A MINLP Optimization Method to Solve Hydraulic Bottlenecks on Existing District Heating Networks

Abstract

The efficiency of district heating networks may be improved by reducing the supply temperature. In existing networks, some technical issues occur especially in pipes, whose velocity increase sometimes over the maximal recommended limit. Two solutions may address the problem, the first one is to replace critical pipes, the second one is to introduce storage downstream critical pipes, named distributed storage, to shift peak demand to off-peak time. To find the best retrofit solution, an exploration with optimization tool is required. In the literature, some approaches compatible with the context of hydraulic congestion are developed. A previous work compares them and concludes that a MINLP optimization initialized with a MILP optimization is appropriate for the introduction of distributed storages. This work proposes to expand this method with the option of replacing pipes. The obtained method is applied on the same case study for a supply temperature between 120°C and 80°C. Distributed storages prove to be more profitable than replaced pipes because they can also do some peak-shaving the rest of the year. When the congestion is too significant, pipes are replaced.