Three Reasons Smaller Utilities Should Adopt Machine Learning to Prep for Severe Weather

Three Reasons Smaller Utilities Should Adopt Machine Learning to Prep for Severe Weather

Extreme weather events are increasing, and utilities and customers alike are feeling the impact. According to Climate Central, the United States has experienced a 67% increase in major power outages from weather-related events since 2000, a trend predicted to continue as extreme weather events increase in frequency.

While using predictive weather analytics can help utilities of all sizes prepare and pre-stage crews, it is the larger utilities that are more likely to use this capability. According to a report from DTN, 72% of utilities that use machine learning serve more than 1 million customers. The majority of small to mid-size utilities still rely on traditional forecasts, severe weather paths and immediate impact assessments to make decisions around operations, safety and resource allocation. But with recent advancements in technology and data insights, smaller utilities have even more incentive to take advantage of machine learning for weather risks.

Improves Resource Allocation

When extreme weather events occur, utilities need an effective storm response to minimize impact to power, infrastructure, and travel. As recent historical weather events have shown, this is often easier said than done. With multiple factors, such as incident types, trouble locations, and how many crews are needed – not necessarily just the number of customers impacted – making it tricky to estimate resources.

This is compounded by utilities requesting restoration crews earlier in the weather preparation cycle than ever before, which limits the pool of available resources. For a smaller utility, that means being judicial in staging crews to make sure resources are in the right spot at the right time.

Using AI and established risk thresholds, incident commanders have clear guidance on the escalation level needed and where to reallocate and source additional restoration crews and materials, as needed, ahead of the event. After the event, it can justify pre-staging costs that could be recovered.

Improves Infrastructure Investments

Intelligent operational expense investments, like AI technology, reduce the need to continue increasing hardening capital expenditure investments. Where hardening programs can take decades to see results, predictive weather technology can be deployed more immediately alongside infrastructure improvements. In addition, they enable utilities to deliver stability in electricity rates to customers and help to improve outage prediction and planning.

AI offers an avenue for smaller utilities that don’t have the resources to complete hardening projects and load and maintain their own data into planning models, but still need to make decisions. AI consumes and collates the enormous amounts of data available and delivers accurate forecasts – without the cost of custom modeling – and provides more utilities with the capability to keep crews and communities safe, protect their infrastructure, minimize outage durations, and avoid potential regulatory penalties.

Reduces Limited Resources

The most cited barriers for smaller utilities not investing in machine learning are limited budget and resources to load and maintain their own data models. Traditionally, the only option for utilities who employed AI was to create a custom model that takes years to calibrate and is supported by an in-house data team. For larger utilities who have assets across multiple regions and geography, these precise insights are beneficial for coordinated risk mitigation and restoration response. But there are new options that allow smaller utilities to have access to predictive weather analytics with the customization.

For example, DTN Storm Risk Analytics combines verified, historical outage data with advanced weather and machine learning models that can be tailored to a utility’s operating region and topology. This allows utilities to predict weather impacts more accurately on their service area up to seven days ahead of an expected weather event. Up until now, this was not an option for smaller utilities.

Technology has also made it easier to overcome resource challenges. Today, cloud-based solutions that can be integrated into existing platforms provide access to dynamic, complex data without investing in additional data science resources. Often, this can be overlaid with other readily available operational insights for enhanced decision-making.

With the compounding pressures of increasing weather events, making grids more resilient, balancing restoration responses and reducing outages, and applying machine learning for weather impacts is a smart – and necessary – investment for smaller utilities. Decision-makers will have increased capability to make agile, confident decisions in the moment about keeping communities and infrastructure safe, while maintaining operations and reducing risk.

Renny Vandewege is the Global Weather Intelligence Leader at DTN and is responsible for developing strategic direction for customers in industries with complex supply chains, such as shipping, transportation, utility, agriculture, and energy.  He started his career as a broadcast meteorologist in Meridian, Mississippi, and holds a master’s degree in meteorology from Mississippi State University. 


#Reasons #Smaller #Utilities #Adopt #Machine #Learning #Prep #Severe #Weather

About admin

Leave a Reply

Your email address will not be published. Required fields are marked *