Optimisation of Cooling Tower for different operating conditions using ANFIS

Abstract

The conventional approach to cooling towers is strenuous and time-consuming as it is tedious to design an efficient cooling tower that works in a wide range of operating conditions. Therefore, the present work aimed at employing an Adaptive Neuro-Fuzzy Inference System (ANFIS), an Artificial Intelligence tool, to optimise the thermal performance of an induced draft cross-flow cooling tower for various working conditions. For this purpose, the water outlet temperature of the cooling tower was estimated and validated with experimental data considered from the literature. Inlet dry bulb temperature and relative humidity of localised climatic conditions, air mass flow rate, water flow rate, and water inlet temperature considered input parameters. For training and testing the model, a Subtractive Clustering approach was utilised. Four ANFIS models with varying cluster radius from 0.5-0.8 constructed and the obtained results compared using the RMSE, MAPE and R2 measurements. The predictions were in good agreement with the data published in the literature with a reasonable level of accuracy. By considering heat load as a performance parameter, the effect of inlet parameters on the heat load of the cooling tower was investigated using a trained ANFIS model in detail. It is observed that the water inlet temperature has a substantial effect on the heat load. The following influencing parameters on the heat load of the cooling tower were relative humidity, air inlet dry bulb temperature, and water mass flow rate, in descending order. The mass flow rate of air had the most negligible impact on the heat load. The optimal working confiditon found to be 42°C intake air and 30°C inlet water temperatures, respectively, corresponding to 30% relative humidity with air and water flow rates of 85 kg/s and 105 kg/s to achieve the maximum possible thermal performance of the cooling tower.