High-temperature heat pumps for geothermal applications in Africa: thermodynamic, economic and environmental evaluation

Research Paper High-temperature heat pumps for geothermal applications in Africa: thermodynamic, economic and environmental evaluation ☆ Zuffi Claudio * , Fiaschi Daniele University of Florence, Department of Industrial Engineering, Florence, Italy ARTICLE INFO Keywords: Geothermal High temperature heat pump Life cycle assessment Africa ABSTRACT Geothermal resources in Africa, from lowto high-enthalpy, remain underutilized despite their vast potential. The Rift Valley is rich in high-enthalpy resources for electricity generation, while the mainland offers abundant mediumand low-enthalpy sources suitable for diverse applications. This study explores the use of hightemperature heat pumps (HTHPs) with geothermal energy. The research develops a predictive model to assess the thermodynamic performance, economic viability, and environmental impact of large-scale HTHP deployment. The metamodels estimate key parameters such as installed capacity, heat output, and the Levelized Cost of Heat (LCOH). Life Cycle Assessment (LCA) quantifies environmental impact using a parametric Life Cycle Inventory (pLCI), linking HTHP construction impacts with thermodynamic performance. A key innovation of this study is its holistic approach, integrating technical, economic, and environmental evaluations to provide a comprehensive sustainability assessment. Environmental aspects focused exclusively on the Climate Change (CC) indicator. Unlike previous research, focused mainly on thermodynamics, this study includes cost analysis and environmental impact assessments using LCA methodologies. It also emphasizes real-world applications in Africa, where geothermal resources remain largely untapped. To bridge this gap, the model is applied to a Malawi case study, assessing hot-spring resources for sustainable cooking and vegetable drying, with direct socioeconomic benefits. Population density maps identify optimal user areas, showcasing HTHP feasibility in offgrid settings. Results highlight the potential of low-enthalpy geothermal energy for cost-effective, sustainable heating and industrial applications, reinforcing its role in Africa’s energy transition. This study provides a replicable framework for advancing geothermal resource utilization and supporting sustainability goals within the LEAP-RE Project. 1. Introduction The African continent is endowed with significant geothermal potential, particularly concentrated in the eastern and southern regions, aligned with the East African Rift System (EARS) [1]. This geological phenomenon presents a promising avenue for the utilization of geothermal energy resources, due to the correlation between geothermal activity and Quaternary volcanism along the rift axis [2,3]. Moreover, Kenya’s Rift Valley stands out as a hotspot for geothermal exploration and exploitation, boasting vast reserves estimated at 10,000 MW spread across 14 sites [4–6]. While conventional geothermal power plants have historically focused on high enthalpy resources, recent advancements have shed light on the untapped potential of medium to low enthalpy resources [7,8]. Notably, the adoption of cascading approaches, leveraging different temperature levels of geothermal energy, emerges as a promising strategy to maximize resource utilization efficiency [9–11]. The direct utilization of geothermal energy is the oldest and most versatile way to harness geothermal energy of medium and low enthalpy. Current trending direct uses are mainly for heating systems working directly or through heat pumps, aquaculture, drying crops, growing plants and vegetables in greenhouses, processes of the paper and the cement industry, food processing, brewing, dyeing of fabrics, snow melting, cooling spaces and balneology, among others [12,13]. In this context, Africa stands at the forefront of harnessing geothermal energy for sustainable development [6]. This underscores the continent’s commitment to transitioning towards cleaner and more reliable energy sources to meet growing energy demands while mitigating environmental impacts [11]. However, the use and the expansion of geothermal energy is ☆ This article is part of a special issue entitled: ‘ECOS2024 - Applied Thermal Engineering’ published in Applied Thermal Engineering. * Corresponding author. E-mail address: [email protected] (Z. Claudio). Contents lists available at ScienceDirect Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng https://doi.org/10.1016/j.applthermaleng.2025.127302 Received 29 November 2024; Received in revised form 5 June 2025; Accepted 23 June 2025 Applied Thermal Engineering 278 (2025) 127302 Available online 25 June 2025 1359-4311/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). hampered by several major challenges. First, the sector demands heavy upfront investments—for exploration, drilling, and building power plants—that can deter investors due to the high initial financial burden [14]. In fact, the cost of drilling an exploratory well can run into millions of dollars, and the uncertainty of discovering commercially viable resources only adds to the risk [15,16]. Furthermore, the long development timeline—from preliminary surveys to operational plants—complicates funding efforts because of extended payback periods and financial uncertainty [17–19]. On top of this, unclear policies and an underdeveloped regulatory framework, compounded by poor coordination between national and local institutions, further impede progress [14]. Additionally, the lack of adequately trained professionals in the geothermal field limits efficient development and maintenance of the infrastructure. Environmental