A Smart Voltage Optimization Approach for Industrial Load Demand Response

This paper proposes a generic and comprehensive Voltage Optimization (VO) strategy for energy savings by industrial customers, to lower operating expenses through the implementation of an optimal process-based Demand Response (DR) program without affecting the real-time manufacturing process. This strategy takes into account the complex nature of industrial loads and their unique set of operating constraints, to reduce energy demand for industrial customers by means of varying the voltage at the utility service entrance to the plant. The proposed approach utilizes a Neural Network (NN) model of the industrial load, trained using historical operating data, to estimate the real power consumption of the load, based on the bus voltage and overall plant process. The NN load model is incorporated into the proposed VO model, whose objective is the minimization of the energy drawn from the substation and the number of switching operations of Load Tap Changers (LTC). The proposed VO framework is tested on a real plant model developed using actual measured data. The results demonstrate that the proposed technique can be successfully implemented by industrial customers and plant operators to enhance energy savings compared to Conservation Voltage Reduction (CVR) approaches, and also as a DR strategy that effectively manages the dependence of industrial loads on time-sensitive and critical manufacturing processes.