The storage of electrical energy is becoming increasingly important to satisfy the demand through renewable energy sources.
In this paper, a continuous and discrete simulation of a pumped thermal energy storage (PTES) system are compared with respect to their computational time and accuracy. The stratified heat storage is modelled using spatial one-dimensional considerations and abstractions.
General setup of the heat storage system
Researchers investigate two different simulation approaches for a PTES with a one-dimensional model for the heat storage:
- a system dynamics model, to simulate the behavior and dependencies of different layers inside the storage according to the simplified differential equations;
- a discrete event simulation model with fixed time steps (DS-FT) with a discretization of continuous flows to only calculate balance equations every time step.
Both simulations include the same operation logic of the storage system and efficiency curves for the heat engines. We compare these approaches regarding the accuracy of heat losses inside the storage and computational efficiency, to determine if a continuous calculation of heat losses is necessary or if a coarser resolution provides sufficiently accurate results.
Finally, a sensitivity analysis is performed for material parameters of the storage fluid and insulation.
Results show that the discrete simulation with fixed time step is about 60 times faster and has neglectable deviation in the resolution of heat loss computations and accuracy, compared to the continuous system dynamics modeling.
The conducted sensitivity analysis shows, that parameters of the storage fluid have little influence on the overall result. The determining factor regarding losses in the storage and possible savings for users is the insulation of the thermal storage. Increasing the number of temperature layers considered in the simulation shows so called deadlayers, which block the storage if not removed during operation.