Prediction Intervals of Cooling System Temperatures Monitoring in Hydropower Generators Using Pretrained LSTM and Sobol Sensitivity Analysis

Energy Engineering Program, Faculty of Engineering and Graduate School, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand
Email: chinachote.dee@egat.co.th (C.D.); tanongkiat_k@yahoo.com (T.K.); thoranisdee@gmail.com (T.D.); attakorn.asana@cmu.ac.th (A.A.)
*Corresponding author
Manuscript received December 10, 2025; accepted February 13, 2026; published April 30, 2026

Abstract—This study presents a data-driven framework that integrates a pretrained Long Short-Term Memory (LSTM) model with Sobol Global Sensitivity Analysis to forecast thermal behavior in hydropower generator cooling systems. The model was developed using real operational data from a hydropower plant in Thailand and demonstrated the capability to predict four key temperature variables: inlet air temperature, outlet air temperature, stator temperature, and outlet water temperature, under both normal and abnormal operating conditions. Monthly Prediction Intervals (PIs) were constructed using uncertainty sampling guided by Sobol analysis, capturing operational variability such as reductions in cooling water flow and elevated inlet water temperatures. Simulation results indicated that the combined approach effectively replicated thermal anomalies observed in historical data, particularly during the dry season. Each thermal variable exhibited distinct sensitivity profiles, especially stator temperature and both inlet and outlet air temperatures which responded critically on the heat exchange performance. These findings highlight the potential for implementing seasonally adaptive maintenance strategies and demonstrate the proposed framework’s practical applicability for early anomaly detection and predictive maintenance planning in hydropower operations.

Keywords—thermal forecasting, Long Short-Term Memory (LSTM) model, sobol global sensitivity analysis, hydropower generators

Cite: Chinachote Deevijit, Tanongkiat Kiatsiriroat, Thoranis Deethayat, and Attakorn Asanakham, "Prediction Intervals of Cooling System Temperatures Monitoring in Hydropower Generators Using Pretrained LSTM and Sobol Sensitivity Analysis," International Journal of Sustainability in Energy and Environment, vol. 3, no. 1, pp. 29-33, 2026.

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited ( CC-BY-4.0).