2.1 Site location and characterization

The site for creating a thermal storage facility is located in the Ruhr coal mining area near the city of Bottrop (Figure 1). The route of the potential closed-loop water reservoir in an underground gallery lies within the altitude interval of 450 and 471 m b.s.l. (below sea level). The on-ground elevations are predominantly between +70 and +80 m a.s.l. (above sea level), with localized elevations of up to +170 a.s.l. raised due to reclaimed rock waste heaps.

In recent years, this mine has been examined as a potential site for hydroelectric energy storage, periodically filled with water during energy surplus periods and emptied when generating electricity during energy shortage periods (Alvarado Montero et al., 2016; Niemann et al., 2018). The machine caverns with the turbine for electricity generation were planned to be installed at approximately 560 m depth.

The total length of the gallery is 16 270 m, and its cross-sectional area is approximately 38.3 m2 (Ewert, 2019). Therefore, storage facility operation with seasonal water replacement requires a free volume capacity on the ground of at least 626 100 m3. The area of a small pond located 400–500 m to the north of the Franz Haniel shaft can be enlarged due to the presence of unused flat territory by up to 9–10 ha allocated for industrial use (Niemann, 2019). This site, without any buildings, is bordered on the west by a road at the foot of the recultivated waste heap, on the east by a forest, and on the south by a parking space. Mining infrastructure for water transportation is already available nearby, which significantly simplifies the operation of a large reservoir. The pond, with an area of up to 8 ha and a depth of 1.0–1.5 m, should be sufficient as an on-ground reservoir to accumulate natural heat and cold in water that fills the underground gallery. Additionally, the pond bottom should be appropriately sealed to minimize leaks and avoid interaction with soil, thereby reducing clogging and silting, both in the pond and in the underground gallery.

The Carboniferous sediments are overlain by an overburden of approximately 350–400 m thickness composed mainly of younger sedimentary rocks of the Cretaceous (Figure 2). A slight tilt of the strata up to 5° causes the inclination of coal seams and increases their depth to the north. Within the Carboniferous deposits, coal and host rocks are intensely interbedded, forming the structures of anticlines and synclines that are disturbed by tectonic faults.

The rocks near the potential reservoir are composed of the Dorsten, Horst, Essen, and Bochum tiers. They include different lithological types, among which are limestones (45% of all occurrences) with a hydraulic conductivity kf of 2 × 10–7 m/s, as well as clayey (25%) and sandy (24%) shales with a kf of 3 × 10–9 and 1 × 10–8 m/s, respectively. In addition, sandstones and other rocks occur, with an average kf of 3 × 10–7 m/s (Ewert, 2019). The average kf of all rocks, 1.1 × 10–7 m/s, may lead to significant leakages from the reservoir during its operation, especially under excess pressure.

Above the 5th level of mining, there are drained workings, whereas the projected mine water level is expected to be between 600 and 640 m b.s.l., with flooding of the 6th level (Niemann, 2019). Hence, the water in the voids of the 5th level will not have hydraulic contact with the mineralized mine water below, and the accessible space of the 5th level can be used for temporary storage of surface water heated in summer.

2.2 Storage facility concept

The idea behind the low-temperature thermal storage facility is to utilize the available mining infrastructure and recover solar heat accumulated during the summer. The water in the closed-loop reservoir is raised and discharged back seasonally through two separate thermally insulated vertical pipes, moving through the gallery according to the approximate schedule (Table 1), and is used thermally (Figure 3).

To minimize energy spent on water circulation, it is suggested to withdraw water from one of the vertical pipes transporting water using the effect of communicating vessels. Sinking the water level by a few dozen meters due to withdrawal from one pipe and maintaining a constant water level in the other pipe creates a head difference sufficient to replace water in the gallery. In contrast to discharge and pumping out, this method enables a significant reduction in the electrical energy required for annual water circulation. Warm- and cold-water flows must be hydraulically and thermally separated at the point of tunnel connection near the entrance to the gallery. To minimize leakages from the gallery, the contact “rock–water” must be isolated, which can be performed using tunnel waterproofing tools (Vorlet & De Cesare, 2024). The detailed assessment of potential leakage is provided in Section 3.4 below.

