Models for Optimization of Energy Consumption of Pumps in a Wastewater Processing Plant

Abstract

In this paper, a data-mining approach for modeling pumps in the wastewater preliminary treatment process is discussed. Data-mining algorithms are utilized to develop pump performance models based on industrial data collected at a municipal wastewater processing plant. The performance of wastewater pumps is described by two parameters, pump energy consumption and water flow rate after the pumps. Two types of models, dynamic and steady state, are established to predict pump energy consumption and water flow rate. The first type of model is developed based on 5-min data and the second type is built based on 30-min data. The accuracy of the models has been validated.