and social concerns also emerge, as the construction and operation of geothermal projects can disturb natural ecosystems and potentially lead to the displacement of local communities [20]. Finally, technological shortcomings—such as outdated equipment and insufficient transmission networks—and a weak political commitment to creating robust regulatory systems continue to obstruct the full exploitation of geothermal resources [20,21]. In this geothermal challenging context, acquiring detailed data on geothermal resource conditions and evaluating geothermal potential using advanced engineering applications is essential. A comprehensive, multidisciplinary approach that integrates energy, economic, and environmental assessments can provide stakeholders with a holistic understanding of the benefits associated with the sustainable utilization of geothermal resources. Such an approach not only supports more informed decision-making but also encourages strategic investments and policy developments that contribute to a resilient and competitive energy sector. For this reason, in this paper the application of High Temperature Heat Pump (HTHP) for steam production is investigated. Integration of HTHP with geothermal systems offers potential energy consumption benefits [22]. A review of the existing literature reveals that the application of HTHP technologies in the geothermal sector has been explored by only a limited number of authors [22–25]. Despite the growing interest in innovative solutions for geothermal energy, research on HTHP remains fragmented and incomplete. Jeßberger et al. [23] investigate the potential enhancement of existing geothermal plants through the integration of large-scale heat pumps. The study assesses the LCOH and conducts sensitivity analyses on geological, design, and economic parameters. Focusing on a district heating case study south of Munich, the system delivers heat at an average temperature of 85 ◦C. Geothermal system costs are excluded, aligning with the approach of upgrading existing plants. Ungar et al. [22] provide a robust thermodynamic basis for evaluating various geothermal HTHP configurations for industrial steam production, highlighting the potential of water-based systems and the limitations of CO 2 systems. However, economic assessments and experimental validation are needed to fully evaluate the practical feasibility of these technologies. Lu et al. [25] presents a thermodynamic analysis of a HTHP system for steam generation using medium–low temperature geothermal water. The study lacks a detailed economic assessment, limiting evaluation of its cost-effectiveness compared to alternative technologies. System optimization focuses on thermodynamic parameters, overlooking factors such as component sizing, costs, and environmental impact. Results are based on specific operating conditions, potentially limiting their generalizability. Kim et al. [24] present a HTHP system design using moderate-temperature geothermal water. The study focuses on a hybrid compression/absorption heat pump capable of producing temperatures above 90 ◦C and as low as 20 ◦C using 50 ◦C geothermal water. The innovative hybrid system combines compression and absorption cycles but introduces greater complexity, potentially increasing maintenance costs and reducing reliability. Like Lu et al. [25], it lacks a detailed economic analysis, limiting assessment of its cost-effectiveness. While the system could reduce global CO 2 emissions by up to 8 %, the study does not provide a comprehensive comparison of greenhouse gas emissions with alternative technologies. Nomenclature General HTHP High temperature heat pump GAA Geothermal atlas for Africa n p number of intervals HFO Hydrofluoroolefins Economic C P0 Cost of machinery reference pressure F P Pressure factor C BM Cost bare module C valve Cost of the valve C well Cost of the well C i Cost of single machinery T EW Total cost of equipment and wells C DPI Total direct permanent investment C TDC Total depreciable capital C TPI Cost total permanent investment C TCI Total capital investment LCOH Levelized cost of heat Environmental LCA Life Cycle Assessment pLCI Parametric life cycle inventory GWP Global warming potential CC Climate change indicator System COP Coefficient of performance Q evap Thermal power evaporator Q cond Thermal power condenser m cycle working fluid flow rate W comp Compressor’s work requirement T geo Geothermal temperature m geo Geothermal mass flow rate Scenarios Hm High mass flow rate Mm Medium mass flow rate Lm Low mass flow rate CS Cooking statio CD Cassava drying LT Low users temperature HT High users temperature Subscript 5 k 5 thousand people 10 k 6 thousand people 50 k 7 thousand people 10 10 thousand tons of cassava 25 25 thousand tons of cassava 80 80 thousand tons of cassava Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 2 In summary, several aspects are noted in the literature: •Many studies focus on thermodynamic performance but omit detailed economic assessments. This includes full cost evaluations—capital, operational, and maintenance costs—which are essential for comparing HTHP with alternative technologies. •Although some studies mention potential CO 2 reductions, comprehensive, comparative analyses of greenhouse gas emissions across different HTHP configurations and alternative systems are lacking. Notably, no studies have investigated the environmental impacts of HTHP technologies using established methodologies such as Life Cycle Assessment (LCA). •The generalizability of results is limited by analyses performed under specific operating conditions. Additionally, issues related to the integration of HTHP systems with existing geothermal infrastructures and