Table of Contents
Fetching ...

Two-Step Blackout Mitigation by Flexibility-Enabled Microgrid Islanding

Philipp Danner, Anna Volkova, Hermann de Meer

TL;DR

This work tackles blackout resilience in highly renewable microgrids by proposing a two-step blackout mitigation framework. It combines a proactive, MPC-based reserve scheduling before outages with a fully distributed islanded control that uses max/min consensus to coordinate load, generation, storage, and grid-forming resources during a blackout. The approach demonstrates substantial reductions in load shedding, robust operation under communication delays, and validation through hardware-in-the-loop and long-horizon simulations on a realistic 13-bus grid. The results indicate that coordinated, distributed control can extend islanded operation time and improve reliability for critical loads in future low-inertia grids.

Abstract

Blackouts are disastrous events with a low probability of occurrence but a high impact on the system and its users. With the help of more distributed and controllable generation and sector-coupled flexibility, microgrids could be prepared to operate in islanded mode during a blackout. This paper discusses a two-step blackout mitigation approach for highly renewable microgrids that utilizes user flexibility and energy storage systems for power balance in islanded grid operation. The proposed method includes a proactive flexibility reservation step, which derives a minimal reservation schedule for microgrid resources under uncertainty considering related operational costs. As a second step, during a blackout, a fully distributed control is implemented to maximize the usage of available resources based on a sequence of max and min-consensus rounds. This paper focuses on the second step, for which the effectiveness of blackstart and long-term coordination is shown. Load shedding can be reduced by 40\% compared to the forecast value. A hardware-in-the-loop simulation of a grid-forming converter further showed a fast convergence toward the optimal operation point.

Two-Step Blackout Mitigation by Flexibility-Enabled Microgrid Islanding

TL;DR

This work tackles blackout resilience in highly renewable microgrids by proposing a two-step blackout mitigation framework. It combines a proactive, MPC-based reserve scheduling before outages with a fully distributed islanded control that uses max/min consensus to coordinate load, generation, storage, and grid-forming resources during a blackout. The approach demonstrates substantial reductions in load shedding, robust operation under communication delays, and validation through hardware-in-the-loop and long-horizon simulations on a realistic 13-bus grid. The results indicate that coordinated, distributed control can extend islanded operation time and improve reliability for critical loads in future low-inertia grids.

Abstract

Blackouts are disastrous events with a low probability of occurrence but a high impact on the system and its users. With the help of more distributed and controllable generation and sector-coupled flexibility, microgrids could be prepared to operate in islanded mode during a blackout. This paper discusses a two-step blackout mitigation approach for highly renewable microgrids that utilizes user flexibility and energy storage systems for power balance in islanded grid operation. The proposed method includes a proactive flexibility reservation step, which derives a minimal reservation schedule for microgrid resources under uncertainty considering related operational costs. As a second step, during a blackout, a fully distributed control is implemented to maximize the usage of available resources based on a sequence of max and min-consensus rounds. This paper focuses on the second step, for which the effectiveness of blackstart and long-term coordination is shown. Load shedding can be reduced by 40\% compared to the forecast value. A hardware-in-the-loop simulation of a grid-forming converter further showed a fast convergence toward the optimal operation point.
Paper Structure (32 sections, 12 equations, 14 figures, 1 table, 2 algorithms)

This paper contains 32 sections, 12 equations, 14 figures, 1 table, 2 algorithms.

Figures (14)

  • Figure 1: Two-step Approach
  • Figure 2: Value Functions for Different Agent Types
  • Figure 3: SimBench Grid "1-LV-rural1--1-sw"
  • Figure 4: Feasible Delay and Graph Diameter for $\Delta t$ with typically delays in 5G networks Christopoulou2021
  • Figure 5: Blackstart
  • ...and 9 more figures