Shortening Quasi-Static Time-Series Simulations for Cost-Benefit Analysis of Low Voltage Network Operation with Photovoltaic Feed-In

Abstract

Executing quasi-static time-series simulations is time consuming, especially when yearly simulations are required, for example, for cost-benefit analyses of grid operation strategies. Often only aggregated simulations outputs are relevant to grid planners for assessing grid operation costs. Among them are total network losses and power exchange through MV/LV substation transformers. In this context it can be beneficial to explore alternatives to running quasi-static time-series simulations with complete input data that can produce the results of interest with high accuracy but in less time. This paper explores two methods for shortening quasi-static time-series simulations through reducing the amount of input data and thus the required number of power flow calculations; one is based on downsampling and the other on vector quantization. The results show that execution time reductions and sufficiently accurate results can be obtained with both methods, but vector quantization requires considerably less data to produce the same level of accuracy as downsampling. In particular, when the simulations consider voltage control or when more than one simulation with the same input data is required, vector quantization delivers a far superior trade-off between data reduction, time savings, and accuracy. However, the method does not reproduce peak values in the results accurately. This makes it less precise, for example, for detecting voltage violations.