district heating networks are not fully explored. These gaps are particularly significant, given the increasing importance of sustainability as a key criterion for developing new energy technologies. Furthermore, literature lacks applied studies analyzing real implementations of HTHP systems, leaving an important research area regarding the technical and operational challenges of this technology unaddressed. These lacks are especially remarkable within the African Continent, where the energy transition is at its infancy and the nexus between several novel energy technologies and territorial resources is undisclosed yet at a large extent. Incorporating an LCA-based approach would allow for a comprehensive evaluation of environmental impacts, fostering the development of more sustainable solutions and supporting the broader adoption of HTHP technology in the geothermal sector. The aim of this paper is to address the gap identified in the scientific literature by proposing a simplified large-scale model for assessing the sustainability of HTHP technologies combined with geothermal energy. The model is structured to evaluate a wide range of thermal capacities and to supply heat to users at different temperature levels. Additionally, it produces specific results related to economic and environmental parameters, providing a holistic and comprehensive sustainability analysis. The environmental model is based on the LCA approach. In addition, to also meet the gap in the African context the model is applied to an African case study, Malawi hot-springs, exploring multiple scenarios involving applications such as sustainable cooking and vegetable drying—applications that could improve the quality of life for local populations. For this reason, population density maps are used to assess the potential user area that could benefit from these solutions. This work was carried out as part of the Long-Term Joint European Union-African Union Research and Innovation Partnership on Renewable Energy, Geothermal Atlas for Africa (LEAP-RE GAA 2020) project. 2. Methodologies The methodology of this study presents a comparative analysis of two High-Temperature Heat Pump (HTHP) system configurations, as shown in Fig. 1. •Cycle A features a single pressure level and includes an evaporator, compressor, condenser, and expansion valve. •Cycle B operates with two pressure levels and incorporates a separator at an intermediate pressure, positioned between the system’s minimum and maximum pressures. The separator plays a key role in recovering heat from the working fluid after condensation and reducing the compressor’s workload, ultimately enhancing the system’s coefficient of performance (COP). Both system configurations are based on the work of Cao et al. [26] and use R152 as the working fluid. To validate the thermodynamic model, the input conditions specified in the study by Cao et al. [26] were replicated. The model was validated by comparing the thermal capacities of the mechanical components, resulting in a deviation of approximately 0.5 % from Cao’s findings. The same refrigerant, R152, was used throughout the validation process. Although hydrocarbons such as R600a and R290 are commonly used in regions where low-cost solutions are preferred, the choice made in this study is primarily driven by environmental considerations. Specifically, lowGlobal Warming Potential (GWP) HFOs, e.g. R1233zd(E) and R1234ze(Z) refrigerants were selected as more sustainable alternatives. Moreover, these refrigerants enable higher condenser temperatures compared to R152a, R600a, or R290, allowing for increased supply temperatures while still operating within subcritical cycle conditions, as required in the model of this work. Fig. 1. Plant layout of the two cycles: A on the left, B on the right. Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 3 These models are designed to evaluate the performance of the two cycles by analyzing several key variables. These include the thermal power of the evaporator (Q evap ), which represents the heat absorbed from the geothermal fluid, and the heat output from the condenser (Q cond ), indicating the system’s useful thermal power. Additionally, the work required by the compressor (W comp ), is considered, along with the coefficient of performance (COP) of the system. The models also assess the flow rate of the working fluid (m cycle ), the flow rate of the generated hot water (m w_out ), and the temperature of the hot water output (T w_out ). Once the key parameters of the thermodynamic model have been identified, it is connected to the economic and environmental models. These models evaluate economic and environmental parameters, enabling a comprehensive assessment of the HTHP system. A detailed description of each model is provided in Sections 2.1 Thermodynamic Modeling, 2.2 Thermo-Economic Modeling, and 2.3 Environmental Modeling. The innovative approach, which allows the developed models to be extended to a wide range of applications in terms of size, energy performance, and economic and environmental outcomes, involves the generation of metamodels for each key parameter. A metamodel is a simplified version of a more complex system, which reduces computational costs [27]. The process began by defining the basic inputs for the thermodynamic model: geothermal temperature (T geo ), geothermal mass flow rate (m geo ), and the extraction well depth. A variation range was established for each parameter: 50–100 ◦C for T geo , 1–50 kg/s for m geo , and 0–15000 m for the well depth. Next, a set number of calculation intervals (n p ) was defined for each range, resulting in three input matrices with dimensions of n p x n p . Then, all possible input combinations were