Improving Fairness in Photovoltaic Curtailments via Daily Topology Reconfiguration for Voltage Control in Power Distribution Networks

TL;DR

This paper introduces a daily topology reconfiguration method for power distribution networks with PV systems, aiming to improve fairness in curtailments caused by voltage issues without significantly increasing total curtailments.

Contribution

It proposes a novel two-stage optimization approach that dynamically adjusts network topology daily to enhance fairness among PV plants while maintaining voltage control.

Findings

Enhanced fairness as measured by Jain Index

Maintained low overall curtailments

Validated on benchmark test cases

Abstract

In PV-rich power distribution systems, over-voltage issues are often addressed by curtailing excess generation from PV plants (in addition to reactive power control), raising fairness concerns. Existing fairness-aware control schemes tackle this problem by incorporating fairness objectives into the cost function. However, such schemes result in increased overall curtailments. This paper proposes a solution through daily topology reconfiguration, ensuring that different PV plants face varying grid conditions each day, leading to different curtailment levels and enhancing fairness. We illustrate that implementing this approach enhances overall fairness without significantly increasing overall curtailments. The optimization problem involves two stages. The day-ahead stage optimizes the network topology using day-ahead forecasts of PV generation and demand, minimizing net curtailment and accounting for fairness based on curtailments from prior days. The real-time stage implements the optimized topology and computes active and reactive power setpoints for the PV plants. Day-ahead grid constraints are modeled using LinDistFlow, and real-time control employs a linearized model with a first-order Taylor approximation. The proposed scheme is numerically validated on several benchmark test cases. Results are compared using the Jain Fairness Index, considering fairness and reconfiguration scenarios.