Experimental study of Lannea microcarpa seed oil as a heat transfer fluid or thermal energy storage material for medium-temperature applications

Research paper Experimental study of Lannea microcarpa seed oil as a heat transfer fluid or thermal energy storage material for medium-temperature applications Mahamadou Maiga , Kokouvi Edem N’Tsoukpoe * , Aboubakar Gomna Laboratoire ´ Energies Renouvelables et Efficacit´ e´ Energ´ etique (LabEREE), D´ epartement G´ enie ´ Electrique, ´ Energ´ etique et Industriel, Institut International d’Ing´ enierie de l’Eau et de l’Environnement (2iE), 01 BP 594, Ouagadougou 01, Burkina Faso ARTICLE INFO Keywords: Heat transfer fluid, Thermal energy storage material Vegetable oil Thermal stability Lannea microcarpa seed oil ABSTRACT In this study, we investigated the suitability of Lannea microcarpa seed oil (LaMSO) as a heat transfer fluid (HTF) and thermal energy storage material (TESM) for medium-temperature applications, focusing on its performance in concentrated solar power (CSP) plants. LaMSO demonstrated higher specific heat capacity than both Dowtherm A and Xceltherm 600 across the temperature range of 30 ◦C to 300 ◦C. It also exhibited a volumetric thermal energy storage density similar to, or approximately 10 % greater than, that of Dowtherm A and Xceltherm 600 at 210 ◦C. The oil was subjected to isothermal aging at 210 ◦C for 500 h and 1000 h, followed by analysis of its thermophysical and chemical properties. Thermogravimetric analysis demonstrated excellent thermal stability, with less than 3 % mass loss after 1000 h of aging. The density remained relatively unchanged, indicating consistent performance as a HTF or TESM. Despite an increase in viscosity at 40 ◦C, the viscosity at 100 ◦C remained comparable between the new and aged LaMSO samples. Furthermore, the flash point also showed no significant change, remaining high even after 1000 h of aging. However, challenges arise from the increase in the melting point due to oil saturation, which poses a risk of solidification in pipelines or tanks in certain climates. From the perspective of thermal storage cost per kWh, LaMSO is cheaper than Dowtherm A and Xceltherm 600, being approximately 8 times and 3 times less expensive, respectively. Overall, the findings suggest that LaMSO holds promise as an effective and sustainable alternative for HTF and TESM applications across a wide temperature range, contributing to advancements in renewable energy technology. 1. Introduction Thermal oils play crucial role in various industrial processes, including concentrated solar power (CSP) plants, where they serve as both heat transfer fluids (HTFs) and thermal energy storage materials (TESMs 1 ) (Al-Barqi et al., 2022). The selection of appropriate thermal oils is essential for ensuring the efficiency and reliability of these systems and involves multiple criteria, many of which overlap between HTFs and TESMs (N’Tsoukpoe et al., 2021). One key parameter influencing the performance of thermal oils is their thermal stability, which refers to their ability to withstand thermal degradation at elevated temperatures (Gomna et al., 2019). The genesis of this research project originated from challenges encountered during the early stages of constructing the CSP4Africa prototype, a small-scale CSP system designed for the Sahelian context with high local content (N’Tsoukpoe et al., 2016). From the local perspective and experience, the procurement process for diphenyl oxide/biphenyl, commonly known under the brand names Dowtherm A or Therminol VP-1, the predominant industrial oil used in CSP plants (National Renewable Energy Laboratory, 2024), has proven complex due to international logistics and bureaucratic hurdles, resulting in significant project delays spanning a year. To address these challenges and explore locally sourced alternatives, vegetable oils have emerged as viable options. However, conventional edible oils such as cottonseed oil, palm oil, coconut oil, and palm kernel oil were excluded to prevent competition with food production. Consequently, research efforts focused on locally available non-edible oils, including studies on Jatropha curcas crude oil (JaCCO) (Kenda et al., 2017) and refined Jatropha curcas oil (RJCO) (Gomna et al., 2020), have shown the * Corresponding author. E-mail addresses: [email protected],[email protected] (K.E. N’Tsoukpoe). 1 In addition to serving as a primary thermal energy storage material (TESM) for storing energy, many fluids also function as heat transfer agents, thereby fulfilling the role of a heat transfer fluid (HTF). For simplicity, throughout this paper, we use the term ’heat transfer fluid’(HTF) to refer to both heat transfer fluids and thermal energy storage materials. Contents lists available at ScienceDirect Energy Reports journal homepage: www.elsevier.com/locate/egyr https://doi.org/10.1016/j.egyr.2024.10.001 Received 1 May 2024; Received in revised form 17 September 2024; Accepted 1 October 2024 Energy Reports 12 (2024) 4129–4142 Available online 9 October 2024 2352-4847/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/bync/4.0/ ). potential of vegetable oils as cost-effective and sustainable substitutes for traditional thermal oils in medium-temperature applications within CSP plants. Subsequently, attention shifted to Lannea microcarpa seed oil (LaMSO) and Lannea kerstingii seed oil (LaKSO), two locally produced vegetable oils recognized for their excellent oxidative stability up to 120 ◦C (Bazongo et al., 2014; Ouilly et al., 2017; Lykke et al., 2021). These oils have emerged as promising candidates for further investigation because of their potential to meet the stringent requirements of CSP plants. Indeed, in response to the increasing demand for sustainable energy solutions, there has been a notable exploration of alternative TESMs and HTFs. Vegetable oils have emerged as promising candidates for these