Predicting the geothermal gradient in Colombia: A machine learning approach

Geothermics, Journal Year: 2024, Volume and Issue: 122, P. 103074 - 103074

Published: June 7, 2024

Accurately determining the geothermal gradient is crucial for assessing energy potential.In Colombia, despite an abundance of theoretical resources, large regions country lack measurements.This study introduces a machine learning approach to estimate in where only global-scale geophysical datasets and course geological knowledge are available.We find that Gradient-Boosted Regression Tree algorithm yields optimal predictions extensively validates trained model, obtaining our model within 12% accuracy.Finally, we present map Colombia serve as indicator potential further exploration data collection.This displays values ranging from 16.75 41.20°C/km shows significant agreement with indicators activity, such faults thermal manifestations.Additionally, results consistent independent findings other researchers specific regions, which supports reliability approach.

Language: Английский

---

A case study on thermal conductivity characteristics and prediction of rock and soil mass at a proposed ground source heat pump (GSHP) site

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 8, 2025

Shallow geothermal energy (SGE) has a wide range of applications in the field building cooling and heating. Ground source heat pump (GSHP) system is technology to extract SGE. The design borehole exchanger (BHE) great impact on transfer performance investment cost, so it important accurately measure thermal conductivity rock soil. Therefore, this study conducted in-situ response test (TRT) laboratory sample based distributed optical fiber temperature sensor (DOFTS) LY research area Changchun, Northeast China. After comparing differences analyzing reasons, an prediction model was established artificial neural network (ANN) algorithm predict basic physical property parameters tests. This used supplement layered lacking CY area. results show that can be obtained calculated by improved combined (ICTRT). average layer about 12.2% lower than test, but two methods same variation trend along depth. mass positively correlated with water content, negatively porosity density. result error calculation mainly within ± 5%, which reliable accurate. CY01 hole. provide new way determine SGE exploration.

Language: Английский

---

Optimizing Sustainable Power Generation with Triplet Deep Borehole Heat Exchangers: A Machine Learning Approach

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 1, 2025

Geothermal energy, a renewable and sustainable resource, has significant potential for meeting global energy demands; most of the study on production generation relies numerical simulation. However, computational intensity physics-based simulations geothermal poses challenges. This explores integration machine learning models with simulation to forecast long-term electricity from triplet deep borehole heat exchanger system. A large dataset generated through COMSOL Multiphysics served as input three models: Decision Tree, XGBoost, Random Forest. The Forest model outperformed others, achieving lowest error metrics Root Mean Square Percentage Error (RMSPE) 0.104, Absolute (MAPE) 0.0539, highest R² value 0.9996. These indicate that RF provides exceptional prediction accuracy generalization capabilities. combined approach significantly reduced time required, enabling forecasting an additional 15 years power using Forest, which makes it easier faster than waiting almost 21 hours before simulating 25 years. results confirm viability optimizing forecasting, ensuring sustainability operational efficiency in generation.

Language: Английский

---

Predicting thermal conductivity of granite subjected to high temperature using machine learning techniques

Environmental Earth Sciences, Journal Year: 2025, Volume and Issue: 84(8)

Published: April 1, 2025

Language: Английский

---

Utilizing the BP Neural Network Model for Comprehensive Assessment of Physical Fitness among College Students

Published: April 12, 2024

Language: Английский

---

A generalized thermal conductivity model of soil-rock mixture based on freezing characteristic curve

Cold Regions Science and Technology, Journal Year: 2024, Volume and Issue: 229, P. 104360 - 104360

Published: Nov. 10, 2024

Language: Английский

---