Probabilistic Quasi-Static Time Series Simulation for Distribution Network Planning Considering Multiple Uncertainties of PV and Load in the Presence of BESS

In this study, we propose a comprehensive and computationally efficient Probabilistic Quasi-Static Time Series (PQSTS) method in which the multiple uncertainties of forecasted load and PV data are employed in the distribution network planning problem. Additionally, inverter reactive power capability is implemented to regulate voltage in response to irradiance variation. Furthermore, a Battery Energy Storage System (BESS) is utilized along with the PV array to smooth and shave substation peak power as well as enhance system security/stability. The Non-dominated Sorting Genetic Algorithm-II combined with Fuzzy Decision-Making Tool (NSGA-II/FDMT) is implemented to solve the multi-objective problem along with three objectives: (a) minimization of distribution network power losses; (b) maximization of system security; and (c) minimization of the total cost. Also, two more indices, i.e., maximum overall voltage deviation and substation peak power, are defined to evaluate DN system performances. A distribution network, including a PV-inverter-battery system with its control functions, is considered to investigate optimal results by an hour-by-hour simulation for the yearly horizon. Since a large computational burden is needed for Monte Carlo (MC) simulation, a robust 8760-hour probabilistic load flow (PLF) method based on the Point Estimate Method (PEM) is implemented to address load/PV uncertainties. Simulation results on the IEEE 33-bus test system through different case studies demonstrate that the detailed distribution network analysis applying an hour-by-hour probabilistic load flow method leads to more realistic PV size and distribution network indices.

I. Introduction

The penetration of distributed energy resources (DERs) into the distribution network (DN) can highly benefit electrical systems in terms of reliability and operation. A significant number of reported researches have focused on different methods to solve the sitting and sizing of DERs in DN [1]–[2]. They have employed heuristic methods like GA or Bee colony to solve multi-objective optimization problems applying various objective functions such as power loss reduction, voltage improvement, stability enhancement, etc. [3]–[7].