applications due to their renewable nature, environmentally friendly properties, non-hazardous characteristics, high availability, and lower cost compared to conventional thermal –mineral, synthetic –oils (Gomna et al., 2019). Studies have highlighted the potential of vegetable oils in medium-temperature applications, particularly in solar plants, where they offer some advantages over traditional thermal oils (Gomna et al., 2019; Kenda et al., 2017; Gomna et al., 2020; Hoffmann et al., 2018; Kahsay and Nydal, 2023; Abedigamba et al., 2023). These advantages include their high specific heat capacity, energy storage density, renewable resources, and relatively low flammability. Additionally, vegetable oils exhibit biodegradability, and relatively low toxicity, further enhancing their appeal for TES and HTF applications. Despite the promising attributes of vegetable oils, challenges remain, particularly concerning their oxidative and thermal stability, especially at mediumand high-temperature. The stability of vegetable oils is influenced by various factors, including temperature, duration of exposure to heat, and the presence of oxygen or moisture (Gomna et al., 2019). Although some research has delved into the suitability of vegetable oils for medium-temperature applications (Gomna et al., 2019; Kenda et al., 2017; Gomna et al., 2020; Hoffmann et al., 2018; Kahsay and Nydal, 2023; Abedigamba et al., 2023), there is limited investigation into their thermophysical properties and long-term performance, including LaMSO, a vegetable oil suggested as particularly stable (Bazongo et al., 2014; Lykke et al., 2021). Hence, further research is imperative to assess the feasibility of vegetable oils, such as LaMSO, as sustainable alternatives for medium-temperature applications, particularly in solar energy systems. This study investigates the thermophysical properties and thermal performance of LaMSO for TES applications, aiming to assess its suitability as a HTF and TESM for small-scale CSP operations, particularly at temperatures below 210 ◦C (Table 1). We aim to evaluate the key physicochemical properties of LaMSO to provide insights into its thermal behavior and ability to fulfill the selection criteria. Through this investigation, we aim to contribute to enhancing the sustainability and efficiency of solar energy systems and informing decision-making processes related to TESM selection. The temperature of 210 ◦C is justified by its alignment with the maximum operational temperature of one of the primary applications targeted in this investigation, namely, the CSP4Africa small-scale concentrating solar power plant (N’Tsoukpoe et al., 2016), where LaMSO is proposed to serve as a potential HTF and TESM. Many small-scale CSP plants operate at or below this temperature (Table 1), as most use an organic Rankine cycle (ORC), which is efficient within 150–250 ◦C (N’Tsoukpoe et al., 2016). However, the utilization of the oil extends beyond CSP applications, encompassing any industrial process that necessitates the use of HTF at temperatures up to 210 ◦C. By evaluating the thermophysical properties and long-term performance of vegetable oils, this study contributes to advancing the use of sustainable and cost-effective TESM for renewable energy technologies. The findings are expected to provide insights into the utilization of vegetable oils as alternative fluids in medium-temperature solar energy processes, including solar cookers, CSP plants, and comparable industrial processes. This will contribute to the development of sustainable energy solutions. Moreover, the findings of this study could have implications for enhancing the environmental sustainability and economic viability of CSP systems while promoting local economic development and energy access. 2. Material and methods 2.1. Lannea microcarpa seed oil extraction Lannea microcarpa, commonly referred to as the African grape, is widely distributed across West Africa (Arbonnier, 2009). While the oil extracted from Lannea microcarpa seeds can potentially serve as an edible oil (Bazongo et al., 2014; Lykke et al., 2021; Ou´ edraogo et al., 2013), the tree is valued primarily for its edible fruits, medicinal properties, honey production, dye extraction, and browsing material. Hence, LaMSO is considered an unconventional oil and is not commonly Table 1 Small-scale CSP projects around the world (N’Tsoukpoe et al., 2016; Seshie et al., 2018; National Renewable Energy Laboratory. n.d.a.). Small-scale CSP Project Technology HTF TESM Electric power [kW e ] Maximum operating temperature [◦C] METU NCC (N’Tsoukpoe et al., 2016) Parabolic trough ––18 107 Cleco solar plant (Raush et al., 2013) Parabolic trough Water No storage 15–50 120 Sun2Power (Dickes et al., 2018) Parabolic trough Pirobloc HTFBASIC Pirobloc HTFBASIC 2 150 Hybrid solar power system (Kane et al., 2003) Linear Fresnel Water –6.5 150 STG solar plant (Quoilin et al., 2011) Parabolic trough Monoethylene glycol Monoethylene glycol 3 160 Reelcoop (Willwerth et al., 2018) Parabolic trough Water Saturated steam 76 170 Sun2Power (N’Tsoukpoe et al., 2016; Seshie et al., 2018) Parabolic trough Syltherm XLT No storage 2.5 175 Holaniku at Keahole Point (Ashcroft and Rawlins, 2013; National Renewable Energy Laboratory. n.d.b.; Power technology, 2015) Parabolic trough Xceltherm 600 –2000 176 Microsol (COSTEA and din Bucuresti, 2012) Parabolic trough Water Water 10 180 LIPI (Pikra et al., 2013) Parabolic trough Palm oil Palm oil 10 200 CSP4Africa (N’Tsoukpoe et al., 2016) Power tower Jacco Jacco 8.6 210 Tresert (Ashcroft and Rawlins, 2013; Krüger et al., 2012) Parabolic trough Water Saturated steam 50 217 CSP Biomass plant (Ashcroft and Rawlins, 2013) Parabolic trough ––256 220 M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4130 consumed by the population (Lykke et al., 2021; Ou´ edraogo et al., 2013). A total of 56 kg of ripe fruits was harvested from various Lannea