AI-Assisted Fault Detection
Overview
AI-assisted Fault Detection (AIFD) is a modern approach to fault protection that leverages artificial intelligence (AI) and machine learning (ML) techniques. The circumstances surrounding a fault and the type of fault that occurs create unique signatures in the voltage waveform, which can be identified using specialized techniques such as neural networks and anomaly detection algorithms. AI and ML are particularly suited for elucidating patterns in the large and noisy data sets that describe fault events. While this technology is still under active development, many protection system vendors have begun deploying AI-based systems to assist with fault detection.
Benefits
Below are a few of the primary grid challenges that AIFD can help to address:
Outage Mitigation:
- AIFD facilitates faster and more accurate restoration of service following outages. By quickly diagnosing the cause of faults, utilities can deploy repair crews more efficiently, minimizing downtime and improving overall grid resilience.
Aging Infrastructure:
- As infrastructure ages, the likelihood of mechanical failures increases, making it essential to rapidly predict and detect faults. AIFD enhances the ability to detect incipient faults, allowing utilities to address issues before they escalate into major outages. This capability is crucial for maintaining reliability in older systems where equipment failures are becoming increasingly common.
Technology Readiness Level (TRL): 7
AIFD technologies have been demonstrated in laboratory settings, pilot projects and deployed in proprietary utility in-house developed implementations. Widespread commercial adoption is still limited.
Adoption Readiness Level (ARL)
Value Proposition
Delivered Cost
Medium Risk
The upfront costs associated with training an AI models can be very large depending on model complexity. Occasionally this training must be updated, which is also costly. This could require substantial R&D investment should highly complex models be necessary. Should these models be less complex, costs will decrease.
Functionality Performance
Medium Risk
AIFD systems must demonstrate high accuracy in fault detection, minimal false positives/negatives, and robust performance under varying operating conditions. Integration with diverse grid topologies and equipment vendors presents additional challenges as each installment may need additional tuning.
Ease of Use/Complexity
Low Risk
Some training will be required to integrate these models into existing systems and processes. Adding AI changes how control rooms operate to some extent. This will likely require some equipment adjustment, but where that is the case it’s expected that this will be relatively simple.
Market Acceptance
Demand Maturity/Market Openness
Medium Risk
Interest in AI is growing across all industries. For grid applications, the incumbent solution has limitations that AIFD could resolve, but the barriers to adoption will limit market size. As grids become more complex and interconnected, and solutions become more standardized AIFD will see greater market demand.
Market Size
Medium Risk
Hypothetically the entire market could use this solution but the market size is dependent on if the solution can become both cost effective and standardized enough to benefit all of the grid.
Downstream Value Chain
Low Risk
The ability for AI models to Increase fault detection provides clear value to grid operators. The incentives for users are increased reliability and reduced maintenance cost.
Resource Maturity
Capital Flow
High Risk
Venture capital and private equity investments in AI/ML and energy storage sectors have been growing. Access to public funding, grants, and other incentives can support AIFD startup and scale-up. Access to facilities to test products are scares.
Project Development, Integration, and Management
Medium Risk
Given these solutions are broadly software-based project development, integration, and management are not major risks. However, the newness of this technology and lack of widespread deployment suggests that derisking is still needed from an implantation standpoint.
Infrastructure
Medium Risk
AI infrastructure (e.g., data centers) constraints could become a limiting factor for AI solutions, but this infrastructure exists at present and is actively being expanded. Gaining access to these expanded resources could be a competitive process given AI’s growing applications.
Manufacturing and Supply Chain
Medium Risk
AIFD solutions primarily rely on software and data processing, with minimal hardware requirements. Dated grid infrastructure may require additional sensors to capture data but these components are off the shell and readily available.
Materials Sourcing
Medium Risk
Materials associated with the AI industry are exposed to some geopolitical risk.
Workforce
Low Risk
Workforce for AI solutions exist in other industries and can be sourced without significant retraining or scaleup.
License to Operate
Regulatory Environment
Low Risk
No regulatory barriers barring the use of AI within power grid systems, but uncertainty exists within the AI regulatory environment broadly.
Policy Environment
Medium Risk
Federal, state, and local policies support grid reliability and resilience. This generally translates into slower adoption of newer technologies which also have a history of creating more problems and grid reliability challenges. The cautious and skeptical approach to new technologies impedes their adoption.
Permitting and Siting
Low Risk
Little to no permitting & sitting are needed for AIFD implementation. Where improvements to the existing grid are needed to add sensors, it will involve upgrading areas that are already permitted for work.
Environmental & Safety
Low Risk
There are very little to no direct environmental or safety risks associated with AI by utilities directly.
Community Perception
Medium Risk
Communities will likely support grid modernization from a reliability standpoint. However, the increased cost to rate payers will not likely be well received. There could be some pushbacks due to the costs associated with these upgrades.
Case Studies & Implementation
Southern California Edison – AWARE
Southern California Edison (SCE) has been awarded the 2024 AEIC Achievement Award for its innovative development of AWARE (Advanced Waveform Anomaly REcognition), a comprehensive suite of advanced fault modeling, signal processing algorithms, and machine learning models designed to enhance fault detection in the electric grid. This in-house software platform utilizes high-fidelity waveform recordings, advanced metering infrastructure data, grid models, and SCADA data in real-time to identify signs of equipment failure and accurately estimate their locations, facilitating rapid and safe field investigations. Since its implementation in early 2023, AWARE has detected over 255 failures, including incipient and underground faults, thereby reducing outage restoration times, promoting safer working conditions, and improving customer experiences.
Key Takeaways:
- Enhanced Fault Detection:
- AWARE has significantly improved the ability to detect and locate faults in real-time, allowing for quicker responses to potential issues.
- Proactive Maintenance:
- The identification of incipient faults enables SCE to address issues before they escalate into major outages, leading to better maintenance practices and reduced downtime.
- Improved Operational Efficiency:
- By leveraging advanced algorithms and machine learning, AWARE has streamlined the fault detection process, leading to faster diagnostics and more efficient resource allocation during repairs.
- Reduced Outage Restoration Times
References: