State estimation of district heating network based on customer measurements

District heating (DH) has been widely used in many European countries since the beginning of 20th century. In Finland approximately half of the heating market was covered by DH by 2011. For better use of district heating, real-time monitoring of the water flows, temperatures and heat losses can be an effective vehicle to manage DH networks. Automated remote meter reading is a new technology that was applied since 2005 and will cover the Helsinki region within 10 years. Automated meter reading enables computing the heat losses more accurately than before. In this paper, a new model is designed to estimate the water flows, temperatures, and heat losses in different parts of the network using automated hourly meter readings for customers. Assuming a connected graph topology for the network, the flow equations form a determined and potentially nonlinear system that we solve iteratively. Then, based on the computed water flows and customer temperature measurements, we express the water temperature and heat loss in each pipe as an over-determined linear system and solve it in the least squares (LSQ) sense which ensures the solution with the smallest 2-norm error vector. We illustrate the method with a real life DH network based on hourly temperature and flow measurements for one week. The model can be applied to arbitrary DH networks.