microcarpa trees in Sapon´ e (12◦06’00’’ N, 1◦36’00”W) in June 2022. Sapon´ e (Burkina Faso) is located in the Sudano–Sahelian zone and is characterized by annual rainfall ranging from 600 mm to 900 mm (Akoudjin et al., 2016). Fig. 1 shows the overall assessment of the oil extraction process from the LaMSO seeds. This process began by manually removing the pulp from the ripe fruits and cleaning them with water to obtain the nuts. The nuts were subsequently shelled to extract the seeds, which were ground into paste using a porcelain mortar and pestle. The Lannea microcarpa seed oil (LaMSO) was extracted using the Soxhlet method as outlined in ISO 734 (International Organization for Standardization, 2023), with the heating mantle set to a temperature of 150 ◦C for 6 h. Hexane was employed as the solvent for extraction. The overall extraction yield obtained was 19 g of LaMSO per kg of ripe fruits. The LaMSO obtained through this process was utilized for the present study. The small-scale production cost of LaMSO in this study is estimated at 6 € per kg of oil (see Appendix A). 2.2. Experimental procedure 2.2.1. Properties of LaMSO (untreated) 2.2.1.1. Thermophysical and chemical properties. The following thermophysical and chemical properties of LaMSO were determined using standardized methods: 2.2.1.1.1. Flash point. The flash point was measured using a closedcup Setaflash series 3 Plus analyzer (model 33000–0, STANHOPE–SETA) in accordance with ASTM D93. The measurement range of the device spans from ambient temperature to 300 ◦C, with an uncertainty of 10 ◦C. It is recalled that the flash point is the temperature at which the oil generates enough volatile substances to allow ignition (Gunstone, 2011). 2.2.1.1.2. Melting point. The melting point was determined by the capillary tube method following ASTM E324. It is recalled that the melting point refers to the temperature at which the oil changes from a solid state to a liquid state at atmospheric pressure. 2.2.1.1.3. Density. The density was measured at 25 ◦C and 40 ◦C using an Anton Paar DMA 4500 M densimeter, as per ASTM D4052. The measurement range of the device spans from 0 to 3000 kg m −3 . 2.2.1.1.4. Kinematic viscosity. The kinematic viscosity was measured from 40 ◦C to 100 ◦C using an OmniTek S–Flow 3000 viscometer following ASTM D7279. The measurement range of the device spans from 0.6 to 3000 mm 2 ⋅s –1 . 2.2.1.1.5. Thermal conductivity. The thermal conductivity was measured at 40 ◦C using a KD2 Pro thermal analyzer, employing the transient hot-wire technique (Ashcroft and Rawlins, 2013). Sensor KS-1, recommended by the device manual for high-viscosity fluids, was utilized for the measurement. The sensor was fully immersed in 30 mL of LaMSO contained in a borosilicate glass tube for approximately 3 min to facilitate accurate measurement of the thermal conductivity. The measurement range of sensor KS-1 spans from 0.02 to 2 W⋅m −1 ⋅K −1 . 2.2.1.1.6. Specific heat capacity. The specific heat capacity was determined directly by the application laboratory of Setaram (Caluire, France), using the C80 Calvet calorimeter with 3D sensors. First, the measurement cell, which was thoroughly cleaned and dried, was filled with the sample and weighed (5551.38 mg of LaMSO). An identical empty cell served as a reference to eliminate the thermal flux signal contribution from the cell. These cells were then inserted into the calorimeter and stabilized at 30 ◦C. The analysis of the samples was subsequently performed via the following steps: –Isothermal conditioning at 30 ◦C for 1 h –Heating from 30 ◦C to 300 ◦C at a rate of 0.4 K⋅min –1 –Isothermal holding at 300 ◦C for 4 h –Return to ambient temperature To ensure the reliability of the results, a second heating cycle was conducted to verify the reproducibility of the observed phenomena. Additionally, a test employing alumina powder was conducted to validate the results (serving as a reference for specific heat capacity measurements). The mass of alumina (9585.42 mg) used was adjusted to achieve a heat capacity (m⋅c p ) similar to that of the oil. The specific heat capacity in J⋅kg –1 ⋅K –1 was determined by Eq. (1) with an uncertainty of 1 %: cp=Ht−Hc ms⋅β(1) where Ht(W) is the total heat flow; Hc(W) is the heat flow contribution from the stainless-steel cell; ms(kg) is the mass of the LaMSO sample and β(K⋅s –1 ) is the heating rate. 2.2.1.1.7. Iodine value. The iodine value was determined by Wijs analytical method according to ISO 3961. The iodine value provides insight into the oxidative stability of a vegetable oil, by indicating its level of unsaturation 2 and its ability to be stored for longer periods (Gunstone, 2011; Vaitilingom, 2013). Saturated vegetable oils are more resistant to oxidation, relatively less viscous, and often solid at ambient temperature (Gunstone, 2011; Vaitilingom, 2013). On the other hand, unsaturated vegetable oils are more siccative (ability to dry at ambient temperature), less resistant to oxidation, and exhibit lower melting points (Gunstone, 2011; Vaitilingom, 2013). 2.2.1.1.8. Acid value. The acid value was determined by potentiometric titration according to ASTM D974. The acid value corresponds to the mass of potassium hydroxide (KOH) required to neutralize the free fatty acids contained in 1 g of vegetable oil (Gunstone, 2011; Vaitilingom, 2013). It provides information on the purity, free fatty acid content, and corrosiveness of the vegetable oil. A low acid value suggests that the oil is of better quality, with less risk of degradation and corrosion at mediumand high-temperatures (Gunstone, 2011; Vaitilingom, 2013). 