DescriptionLarge-scale heat pumps allow for efficient conversion of electricity and low-grade heat into heat at higher temperatures that can be used to supply district heating networks. In systems that are characterised by high shares of renewable power, heat pumps may also act as a flexible load for the power grid. This requires that these systems are designed to operate flexibly and according to variable boundary conditions. The most common type of heat pumps supplying district heating in Denmark is two-stage ammonia heat pumps. The influence of the size of the heat exchangers, vessels and pipes in the system on the regulation time was assessed using a dynamic model and Monte Carlo method. The dynamic model of a two-stage ammonia heat pump with variable speed drive piston compressors was implemented in Modelica. The sensitivity of the regulation time to changes in the geometric input parameters was determined using a linear regression. The size of the compressors was kept constant. The regulation time was assessed for a change in load of 30 %. It was found, that the regulation time depends strongly on the overall refrigerant mass in the system, while the ratio of evaporator heat transfer area to the total heat transfer area of all heat exchangers in the system had a minor influence, as well as the total heat exchanger area. The analysis showed that the size of the intercooler had a dominant influence on the regulation time of the system. This is due to the comparably large amount of refrigerant mass in this component. We conclude that a considerable reduction in refrigerant mass in the system can help to allow faster regulation of the system, and that the design of low-charge intercoolers is important for flexibly operated two-stage ammonia heat pumps.
|Period||29 Jun 2020 → 3 Jul 2020|
|Event title||33rd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems|
|Degree of Recognition||International|
- Heat pump
- frequency regulation
- system design
- dynamic simulation