Research on Fresh and Hardened Sealing Slurries with the Addition of Magnesium Regarding Thermal Conductivity for Energy Piles and Borehole Heat Exchangers

Currently, renewable energy is increasingly important in the energy sector. One of the so-called renewable energy sources is geothermal energy. The most popular solution implemented by both small and large customers is the consumption of low-temperature geothermal energy using borehole heat exchanger (BHE) systems assisted by geothermal heat pumps. Such an installation can operate regardless of geological conditions, which makes it extremely universal. Borehole heat exchangers are the most important elements of this system, as their design determines the efficiency of the entire heating or heating-and-cooling system. Filling/sealing slurry is amongst the crucial structural elements. In borehole exchangers, reaching the highest possible thermal conductivity of the cement slurry endeavors to improve heat transfer between the rock mass and the heat carrier. The article presents a proposed design for such a sealing slurry. Powdered magnesium was used as an additive to the cement. The approximate cost of powdered magnesium is PLN 70–90 per kg (EUR 15–20/kg). Six different slurry formulations were tested. Magnesium flakes were used in designs A, B, C, and magnesium shavings in D, E and F. The samples differed in the powdered magnesium content BWOC (by weight of cement). The parameters of fresh and hardened sealing slurries were tested, focusing mainly on the thermal conductivity parameter. The highest thermal conductivity values were obtained in design C with the 45% addition of magnesium flakes BWOC.


Introduction
An increased share of renewable energy in the total energy balance of European countries can be observed in recent years. This is mainly due to the new energy guidelines included in the 2018 Directive of the European Parliament and of the Council on the promotion of the use of energy from renewable sources. The Directive determines the share of renewable energy to be at least 32% of the European Union's total energy balance by 2030, referring to the need for reducing emissions related to fossil fuels combustion and the implementation of other environmental objectives [1].
The new energy policy of Poland described in the Annex to the Resolution No. 22/2021 of the Council of Ministers of 2 February 2021 entitled "Energy Policy of Poland until 2040" the so-called EPP2040 assumes the optimal, longest possible use of own energy resources through i.a. the development of renewable energy sources, as well as energy efficiency improvement. From the renewable energy perspective, particularly important are the provisions concerning the renewable energy sources share (RES) in the final gross energy consumption, which has to be at least 23% by 2030, and the reduction of greenhouse gas (GHG) emissions by approximately 30% in comparison to 1990, also until 2030. In addition, it is planned that by 2040 all households' thermal needs will be covered by system heat The exploitation of geothermal water is the most effective method of heat extraction, but such a solution is strongly conditioned by the presence of aquifers with high water temperature. The most popular solution, available to everyone, which can be performed with any lithology, are borehole heat exchangers [12,34], as well as energy piles [16,35,36]. The so-called shallow geothermal energy has been described for a long time [10,[37][38][39][40]. Borehole heat exchangers are undoubtedly an increasingly common method of obtaining energy from the rock mass. They enable heat provision to both large facilities (such as shopping centers, schools, office buildings) and small single-family houses, which enhances their multidimensional ability to operate. The greatest increase in the use of the described system is mainly observed in highly developed countries such as Sweden, Germany or Switzerland [41][42][43][44][45][46][47], but also in other countries [48]. In 2018 alone, around 23,500 geothermal heat pumps were sold in Germany [46]. This progress is also noticeable in Poland. Based on data from two drilling companies, the amount of borehole heat exchangers performed (in meters) increased from approx. 4000 m to over 60,000 m over the years 2004-2010 [41,49]. Moreover, as of 31 December 2018, there are estimated to be over 56,000 ground source heat pumps in Poland (with a heating capacity between 10 and 200 kW). Their total capacity was at least 650 MWt, and heat production was 3100 TJ/year [50]. Proposals for using various types of hybrid systems are also increasingly popular [51][52][53].
Borehole exchangers are made by placing an appropriate pipe structure (single U-tube, double U-tube, coaxial system) in boreholes specifically drilled for this purpose [14,41,54,55]. There are also analogous systems with three U-tubes [56]. Discussions on using multiple U-tubes (multi U-tube) in one borehole [57], or W-type and coil-type constructions in concrete energy piles [58] are presented in international publications.
It is common practice to construct installations in the form of U-tubes at depths up to 150 m [49] or even 200 m [59,60], and coaxial systems in deeper boreholes [59,60].