2.2.1.1.9. Peroxide value. The peroxide value was determined by potentiometric titration according to the method presented by NFT 60–220. The peroxide value is used to assess the level of oxidation of a new (untreated) vegetable oil (Gunstone, 2011). It corresponds to the amount of peroxide (formed as a result of an oxidation process) present in the oil. However, the peroxide value is not considered a reliable indicator for assessing oxidation after a medium-temperature process (Mendonça et al., 2015; Pokorny et al., 2001). This is because peroxides appear when the oil begins to oxidize and generally decompose around 150 ◦C (Gunstone, 2011; Mendonça et al., 2015; Pokorny et al., 2001). Above this temperature, no accumulation of peroxides is observed in the oil. Apart from the specific heat capacity determination, which was conducted twice, all other thermophysical and chemical properties were measured three times. Data are presented as the mean ±standard deviation for all measured thermophysical and chemical properties, except for the flash point and the specific heat capacity, whose uncertainties were presented earlier (±10 ◦C for the flash point and ±1 % for the specific heat capacity). Eqs. (2) and (3) determine the statistical mean (X) of the measured values (X i ) and their standard deviation ( σ ) for a specific property, respectively. 2 Vegetable oils are characterized as (i) saturated when their iodine value is between 5 g I 2 ⋅100 g −1 and 50 g I 2 ⋅100 g −1 , (ii) monounsaturated when their iodine value is between 50 g I 2 ⋅100 g −1 and 100 g I 2 ⋅100 g −1 , (iii) diunsaturated when their iodine value is between 100 g I 2 ⋅100 g −1 and 150 g I 2 ⋅100 g −1 , and (iv) triunsaturated when their iodine value is higher than 150 g I 2 ⋅100 g-1 (Gunstone, 2011; Vaitilingom, 2013). M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4131 X=∑ 3 i=1 Xi 3(2) σ = ∑ 3 i=1 (Xi−X)2 3 √ √ √ √(3) 2.2.1.2. Thermogravimetric analysis. The thermal degradation of LaMSO was examined through thermogravimetric analysis (TGA) conducted under inert gas conditions. The SETARAM Setsys 2000 calorimeter equipped with an alumina cell was utilized for this purpose. Approximately 100 mg of the LaMSO sample was carefully deposited into the alumina cell. The sample subsequently underwent controlled heating from 25 ◦C to 600 ◦C at a rate of 10 K⋅min –1 under a nitrogen atmosphere. 2.2.1.3. Fourier transform infrared spectroscopy analysis. The functional groups of LaMSO were analyzed using Fourier transform infrared (FTIR) spectroscopy. The FTIR spectrum was obtained with a Bruker Alpha II compact spectrometer, equipped with a platinum attenuated total reflectance (ATR) accessory. Approximately 10 μ L of LaMSO was placed on the lens of the instrument, and the spectrum was recorded across a wavenumber range of 4000 cm⁻ 1 to 500 cm⁻ 1 , at a resolution of 2 cm⁻ 1 . Each result represents the average of 24 scans. 2.2.2. Isothermal aging and characterization of aged LaMSO 2.2.2.1. Isothermal aging of LaMSO. Currently, there are no specific standards for characterizing vegetable oils as HTF or TESM. However, ASTM (ASTM, 2020) offers a standard method (ASTM D6743) for analyzing the thermal stability of organic HTFs, which can be adapted for use with vegetable oils. This standard involves heating the fluid in a sealed 316 L stainless steel reactor at a specified temperature for a minimum of 500 h. Gas chromatographic analysis before and after the thermal stress helps identify degradation products resulting from the heating process. Conducting tests at various temperatures allows for the construction of degradation curves as a function of temperature. In this study, we adapted the ASTM D6743 standard to investigate LaMSO as both HTF and TESM. The experimental setup involved isothermally aging LaMSO at 210 ◦C for various durations in 30 mL borosilicate glass tubes. The choice of borosilicate glass tubes as test reactors was based on their thermal stability and chemical resistance at medium and high temperatures (Lima and Monteiro, 2001; AZoM, 2009), justifying their use in various studies on vegetable oil for similar experiments (Gomna et al., 2020; Akoh, 1994; Tam´ asi and Marossy, 2022; Borsodi et al., 2020). The choice of 210 ◦C is justified by its alignment with the operational temperature of one of the primary applications targeted in this investigation, as explained in Section 1. Introduction. The isothermal aging process commenced by heating the LaMSO from room temperature to 210 ◦C in an oven (Fig. 2). The LaMSO was subsequently maintained at this temperature for the specified duration and then allowed to cool naturally to room temperature. It is noteworthy that the oven displays the temperature of the heating room, so the aging duration effectively started to be counted when the oven reached the set temperature of 210 ◦C. For each isothermal aging test at 210 ◦C, three samples were prepared. The tests were conducted for durations of both 500 h and 1000 h. The nomenclature for the samples is outlined in Table 2. To minimize the presence of oxygen during the aging process, the test tubes were fully filled with LaMSO. Glass lids were then securely placed over the tubes in contact with LaMSO, and both the tubes and lids were wrapped with Teflon tape. This sealing method, which has been employed in numerous studies (Fitzgerald and Gill, 1979; Hossner, 2018; Yang et al., 2007), including those involving the heating of vegetable oils at 210 ◦C (Gomna et al., 2020), ensured a satisfactory seal at medium temperatures. 