Climate change has led to more extreme weather conditions that have caused extended outages. As grid resiliency becomes increasingly important and costly for utilities, AI can help improve data quality, reduce inspection cycle times and costs, and enhance vegetation management.
Hurricane Isaias struck the east coast on Tuesday, August 4, 2020, leaving over 2 million customers without power across NY, NJ and CT. A week later, over tens of thousands of customers are still without power. After Isais, NOAA predicts 4 to 10 more named storms for the remainder of 2020.
Simultaneously, on the west coast, wildfire season is just beginning. Extreme heat, dry vegetation and high winds are expected to cause significant fires across much of the Great Basin, northern California, Pacific Northwest, and northern Rockies.
These extreme weather conditions have led to extended outages, upset customers, poor reliability, and angry politicians who are calling for increased regulatory scrutiny and potential penalties for utilities.
To adapt to this “new normal,” utilities are designing grid resiliency programs that include two types of mitigations:
Preventative mitigations can include increased inspections, vegetation management and grid hardening investments.
Traditionally, utilities use ground-based inspectors to manually evaluate grid assets. However, this process can be slow, costly, and lack a “top-down” view of the assets. Increasingly, utilities have been pairing ground-based inspections with aerial inspections to get both “top-down” as well as “bottom-up” images and video, producing significantly more data that needs to be analyzed in the back office. AI can help quickly highlight areas that require attention, such as defective or damaged equipment. By speeding up the back-office work, utilities can reduce inspection costs, do more inspections in less time and focus on critical remediation work.
Fallen trees or branches is the leading cause of power outages during storms and can also ignite devastating wildfires. AI can quickly analyze aerial imagery and video footage to identify vegetation that is too close to power lines or to confirm new trimming standards, such as ground-to-sky clearance.
One of the critical inputs to making optimal grid hardening investment decisions is good quality data so that utilities know where and how much to invest. However, poor data quality is a key challenge for many utilities - while some utilities have multiple databases with conflicting data (such as the same asset tagged to multiple locations); others are finding that they are missing or have inaccurate information (such as whether a transformer is mounted to the pole or not). AI can help improve data by extracting equipment ID tags and identify where assets are located. This information can be used to corroborate or invalidate existing data. AI can also help identify the type of asset that it is (e.g., wood pole vs. steel pole) and what assets are mounted on that pole (e.g. transformer, communication lines). At the end of the day, utilities need a “single source of truth” so that they can make optimal decisions for grid hardening investments.
As the number of extreme weather events increase, the need for a more resilient grid is increasing as well. AI can support mitigations that increase grid resiliency by speeding up inspections, enhancing vegetation management, and improving data quality.
In the next article, we will discuss how AI can support grid resiliency by mitigating the impact of interruption events.