Advanced Algorithms for Operational Optimization and Predictive Maintenance of Large District Heating Systems

District heating systems have served domestic and commercial needs in population centers for decades. Modern IT tools have been designed to optimize their operation, for economic and environmental benefit. Recently, the smart system concept has gained popularity. District heating systems should not be simply viewed as three distinct components - for more efficient operation pipelines, combined heat and power plants and customers should be linked by appropriate control structures. The paper contains an overview of the implementation of state-of-the-art algorithms in the various elements of a heating system and the wide spectrum of related issues. The implementation of algorithms for predictive maintenance and optimization of coal mills, coal combustion, or operation of the entire heat and power plant, with additional heat accumulation, are briefly discussed and described. Methods directly connected with the distribution network such as predictive maintenance of pipeline failures, forecasting of heat demand, network optimization and anomaly detection are also outlined. Algorithms are now emerging to demonstrate how to make the heating system worthy of the name smart or intelligent by tapping information sources outside the classical control loop architecture.