2.2.2.2. Characterization of aged LaMSO properties. After the isothermal aging process for a given duration, the properties of the aged LaMSO samples were determined. These assessments facilitated monitoring of the evolution of both the thermophysical and chemical properties of LaMSO with aging. All the samples were thoroughly characterized at the conclusion of the aging tests, and the resulting data are presented as the mean values along with their standard deviations. The determination of thermophysical and chemical properties, as well as thermogravimetric and Fourier transform infrared spectroscopy analyses, was conducted following the same procedures as those employed for the untreated LaMSO samples (2.2.1. Properties of LaMSO (untreated)). However, the peroxide value was not determined for the aged LaMSO sample, as it is not considered a reliable indicator for oil degradation assessment in medium-temperature applications (Gomna et al., 2019; Mendonça et al., 2015; Pokorny et al., 2001; Hoffmann, 2015). Fig. 1. Overall assessment of the oil extraction process from Lannea microcarpa seeds. Fig. 2. Representation of the thermal stress experienced by LaMSO during the isothermal aging test. Table 2 Nomenclature of the samples. Nomenclature Designation LaMSO LaMSO (untreated) LaMSO-500h Aged LaMSO for a duration of 500 h LaMSO-1000h Aged LaMSO for a duration of 1000 h M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4132 3. Results 3.1. Properties of LaMSO (untreated) 3.1.1. Thermophysical and chemical properties The specific heat capacity of LaMSO, as depicted in Fig. 3, increases progressively with temperature, rising from 2057±21 J⋅kg –1 ⋅K –1 to 2853 ±29 J⋅kg –1 ⋅K –1 within the range of 30 ◦C to 300 ◦C. A polynomial equation (Eq. (4)), was derived to model this trend, providing the specific heat capacity (c p ) as a function of temperature (T) [◦C] within the specified range. The equation is expressed as: cp[J•kg−1•K−1]= − 2.9669⋅10−5⋅T3+1.2851⋅10−2⋅T2+1.6541⋅T +1995.8 (4) The kinematic viscosity of LaMSO decreases from 30.3±0.1 mm 2 ⋅s –1 to 6.7±0.1 mm 2 ⋅s –1 in the temperature range of 40 ◦C to 100 ◦C, as presented in Fig. 4. The observed data were fitted using Eq. (5) to describe the change in kinematic viscosity ( η ) of LaMSO with temperature T [◦C] within the temperature range of 40 ◦C to 100 ◦C, yielding a coefficient of determination r 2 =0.999993: η [mm2•s−1]=exp (356953.45 (T+273.15)2+831.0166 T+273.15 −2.8849 )(5) The adopted form for viscosity, suggested as most appropriate for vegetable oils by Esteban et al. (2012), is employed. The other measured thermophysical and chemical properties of LaMSO are presented in Table 3. Assuming that the density changes linearly with temperature, as is commonly observed for most oils, including vegetable oils (Hoffmann et al., 2018), the density may be determined as a function of temperature T [◦C] by Eq. (6): ρ [kg •m−3] = − 0.5067⋅T+912.87 (6) 3.1.2. Thermal degradation The results of the thermogravimetric (TG) and derivative thermogravimetric (DTG) analyses of LaMSO are presented in Fig. 5. The mass loss was assimilated to the thermal degradation (ASTM, 2021; Garcia et al., 2007) of LaMSO and occurred in one step, starting at 130 ◦C with onset and offset points observed at 402 ◦C and 454 ◦C, respectively. The peak of degradation was observed at 432 ◦C, with 60 % mass loss. 3.2. Characterization of the aged LaMSO 3.2.1. Evolution of the thermophysical and chemical properties Fig. 6 presents the evolution of the thermophysical and chemical properties with aging. The flash point of LaMSO remained stable at 220±10 ◦C after 500 h of aging at 210 ◦C (Fig. 6a). However, after 1000 h of aging, it decreased to 185±10 ◦C. The acid value of LaMSO increased linearly from its initial value of 8.8 g KOH⋅g –1 to 12.2 g KOH⋅g –1 after 1000 h of aging at 210 ◦C, representing a 40 % augmentation (Fig. 6b) The iodine value of LaMSO decreased linearly from its initial value of 13.8 g I 2 ⋅100 g –1 to 6.1 g I 2 ⋅100 g –1 after 1000 h of aging at 210 ◦C, representing a 50 % reduction (Fig. 6c). The saturation of the aged LaMSO was even visually observable, as demonstrated by the increase in the melting point (Fig. 6d). Notably, the aged LaMSO samples tended to solidify at room temperature. The density of LaMSO showed a negligible change with aging, with only an approximately 2.5 % increase observed after 1000 h at 210 ◦C (Fig. 6e). The trend of kinematic viscosity as a function of temperature is depicted in Fig. 6f. The kinematic viscosities of LaMSO after 500 h and 1000 h of aging at 210 ◦C are very similar. However, the kinematic viscosity at 40 ◦C practically doubled after 500 h of aging at 210 ◦C. But, with increasing temperature, the kinematic viscosity of the aged LaMSO sample tended to decrease and approach that of the fresh LaMSO sample. 3.2.2. Thermal degradation of the aged samples The TG and DTG curves of both the new and aged LaMSO samples are depicted in Fig. 7a and b, respectively. Analysis reveals that the LaMSO sample aged for 500 h (LaMSO-500h) exhibits a TG profile closely resembling that of the new LaMSO sample. Like that of the new sample, its mass loss occurs in a single step. In contrast, the sample aged for 1000 h (LaMSO-1000h) shows a distinct behavior, with mass loss occurring in two steps: from 240 ◦C to 360 ◦C and from 374 ◦C to 490 ◦C. Additionally, all the samples exhibited a release of less than 3 % of volatile compounds at 210 ◦C (Fig. 7b). However, above 300 ◦C, a notable increase in volatile compound emission was observed for all samples, reaching approximately 50 % at 430 ◦C. The samples ultimately completely disappeared at 490 ◦C, following a convergent trend from 430 ◦C. Fig. 3. Specific heat capacity of LaMSO as a function of temperature. Fig. 4. Kinematic viscosity of LaMSO as a function of temperature. Table 3 Thermophysical and chemical properties of Lannea microcarpa seed oil. Thermophysical and chemical properties Value Flash point (◦C) 220 ±10 Melting point (◦C) 20.5 ±0.6 Density at 25 ◦C (kg⋅m –3 ) 900.2 ±0.1 Density at 40 ◦C (kg⋅m –3 ) 892.6 ±3.3 Thermal conductivity at 40 ◦C (W⋅m −1 ⋅K −1 ) 0.329 ±0.004 Iodine value (g I 2 ⋅100 g –1 ) 13.8 ±0.7 Acid value (mg KOH⋅g –1 ) 8.8 ±0.4 Peroxide value (mEq O 2 ⋅kg –1 ) 40.8 ±0.5 M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4133 The FTIR spectra of the new and aged LaMSO samples are depicted in Fig. 8. A detailed report of the FTIR analysis is provided in Appendix B. The band assignment used in this study is based on previous FTIR analyses of vegetable oils (Gomna et al., 2020; Rohman