An Automated Approach to Building and Simulating Dynamic District Heating Networks

Abstract

In Nordic countries, district heating accounts for a large share of the consumers’ heat demand. In Sweden, roughly 50% of the total heat demand is attributed to district heating. Which, over the past few years, is equivalent to around 50 TWh, and imposes a difficult balance between supply and demand for the suppliers of district heating. For large networks the propagation of heat from supplier to end-user can vary several hours. Further complexities of large networks, which can consist of multiple overlapping rings, is that during transient conditions the flow can actually change direction. A dynamic modeling library has been developed in Modelica using OpenModelica for district heating networks. Methods for modeling, handling data, simulating and the visualization of results has been developed using Matlab. The model has been validated using data from Mälarenergi AB, a local provider of district heating in Västerås, Sweden. The model provides to an acceptable degree in predicting the heat propagation and temperature distribution in a localized case study. Adding a higher level of robustness, the model has the capacity to handle bi-directional and reversing flows in complex ring structures. Through this work, the combination of OpenModelica and Matlab, a framework for automating the building and simulation of district heating networks is obtainable. The implications of automating network modeling from computer-aided design drawings allows for a quick robust overview of how the network is working and how prospective additions to the network could impact the end-users. Furthermore, incorporating visual aspects for heat propagation in a network contributes to a higher understanding of complex network structures.