Thermal conductivity of bentonite-sand/fly ash composites for underground thermal infrastructure using neural network and ensemble tree-based machine learning approaches

Abstract

The thermal conductivity of bentonite-based mixtures significantly influences the performance and life-span of underground thermal energy systems and radioactive waste geological repository. The Machine Learning (ML) predictive models provide precise and accurate estimation of soil’s thermal conductivity over costly measurements. This investigation evaluates the predictive performance of multilayer perception structures, adaptive neuro-fuzzy architectures and regularized gradient boosting framework for evaluating heat transmission matrices for bentonite-sand (B-S) and bentonite-fly ash (B-F) mixes. The studied input (influencing) parameters included dry density, water content, particle compositions (sand, silt and clay), quartz content, and plasticity index. To perform ML, total 216 dataset of thermal conductivity of B-S and B-F mixtures were measured using KD-2 pro dual probe. The XGBoost model predicts thermal conductivity more accurately than ANN and ANFIS with overall R² = 0.996, MAE = 0.0113 and RMSE = 0.0246. Furthermore, the study involved a comparison of the proposed ML models with the four different empirical equations for thermal conductivity. It can also be noticed that ML techniques are more superior to the empirical equations for the present dataset.

Keywords

Bentonite-sand-fly ash / Thermal conductivity / ANN / ANFIS / XGBoost / Statistical comparison

References

[1] Zhang XR, Kong GQ, Chen YH, Chen L, Chen G. Measurement and prediction of the thermal conductivity of fused quartz in the range of 5–45 ℃. Int J Thermophys, 2021, 42(81-21

[2] Park K, Lee J, Yoon HK, Kim D. Hydraulic and thermal conductivities of kaolin–silica mixtures under different consolidation stresses. Mar Georesour Geotechnol, 2016, 34(6): 532-541

[3] Yoon S, Kim MJ, Park S, Kim GY. Thermal conductivity prediction model for compacted bentonites considering temperature variations. Nucl Eng Technol, 2021, 53(10): 3359-3366

[4] Patwa D, Chandra A, Ravi K, Sreedeep S. Influence of biochar particle size fractions on thermal and mechanical properties of biochar-amended soil. J Mater Civ Eng, 2021, 33(9): 1-15

[5] Wang H, Cui Y, Qi C. Effects of sand–bentonite backfill materials on the thermal performance of borehole heat exchangers. Heat Transf Eng, 2013, 34(1): 37-44

[6] Dong Y, McCartney JS, Lu N. Critical review of thermal conductivity models for unsaturated soils. Geotech Geol Eng, 2015, 33(2): 207-221

[7] He H, Li M, Dyck M, Si B, Wang J, Lv J. Modelling of soil solid thermal conductivity. Int Commun Heat Mass Transf, 2020, 116: 104602

[8] Yoon S, Kim MJ, Lee SR, Kim GY. Thermal conductivity estimation of compacted bentonite Buffer materials for a high-level radioactive waste repository. Nucl Technol, 2018, 204(2): 213-226

[9] Wang J, He H, Li M, Dyck M, Si B, Lv J. A review and evaluation of thermal conductivity models of saturated soils. Arch Agron Soil Sci, 2021, 67(7): 974-986

[10] Abu-Hamdeh NH. Thermal properties of soils as affected by density and water content. Biosyst Eng, 2003, 86(1): 97-102

[11] Zhang N, Zou H, Zhang L, Puppala AJ, Liu S, Cai G. A unified soil thermal conductivity model based on artificial neural network. Int J Therm Sci, 2020, 155: 106414

[12] Bruno AW, Alamoudi D. A simple thermal conductivity Model for unsaturated geomaterials accounting for degree of saturation and dry density. Int J Geosynth Ground Eng, 2020, 6(4): 1-7

[13] Zhang N, Wang Z. Review of soil thermal conductivity and predictive models. Int J Therm Sci, 2017, 117: 172-183

[14] Bruno AW, Najdi A, Balzano B. A thermal conductivity Model for deformable and unsaturated soils to assess the thermal behaviour of energy piles. Int J Geosynth Gr Eng, 2023, 9(5): 1-12

[15] Zhao Y, Si B, Zhang Z, Li M, He H, Hill RL. A new thermal conductivity model for sandy and peat soils. Agric Meteorol, 2019, 274: 95-105

[16] Jia GS, Tao ZY, Meng XZ, Ma CF, Chai JC, Jin LW. Review of effective thermal conductivity models of rock-soil for geothermal energy applications. Geothermics, 2019, 77: 1-11