The amount of heat exchanged with the rock mass is mainly influenced by the thermal conductivity of rocks [61][62][63]. The presence of groundwater flow is also a very important element that influences the heat transfer in aquifer caused by the operation of a BHE [9]. Another very important element improving the heat transfer is the filling/sealing slurry with appropriate parameters [12,64,65]. The selection of the sealing slurry affects the efficiency of the borehole heat exchanger's operation. Currently, the slurry is selected based on four criteria [41]: -Physicochemical compatibility with the environment (no negative impact on the natural environment); -Appropriate rheological properties (the slurry can be pumped); -Economic factors (minimization of slurry cost); -Highest possible thermal conductivity. Appropriately designing the slurry is not easy, it requires long and meticulous research in terms of both the additives used and the ingredients' proportions. The current state of the art includes the use of ready-made industrial mixtures, as well as cement with additives such as graphite, as sealing slurries for borehole heat exchangers [66][67][68]. According to the manufacturers, the thermal conductivity of ready-made industrial mixtures is approx. 2.0 W·K −1 ·m −1 . However, they are expensive and therefore rarely used in Poland, hence the legitimacy of searching for alternative fillers.
The research aims to design a sealing slurry with increased thermal conductivity, enabling more effective heat extraction from the rock mass. Powdered magnesium was selected for testing due to its high thermal conductivity. The thermal conductivity of magnesium at room temperature is approximately 156 W·K −1 ·m −1 [69]. The use of magnesium for energy exchange purposes is described, among others, by Tian et al. [70]. They describe the effect of magnesium addition on thermal conductivity and the upper temperature limit of thermal stability. They describe a highly thermally conductive composite phase change material created by mixing magnesium particles with a eutectic ternary carbonate salt (Li 2 CO 3 -Na 2 CO 3 -K 2 CO 3 ). The designed material was used as a heat transfer medium and/or energy carrier in advanced high-temperature concentrating solar power plants [70]. The use of this material, but in the form of magnesium oxide, is reported by Du et al. [71]. They describe the Thermal Conductivity of Epoxy Resin Reinforced with Magnesium Oxide-Coated Multiwalled Carbon Nanotubes studies. Multiwalled carbon nanotubes coated with the magnesium oxide (MgO@MWNT) were fabricated and dispersed into an epoxy matrix. The thermal conductivity of the epoxy resin was increased due to the increased content of MgO@MWNT [71].
The use of such slurry in borehole heat exchangers will allow reduction in the number of exchangers in installations for facilities with high demand for heat or cold, while limiting the area of land required for drilling [72][73][74][75].

Research Methodology
The design was based on the common cement CEM I 42.5R according to the PN-EN 197-1: 2012 standard. The choice of the binder was dictated by good strength properties (high value of early strength, greater than or equal to 20 MPa after 2 days), high availability on the market and, above all, low price. CEM I cement consists of 95-100% Portland clinker with 0-5% of secondary component admixtures [76].
The tested cement additive is magnesium. It is a silver-grey alkaline earth metal, used for research in the solid-state, in the form of flakes with a grain size below 0.25 mm, and in the form of shavings ( Figure 1 Appropriately designing the slurry is not easy, it requires long and meticulous research in terms of both the additives used and the ingredients' proportions. The current state of the art includes the use of ready-made industrial mixtures, as well as cement with additives such as graphite, as sealing slurries for borehole heat exchangers [66][67][68]. According to the manufacturers, the thermal conductivity of ready-made industrial mixtures is approx. 2.0 W•K −1 •m −1 . However, they are expensive and therefore rarely used in Poland, hence the legitimacy of searching for alternative fillers. The research aims to design a sealing slurry with increased thermal conductivity, enabling more effective heat extraction from the rock mass. Powdered magnesium was selected for testing due to its high thermal conductivity. The thermal conductivity of magnesium at room temperature is approximately 156 W•K −1 •m −1 [69]. The use of magnesium for energy exchange purposes is described, among others, by Tian et al. [70] They describe the effect of magnesium addition on thermal conductivity and the upper temperature limit of thermal stability. They describe a highly thermally conductive composite phase change material created by mixing magnesium particles with a eutectic ternary carbonate salt (Li2CO3-Na2CO3-K2CO3). The designed material was used as a heat transfer medium and/or energy carrier in advanced high-temperature concentrating solar power plants [70]. The use of this material, but in the form of magnesium oxide, is reported by Du et al. [71]. They describe the Thermal Conductivity of Epoxy Resin Reinforced with Magnesium Oxide-Coated Multiwalled Carbon Nanotubes studies. Multiwalled carbon nanotubes coated with the magnesium oxide (MgO@MWNT) were fabricated and dispersed into an epoxy matrix. The thermal conductivity of the epoxy resin was increased due to the increased content of MgO@MWNT [71].
