| Title | Application of Artificial Neural Network to Predict Cooling Tower’s Performance at Lahendong Unit 5&6 Geothermal Power Plant |
|---|---|
| Authors | Fikrul Jihad |
| Year | 2026 |
| Conference | World Geothermal Congress |
| Keywords | Artificial neural network, cooling tower’s performance, Lahendong, Geothermal Power Plant |
| Abstract | This paper describes an application of an artificial neural network to predict the cooling tower’s performance at Lahendong Unit 5&6 Geothermal Power Plant. The cooling tower is one of the many critical equipment of a GPP. Its purpose is to cool hot water that exits the condenser. To ensure the optimal efficiency of the cooling tower, PGE Lahendong regularly does scheduled performance tests with a certified third-party company. The ANN method had been used to make a model and simulate the cooling tower’s performance. The input process for the ANN model was designed using five parameters, such as hot water temperature, cold water temperature, dry bulb temperature , wet bulb temperature, and water flow to the inlet cooling tower. The data utilized for the training of the ANN was taken from the early operation of the GPP and the present data. We also take the measurement data based on CTI-ATC-105 taken by a third-party company for model generation and analysis. The predictive capability of the model was evaluated in terms of root mean square error. It was found that the ANN model can predict the cooling tower’s performance with good results. The ANN provides an effective prediction tool to simulate the cooling tower’s performance that can be used as a valuable decision tool for better operation and maintenance of the cooling tower. |