[17] Johansen OE (1975) Thermal conductivity of soils. Corps Eng US Army, cold Reg Res Eng Lab PhD Thesis, 637: 1977, ADA 044002

[18] Côté J, Konrad JM. A generalized thermal conductivity model for soils and construction materials. Can Geotech J, 2005, 42(2): 443-458

[19] Lu S, Ren T, Gong Y, Horton R. An improved Model for predicting soil thermal conductivity from water content at room temperature. Soil Sci Soc Am J, 2007, 71(1): 8-14

[20] Nikoosokhan S, Nowamooz H, Chazallon C. Effect of dry density, soil texture and time-spatial variable water content on the soil thermal conductivity. Geomech Geoengin, 2016, 11(2): 149-158

[21] He H, Dyck M, Lv J. A new model for predicting soil thermal conductivity from matric potential. J Hydrol, 2020, 589: 125167

[22] Barry-Macaulay D, Bouazza A, Wang B, Singh RM. Evaluation of soil thermal conductivity models. Can Geotech J, 2015, 52(11): 1892-1900

[23] Tarnawski VR, McCombie ML, Leong WH, Coppa P, Corasaniti S, Bovesecchi G (2018) Canadian field soils IV: modeling thermal conductivity at dryness and saturation. Int J Thermophys 39(3). https://doi.org/10.1007/s10765-017-2357-9

[24] Liu L, He H, Dyck M, Lv J. Modeling thermal conductivity of clays: a review and evaluation of 28 predictive models. Eng Geol, 2021, 288: 106107

[25] Zhang N, Yu X, P A, Puppala AJ. A new generalized soil thermal conductivity model for sand–kaolin clay mixtures using thermo-time domain reflectometry probe test. Acta Geotech, 2017, 12(4): 739-752

[26] He H, He D, Jin J, Smits KM, Dyck M, Wu Q, Si B, Lv J. Room for improvement: a review and evaluation of 24 soil thermal conductivity parameterization schemes commonly used in land-surface, hydrological, and soil-vegetation-atmosphere transfer models. Earth-Sci Rev, 2020, 211: 103419

[27] Wang Z, Zhang N, Ding J, Li Q, Xu J. Thermal conductivity of sands treated with microbially induced calcite precipitation (MICP) and model prediction. Int J Heat Mass Transf, 2020, 147: 118899

[28] Peng F, Tan Y, Sun D. Thermal conductivity of bentonite-graphite mixture and its prediction for high-level radioactive waste repository. Ann Nucl Energy, 2021, 154: 108142

[29] Baghbani A, Choudhury T, Costa S, Reiner J. Application of artificial intelligence in geotechnical engineering: a state-of-the-art review. Earth-Sci Rev, 2022, 228: 103991

[30] Zhang W, Li H, Li Y, et al.. Application of deep learning algorithms in geotechnical engineering: a short critical review, 2021, Netherlands, Springer

[31] Taherdangkoo R, Tyurin V, Shehab M, Doulati Ardejani F, Minh Tang A, Narmandakh D, Butscher C. An efficient neural network model to determine maximum swelling pressure of clayey soils. Comput Geotech, 2023, 162: 105693

[32] Shehab M, Taherdangkoo R, Butscher C. Towards reliable barrier systems: a constrained XGBoost Model coupled with Gray Wolf optimization for maximum swelling pressure of bentonite. Comput Geotech, 2024, 168: 106132

[33] Hanandeh S, Ardah A, Abu-Farsakh M. Using artificial neural network and genetics algorithm to estimate the resilient modulus for stabilized subgrade and propose new empirical formula. Transp Geotech, 2020, 24: 100358

[34] Jangir H, Rupali S. in: Venkata Rao, R, Taler, J (eds.) (2023) Evaluation of adaptive neuro-fuzzy inference System for forecasting Soil’s unconfined compressive strength. Springer Nature Singapore, Singapore, pp 377–387

[35] Wang C, Cai G, Liu X, Wu M. Prediction of soil thermal conductivity based on intelligent computing model. Heat Mass Transf, 2022, 58(10): 1695-1708

[36] Rizvi ZH, Zaidi HH, Akhtar SJ, Sattari AS, Wuttke F. Soft and hard computation methods for estimation of the effective thermal conductivity of sands. Heat Mass Transf, 2020, 56(6): 1947-1959

[37] Li KQ, Liu Y, Kang Q. Estimating the thermal conductivity of soils using six machine learning algorithms. Int Commun Heat Mass Transf, 2022, 136: 106139

