Surrogate model based optimization of industrial cogeneration system under real-time energy commodity market

Chuan Zhang, Alessandro Romagnoli, Markus Kraft

Abstract

A novel framework for the optimization of industrial cogeneration system under real-time energy commodity market is proposed in this paper. Surrogate model is used in the proposed framework: surrogate model of industrial process is used to forecast the heat and electricity demand of industrial process; surrogate model of the cogeneration plant is used to mimic the combined heat and power (CHP) plant dynamics. Design parameters and operational parameters are described separately in the formulation so that system design problem and system operation problem can be differentiated; operational parameters are further divided into process parameters (e.g. pressure, temperature and flow rate of raw material streams) and economic parameters (e.g. price of raw material). Economic parameters are connected to the real-time energy commodity market to get up-to-date price information. The proposed formulation could accommodate both system synthesis problem at the design stage and system scheduling problem at the operation stage; moreover, the proposed optimization formulation can dynamically react to real-time energy prices, which could significantly increase the cogeneration system operational efficiency economically.