evaluated using the thermodynamic, economic, and environmental models, with the thermodynamic model being optimized each time to achieve the maximum thermal capacity at the evaporator. The results from the various models were compiled into output matrices, covering mechanical component sizes (kW), basic cycle parameters (pressure, temperature, flow rate, etc.), component costs ($), thermoeconomic costs ($/kWh), and environmental indicators (kg CO 2 eq/ kWh). Finally, by performing multidimensional interpolation on each output matrix, the metamodels were generated for the following parameters: Q evap , Q cond , W comp , m cycle , COP, LCOH, and CC. The number n p was determined through an iterative process, meaning it is set when the metamodel achieves a relative mean prediction error of 0.5 % compared to the initial model. Based on the structure of the model, it allows for the evaluation of scenarios that either use surface geothermal water or involve extraction from shallow wells. It is important to note that the thermodynamic metamodel’s efficacy is independent on well depth. However, the metamodel’s thermo-economic and environmental impact are significantly influenced by the presence and depth of geothermal wells, as highlighted in recent studies [28,29]. Accordingly, this study engages on a comprehensive thermo-economic and environmental analysis across three scenarios: surface geothermal water extraction (well 00 ), extraction from a 500 m well coupled with reinjection into a 300 m well (well 500 ), and extraction from a 1000 m well with reinjection into a 600 m well (well 1000 ). 2.1. Thermodynamic modeling In the first configuration, marked as Cycle A operating at a single pressure level, the geothermal fluid is introduced into the evaporator. Here, it transfers heat to the working fluid, which flows in countercurrent. On the geothermal side, the outlet temperature (T geo,out ) is fixed at 45 ◦C, while the inlet temperature (T geo,in ), as specified in Section 2, varies between 50 ◦C and 100 ◦C. Therefore, the temperature difference (ΔT geo ) is variable, ranging from a minimum of 5 ◦C to a maximum of 55 ◦C. On the HTHP side, the temperatures T a1 and T a2 , as well as T b4 and T b5 , are set, with the temperature differences at the evaporator inlet (ΔT evap,in ) and outlet (ΔT evap,out ) being dependent on the geothermal conditions, as described in Eqs. (1)–(3). ΔTgeo =Tgeo,in −Tgeo,out (1) Ta1,Tb4=Tgeo,out −ΔTevap,out (2) Ta2,Tb5=Tgeo,in −ΔTevap,in (3) After this step, the evaporator and mass flow rate in the cycle (m cycle ) can be evaluated using Eqs. (4) and (5), where h geo,in and h geo,out represents the enthalpy of the fluid at the inlet and outlet conditions, respectively Qevap =mgeo • (hgeo,in −hgeo,out)(4) mcycle =Qevap/(ha2−ha1)(5) The working fluid, which exists either as saturated vapor (x =1) or superheated steam, enters the compressor. The compressor is responsible for increasing the fluid pressure to match that of the condenser and it is calculated as Eq. (6). The power required by the compressor is evaluated by considering the pressure conditions at the inlet and outlet streams of the compressor, while also accounting for the device’s efficiency, set at etacomp. Wcomp =mcycle • (ha3−ha2)(6) The operating conditions of the condenser have been set. Specifically, the pressure corresponds to the saturation pressure of the fluid at the condensation temperature (T cond ). This temperature is selected based on the intended application. In the model, two different temperature levels are considered: •low-temperature (LT) level set at 105 ◦C •high-temperature (HT) level set at 140 ◦C On the user side, the water is assumed to be drawn at an inlet temperature of 50 ◦C (T us,in ) from various sources such as residential, commercial, or industrial applications. The water is then heated to an outlet temperature (T us,out ) which is determined using Eq. (7). Thus, the condenser is evaluated using Eq. (8): Tus,out =Tcond −ΔTcond (7) Qcond =mcycle • (ha3−ha0)(8) By utilizing the heat output provided by the condenser, the flow rate of hot water that can be produced at this temperature is determined. Finally, the expansion valve reduces the working fluid pressure back to the evaporator conditions, thereby completing the thermodynamic cycle. Finally, the COP is evaluated using Eq. (9). COP = (Qcond/Wcomp)(9) In configuration (B), a separator is integrated into the cycle. Following the initial expansion undergone by the fluid in the highpressure expansion valve, the separator segregates saturated liquid from saturated vapor. The former undergoes further expansion in a second expansion valve until it reaches the evaporator pressure. Meanwhile, the latter is blended with the fluid exiting the low-pressure compressor. This mixture is subsequently pressurized to match the condenser pressure by the high-pressure compressor. The intermediate pressure at the separator has also been optimized to achieve the maximum COP in configuration B. The optimized values are reported in Table 1 for each fluid and for both LT and HT scenarios. Table 1 reports the main parameters for both cycles. One important aspect to highlight is the influence of different refrigerants in the model. The selected fluids, R1234ze(z) and R1233zd(e), have relatively similar characteristics in terms of critical pressure (P c,R1234ze(z) =39.7 bar; P c, Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 4 R1234zd(e) =35.7 bar) and critical temperature (T c,R1234ze(z) =153 ◦C; T c, R1234zd(e) =165.5 ◦C). As a result, their performance differs from other refrigerants, such as R152, under these conditions, particularly in terms of system efficiency and the temperature level at which heat is delivered to the user. This aspect is further examined in Section 4.1. The final output of the metamodel is a surface which depends on the two inputs (T geo , m geo ) and represents a thermodynamic parameter of the cycle. For better understanding, the metamodel derived from linear multidimensional interpolation is shown in Fig. 2, with Cycle A featuring fluid R1233zd(E) as an example. 