Study of the combined impact of auto-consumption behaviour and correlation level between prosumers on overvoltage probabilities in low voltage distribution grids

Connections of distributed generation (DG) units based on the use of photovoltaic (PV) cells are highly increasing in low voltage distribution grids. In that way, one of the major problems met by the Distribution System Operators (DSO) comes from overvoltage in the neighbourhood of dispersed units. In previous studies, it has been shown that probabilities of overvoltage in such grids could be obtained by using a Probabilistic Load Flow based on analytical techniques or Monte Carlo methods. In this paper, given its simplicity of implementation, a pseudo-chronological Monte Carlo simulation is used and the statistical behaviour of prosumers (consumers with PV units) is directly based on smart meters measurements. Thanks to this tool and, in the context of alleviating the impact of photovoltaic generation on the recorded voltage profiles, smart meters data are employed in order to not only estimate the auto-consumption potential of prosumers connected to the grid but also to quantify the impact of the correlation level between those prosumers.