Abstract
•Model uncertainty of water source heat pumps is investigated and quantified.•13 commonly-used models of WSHPs are involved and validated using manufacture data.•Impact of model uncertainty on the energy consumption of WSHP system is significant.•Model uncertainty can result in very different conclusions and affect decision making.•Model uncertainty should be considered in design optimization and performance assessment.
Water source heat pump systems are widely used for cooling and heating due to high efficiency. The energy source can be surface water or underground water. Accurate modeling of water source heat pump systems is the basis for performance prediction, design and control optimization. However, various heat pump models are available to predict the system performance and results obtained can be very different under different models due to the model uncertainty. Without considering these uncertainties, the performance of water source heat pump systems would be overestimated or underestimated and the decision making would be affected. This paper attempts to quantify the model uncertainty of water source heat pump and its impact on the system performance. Thirteen commonly-used models are selected and validated using the manufacture data. By importing these models into the water source heat pump system, the impact of heat pump model uncertainty on the system performance is quantified. It shows that the model uncertainty can result in a deviation of up to 30% in the annual energy consumption. The energy saving potential of water source heat pump systems can vary from -18.43% to 14.78% compared with the chillers & boiler in the hot summer and cold winter area. The priority of water source heat pump systems is also controversial. The difference in the annual average coefficient of performance of the system can be up to 1.22. It demonstrates that the model uncertainty of water source heat pumps affects the system performance significantly and should be taken into account in building energy prediction and design optimization. The models considering the correction of part load ratio are recommended.