A Hybrid Optimization Algorithm for Water Volume Adjustment Problem in District Heating Systems

Abstract

Nowadays, the winter is getting harsher and harsher in Northern China. Thus, the centralized heating systems (CHSs) are playing even more irreplaceable, essential and critical roles in ensuring general public’s livelihood than never ever before. CHSs are normally composed of one or several combined heat and power (CHP) plants (units) and an extensive vein like district heating networks (DHNs) connecting with chemical plants, paper mills, food processing factories, hospitals, hotels, universities, prisons and residential complexes. A CHP plant in Northern China usually consumes coal to heat the cold water into steam to drive high-pressure turbines and low pressure turbines to generate electricity. Then the low-temperature steam is used to heat up the cold water in a main pipe into hot water travelling through the DHNs to provide heat to each end nodes. The returned water will be heated again for reuse and the surplus steam will be released into air through cooling towers. In 2020, China promised to the world that carbon dioxide will peak in 2030 and net-zero emission will happen in 2060. On the one hand, CHP plants need to guarantee enough hot water flowing within each household’s heating radiator. On the other hand, they should cut down on the consumption on non-renewable resources. Lowering water temperature, adjusting water volume and reducing water pressure will all contribute to energy-saving purpose. Lowering water temperature and reducing water pressure may cause too much heat losses during long-distance transmission in frigid winter. Therefore, a reasonable water volume adjustment becomes an advisable action comparatively. Here, we transfer the hot water supply volume optimization problem (HWSVOP) into a heat exchange station (HES) valve angle adjustment problem (SHWESVAAP). Then, a multi-objective mathematical model is established considering balancing the satisfactory degree of each household in residential quarters and the hot water volume (HWV) in the main pipe. And a hybrid polar bear optimization algorithm integrated with chemical reaction optimization (HA-PBO-CRO) is designed to optimize the valve angle (VA) in each HES. The comparative results between HA-PBO-CRO and non-dominant sorting genetic algorithm (NSGAII) demonstrate HA-PBO-CRO is superior to NSGAII with better Pareto frontiers on one hand and provide a critical reference supporting the management in a CHP plant to make a right decision on what to do to cut energy consumption while satisfying customers’ needs.