The storage facility operation is accompanied by alternating displacements of cold water with warm water and vice versa as per the annual cycle (Table 1). Water movements lead to the appearance and expansion of the transitional zone between the cold and warm water. The density gradient between the warm and cold parts of the gallery at 0.002 does not noticeably contribute to the extension of this zone. The length of this zone is evaluated in Section 3 below.

2.3 Climate conditions

The highest annual temperatures near the site occur in July and August, while the lowest occur in December and January. The period for heating the indoor spaces lasts from November to March (Figure 4). The mean air temperature for the period November–March from 1991 to 2021 was 4.7°C and that for July–August was 19.0°C. Studies on temperature trends for this region (Effects, 2017) demonstrated an increase in the mean air temperature of 1.5°C from 1974 to 2015. For the more extended operation period until 2050, one should consider a scenario with a higher temperature, most likely by up to 1°C, which yields 5.7°C for November–March and 20.0°C for July–August on average.

For water temperature calculations, it is essential to consider the difference between the air temperature and the temperature of the shallow layer in a water body. To assess the water temperature in summer, the correlation for the ponds in the form Tw = aTa + b can be used, where Tw is the water temperature at some depth and Ta is the air temperature at an elevation of 2 m above the water level (Abis & Mara, 2006; Aldomany et al., 2018; Ali, 2013; Jacobs et al., 1998). Jacobs et al. (1998) estimated the regression factors as a = 0.84 and b = 4.9°C, while Ali (2013) estimated a = 0.9 and b = 4.2°C. Based on these assessments, the mean temperature in the upper water layer for July–August is evaluated at 21.1°C and is expected to increase to 22.0°C during the next two to three decades. During the cold season, we applied the equality of air and water temperatures.

Discharging water from the pond in the summer afternoons allows to achieve higher temperatures. Likewise, discharging water in the morning during the winter can result in a lower temperature. Thus, schedule flexibility may allow for a slight increase in the temperature difference between the two seasons, thus enhancing the overall energy efficiency. One should also consider that the projected water pond is a relatively small and shallow water body compared to lakes and rivers, and is expected to respond more quickly to variations in air temperature. The more detailed analysis should also include the other climate data affecting the pond temperature (evaporation, sunshine, wind, etc.).

Cold water stored since the winter period can be used for indoor air conditioning in the summer. This may enhance the energy efficiency of the storage facility.

2.4 Summary of previous studies and the tasks of modeling

The most significant studies related to the study site are compiled in Table 2.

3.1 Modeling assumptions

3.2 Energy balance

3.3 Hydraulic parameter assessment

Substituting Dt = 10–4 m2/s, we obtain Lt = 240 m after displacement in the heating period, which may last up to 5 months. For the period of replacing water in the summer, which lasts for 2 months, this length should not be longer. In the event of warm water displacement during the 120-day heating season, the flow velocity is expected to be below 140 m/day. Therefore, the appearance and extension of the transitional zone up to 500 m can decrease heat output only in the last 3–4 days of water replacement. This may reduce the amount of thermal energy extraction for the heating season from 115 to 155 days, by 2.6%–3.5%.

3.4 Lateral heat transfer and heat losses

The crucial issue for thermal storage facility efficiency is water temperature alteration during the annual cycle. This causes the cooling of warm water when it is contacted with rocks cooled by cold water stored for several months, and the opposite process. Heat waves in the rock–water system need to be evaluated by modeling heat transfer in rocks in contact with water in the gallery.

Equation (9) is solved in the interval $$ . The exact analytical solution of the boundary value problem formulated by Equations (9) and (10) is quite complicated, especially for time-dependent boundary conditions. Hence, Equation (9) was solved using the finite-difference method, whose accuracy is sufficient for practical purposes.

