Improving heat measurement accuracy in district heating substations

Abstract: The idea of district heating is to heat up a whole district from a central source through a distribution network. The heat is extracted from the distribution network by heat exchangers and the water is then subsequently returned to the central. The heat exchange between the district heating network and the building occurs in district heating substations. Heat meters are located in such substations and are divided into two main categories depending on their heat energy estimation frequency modes which is either constant or flow rate dependent. The Swedish district heating industry is a business with a revenue of appreciatively 19.8 billions SEK (as of the year 2002). Considering an error of 1% in energy delivered, that is a loss of 198 million SEK, which justifies the current research. The accuracy in heat measurement for billing purposes is then one of the major reasons for conducting this research. There are few works done in this area. A Swedish study shows that the major causes to the errors are: flow meters, temperature sensors, integrating units, lightnings, control systems, valves, leaks in heat exchangers and the dynamic heat demand imposed on the district heating substation. I have chosen to study the heat measurement errors due to dynamic load imposed on the district heating substation because it is the least investigated and stands for a substantial portion of the total error. A major tool in the investigation have been simulations based on a Simulink model of a district heating substation and a house. For the purpose the simulation model has been extended to handle new heat measurement strategies. Based on analysis of the measurement strategies an adaptive algorithm and a feed-forward method are proposed in this thesis to reduce the heat measurement errors due to the dynamic heat demand imposed on the substation. Simulations conducted show that the adaptive algorithm has a higher measurement accuracy than both kinds of existing heat meters. The feed-forward method has the highest measurement accuracy compared to both kinds of existing heat meters and the adaptive algorithm.