2.2. Thermo-economic modeling The economic evaluation has two main stages. First, the cost of each individual plant component is estimated using the thermo-economic correlations proposed by Turton [30]. Then, the initial investment cost, along with total operating and maintenance costs, is assessed to calculate the Levelized Cost of Heating (LCOH). The thermodynamic analysis provides key parameters needed for the thermo-economic correlations. These include the heat transfer surface area required for the condenser and evaporator, the flow rates handled by the compressor and valve, and the volume flow rate in the separator. For heat exchangers, the parameter S, which represents the size of the mechanical component, corresponds to the surface area. This is determined using the thermal potential obtained from the thermodynamic model (expressed in kW), the logarithmic mean temperature difference (evaluated using Eq. (10)), and the global heat transfer coefficient U, which is set to 1.8 kW/(m 2 K) for the condenser and 2.5 kW/(m 2 K) for the evaporator. In Eq. (11), Q HE represents the nominal heat power of either the condenser or the evaporator. Table 1 Main thermodynamic parameter for the cycles. Parameters Value Unit Geothermal fluid outlet temperature T geo,out 45 ◦C Evaporator inlet temperature difference ΔT evap,in 5◦C Evaporator outlet temperature difference ΔT evap, out 9◦C Compressor isentropic efficiency η comp 0.8  Inlet temperature of the supplied water T us,in 50 ◦C Temperature difference between critical point and condenser ΔT cond 10 ◦C Condenser outlet temperature difference ΔT cond, out 5◦C Separator pressure LT for R1234ze(z) P LT 7.6 bar Separator pressure HT for R1234ze(z) P HT 13.4 bar Separator pressure LT for R1233zd (e) P LT 5.8 bar Separator pressure LT for R1233zd(e) P HT 10.4 bar Fig. 2. Cycle_a fluid R1233zd(E). Table 2 Main equations for the thermo-economic model. Cost Correlation Evaporator, condenser, compressor CP0=10K1+K2*log10 S+K3*log10 S2FP= 10C1+C2*log10P+C3*log10 P2CBM = CP0*(B1+B2*FM*FP)  All variables are indicated in the tables of Turton 2008  Valve Cvalve =114.5*mcycle [47] Wells Cwell =2500*mdrilled [28] Total cost of equipment and wells TEW =∑nc iCi+Cwell C i cost of single component C well cost of the wells Total direct permanent investment CDPI =TEW +Csite +Cserv C site cost of site preparation Cserv cost of service facilities C serv cost of service facilities Total depreciable capital CTDC =CDPI +Ccont C cont cost of contingencies Cost total permanent investment CTPI =CTDC +Cland +Croyal+Cstartup C royal cost of royalties C land cost of land C startup cost of plant startup Total capital investment CTCI =CTPI +CWC C WC cost of working capital CP 0 =Purchase component to ambient pressure; S =capacity or size parameter for the equipment; K n =Turton table value; F p =pressure factor, F M =Material factor; BM =Bare module; C i =Cost of single machinery; B n =Turton Costant. Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 5 For the separator, present in the Cycle B only, the thermo-economic correlation referenced from Mosaffa et al. [31] is adopted. Table 2 outlines the component types and equations used for each one. Various economic equations exist for geothermal wells, often following linear correlations. In this study, the correlation proposed by Shamoushaki et al. [28] is adopted. LCOH is a techno-economic parameter fundamental for feasibility assessment, enabling the economic comparison of thermal systems. As reported in Eq. (12), it quantifies the cost of heat produced per kilowatt-hour ($/kWh), as discussed by de Sim´ on-Martín et al. [32]. ΔTml =ΔThot −ΔTcold log((ΔThot/ΔTcold)) (10) SHE = (QHE/(U•ΔTml)) (11) LCOH =CTCI +∑n i=1CO&M (1+r)i ∑n i=1Ei (1+r)i (12) C TCI is the total capital investment cost, C O&M is the annual operating and maintenance cost, E i is the annual produced heat, r is the discount rate and n is the lifetime of the energy system. For the calculation of C TCI , the method and parameters used in this work were taken from Karimi and Mansouri, [33] and Shamoushaki et al. [34] and readapted to the case of HTHP, as shown in Table 2. On the other hand, the yearly O&M cost was taken from Beckers et al. [35] for the general part. The cost of electricity, set at $0.15/kWh and the total energy consumed by the compressors are taken from the above discussed thermodynamic metamodels. Furthermore, a 7 % discount rate and a useful life of 25 years were assumed. 2.3. Environmental modeling The environmental modeling approach is based on the Life Cycle Assessment (LCA) methodology, outlined by ISO 14040 and ISO 14044 standards [36,37]. The approach used is cradle-to-gate, so the system boundaries include the production and reinjection geothermal wells (if present), the piping system, the surface plant, and the utilization phase. The functional unit for the analysis is the thermal kWh produced by the HTHP system. Specifically, to facilitate the assessment across numerous scenarios, emphasis was placed on developing a Parametric Life Cycle Inventory (pLCI). pLCI links material and energy consumption throughout the production, O&M of the HTHP system with its size, as determined by the thermodynamic metamodel. For the construction phase of the system, primary data were collected from commercially available HTHP models