and Man, 2010; Vongsvivut et al., 2014; Guill´ en and Cabo, 1997; Timilsena et al., 2017). The results show a progressive change in the composition of LaMSO with aging, manifested by the slight variation in the absorption level in spectra (Fig. 8). In particular, the absorption peaks representing the triglycerides (1743 cm −1 ,1158 cm −1 , 1116 cm −1 and 1098 cm −1 ) (Gomna et al., 2020; Rohman and Man, 2010; Vongsvivut et al., 2014; Fig. 5. TG and DTG curves of LaMSO. Fig. 6. Evolution of the thermophysical and chemical properties of LaMSO: a) flash point, b) acid value, c) iodine value, d) melting point, e) density at 25 ◦C, and f) kinematic viscosity. M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4134 Guill´ en and Cabo, 1997; Timilsena et al., 2017) and the unsaturated fatty acids (3005 cm −1 et 720 cm −1 ) (Gomna et al., 2020; Rohman and Man, 2010; Vongsvivut et al., 2014; Guill´ en and Cabo, 1997; Timilsena et al., 2017) slightly decreased after 1000 h of aging, while the absorption peaks representing the methylene and ethyl groups (2921 cm −1 , 2851 cm −1 , 1463 cm −1 , and 1236 cm −1 ) (Rohman and Man, 2010; Vongsvivut et al., 2014; Guill´ en and Cabo, 1997; Timilsena et al., 2017), increased. Moreover, a spectral change in the range of 1720 cm −1 to 1700 cm −1 is progressively observed with aging (Fig. 8), while well-pronounced absorption peaks, indicative of alkene compounds (Gomna et al., 2020; Rohman and Man, 2010; Guill´ en and Cabo, 1997; Yuzhen et al., 2014), were observed at 965 cm −1 in the aged LaMSO samples. 4. Discussion 4.1. Properties of LaMSO In this section, the chemical properties of untreated LaMSO measured in the present study are discussed and compared with data from the literature (Table 4). This evaluation includes comparisons with other non-edible vegetable oils, namely RJCO (Gomna et al., 2020) and JaCCO (Kenda et al., 2017), diphenyl oxide/biphenyl (Dowtherm A) (The Dow Chemical Company, n.d), and Xceltherm 600 (Radco Industries, 2024). Dowtherm A, a widely used industrial synthetic organic oil in CSP plants, is included in the comparison due to its established presence and performance in the CSP industry (National Renewable Energy Laboratory, 2024). Xceltherm 600, a common mineral oil in CSP applications, is known for its high safety profile and cost-effectiveness, despite thermal stability limitations, which generally confine mineral oils use to temperatures below 300 ◦C (Wang et al., 2023). Notable applications of Xceltherm 600 in commercial CSP plants in the USA include its use in the Holaniku at Keahole Point CSP (Ashcroft and Rawlins, 2013; National Renewable Energy Laboratory. n.d.b.; Power technology, 2015), operating at temperatures below 180 ◦C (2 MW e ), and in the Saguaro Power Plant (1.16 MW e ) (Bilal Awan et al., 2020; National Renewable Energy Laboratory, n.d.c.) and Thermesol 2 (2 MW e ) (Bilal Awan et al., 2020), which operate up to 300 ◦C. Both the measured acid value and peroxide value of LaMSO are notably high, especially when compared to the values reported by Bazongo et al. (2014) (Table 4). This difference in chemical properties may be attributed primarily to the solvent used during the extraction process. Bazongo et al. (2014) used petroleum ether, which has a lower evaporation temperature (40 ◦C), compared to the hexane (69 ◦C) used Fig. 7. a) TG and b) DTG curves of all samples. Fig. 8. FTIR spectra of all samples. M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4135 in the present study. The elevated peroxide value suggests that the oil underwent pre-oxidation, likely during the extraction process (Choe and Min, 2006; Ndiaye et al., 2022), as certain peroxide compounds, such as hydroperoxides, begin forming within the temperature range of 80 ◦C to 90 ◦C. Notably, during the extraction of LaMSO in the current study, although the heating mantle was set to 150 ◦C, subsequent measurements revealed that the temperature range of the LaMSO and hexane mixture was between 73 ◦C and 96 ◦C. Moreover, the elevated acid value could be attributed to the exposure of the oil to oxygen or water vapor during the extraction process, leading to the formation of free fatty acids. In contrast, the iodine value decreases with oxidation, resulting in the low iodine value of LaMSO in the present study compared with that of Bazongo et al. (2014). The measured iodine value categorizes LaMSO as a saturated and non-drying oil (Hoffmann, 2015), indicating its potential for long-term storage without significant deterioration at room temperature. The TG curve of LaMSO (Fig. 5) is similar to those of various vegetable oils including sunflower oil (Abedigamba et al., 2023), Roki oil (Abedigamba et al., 2023), JaCCO (Kenda et al., 2017) and RJCO (Gomna et al., 2020). It shows that the thermal decomposition of the oil occurs at a temperature well above the level planned for the target application (210 ◦C). However, at 210 ◦C, a mass loss of 3 % has already been observed, suggesting the release of elements that could be the oil vapor or any other vapors, possibly including water or degradation products, due to oil preoxidation (Kenda et al., 2017). The latter is difficult to confirm with the FTIR spectrum of LaMSO (Fig. 8), which is also similar to that of various vegetable oil (Gomna et al., 2020; Rohman and Man, 2010; Vongsvivut et al., 2014; Guill´ en and Cabo, 1997; Timilsena et al., 2017). Indeed, in the band range of 1500 cm −1 to 1400 cm −1 , there are several negligible absorption peaks, not well pronounced, which could include the absorption peak of hydroperoxides that is generally observed at 1444 cm −1 (Guill´ en and Cabo, 1997). Further investigations could provide more insight into these phenomena. Among the vegetable oils, LaMSO stands out for its thermal conductivity at 40 ◦C, which is twice that of JaCCO (Table 4). Its other measured thermophysical properties, including viscosity, density, and