Predicting Heat Demand for a District Heating Systems

Abstract

Poland is one of the heaviest users of district heating systems in Europe, and those district heating systems are heated mainly by coal. Sustainable development of district heating systems in Poland including improving quality of environment, economic of heat production and security of heat supply is in close connection with increasing of energy efficiency. Heat production and heat distribution plays important role in national energy balance. Additional increasing of energy efficiency in district heating systems need detail forecasts for future heat consumption in scale of individual district heating system and for systems in whole country. Accurate forecast give possibility for increasing efficiency of heat production, decreasing fuel consumption and connected with it emission decreasing from combustion products to the atmosphere. Heat production efficiency can be optimized through the use of appropriate procedures for running heat sources alongside short-term heat demand forecasting combined with preparation for adjusting heat source work parameters to the predicted heat load for a few hours hence. The artificial neural networks model delivers good forecasting results. The accuracy of the results depends on the kind of network, its architecture, the size and type of input data as well as the forecasting period. Forecasting accuracy within a 3-5% margin of error is sufficient to steer heat source operations. Described forecasting methods can be use as a good tool to regulate district heating networks and heat sources.