affected by frost heaving thawed completely once the average temperature rose to above 0 °C.

6. Results

This study presented an innovative active artificial heating method for addressing the issue of the frost heaving of embankments in cold regions. A DE-GSHP was designed and developed to meet space and performance requirements. An experimental test was conducted to evaluate its heating performance during cold seasons. Based on the obtained results, the main achievements of this article can be summarized as follows:

(1)

Active heating was essential to maintaining the thermal stability of the embankments in extremely cold regions. The application feasibility of the existing heating methods, including boilers, electric heaters, solar thermal collectors, and GSHPs, in embankment engineering, in terms of the location conditions along railways, initial costs, and efficacy, etc., were compared. Subsequently, the GSHP technique was appropriate for embankment heating due to its high heating capacity and flexibility with the ambient temperature.

(2)

A prototype for testing the DE-GSHP’s performance was successfully constructed. Through extensive experimentation, the performance parameters, such as the heat-supplying temperature, heat-absorbing temperature, and the COP, indicated that the DE-GSHP was highly efficient for embankment heating. The heat-supplying temperature in cold seasons was automatically controlled at 30, 45, 60, and 75 °C, which corresponded to negative heat-absorbing temperatures below 0 °C that facilitated geothermal energy utilization. In addition, the novel GSHP consumed low energy with a COP value above 3.5, while high COP values were acquired for low predetermined heat-supplying temperatures. Therefore, the novel technique was preferred during the construction and maintenance of high-speed railways and expressways in extremely cold regions due to it preventing frost heaving in embankments.

(3)

The results of numerical computations demonstrated that, under the artificial heating effect, regions of significant warming and cooling were formed in the roadbed and in the foundation beside the foot of the slope, respectively, which indicated that the heat pump could realize an efficient collection, transformation, and transfer of geothermal energy. The freezing depth of the roadbed under artificial heating conditions was significantly reduced, the melting process took place earlier, and there was no bidirectional thawing phenomenon.

(4)

There are operational restrictions on this device. Due to the solidification of lubrication oil, the compressor performance is limited by the ambient conditions during the operation of the DE-GSHP. More specifically, it cannot operate properly in ambient temperatures below −40 °C. Under this severe condition, thermal insulation should be adopted.