specific heat capacity, are similar to those of other vegetable oils (Hoffmann et al., 2018; Kahsay and Nydal, 2023; Abedigamba et al., 2023; Hoffmann, 2015; Zongo et al., 2018), including non-edible oils such as JaCCO (Kenda et al., 2017) and RJCO (Gomna et al., 2020), which have been proposed for use as HTF and TESM in medium-temperature applications. However, JaCCO and RJCO have higher iodine values (Table 4), indicating greater unsaturation and susceptibility to oxidation. The relatively low acid value of RJCO is probably due to neutralization of the acidity of the oil used by Gomna et al. (2020). This was not the case for JaCCO (Kenda et al., 2017) and LaMSO, and it seems that the neutralization of JaCCO did not have a significant impact on its thermophysical properties when we compare the properties of JaCCO to those of RJCO. Therefore, the neutralization of LaMSO could be imagined if the acid value was seen as an issue. At the moment, the main advantage of JaCCO and RJCO over LaMSO is their availability in quantity due to the mastery and circular economy already present locally for their production. In short, the three oils are sustainable, renewable, environmentally friendly, and possess similar and relevant thermophysical properties for use as HTF or TESM, with LaMSO Table 4 Thermophysical and chemical properties of LaMSO (from the present study and others (Bazongo et al., 2014;Yunus et al., 2013), Dowtherm A, Xceltherm 600, RJCO and JaCCO. Thermophysical and chemical properties LaMSO (Present study) LaMSO ( Bazongo et al., 2014) LaMSO ( Yunus et al., 2013) Dowtherm A (The Dow Chemical Company, n.d; Thermal Fluids Hub, 2024) Xceltherm 600 ( Unoclean, 2024; Radco Industries, 2024) JaCCO (Kenda et al., 2017; Hoffmann et al., 2018; BELWET, 2024, 2024) RJCO (Gomna et al., 2020; BELWET, 2024, 2024) Flash point (◦C) 220±10 – – 113 193 220–240 220±10 Melting point (◦C) 20.5±0.6 22.60±0.75 –12 −29 3 – Density at 25 ◦C (kg⋅m −3 ) 900.2±0.1 – – 1055.54 851 – – Density at 40 ◦C (kg⋅m −3 ) 892.6±3.3 – – 1044 841 926 – Density at 210 ◦C (kg⋅m −3 ) 806.463 a – – 898 736 802 – Kinematic viscosity at 40 ◦C (mm 2 ⋅s −1 ) 30.3±0.1 –33.10±0.41 2.45 14.13 30–35 36.54±0.48 Kinematic viscosity at 100 ◦C (mm 2 ⋅s −1 ) 6.7±0.1 – – 1 2.24 6.45±0.65 7.97±0.20 Thermal conductivity at 40 ◦C (W⋅m −1 ⋅K −1 ) 0.329 ±0.004 – – 0.1355 0.1343 0.165±0.002 – Specific heat capacity at 40 ◦C (J⋅kg −1 ⋅K −1 ) 2081±21 – – 1630 2062 2150±81 2110±147 Specific heat capacity at 210 ◦C (J⋅kg −1 ⋅K −1 ) 2638±26 – – 2107 2643 2509 2375±166 Thermal storage capacity at 40 ◦C (kJ⋅m −3 ⋅K −1 ) b 1858±19 – – 1702 1734 1 991 – Thermal storage capacity at 210 ◦C (kJ⋅m −3 ⋅K −1 ) b 2127±21 a – – 1892 1947 2012 – Iodine value (g I 2 ⋅100 g −1 ) 13.8±0.7 61.33±0.25 – – – 107±4.8 101.93±0.98 Acid value (mg KOH⋅g −1 ) 8.8±0.4 1.21±0.01 –<0.2 –11 1.73±0.04 Peroxide value (mEq O 2 ⋅kg −1 ) 40.8±0.5 1.48±0.11 – – – 23±1.4 68.34±0.42 Unit cost ( € ⋅t −1 ) c 6000 – – 37026 c 15527 c 1220 1520 Cost of energy storage normalized per ◦C ( € ⋅◦C⋅kWh −1 ) d 8188 a – – 63262 21149 1750 2304 a Assuming the linear evolution of the density, as suggested by Eq. (6). b The thermal storage capacity is determined by multiplying the specific heat capacity by the density. c The cost of procuring oil, as observed from the perspective of Burkina Faso, which is fairly representative of Sahelian countries, is detailed in Appendix A. d This represents the cost per unit of thermal energy stored for a temperature variation of 1 ◦C. For any sensible thermal energy storage over a temperature range ΔT, the cost per unit of thermal energy stored ( € ⋅kWh −1 ) could be calculated by dividing the cost of energy storage normalized per ◦C ( € ⋅◦C⋅kWh −1 ) by ΔT. M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4136 being less susceptible to oxidation and more expensive. The thermophysical properties of LaMSO demonstrate its potential for use as both as HTF and TESM. At 40 ◦C, LaMSO exhibits a volumetric heat capacity of 1858±19 kJ⋅m −3 ⋅K −1 , which is about 10 % higher than that of Dowtherm A at the same temperature. Assuming that the density of LaMSO follows the linear trend proposed in Eq. (6), the volumetric heat capacity of LaMSO at 210 ◦C is calculated to be 2127 ±21 kJ⋅m −3 ⋅K −1 . This indicates that LaMSO offers a slightly higher volumetric thermal storage density than Dowtherm A within the temperature range relevant to the current study. This superiority is attributed mainly to LaMSO’s specific heat capacity, which surpasses that of Dowtherm A by 20 – 30 %. Furthermore, LaMSO presents a flash point that is 100 ◦C higher than that of Dowtherm A, thus, exhibiting a lower risk of fire or ignition in the presence of a heat source. Although the melting point Dowtherm A is well below that of LaMSO, this is not a significant concern given the context of the target application. Ambient temperatures rarely drop below 20 ◦C in the African intertropical region and do not persist below this level for extended periods. However, LaMSO exhibits higher kinematic viscosity than Dowtherm A in the temperature range of 40 ◦C to 100 ◦C. Consequently, more energy would be required to circulate the same volume of LaMSO compared with Dowtherm A, particularly during cold starts. Nevertheless, it is anticipated that the difference in viscosity decreases significantly with increasing temperature, as observed with other vegetable oils (Gomna et al., 2019; Kenda et al., 2017; Hoffmann et al., 2018; Zongo et al., 2018). This convergence leads to similar power consumption (N’Tsoukpoe et al., 2021), especially when considering that the circulated volume of LaMSO would be lower due to its higher volumetric energy storage density. In a nutshell: (i) LaMSO offers a thermal storage density similar to that of Dowtherm A. (ii) The flow rate required to transfer a given amount of heat with LaMSO is quite similar to, or even lower than, that of Dowtherm