[38] Zhao T, Liu S, Xu J, He H, Wang D, Horton R, Liu G. Comparative analysis of seven machine learning algorithms and five empirical models to estimate soil thermal conductivity. Agri For Meteorol, 2022, 323: 109080

[39] Liu W, Li R, Wu T, Shi X, Zhao L, Wu X, Hu G, Yao J, Xiao Y, Ma J, Jiao Y, Wang S. Simulation of soil thermal conductivity based on different schemes: an empirical comparison of 13 models. Int J Therm Sci, 2023, 190: 108301

[40] Ibraheem S, Al S, Iskandar H (2023) A comparative study of MARS, GEP, and GMDH methods for modeling soil thermal conductivity. Int J Thermophys. https://doi.org/10.1007/s10765-023-03215-0

[41] Wang X, Gao Y, Hou J, Yang J, Smits K, He H. Machine learning facilitates connections between soil thermal conductivity, soil water content, and soil matric potential. J Hydrol, 2024, 633: 130950

[42] Zhang T T, Wang C-J, Liu S-Y, Zhang N, Zhang T-W. Assessment of soil thermal conduction using artificial neural network models. Cold Reg Sci Technol, 2020, 169: 102907

[43] Yuan X, Xue X. Prediction of soil thermal conductivity using artificial intelligence approaches. Geothermics, 2023, 113: 102769

[44] Mustafa R, Kumari K, Kumari S, Kumar G, Singh P (2024) Probabilistic analysis of thermal conductivity of soil. Arab J Geosci 17(1). https://doi.org/10.1007/s12517-023-11831-1

[45] Malek K, Malek K, Khanmohammadi F. Response of soil thermal conductivity to various soil properties. Int Commun Heat Mass Transf, 2021, 127: 105516

[46] Lee GJ, Yoon S, Kim T, Chang S. Investigation of the various properties of several candidate additives as buffer materials. Nucl Eng Technol, 2023, 55(3): 1191-1198

[47] Kong D, Wan R, Chen J, Kang J, Jiao X. Effect of gradation on the thermal conductivities of backfill materials of ground source heat pump based on loess and iron tailings. Appl Therm Eng, 2020, 180: 115814

[48] Dinh BH, Kim Y-S, Kang G-O. Thermal conductivity of steelmaking slag-based controlled low-strength materials over entire range of degree of saturation: a study for ground source heat pump systems. Geothermics, 2020, 88: 101910

[49] Do TM, Do AN, Kang GO, Kim YS. Utilization of marine dredged soil in controlled low-strength material used as a thermal grout in geothermal systems. Constr Build Mater, 2019, 215: 613-622

[50] Xu Y, Yao Z, Huang X, Fang Y, Shu S, Zhu H. Heat transfer characteristics and heat conductivity prediction model of waste steel slag–clay backfill material. Therm Sci Eng Prog, 2023, 46: 102203

[51] Yi F, Gao J, Qi X. Influence of granite powder addition on thermal conductivity of bentonite-based backfill material in ground source heat pump. Geothermics, 2023, 107: 102613

[52] Bang HT, Yoon S, Jeon H. Application of machine learning methods to predict a thermal conductivity model for compacted bentonite. Ann Nucl Energy, 2020, 142: 107395

[53] Da T-X, Chen T, He W-K, Elshaikh T, Ma Y, Tong Z-F. Applying machine learning methods to estimate the thermal conductivity of bentonite for a high-level radioactive waste repository. Nucl Eng Des, 2022, 392: 111765

[54] Mishra PN, Surendran S, Gadi VK, Joseph RA, Arnepalli DN. Generalized approach for determination of thermal conductivity of Buffer materials. J Hazard, Toxic, Radioact Waste, 2017, 21(4): 1-11

[55] Xiuling R, Fujun N, Qihao Y, Guoan Y. Research progress of soil thermal conductivity and its predictive models. Cold Reg Sci Technol, 2024, 217: 104027

[56] Cakiroglu C, Batool F, Islam K, Nehdi ML. Explainable ensemble learning predictive model for thermal conductivity of cement-based foam. Constr Build Mater, 2024, 421: 135663

[57] Fan H, Hang T, Song Y, Liang K, Zhu S, Fan L. Assessment of small strain modulus in soil using advanced computational models. Sci Rep, 2023, 13(1): 22476

[58] Erzin Y, Gumaste SD, Gupta AK, Singh DN. Artificial neural network (ANN) models for determining hydraulic conductivity of compacted fine-grained soils. Can Geotech J, 2009, 46(8): 955-968