sourced from TRANE [38]. A correlation was established between system thermal power (kW) and total weight, as reported in Fig. 3, based on the collected data. Subsequently, the production process −heat pump production, brine-water, 10 kW (heat pump, brine-water, 10 kW) −from the Ecoinvent database was selected as a reference model [39]. Inputs and outputs of this process were categorized into materials and energy utilized during construction. Percentages for each material and energy consumption were determined, with energy normalized to the total mass of the element, as reported in Table 3. This allowed the creation of a new environmental model for the HTHP, incorporating adjusted input and output values multiplied by the correlation obtained in Fig. 3. The modeling of geothermal wells was approached linearly, utilizing a shallow well drilling process as described by [40]. The operation phase accounted for electricity consumption using a reference process for electricity production (market for electricity, medium voltage | electricity, medium voltage, KE) while maintenance was evaluated as a 20 % material replenishment throughout the whole life cycle. For the analysis, the Ecoinvent 3.7.1 database [39] and the Brightway software, integrated with activity-browser software via Python, were employed. The environmental impact, primarily focused on the Climate Change (CC) indicator, is expressed through the Environmental Footprint 3.1 methodology. 3. Case studies: Malawi hot springs The case study evaluated in this paper is the Malawi hot springs. Davalos et al. [41] performed an extensive study on the geochemical characteristics of 27 hot springs in Malawi, distributed along north to south of the Malawi Rift Zone (MRZ). The Malawi Rift Zone (MRZ) is a magma-poor rift where geothermal potential is indicated by high heat flow and the presence of hot springs. Davalos et al. [41] examined 27 hot springs to estimate reservoir temperatures, investigate geochemical processes influencing water composition during ascent, and identify the most promising sites for geothermal energy development. During the dry season in July 2013, water samples were collected directly from the spring sources using the grab technique. These samples were filtered and stored in designated containers for anion, cation, and isotope analysis. In addition to the characteristics related to the chemistry of the geothermal resource, the surface temperature assessment is carried out by reporting the temperature of all 27 sites. It ranges between 35–80 ◦C, but only hot springs with a temperature above 50 ◦C are taken into consideration in this work, as reported in Table 4. Unfortunately, the required mass flow rate are not available, for this reason, three flow rate scenarios are defined −High 50 kg/s (Hm), Medium 20 kg/s (Mm), Low 8 kg/s (Lm). Final applications have been defined for possible users of the heat produced at the respective temperatures. In particular, two possible uses have been selected: •Cooking station (CS) supplying steam at 140 ◦C (HT condition) for cooking food, which requires about 2.89 GWh/year heat load for a Fig. 3. Development of correlation between total system weight and nominal HTHP heating capacity. Table 3 Material and energy flows relative to total weight evaluated by the correlation in Fig. 3. Material / energy flows Amount Unit copper 16.96 % lubricating oil 1.31 % polyvinylchloride 0.77 % reinforcing steel 57.83 % steel, low-alloyed, hot rolled 15.42 % tube insulation, elastomere 7.71 % Total weight (Tw) Tw Tons Refrigerant R134a 0.02 kg/Tw Electricity, medium Voltage 1.079414 kWh/Tw heat, district or industrial, natural gas 10.25443 MJ/Tw Water, unspecified natural origin 0.005459 m3/Tw Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 6 group of 5,000 people, (CS 5k ); 5.79 GWh/year for 10,000 people (CS 10k ); 29.00 MWh/year for 50,000 people (CS 50k ) [42]. •Cassava drying (CD) supplies steam at 100 ◦C (LT condition) for drying vegetables. This process requires about 306 kWh/(ton), so three scenarios are assumed in which a company produces about 10 k tons/year CD 10 , 25 k tons/year CD 25 , and 80 k tons/year CD 80 [43]. 4. Results 4.1. Parametric analysis This section presents the results obtained from the parametric analysis of the thermodynamic model introduced in Section 2.1. From the analysis conducted on the performance of the system depicted in Fig. 4, the COP in all scenarios within Cycle A and Cycle B are compared, referred to the above mentioned HFO and R152 fluids. COP is not m geo dependent but only on T geo , as the m cycle increases and decreases at the same rate as m geo , thus the COP does not change at variable m geo . This analysis reveals how the results are affected by different refrigerants. As shown in the figure, high-temperature conditions for R152 are not reported. This is because the critical temperature of R152 is below 140 ◦C (T c,R152 =113 ◦C), which prevents maintaining a subcritical cycle, making this refrigerant unsuitable for the working code. The general trend is increasing because, as T geo increases, the difference between highand low-pressure decreases, so the compressor work is reduced and the COP increases. Focusing on the case of LT, the COP for HFO fluids ranges between 3.67 and 4.14 for Cycle A and between 3.53 and 4.13 for Cycle B. Despite the difference of pressure levels in this case, there is no remarkable difference between the two cycles at the thermodynamic level. Conversely, there is a notable performance difference with fluid R152, exhibiting consistently lower COP across all T geo values. Considering that HFO’s environmental performance is better than R152 due to its lower GWP and the better thermodynamic