3.5 Leakage assessment

Calculations using Equation (11) at ρr = 2400 kg/m3, Hrd = 560 m, and ηl = 0.75 yield the minimum permissible depth of reservoir Hmin of approximately 310 m, whereas its projected depth is 560 m. Thus, Hrd ˃ Hmin for this case, which indicates that the overlying rock pressure balances the possible water pressure in the gallery.

Artificial sealing of the gallery requires leveling by applying an additional layer of shotcrete to remove local peaks, cavities, and achieve a smooth surface, following the best practices reviewed by Vorlet and De Cesare (2024). After that, a geotextile regularizing layer is laid on the shotcrete surface with a slight sag using fasteners. Next, the geotextile is covered with the sheets of a waterproofing geomembrane (PVC-P + PP material) joined together with an overlap using soldering. Then, the membrane is covered by an additional protective layer. At the last stage, the tightness of the waterproofing and joints is checked.

Considering that the water pressure in the gallery at the Prosper Haniel mine can be balanced by the overlying rock pressure, possible water leakages can be calculated by multiplying the hydraulic conductivity of the geomembrane, kf,gm, by the surface area of the reservoir Sg. According to Geomembrane (2010), kf,gm is estimated at 10–12 m/s. However, the results of Zhang, Ma, et al. (2024) demonstrated an increase in the pore radius in the peripheral parts of the membrane by almost 10 times after elongation of up to 125% due to stretching. Considering the parabolic law of flow in microchannels under a high water pressure Pw of 5.5 MPa, we increased kf,gm by 102 times, thus setting kf,gm = 10–10 m/s. For the inner surface area of the gallery of approximately 3.6 × 105 m2, the leakage is expected to be 3.11 m3 daily or 1135.2 m3 per annum, which is below 0.2% of the gallery volume of 6.26 × 105 m3. The variation of rock hydraulic conductivity becomes noncritical for leakage because the increased geomembrane conductivity is expected to be 3.3 times lower than the assumed lowest rock conductivity. The estimated water leaks from the properly sealed gallery in the Prosper Haniel mine are expected not to affect the storage efficiency. Therefore, for the calculations below, we did not evaluate the potential impact of leakages on the performance efficiency. However, this risk must be assessed in more detail for the feasibility studies.

4.1 Input parameter identification

To identify the input parameters, we first calculated the temperature changes in the gallery (Figure 5) based on the schedule (Table 1) and climate data (Figure 4), assuming that the annual cycle begins on July 1st (t = 0) and ends on June 30th, and the point x = 0 meets the warm water entrance to the gallery, and x = Lg is the same for cold water. The transitional zone length, as assessed by Equation (8), may reach 150–240 m by the end of the 5-month heating period, which is 1.5% of the total gallery length.

To assess the impact of rock–water heat exchange, we calculated the temperature in the rock layer around the gallery by solving the boundary value problem formulated by Equations (9) and (10). The water temperature, Tw, was identified from simulations demonstrated in Figure 5. Considering data from Geologischer Dienst (2018), we evaluated the background rock temperature at a depth of 560 m, which was found to be 27.5°C. It is higher than the temperature of warm water pumped in the gallery during July–August.

To estimate water temperature changes along the gallery, we calculated heat transfer at two positions: the first, close to the warm water entrance (x = 100 m), and the second near the cold water entrance (x = Lg – 100 m). The temperature waves are expected to occur primarily within the approximately 1 m thick rock layer (Figures 6 and 7). To evaluate the influence of varying rock thermal conductivity λr along the gallery, we simulated temperature changes in the surrounding rocks for the minimum and maximum values of λr. Considering the occurrence of limestones, clayey and sandy shales, and sandstones along the tracing (Section 2.1) and their properties (Robertson, 1988), we assessed λr,min = 1.5 W/(m·K) and λr,max = 2.7 W/(m·K). The difference between the rock temperatures calculated for the same points at the two extreme values of λr may reach 3°C within a 1 m thick rock layer around the gallery. However, beyond the rock layer of more than 1.5 m, these differences became negligible. The evaluated decrease in warm water temperature and increase in cold water temperature during the storage periods are estimated to be below 1°C along the gallery. Regarding the uncertainties in temperature fluctuations, the transitional zone between the zones of cold and warm water, and climate change, we used the range 18 to 22°C for warm water and 5 to 7°C for cold water in the energy efficiency calculation.

