Comparison of Adapted PSO Methods Regarding the Determination of Grid Operation Constraints in the Vertical Active and Reactive Power Plane

With the ongoing electrification of heat and mobility a large amount of high-power devices is added to the distribution system level. Their so far assumed power behavior is expected to change in case of an active participation in emerging power flexibility markets. Thus, conventional grid planning approaches, that use simultaneity factors and determined power values, incompletely map the utilization potentials of the grids utilities due to the new active and reactive power flexibilities. To derive possibly needed grid reinforcement measures, knowledge about the occurring grid operational constraints is required. One approach to enable system operators to use these flexibilities in the context of vertical system level cooperation is the aggregation of power flexibilities at the vertical interconnection as a feasible operation region (FOR). This paper focusses on the comparison of methods, that emerged from the adaption of a particle swarm optimization (PSO) algorithm, which aims on identifying the FOR but also the grid operation constraints, i.e. voltage band, thermal line currents and transformer rating, in the active and reactive power plane. This is done to improve the integration of new devices in the planning process towards active distribution grids by comprehensively identifying constraining limits. As an exemplary case-study a generic low-voltage network with simple string topology is used. Results show, that post-processing data driven methods are significantly faster in regards to computation time, but need significantly higher data storage in contrast to an additional sampling strategy with the PSO algorithm. As key factor, the PSO algorithm only finds solutions that comply with the grid operational constraints, whereas the data driven approaches can over and underestimate the resulting polygonal area. The main contribution of this paper is the methodological comparison of PSO based approaches towards an improved system planning process of active distribution grids.