Simulation, optimization and design of a heating network at an industrial plant

Heating networks are a particularly interesting method to recover industrial waste heat (i.e. heat at temperature levels too low for electricity generation or direct recovery) for heating purposes. This study optimizes a heat network to be built at a steel plant, where the heat originates from the cooling loop of a walking beam furnace. The decision variables are the pipe sizes, pipe insulation thicknesses, supply and return temperatures, amount of consumers, size of the thermal energy storage system and size of the heat pump. The latter is not always required as different configurations are tested and only certain configurations includes higher temperature consumers. A thermal energy storage system is also optional as a steam back-up, in order to cope with potential heat shortages, is available. The optimization is done based on maximizing the net present value (NPV). Both a genetic algorithm as a brute force algorithm are used for the maximization of the NPV. However, in cases where the complexity of the model is high the brute force algorithm is disregarded. The results show that the heat network is economically feasible and exposes a clear trade-off for each decision variable. The introduction of a thermal energy storage system and/or heat pumps however are not feasible for this particular case study. In the optimized and best case scenario a NPV of 4 059 694 € and pay-back period of 4.01 year was found. Finally, a parametric sensitivity analysis is performed. In this analysis the sensitivity of the electricity cost, absolute pipe roughness, fuel costs, heat pump investment cost and imposed velocity limitations on the NPV are analysed