A. (iii) LaMSO stores and recovers thermal energy approximately 2.5 times faster than Dowtherm A at 40 ◦C, and this is expected to be roughly the case up to 210 ◦C, as with various vegetable oils (Abedigamba et al., 2023; Hoffmann et al., 2016). (iv) There is no risk of inflammability when LaMSO is used at temperatures up to 210 ◦C. (v) Despite its high viscosity at 40 ◦C, the power required to pump LaMSO may be similar to that of Dowtherm A at 210 ◦C. (vi) The risk of new LaMSO solidifying in pipelines or tanks at ambient temperature is very low in the Sahel region. (vii) The cost of procuring LaMSO is approximately 8 times lower than that of Dowtherm A in the context of Sahelian countries On the other hand, LaMSO exhibits similar specific heat capacity to that of Xceltherm 600 in the temperature range of 30 ◦C to 300 ◦C (Table 4). However, due to its higher density, the volumetric thermal energy storage density of LaMSO is similar to, or approximately 10 % greater than, that of Xceltherm 600 at 210 ◦C. The main advantages of Xceltherm 600 over LaMSO are its relatively low melting point (-29 ◦C) and low viscosity. The latter is 3 times lower than that of LaMSO at 100 ◦C, but both are expected to be similar at 210 ◦C (Hoffmann et al., 2018). Furthermore, LaMSO has the highest flash point, the highest thermal conductivity at 40 ◦C, and the lowest procurement cost compared to Xceltherm 600. Although this paper does not delve into the environmental aspects of using LaMSO, it is crucial to note that importing and using conventional oils like Dowtherm A in a country like Burkina Faso presents several environmental and sustainability challenges (Olmedo-Torre et al., 2018). The adoption of locally sourced vegetable oils, such as LaMSO, mitigates these issues and offers environmental, social, and economic benefits (N’Tsoukpoe et al., 2016). 4.2. Analysis of the thermal stability with aging Thermogravimetric analysis (TGA) conducted after 1000 h of isothermal aging at 210 ◦C revealed notable thermal stability, with less than 3 % of volatile compounds released by the samples at this temperature. However, significant differences were observed in the behavior of the aged samples during TGA, particularly above 300 ◦C, suggesting changes in the chemical composition of the oil and the formation of degradation products like alkene compounds, as evidenced by the well-pronounced peak at 965 cm −1 in FTIR spectra of the aged LaMSO samples (Fig. 8). This was also corroborated by alterations in various thermophysical and chemical properties, including the drop in flash point, oil saturation, and increases in the melting point, viscosity, and acid value. Despite the aging process, the density did not undergo a significant change, indicating minimal impact on the oil’s performance as a HTF or TESM, similar to observations with RJCO (Gomna et al., 2020). Despite the increase in kinematic viscosity at low temperature with aging, the viscosity of the aged oil samples remains close to that of the new oil sample at 100 ◦C and is expected to be quite similar at higher temperatures, such as 210 ◦C. Similar observations were made for RJCO (Gomna et al., 2020) with aging and many other vegetable oils, as reported by Gomna et al. (2019). The increase in viscosity with aging is the consequence of the degradation of peroxides into polymeric materials in the oil (Choe and Min, 2006, 2007), since peroxides appear when the oil begins to oxidize and generally decompose around 150 ◦C (Gomna et al., 2019). Above this temperature, no accumulation of peroxides is observed in the oil. The flash point remained relatively high (185±10 ◦C) even after 1000 h of aging and is still above the flash point of Dowtherm A (113 ◦C). This decrease suggests the breakdown of triglycerides into free fatty acids due to hydrolysis, oxidation, or thermal degradation, which is consistent with the observed increase in acid value with aging (Wang, 2011) (Fig. 6b). This was corroborated by the decrease in absorption peaks representing the triglycerides (1743 cm −1 ,1158 cm −1 , 1116 cm −1 and 1098 cm −1 ) with aging (Fig. 8), while the band range (from 1720 cm −1 to 1700 cm −1 ) representing the free fatty acids (Guill´ en and Cabo, 1997) recorded a slight increase in absorption. Previous studies (Gomna et al., 2019; Kenda et al., 2017; Hoffmann, 2015) on vegetable oils have also reported a drop in flash point with aging, whereas the acid value increases with aging. The increase in the melting point (38.8±0.6 ◦C at the end of 1000 h of aging) is attributed to oil saturation, confirmed by the decrease in the iodine value (Gomna et al., 2019; Hoffmann, 2015; Karak, 2012), and might be explained by the migration of unsaturated fatty acids to saturated fatty acids (Hoffmann, 2015) when we already know that LaMSO contains more than 50 % unsaturated fatty acids (Bazongo et al., 2014). This was corroborated by the increase with aging in absorption peaks representing the methylene and ethyl groups (2921 cm −1 , 2851 cm −1 , 1463 cm −1 , and 1236 cm −1 ), which probably include the ones that form the saturated fatty acids contained in the oil (Bazongo et al., 2014; Vongsvivut et al., 2014; Guill´ en and Cabo, 1997). The oil saturation with aging also partly explains the increase in viscosity with aging at low temperatures. While the specific heat capacity and thermal conductivity of the aged oil were not directly measured, they are expected to be similar to those of the new LaMSO, on the basis of the typical behavior observed in vegetable oils (Gomna et al., 2019; Hoffmann et al., 2018; Abedigamba et al., 2023; Hoffmann, 2015; Hoffmann et al., 2016). Therefore, LaMSO is anticipated to exhibit a consistent specific heat capacity and thermal conductivity even after aging, as demonstrated in studies on RJCO (Gomna et al., 2020) and colza oil (Hoffmann, 2015), respectively. M. Maiga et al. Energy Reports 12 (2024) 4129–4142 4137