The use of such slurry in borehole heat exchangers will allow reduction in the number of exchangers in installations for facilities with high demand for heat or cold while limiting the area of land required for drilling [72][73][74][75].

Research Methodology
The design was based on the common cement CEM I 42.5R according to the PN-EN 197-1: 2012 standard. The choice of the binder was dictated by good strength properties (high value of early strength, greater than or equal to 20 MPa after 2 days), high availability on the market and, above all, low price. CEM I cement consists of 95-100% Portland clinker with 0-5% of secondary component admixtures [76].
The tested cement additive is magnesium. It is a silver-grey alkaline earth metal, used for research in the solid-state, in the form of flakes with a grain size below 0.25 mm and in the form of shavings ( Figure 1). The approximate cost of powdered magnesium is PLN 70-90 per kg. PLN is the official currency and legal tender of Poland. According to the exchange rate from the National Bank of Poland as of 14 June 2021, one US dollar costs PLN 3.7185. The material is chemically stable under normal environmental conditions (1 atm and 20 • C). According to the classification system of the Globally Harmonized System of Classification and Labelling of Chemicals (GHS), it is a flammable solid substance (hazard class and category-Flam. Sol. 2), and a substance which releases a flammable gas in contact with water (Water-react. 2). Considering the fact that one of the extinguishing agents in the event of the selected additive's ignition is cement, it can be assumed that the design based on cement, water and magnesium will not be flammable [77].
Analyzing the toxicological hazards, magnesium is not acute toxic and is not classified as a sensitizing, corrosive or irritant substance, and therefore no special precautions or protective clothing are required. According to Regulation (EC) No 1272/2008 of the European Parliament and of the Council of 16 December 2008 on the classification, labelling and packaging of substances and mixtures, it is not classified as a substance hazardous to the aquatic environment. Despite that, it is recommended to prevent the substance from entering the sewage system, surface water and groundwater. Therefore, it is planned to perform a washout treatment in the future to reduce potential adverse effects [77].
The research on sealing slurries can be divided into two main categories: research on fresh slurries and on hardened ones [14]. Testing fresh slurry determined its liquidity, density, viscosity, and rheological parameters. For the hardened slurry, the most important parameter, its thermal conductivity, was investigated.
The FOX50 instrument ( Figure 2) is designed to test the thermal conductivity of materials in the range from 0.1 to 10 W·K −1 ·m −1 . It contains a set of two round plates covered from the outside with a cylinder equipped with an insulation layer. The instrument complies with the ISO 8301 standard for Thermal insulation-Determination of steadystate thermal resistance and related properties-Heat flow meter apparatus. The upper plate remains stationary, while the lower plate can move vertically due to the pneumatic mechanism. The instrument's body houses electronics, control devices and a liquid crystal display. Additionally, the device is equipped with a digital sample thickness reader with an accuracy of ±0.025 mm. The kit also includes a cooling module and a compressor ( Figure 2). Thermal conductivity studies were based on Pyrex calibration to standardize the results [78]. The material is chemically stable under normal environmental conditions (1 atm and 20 °C). According to the classification system of the Globally Harmonized System of Classification and Labelling of Chemicals (GHS), it is a flammable solid substance (hazard class and category-Flam. Sol. 2), and a substance which releases a flammable gas in contact with water (Water-react. 2). Considering the fact that one of the extinguishing agents in the event of the selected additive's ignition is cement, it can be assumed that the design based on cement, water and magnesium will not be flammable [77].
Analyzing the toxicological hazards, magnesium is not acute toxic and is not classified as a sensitizing, corrosive or irritant substance, and therefore no special precautions or protective clothing are required. According to Regulation (EC) No 1272/2008 of the European Parliament and of the Council of 16 December 2008 on the classification, labelling and packaging of substances and mixtures, it is not classified as a substance hazardous to the aquatic environment. Despite that, it is recommended to prevent the substance from entering the sewage system, surface water and groundwater. Therefore, it is planned to perform a washout treatment in the future to reduce potential adverse effects [77].
The research on sealing slurries can be divided into two main categories: research on fresh slurries and on hardened ones [14]. Testing fresh slurry determined its liquidity, density, viscosity, and rheological parameters. For the hardened slurry, the most important parameter, its thermal conductivity, was investigated.
