Simple models of district heating systems for load mangement and operational optimization

Main conclusions: To reduce the simulation time, aggregated models of the network and consumers are an attractive alternative to a full description of the DH system. In this work the thermal aggregated model developed in 1999 has been further validated on new DH systems and on large data sets. Moreover, a new aggregated model for pressure losses has been developed in the present project. For both the thermal and the pressure aggregated models we found that the number of pipes and consumers tan be reduced to three to five pipes without seriously affecting the accuracy of the simulation. In an online application, the uncertainty associated with the weather and heat load forecasts is expected to be much greater than the inaccuracy from aggregation of the DH system. The optimum temperature level is specific for each and every DH system. For DH systems with high line heat demands and low fuel/electricity cost ratio like in Ishoej and Roedovre R01 Madumvej, the optimum temperature level is high (90-105 deg. C in the winter period). For DH systems with large heat losses (in percentage) the temperature level is lower. Operational optimisation of DH systems is expected to be more complicated in the future, not only for togeneration plants but also with respect to heat production only. Emission taxes (C02) and variable gas and electricity prices need to be taking into account to find an optimum solution. The incentive tariff between VEKS and the distribution companies applies. An incentive for daily load variations and the return temperature level at the substation must be taken into account. This incentive has a big influence on the optimum supply temperature curve. Demand Side Management should be taken into account to find the optimum temperature level. It has been documented in the report that DSM in Ishoej and Roedovre R01 Madumvej tan reduce the operational costs bv reducina the daily load variations