Comparison of COP estimation methods for large-scale heat pumps in energy planning tools

Energy planning tools often consider a constant coefficient of performance (COP) for large-scale heat pumps (HPs) over the entire year, even though the optimization is usually performed on hourly basis. Because the HP performance is highly dependent on ambient conditions, the constant COP may not represent real implementation well.

Ambient temperature levels can be included by assuming simple representations of COP based on a Carnot or a Lorenz cycle. These simple representations of COPs are used, because implementing a thermodynamic HP model in the energy planning software may be complicated or would result in values limited to a specific HP application.

Other options of estimating the COP are the use of more advanced COP estimation methods. This paper compares COP estimations based on four different methods to the COP obtained using a thermodynamic HP model. Four different heat sources and varying district heating supply temperatures were considered.

Furthermore, the results of implementing the different COP estimations in an energy planning tool were analysed. As a case, the planning tool was used to assess HP implementation in a new development district in Copenhagen, Denmark. The change in seasonal COP of the HPs as well as the economic results and optimal HP capacities were compared.

The results show that the Jensen COP estimation method for design conditions provide good approximations compared to a thermodynamic HP model. The results indicate that this method is suitable for the use in energy planning tools. Assuming a constant Lorenz efficiency, constant exergy efficiency or constant COP over the year resulted in very large deviations in COP, especially when the operating conditions differ considerably from the design conditions.

Consequently, different heat sources were found for the most economical optimum. The accuracy of the three methods was highly dependent on the initial guesses of constant COP and efficiencies.