The FOX50 instrument ( Figure 2) is designed to test the thermal conductivity of materials in the range from 0.1 to 10 W•K −1 •m −1 . It contains a set of two round plates covered from the outside with a cylinder equipped with an insulation layer. The instrument complies with the ISO 8301 standard for Thermal insulation-Determination of steady-state thermal resistance and related properties-Heat flow meter apparatus. The upper plate remains stationary, while the lower plate can move vertically due to the pneumatic mechanism. The instrument's body houses electronics, control devices and a liquid crystal display. Additionally, the device is equipped with a digital sample thickness reader with an accuracy of ±0.025 mm. The kit also includes a cooling module and a compressor ( Figure 2). Thermal conductivity studies were based on Pyrex calibration to standardize the results [78].  Six different formulations of the sealing slurry were prepared based on the addition of magnesium in two granulations, CEM I 42.5 R cement, and water as a mixing liquid. The amount of the additive was selected in concentrations of 15%, 30% and 45% BWOC (by weight of cement-based on the dry weight of cement). This choice was dictated by previous research observations, in which various concentrations were tested (1%, 2%, 5%), but the results were not satisfactory. A constant water-mixture ratio (W/M) of 0.5 for each formula was established to eliminate the potential impact of changing the coefficient W/M on the obtained results. Sealing slurry samples were prepared in accordance with the applicable standard. For this purpose, cylinder-shaped molds with a diameter of 55 mm were prepared and filled with slurry. After hardening, the discs were stored completely immersed in water, which corresponds to the conditions present in the borehole heat exchangers. Figure 3 shows an exemplary test sample. The composition of individual designs is presented in Table 1.
Energies 2021, 14, x FOR PEER REVIEW 5 Six different formulations of the sealing slurry were prepared based on the add of magnesium in two granulations, CEM I 42.5 R cement, and water as a mixing liq The amount of the additive was selected in concentrations of 15%, 30% and 45% BW (by weight of cement-based on the dry weight of cement). This choice was dictate previous research observations, in which various concentrations were tested (1%, 5%), but the results were not satisfactory. A constant water-mixture ratio (W/M) of 0. each formula was established to eliminate the potential impact of changing the coeffic W/M on the obtained results. Sealing slurry samples were prepared in accordance the applicable standard. For this purpose, cylinder-shaped molds with a diameter o mm were prepared and filled with slurry. After hardening, the discs were stored c pletely immersed in water, which corresponds to the conditions present in the bore heat exchangers. Figure 3 shows an exemplary test sample. The composition of indi ual designs is presented in Table 1.    Test results for the fresh sealing slurries are presented in Table 2. An increase in the additive concentration causes a decrease in the sealing slurry density, as magnesium (ρ ≈ 1.74 g·cm −3 ) has a much lower density than cement (ρ ≈ 3.05 g·cm −3 ). The liquidity decreases with increasing additive concentration, which makes the slurry more difficult to pump. With the increase in magnesium concentration, the dynamic viscosity of the slurry increases. It should be noted that cement slurry is a complex system that changes its properties over time, under the influence of both internal and external factors. An increase in the slurry's viscosity may lead to difficulties related to its injection. Therefore, in the future, prior to the potential industrial application, it is planned to test the designs of slurries enriched with admixtures of agents regulating the slurry's viscosity and increasing its liquefaction.
During the research practice, samples behaved differently from the moment they were prepared. For some additives, a height shrinkage (reduction of the sample height), while for others swelling, of even a few millimeters, was observed.

Results and Discussion
Each design was tested at least five times on a Pyrex calibration, and baseline descriptive statistics for the thermal conductivity and thickness were assessed for the individual tests. Table 3 presents the measured values of the thermal conductivity for the base sample, consisting only of cement and water with w/c = 0.5. Table 4 presents the results for designs A, B, and C. Figure 4 shows the effect of magnesium flakes concentration on the sample's average thermal conductivity value.
