Artificial Intelligence for Predicting District Heating Load from Different Sources Based on Environmental Temperature

Technological and ecological environmental regulations in the European Union are increasingly leaning towards reducing greenhouse gas emissions and improving the operational efficiency of production facilities. In line with this aim, a simulation model for predicting district heating loads using artificial intelligence with real process data has been developed. The heat for district heating is supplied from two sources depending on the external temperature and electrical demand. During the winter months, a gas-steam power plant is predominantly used for heating, while in the summer months, when heating needs are lower, a hot water boiler is utilised. Additionally, the hot water boiler serves as a backup during the winter months in case of failures or outages of the main heating system. The district heating system is equipped with a heat accumulator, which serves to cover peak demands when heating needs are at their highest. It also helps maintain the static pressure in the district heating system and stores heat when the price of natural gas is at its lowest. The results of the simulation model are used to predict the amount of natural gas that needs to be purchased several days or even a week in advance. The timing of natural gas purchases is closely linked to the price of gas, thereby contributing to the more economical operation of the district heating system.