performance of the cycle, the selection of HFO fluids is advised. In addition, in the case of HT with HFOs, it is possible to maintain the subcritical cycle by increasing the output temperature. The comparative analysis between Cycle A and Cycle B unveils a notable downturn in COP for both cycles when juxtaposed with the LT scenario. Comparing LT and HT, COP of cycle A shows a sharp drop from 1.90 to 2.47. In Cycle B, albeit exhibiting a decrease compared to the preceding scenario, the COP ranges between 2.59 and 2.89. The rise of T cond leads to a growth of pressure, consequently increasing the workload for the compressor. In spite of the higher temperature steam produced for the end user, at the same time the COP of the cycle is reduced. This behavior is enhanced in Cycle A, exerting a lower impact on Cycle B due to the presence of two compression levels, which mitigates the overall work required during the compression stage. For the following analyses, Cycle A for the LT scenario and Cycle B for the HT scenario are taken as references. 4.2. General results of metamodels: economic and environmental This section presents the results obtained by applying the economic and environmental models introduced in Sections 2.2 and 2.3 for the generation of metamodels, as described in Section 2. Specifically, the results of the metamodels for the economic parameter LCOH and the environmental parameter CC are presented. The thermo-economic evaluation reveals a significant variability in Table 4 Hot spring geothermal sites in Malawai [41]. Hot spring Surface Hot spring Surface T◦C  T◦C Ngala 54.9 Chombo 66.8 Chiweta 79.7 Madzimawira 63.7 Mphizi Stream 78.7 Ling’ona 58.1 Mtondolo 64.8 Chikwizi 52.2 Mtondolo 2 72.9 July Borehole 51.4 Chiwe 74.9    Fig. 4. Comparison of COP for the different analyzed cycles. Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 7 LCOH, depending on the resource conditions and well characteristics. Fig. 5a–c shows the LCOH of well 1000 , well 500 , and well 00 respectively. Additionally, the red dashed line represents the reference LCOH value provided by IEA [44] for heat generated from natural gas-fired boilers, hovering around 17c$/kWh, while the cost obtained for HTHP is indicated as 22c$/kWh [45]. It is clear the relevant decrease in LCOH from 5a to 5c, highlighting the significant influence of drilling costs on this parameter. In Fig. 5a, it is observed that LCOH is exceedingly high for low value of T geo and m geo , reaching a maximum of 163.4c$/kWh, with economic feasibility only apparent at higher temperatures or increased mass flow rates, where LCOH drops to about 12.8c$/kWh. In Fig. 5b, where the well depth is shallower, the maximum LCOH reduces to 100c $/kWh, and more combinations of T geo and m geo below the red dashed line LCOH values. Finally, in Fig. 5c, only a narrow range fails to yield an economic advantage compared to literature data. In fact, for all combinations of T geo and m geo above 62◦C or above 20 kg/s, an LCOH below the boiler heat output is obtained. This remarks the strongly affected dependence of LCOH on the cost of wells. Therefore, for HTHP applications, cases where the geothermal reservoir is superficial, hot waste fluids from geothermal power plants or other situations without drilling wells would offer significant economic benefits. However, if a greater number of wells or deeper drilling is necessary, the conditions for LCOH advantage are reduced considerably. Therefore, a careful evaluation is needed for each specific case, according to the depth of the available resource. The same contours for the LT case are not displayed due to the similar trend, with minimal change in LCOH range, (±3–6c$/kWh). Fig. 6 illustrates the metamodel of the environmental impact indicator (CC) for the same cases presented in Fig. 5. As noted in section 1, the literature on geothermal HTHP systems is extremely limited, with few studies available on their environmental impacts. Consequently, the reference values used for comparison are adopted from standard heat production systems provided by the Ecoinvent database and derived from LCA. Specifically, one reference value corresponds to a natural gas modulating condensing boiler producing 1 kWh at 264.16 g CO 2 eq/ kWh (red line), while the other represents a borehole heat exchanger with a brine-water system, evaluated at 223.88 g CO 2 eq/kWh (blue line) [39]. The behavior is similar to the LCOH, where CC has a higher impact for case 6a, whereas for low m geo and T geo , 329 gCO 2 eq/kWh is found which very similar to normal boiler emissions. For this reason, temperatures above 67 ◦C or m geo above 25 kg/s are preferred to have a low environmental impact, compared to other low-impact systems. Also, all the combination of m geo and T geo above the dashed blue line are favorable, as they correspond to a lower environmental impact. It should be considered that, also in this case, the impact of drilling has a major contribution, even though it is higher compared to the effect on LCOH, because CC is also strongly affected by electricity production. As shown in both Fig. 6b and 6c, the potential for achieving low environmental impact is even larger, confirmed by a decrease in equivalent emissions. The borehole heat exchanger brine-water limits given in Fig. 6b only cover a small range for temperatures below 53◦C. On the other hand, from Fig. 6c all cases are environmentally advantageous. Combining the information obtained from the economic (LCOH) and environmental (CC) metamodels, a cost-benefit or impact-benefit approach can be defined. Using reference values from the literature—17c$/kWh for a gas boiler and 223.88 g CO 2 eq/kWh for a borehole heat exchanger with a brine-water