The density of geothermal heat flux, estimated at around 0.05 W/m2, based on local measurements (Geologischer Dienst, 2018), yields a total heat flux of up to 10 kW across a planar area of the gallery, equivalent to approximately 200 000 m2. The amount of heat gained at this flux annually is below 9 × 10–6 of the amount of heat stored in the gallery, so that the geothermal impact can be neglected.

The input parameters for the calculations include storage facility characteristics and parameters related to water thermal use, as listed in Table 3.

4.2 Calculations and discussion

The hydraulic parameters of the storage performance are summarized in Table 4.

A higher water temperature and a shorter heating period increase the thermal output (Figure 8). The expected maximum thermal output of the system, approximately 4.8 MW, is achieved at a water temperature of 24°C and a heating period of 115 days; the minimum, about 2 MW, is achieved at a water temperature of 16°C and a heating period of 155 days. A change in Tw by 1°C within the expected summer temperature range of 18–22°C causes a change in the thermal output of 6.3%–8.4%, with the more minor relative changes at higher values of Tw. The amount of thermal energy accumulated in the storage depends only on the temperature of the water accumulated and varies from 26.3 GJ at Tw = 16°C to 47.33 GJ at Tw = 24°C.

The total thermal energy stored in the gallery and the energy recovered, considering the electricity spent on operation at Tw = 22°C, are 60% higher than those at Tw = 18°C (Figure 9). 38%–45% of the thermal energy stored in July–August can be recovered during the heating season, which quantifies the thermal efficiency of the storage facility (line 3 in Figure 9). One should account for heat losses of up to 1°C caused by water–rock thermal interaction (Figures 6 and 7). Based on the COP formula, a relative change in heat pump efficiency h causes an almost proportional relative change in COP, which in turn leads to a proportional change in the energy efficiency of the UTES.

Equation 12 facilitates a preliminary assessment of the storage facility efficiency with the deviation of Tw by a few degrees from the range of 18–22°C without additional calculations. Utilizing a renewable heat source, such as solar panels, to produce extra heat at the inlet for the warm water filling the gallery in July and August, may increase energy efficiency. For example, the increase in temperature from 22 to 23°C increases storage efficiency by 9.3%.

The amount of thermal energy that can be extracted from the storage (Figure 10) depends on the water temperature, the duration of the heating period, and the temperature of the heat transfer fluid carrier at the heat consumers, which ranges from 30 to 60°C. The amount of recoverable heat increases with increasing water temperature and slightly decreases with the lengthening of the heating period. A higher temperature of the fluid at heat consumers leads to a decrease in the extracted thermal energy due to a lower coefficient of performance of heat pumps. The maximum thermal energy is extracted at the highest water temperature, the shortest heating period, and the lowest temperature of the heat transfer fluid at the consumer Tc. A change in Tw by 1°C within the expected summer temperature range of 18–22°C causes a change in the thermal output of 9.2%–12.4% for a heating period duration of 115–155 days, with the more minor relative changes at higher values of Tw (Figure 10a). A change in Tw by 1°C within the expected summer temperature range of 18 to 22°C causes a change in the thermal output of 8.5%–14.5% for the range of temperatures Tc from 30 to 60°C, with the more minor relative changes at higher values of Tw (Figure 10b). Note that the range of Tc matches the real values: a fluid with a temperature of 30°C can be used in underfloor heating systems and a fluid with a temperature of 60°C in radiators at a very low outdoor temperature.