[59] Li KQ, Kang Q, Nie JY, Huang XW. Artificial neural network for predicting the thermal conductivity of soils based on a systematic database. Geothermics, 2022, 103: 102416

[60] Vadyala SR, Betgeri SN, Matthews JC, Matthews E. A review of physics-based machine learning in civil engineering. Results Eng, 2022, 13: 100316

[61] Erzin Y, Rao BH, Singh DN. Artificial neural network models for predicting soil thermal resistivity. Int J Therm Sci, 2008, 47(10): 1347-1358

[62] Ansari MA, Roy LB. Experimental and ANN-based model of footing pressure of a layered soil reinforced with geogrid. Innovative Infrastruct Sol, 2023, 8(3): 100

[63] Takagi T, Sugeno M. Fuzzy identification of systems and its applications to modeling and control. IEEE Trans Syst Man Cybern SMC-, 1985, 15(1): 116-132

[64] Yashawantha KM, Vinod AV. ANFIS modelling of effective thermal conductivity of ethylene glycol and water nanofluids for low temperature heat transfer application. Therm Sci Eng Prog, 2021, 24: 100936

[65] Chen T, Guestrin C (2016) Xgboost. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ACM, New York, NY, USA, 785–794

[66] Chen Y, Xu Y, Jamhiri B, Wang L, Li T. Predicting uniaxial tensile strength of expansive soil with ensemble learning methods. Comput Geotech, 2022, 150: 104904

[67] Decagon Devices, I (2016) KD2 pro thermal properties analyzer 1–71

[68] Sah PK, Kumar SS. Thermal conductivity and shrinkage characteristics of bentonite-fly ash and bentonite-sand backfill material. Granul Matter, 2025, 27(1): 1-13

[69] Sah PK, Sreedeep S. Evaluation of bentonite-based thermal backfill materials. Environ Geotech, 2014, 1(3): 179-188

[70] Sah PK, Kumar SS, Sreedeep S. Thermal conductivity and specific volume heat capacity of bentonite–fly ash-based fluidized thermal backfill. J Therm Anal Calorim, 2023, 148(21): 11607-11617

[71] Sah PK, Kumar SS, Sreedeep S. Thermophysical properties of bentonite–Sand/Fly ash-based backfill materials for underground power cable. Int J Thermophys, 2023, 44(4): 57

[72] Sah PK, Kumar SS. Thermal diffusivity of bentonite–sand based engineered backfill material. Sādhanā, 2024, 49(1): 30

[73] Tsui APM, Jones AJ, Guedes de Oliveira A. The construction of smooth models using irregular embeddings determined by a gamma test analysis. Neural Comput Appl, 2002, 10(4): 318-329

[74] Nil N, Baranwal A, Das BS. Clear-water and live-bed scour depth modelling around bridge pier using support vector machine. Can J Civ Eng, 2023, 50(6): 445-463

[75] Wang C, Cai G, Wu M, Zhao Z. Prediction of soil thermal conductivity based on multivariate probability distribution models. Int Commun Heat Mass Transf, 2022, 138: 106355

[76] Tang AM, Cui YJ, Le TT. A study on the thermal conductivity of compacted bentonites. Appl Clay Sci, 2008, 41(3–4): 181-189

[77] Sah PK, Kumar SS. Assessment of thermal conductivity prediction models for compacted bentonite-based backfill material. Int J Environ Sci Technol, 2025, 22(6): 4571-4582

[78] Sandilya SS, Das BS, Proust S, Shekhar D. Discharge estimation in compound channels with converging and diverging floodplains using an optimised gradient boosting algorithm. J Hydroinf, 2024, 26(5): 1122-1149

[79] Nguyen MH, Mai HVT, Trinh SH, Ly HB. A comparative assessment of tree-based predictive models to estimate geopolymer concrete compressive strength. Neural Comput Appl, 2023, 35(9): 6569-6588

[80] Pratap B, Shubham K, Mondal S, Rao BH. Exploring the potential of neural network in assessing Mechanical properties of geopolymer concrete incorporating fly ash and phosphogypsum in pavement applications. Asian J Civ Eng, 2023, 24(8): 3575-3589

[81] Mahmoud M, Ramadan M, Pullen K, Abdelkareem MA, Wilberforce T, Olabi A-G, Naher S. A review of grout materials in geothermal energy applications. Int J Thermofluids, 2021, 10: 100070

[82] Javadi H, Ajarostaghi SSM, Rosen MA, Pourfallah M. A comprehensive review of backfill materials and their effects on ground heat exchanger performance. Sustain, 2018, 10(12): 4486

Rights & permissions

The Author(s)