Development of EMT/TS Co-simulation Using PowerFactory and PSS/E

Abstract

As the scale and complexity of power systems increase, simulating them in efficient and accurate ways continues to
be a challenge in power systems engineering. Electromagnetic Transient (EMT) and Transient Stability (TS) simulation are the two main dynamic power system simulation methods. To simulate large and complex power systems in sufficient detail without sacrificing execution time, one of the idea is to perform a co-simulation that couples EMT and TS simulator. Although several attempts have been made to couple both simulator, only on rare occasions do these hybrid EMT-TS simulators couple two or more industry-adopted simulation tools.

The objective of this thesis is to to develop and study the benefits and limitations of the Electromagnetic Transient – Transient Stability co-simulation based on PowerFactory and PSS/E, both of which are among the most extensively used simulation tools in industry and academia alike. With regards to the objective, the EMT-TS co-simulation using PowerFactory and PSS/E has been developed. Then, several tests are performed to evaluate the function of each composing part of the EMT-TS co-simulation, and to test the integration between all its component. Next, The developed co-simulation is applied to study cases and the results are compared to a monolithic EMT simulation to evaluate its accuracy and execution time. Furthermore, the effect of TS and EMT time step to the accuracy and execution time of EMT-TS co-simulation have also been investigated.

The study case results show that the developed EMT-TS co-simulation has not been beneficial yet in terms of accuracy and execution time. Although the active power result shows a similar tendency with the monolithic EMT result, the difference between both are visible. The difference between both are more prominent in the reactive power result. The total execution time of the developed co-simulation in the study cases are in the range of 23-24 minutes, significantly larger than the total execution time obtained from the monolithic EMT simulation which is around of 12s. Also, it is found that reducing the TS time step from 0.02 s to 0.01 s slightly increases the total simulation time from 23 to 26 minutes. However, it does not contribute a significant improvement on the accuracy of the developed EMT-TS co-simulation. The result obtained from reducing the EMT time step to $ 25 \mu s$ is the same with the result obtained using $ 50 \mu s$ EMT time step. Moreover, the reduction of the EMT time step significantly increases the total simulation time from 23 to 42 minutes. The developed co-simulation still has a lot of room for improvement and further developments in this topic might increase its performance.