The energy efficiency of the storage facility is defined as the ratio of the recovered thermal energy to that accumulated. Energy efficiency increases with increasing water temperature and decreases with the lengthening of the heating period and the temperature of the heat transfer fluid at the consumer (Figure 11). Thus, a maximum efficiency of more than 85% is expected at the highest water temperature, the shortest heating period, and the higher temperature Tc at the consumer of 60°C. A change in Tw by 1°C within the expected summer temperature range of 18–22°C causes a change in the energy efficiency of 1.8%–2.1% for a heating period duration of 115–155 days and the range of Tc from 30 to 60°C, with more minor relative changes at higher values of Tw. A change in Tc by 1°C within the expected summer temperature range of 18–22°C and the range of Tc from 30 to 60°C results in a change in the energy efficiency of 1.3%–1.6%, with more minor relative changes at higher values of Tc.

The closest potential consumers are located nearby in the city of Bottrop, within several hundred meters of the mine shaft, making heat supply from the storage economically viable for local communities. In case potential consumers of cold in the summer are available nearby, storage efficiency can be increased. The energy efficiency of heat extraction, as quantified by Equation (1), is expected to range from 1.62 to 1.84 during the heating season, depending on the water temperature, the duration of water displacement, and withdrawal from the gallery.

An annual cycle of storage facility operation allows for saving approximately 1100–1800 tons of CO2 emissions if the electrical energy spent on operation is generated from coal. In the case of using the gas as fuel, these figures are approximately 1/3 lower.

The heat capacity of the potential low-temperature storage ranges from 14.0 to 18.7 kW·h/m3, which is lower than that of ATES facilities. However, the potential storage facility considered in this paper allows for a significant reduction of energy spent on accessing the thermal use of water stored underground.

4.3 Economic and environmental assessments

At the current stage of research, only preliminary economic assessments can be carried out. They should include the operational income (OI) and the capital investment costs (IC). The OI can be evaluated by comparing the costs of thermal energy produced and the electricity consumed. The thermal energy that can be recovered from the UTES during the 135-day heating season, with a temperature range of 18–22°C, is estimated to be 8.75–11.66 GW·h. The cost of this amount of heat for the tariffs in effect in 2025 is €1.14–1.52 million. Assuming a maintenance cost of 50% of the electricity costs, the annual OI may range from €0.51 to €0.81 million.

The IC for the heat conversion system on the ground, along with the auxiliary equipment, can be assessed by the data of the project implemented at the Ostrava-Karvina coal basin mine in the Czech Republic (Petričko et al., 2018). Heat is recovered from the water raised from flooded deep mine workings, with a total volume of 3.6 million m³. Investment costs associated with the installation of a geothermal system with a capacity of 10 MW, which includes heat pumps, accessories, tanks, heat exchangers, circulation pumps, heating and control, installation, etc., amounted to €2.86 million, so the specific cost of the equipment for a thermal output of 1 MW amounts to €286 000. Based on this estimate, creating a thermal capacity of 2.7–3.6 MW may require €0.77–1.03 million. Hydraulic sealing, considering the gallery inner surface of 3.6 × 105 m² and the geomembrane cost of €5/m² (Luciani & Peila, 2018; Method Statement, 2025), is estimated to be €1.78 million. A total IC should include the costs for pond creation, which can be estimated only preliminarily. Assuming the costs for pond creation of €3 million yield the total IC of €5.55–5.81 million, and the payback period of 7.2–10.9 years for a temperature range of 18–22°C.

The environmental impact of CO2 emission reduction can be calculated using specific reduction rates estimated in (Potentialstudie, 2018) for coal (93.5 t/TJ of produced heat), oil products (74 t/TJ), gas (55.9 t/TJ), and the amount of heat recovered from the UTES, considering the thermal equivalent of energy spent on operation. The CO2 emission reduction for the range of water temperatures from 18 to 22°C is estimated to be 1529–2343 t for coal, 1210–1854 t for oil products, and 914–1401 t for gas.

The issues related to the new pond, including land use, biodiversity, and water quality, have been addressed in the projects considered in previous studies (Alvarado Montero et al., 2016; Niemann, 2019; Niemann et al., 2018).