Performance optimization of a large-scale ammonia heat pump in off-design conditions.

A thermodynamic model of a 5 MW test and demonstration heat pump, implemented in the district heating network of Copenhagen, is presented. A genetic algorithm was applied to the model for 27 different operating conditions, optimizing the system Coefficient of Performance (COP) by adjusting the available set points (SP) in the heat pump system. Analyzing the optimization result, new SP values depending on the operating conditions were proposed. The potential improvement in COP was between 0.4 % and 2.9 %. A performance map for all operating conditions was presented and discussed. In general the COP increased with decreasing forward temperatures, increasing source temperatures and further an optimum at 80 % heating load was observed within the investigated span of operating conditions.