Tech Insights

Improving Microgrid Resilience With New Sensitivity Analysis Method

May 20, 2024 by Liam Critchley

Scientists have created a model for measuring how quickly a microgrid can recover after a disturbance.

Microgrids have become ideal networks for connecting various energy sources, such as renewable sources and storage systems, with high flexibility and control. If effective management protocols are implemented, microgrids can optimize the power supply network, load, and energy storage and easily control operations.

However, as energy demand increases, microgrids face technical challenges with malfunctions, power outages, load switching, and backup capacity. Outside disturbances, such as weather conditions, can also stress microgrids. These issues can compromise grid resilience, which refers to how well the power grid bounces back from a disturbance to continue its normal operations. 

Researchers at Wuhan University in China have developed resilience strategies to improve current analysis methods and build more robust microgrids that can bounce back quickly after disturbances. 


Microgrid resources in California.

Microgrid resources in California. Image used courtesy of Consortium for Electric Reliability Technology Solutions 


Challenges in Current Resilience Analysis Strategies

Measuring power system resilience includes analyzing the factors affecting it and evaluating its performance during disturbances. Analyzing resilience strategies involves examining the degree and speed of system performance degradation and measuring system resilience recovery time.

Resilience strategies must consider both static and dynamic indicators within the system. Static indicators include key node vulnerability and the connectivity of the system's topological structure, whereas dynamic indicators quantify the system's real-time response capability under extreme events.

Current strategies and analysis methods for improving microgrid resilience rely on using large amounts of configuration resources. These include selecting energy storage sites, reconstructing grids, adjusting capacity, and designing load switching and demand response systems.

However, since microgrids contain differing power sources, loads, and architectures, it’s difficult to accurately analyze them from a real-world, practical perspective.

Additionally, many methods use emergency control protocols to replan the operation mode after a grid failure. However, during these times, a power imbalance can cause safety concerns and the failure to safely re-establish the grid’s normal workings. These technical factors limit the effectiveness of many microgrid resilience enhancement strategies.


Method Aims To Cover Resilience Analysis Shortcomings

Current methods for analyzing resilience focus on system indicators, but researchers from China studied measurement methods to enhance resilience from qualitative and quantitative standpoints.

Unexpected disturbances within microgrids affect the wider network system and the grid’s performance. Examining the network resilience through the system's basic inherent parameters alone is insufficient to assess the grid's performance after a disturbance—the system’s entire process and uncertainty must be evaluated.

The researchers looked at the impact of system disturbance—which they named a “fault”—and the recovery process for these faults. During the time between the fault’s occurrence and the performance recovered post-fault, the grid system undergoes parameter changes. The researchers used these parameter changes within the grid components to quantify the network’s resilience development process. Unlike other methods, this led to a detailed measurement analysis model that could provide a quantitative output.


Grid resilience analysis

Grid resilience analysis. Image used courtesy of the authors


The scientists used a sensitivity analysis method to examine the parameters in detail under varying conditions. This enabled them to study the parameters’ influence on grid reliance. The approach allowed the scientists to adjust the parameters to change and raise the system’s resilience.

The sensitivity analysis examined the relationship between components and parameter configurations to construct the overall system resilience. The researchers could evaluate each component’s importance.  Some specific parameters they analyzed for resilience included failure and repair rate, post-failure performance, post-repair performance, repair time, and the average time between failures. The analysis also considered the challenges of wind power and photovoltaic grid connections.

The analysis found some parameter changes, such as repair rate, positively affected grid resilience. However, some parameter changes, like fault repair time, had a negative impact. The analysis method can be used to propose an optimization strategy to enhance system resilience and consider the economic factors of adjusting the grid parameters.


Recovery time was faster using the researchers’ method.

Recovery time was faster using the researchers’ method. Image used courtesy of the authors


Finally, the researchers verified their model using a simulation based on the IEEE 37-node microgrid system. The simulation results showed the analysis can measure a microgrid’s resilience scientifically and reasonably. Researchers found that the post-analysis improvement strategies implemented after the results significantly improved the grid’s operational resilience, further verifying the effectiveness and robustness of their method.


Other Factors to Consider

The Chinese researchers used practical scenarios to demonstrate their method as effective in analyzing resilience levels within a grid configuration and implementing changes to improve grid resilience. The method builds on its predecessors and considers more relevant features impacting grid resilience after a disturbance. However, it’s rare for a model to be 100% perfect and consider every available scenario. While the model, in its current form, effectively analyzes microgrid resilience, it assumes the microgrid system's repair rate is 100% after a failure occurs, which is not always the case. In cases where the failure cannot be fully repaired, the grid’s resilience will still be damaged in some form. This issue needs to be considered in the analysis model’s next iteration.