Based on the results, a linear relationship between the additive concentration and the slurry's thermal conductivity was determined. Additionally, in comparison to the base sample, for design A there was a 23% increase in thermal conductivity, for design B-42%, and for design C-as much as 68%. More detailed studies on the pumpability of slurries with the 30-45% magnesium addition should be carried out. The recommendation results from Table 2 data, where the slurry with the addition of 30% magnesium is pumpable, while the slurry with the addition of 45% magnesium is not. Table 5 presents the results for designs D, E, and F. Figure 5 shows the effect of magnesium shavings concentration on the average thermal conductivity values.   Thermal conductivity l, W•K −1

•m −1
The percentage concentration of magnesium, C A % Series1 Serie2   The disturbance of the linear trend may be related to the uneven additive distribution in the sample, as well as a different thickness of the samples (shrinkage) compared to the others. Moreover, in comparison to the base sample, for design D there was a 26% increase in thermal conductivity, for design E-22%, and for design F-38%.
Comparing Figures 4 and 5, it can be seen that with the same percentages of additives BWOC (30% and 45%), the higher thermal conductivity values were obtained for magnesium flakes. For a 15% BWOC addition, magnesium shavings resulted in a higher thermal conductivity value. For Designs A and D, which had the same BWOC content (15%) but different types of magnesium, similar thermal conductivity results were obtained. In the case of Designs B and E, as well as C and F, with the same BWOC content (30 and 45%) but different types of magnesium, these values are not similar. The results indicate that the form of the additive (either shavings or flakes) influence the thermal conductivity value of the tested samples. Figures 4 and 5 present the equations describing the studied phenomena, taking into account the coefficient of determination (R 2 ) as the relationship between the regression model and the studied phenomenon. For both additives, the polynomial function is a better description, as the coefficient of determination for the model is equal to one. In the case of the linear equation, a much better fit occurs in the case of magnesium flakes (R 2 = 0.9968) than in the case of magnesium shavings (R 2 = 0.7899). According to the literature [79], the stepwise regression family is not suitable for the approximation of thermal conductivity, which was confirmed by the conducted calculations. The linear regression has a much lower R 2 value compared to the polynomial regression. Other regression methods can also be applied, such as the regression based on artificial neural network or group method of data handling. Regression models always require validation, and the test of R 2 is most often used for this purpose. The higher value of R 2 indicates that the empirical model is highly prognostic for the original model [79]. In many fields of science, various regression models are developed and applied with the use of analytical data [80]. The authors used the most popular method for determining the correlation of the interdependence of phenomena. With polynomial regression, a functional correlation (R 2 = 1) was obtained due to the small number of variants of the examined slurry recipes.
Energies 2021, 14, 5119 9 of 13 base sample, for design A there was a 23% increase in thermal conductivity, for design B-42%, and for design C-as much as 68%. More detailed studies on the pumpability of slurries with the 30-45% magnesium addition should be carried out. The recommendation results from Table 2 data, where the slurry with the addition of 30% magnesium is pumpable, while the slurry with the addition of 45% magnesium is not. Table 5 presents the results for designs D, E, and F. Figure 5 shows the effect of magnesium shavings concentration on the average thermal conductivity values.  The percentage concentration of magnesium, C A % Series1 Serie2

Conclusions
The research aimed to find a cement slurry with the highest possible thermal conductivity while maintaining the lowest possible production costs. The use of a sealing slurry with an increased thermal conductivity improves the heat exchange with the rock mass. The innovation in the study was the addition of magnesium in the form of flakes and shavings. The addition of magnesium lowers the slurry's density but increases its viscosity. Magnesium tends to slightly increase in volume during the setting of the slurry in comparison to its fresh state.
An important factor influencing the results is the thorough mixing of the additive in the slurry, and the distribution of the additive particles in the hardened sample. The samples were tested on both sides in order to eliminate the influence of the uneven mixing, and the distribution of the additive in the hardened sealing slurry, on the thermal conductivity results. Thermal conductivity is strongly dependent on environmental conditions (including temperature and humidity). The humidity of the tested samples was at maximum due to the maturation conditions-full immersion in water as an environment similar to borehole conditions was assumed.
For slurries with the magnesium flakes addition, the thermal conductivity of the tested samples increases together with the percentage of additive in the sample, relative to the weight of dry cement. The highest thermal conductivity values were obtained for design C, where the thermal conductivity increased by 68% in relation to the base sample made of cement alone. For design A, a 23% increase was noted, and for design B-42%. In the case of samples with magnesium shavings addition, the highest increase in thermal conductivity compared to the base sample was recorded for design F (by 38%). For design E, it increased by 22%, while for design D by 26%.
In the future, it is planned to test samples with lower humidity and at different temperatures. The operation of a borehole heat exchanger, as well as an energy pile, can cause drying of the hardened sealing slurry, and can take place in a fairly wide temperature range, usually from −5 to +35 • C.