system—as normalization thresholds for LCOH and CC, respectively, the metamodels can be normalized to derive a cost-benefit or impact-benefit ratio. If this ratio is less than 1, the resource conditions provide an economic or environmental benefit; if it is greater than 1, no benefit is obtained. Naturally, the lower the ratio, the more advantageous the system becomes. As shown in Fig. 7, shallower geothermal resources allow for a broader range of T geo and m geo combinations that yield a cost-benefit or impact-benefit ratio below 1, orange dot-line. This implies that, under such conditions, the system delivers economic and environmental advantages relative to the reference thresholds. Furthermore, among the constraints examined, economic feasibility stands out as the most critical factor, overshadowing environmental considerations. In practical terms, this means that parameters such as initial capital expenditure, operational costs, and maintenance expenses have a greater influence on system viability than environmental impacts like CO 2 emissions. Consequently, when evaluating HTHP installations, it is advisable to Fig. 5. Techno-economic metamodel of the LCOH parameter for the three well depth scenarios: 5a well 1000 , 5b well 500 , 5c well 00. Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 8 prioritize economic viability in the decision-making process. The optimal conditions derived from the economic metamodels—indicating the best balance of cost-efficiency and performance—should serve as the primary benchmark for project design and implementation. 4.3. Results of Malawi case studies This section presents an in-depth analysis of the Malawi case study, where the complete metamodel—encompassing energy, economic, and environmental dimensions—is applied to evaluate the performance of geothermal heat utilization. As introduced in Section 3, this study focuses on Malawi’s hot springs as a potential renewable energy source. The results for Cycle B, using R1234ze(Z) as the working fluid, are illustrated in Fig. 8, which provides a comprehensive overview of both thermal energy production and the associated cost and environmental impact metrics. On the left side of Fig. 8, the graph displays the thermal energy produced by the system, with dashed red lines indicating the specific heat demand levels required for three different cooking station (CS) scenarios: CS 5k , CS 10k , and CS 50k . These values represent progressively larger community cooking demands, with CS 5k corresponding to smallscale operations, CS 10k representing medium-scale cooking stations, and CS 50k indicating large-scale community cooking facilities. The analysis of thermal performance under different operational conditions provides key insights into the feasibility of utilizing hot spring resources for cooking applications. Under Low (Lm) conditions, nearly all hot springs successfully achieve and sustain the CS 5k level, with the exception of July Borehole, Chikwizi, and Ngala. These three hot springs exhibit lower thermal energy output, making them unsuitable for small-scale cooking applications under minimal flow rate conditions. Some of the most productive hot springs, such as Chiwe, Motondolo 2, and Mphizi Stream, demonstrate the ability to meet the thermal demand of CS 10k , even under Lm conditions. This finding is particularly significant, as it suggests that these hot springs have sufficient energy output to support medium-sized cooking stations even when operating at lower flow rates. Under Medium (Mm) conditions, the thermal output of all hot springs increases, allowing all wells—except for July Borehole—to fully satisfy the CS 10k thermal load. In the case of July Borehole, the thermal energy falls just short of the requirement by approximately 1.6 %, indicating that a minor optimization or slight increase in operational flow rate could enable this hot spring to meet the CS 10k demand. Finally, at the highest heat demand level (CS 50k ), only a selected group of high-performing hot springs—Chombo, Chiwe, Motondolo, Motondolo 2, Mphizi Stream, and Chiweta—can reach or exceeding this threshold. The ability of these wells to supply sufficient energy for largescale cooking stations is a crucial finding, highlighting their potential for supporting large communities. A particularly noteworthy observation is that some hot springs—such as Chiwe, Motondolo 2, and Mphizi Stream—produce thermal energy well above the CS 50k requirement. In certain cases, their output exceeds the assigned threshold by more than 50 %, emphasizing their potential as reliable, high-capacity geothermal energy sources. These findings reinforce the viability of geothermal heat for large-scale cooking applications, particularly in regions where access to affordable and sustainable energy sources is essential. Beyond thermal performance, the economic feasibility of using geothermal energy for cooking applications is assessed by analyzing the LCOH. The right side of Fig. 8 presents the LCOH values for all evaluated hot springs. The LCOH across all hot springs remains within a highly competitive range of 11.2c$/kWh to 12.5c$/kWh, positioning geothermal heat as an economically viable alternative to conventional heating sources. The primary reason for these notably low costs is the shallow depth of the geothermal resources being utilized. Since the hot springs are relatively close to the surface, the cost associated with well drilling is significantly reduced, minimizing its impact on the overall cost per kWh Fig. 6. Environmental metamodel of the CC parameter for the three well depth scenarios: 5a well 1000 , 5b well 500 , 5c well 00. Z. Claudio and F. Daniele Applied Thermal Engineering 278 (2025) 127302 9