Retraction Note: This paper was formally retracted on March 14, 2025. The following summary describes the research as presented in the original publication for informational purposes, but the findings should be viewed in light of the retraction.
Introduction
As climate change intensifies the frequency and severity of extreme weather events, traditional centralized power grids face increasing vulnerability. Microgrids—localized energy systems capable of operating independently from the main grid—have emerged as a critical solution for enhancing energy security. However, maintaining stability during "islanded" mode (when disconnected from the main utility) remains a significant technical challenge. This research addresses the problem of optimal planning: determining the necessary energy storage capacity and managing demand to minimize the "cost of darkness" during critical conditions.
The Proposed Microgrid Model
The study proposes a comprehensive microgrid architecture that integrates diverse energy resources and communication infrastructure to ensure reliability.
As shown in Figure 1, the system incorporates wind turbines, micro-turbines, solar energy, and battery storage systems. A central operator manages the energy flow and communication data between these resources and the end-users, while also interfacing with the broader energy market to optimize costs during normal operation.
Two-Level Optimization Framework
To address the complexities of both long-term planning and real-time operation, the authors developed a two-level mathematical modeling approach.
- Level 1 (Planning): This stage focuses on normal operating conditions. It determines the optimal sizing and installation of energy storage systems (ESS) to minimize initial investment and daily operational costs.
- Level 2 (Operation): This stage activates during "critical" scenarios, such as grid outages caused by extreme weather. It optimizes the dispatch of resources and implements Demand Response (DR) programs to minimize load shedding and operational vulnerability.
Methodology and Data
The model employs stochastic programming to account for the inherent uncertainty of renewable energy. Solar and wind profiles were modeled using Beta and Weibull probability functions, respectively. The researchers tested their framework using a simulated 24-hour load curve peaking at approximately 15 MW, alongside market price data and technical specifications for storage efficiency and investment costs sourced from industry literature.
Key components of the methodology include:
- Time of Use (TOU) Load Response: A program that shifts electricity usage from high-cost peak hours to more affordable low-load hours.
- Objective Functions: Mathematical equations designed to minimize the "Expected Energy Not Served" (EECC) and total operational costs (OC).
Key Insights and Results
The study demonstrates that the synergy between energy storage and demand response is vital for cost-effective resilience. Key findings include:
- Synergistic Resilience: Combining storage with a 10% demand response program allows the microgrid to withstand longer isolation periods with significantly smaller battery investments.
- Optimal Sizing: For the 15 MW peak scenario, the model identified an ideal storage capacity of 19.731 MWh and a rated power of 4.44 MW.
- Reduction in Load Shedding: The "Expected Energy Not Served" (EECC) dropped dramatically from 16.33 MW in the baseline scenario to 3.65 MW in the optimized scenario.
- Cost Efficiency: Total daily operational costs were reduced from $24,863.3 to $24,654.4, even when accounting for the amortized cost of storage infrastructure.
Limitations and Future Work
While the proposed model offers a robust framework for microgrid resilience, the authors identified several areas for future research:
- Electric Vehicle (EV) Integration: Utilizing EVs as "mobile storage" units to further stabilize the grid during crises.
- Load Prioritization: Refining the model to prioritize critical infrastructure (e.g., hospitals and industries) over residential loads during emergency shedding.
- Network Restoration: Developing more efficient algorithms for "black start" capabilities and rapid reconnection to the main grid following a disaster.
Summary updated