Steps Utilities Can Take Now to Prepare for Future Extreme Weather Events

How can a utility, or any organization for that matter, prepare for the unexpected, especially when it comes to the volatility of weather? This question is increasingly coming to the forefront of risk management as emergency preparedness managers help their utilities prepare for a multitude of risk scenarios.

Weather disruption is not unusual for the utility sector. According to Climate Central, about 83% of reported major power outages between 2000 and 2021 in the U.S. are attributed to weather-related events. What is unusual is the frequency, intensity, and type of weather events. Last year was a historic year of U.S. billion-dollar weather and climate disasters. Other countries also faced extreme weather, particularly wildfires, floods, and historic heatwaves.

While these weather impacts may be unprecedented, utilities are still responsible for supplying uninterrupted power and demonstrating action plans that show how they will be weather resilient in the future. As a global data, analytics, and technology company that works with utilities worldwide, DTN has helped power companies of all sizes facing different extreme weather risks become more weather resilient. Here are steps that can help utilities better plan for future weather events.

Assess Risk Threats and Tolerances

It used to be that energy companies could count on certain weather patterns to build risk mitigation plans. Utilities on the West Coast anticipated wildfire responses from late summer through fall. Utilities in the northern U.S. were geared for winter hardening. Power companies along the Eastern Seaboard paid close attention to tropical storm forecasts. While these are still threats for power disruption, other weather events that were once a remote risk must also be considered. The 2021 Arctic freeze in Texas and the 2023 Hawaii wildfires compounded by winds from Hurricane Dora are two recent examples. As extreme weather continues, predictions must go beyond what the weather will be, or used to be, to include what the weather could do.

Climate modeling can point to how the weather may evolve over years or decades into the future. It uses computational analysis of complex earth systems to recreate the past climate, predict future climate, and test whether extreme weather events are related to climate change. The models are surprisingly accurate. In 2020, an evaluation of global climate models used to predict Earth’s future global average surface temperatures over the past half-century proved that most of the models have been correct. Understanding how climate could impact different regions and the type of weather risks associated with it would allow utilities to better plan for weather resilient measures, such as long-term infrastructure investments, or expanded response processes.

Another tool that can help with risk planning closer in time is incorporating longer-range forecasts into annual risk assessments. Weighted with a probability of a weather event happening in a region, a utility can use the outlook to assess additional weather risks such as soil moisture for elevated wildfire risk, or jet stream variations that influence extreme flooding and temperatures.

Historically, these forecasts have been considered an indicator of how weather may act during a three- to five-month period to better plan for load forecasting. But as the science of weather and technology has advanced, so has the accuracy of long-range forecasts. Even “unprecedented” events like the 2021 plunging temperatures in Texas that left nearly 10 million people in the dark were indicated in the long-range forecast well before winter arrived.

After assessing weather risks, common, low-frequency, and now potentially extreme scenarios, control centers can build an operational response plan based on a matrix of risks (Figure 1). Include both regional and hyper-focused risks for assets, networks, and service. For example, one area may be more susceptible to drought and high winds, while another may have a higher threat from dense vegetation. Determine a level of risk tolerance for each risk and what triggers the next level of operational response.

1. Identifying risks and establishing a tolerance per risk is beneficial to streamlining response management. The DTN Energy Event Index combines real-time forecasts, identified weather risks, and alerts when exceeding tolerances to help incident commanders monitor multiple inputs. Courtesy: DTN

Establish One Source of Truth for Weather Intelligence

Most utilities have assets and infrastructure spanning diverse geographies and topography exposing them to potentially different weather risks at the same point in time. This diversity can create another challenge—multiple weather forecasts that often overlap or misalign. The forecasts are often derived from field operators who must become “ad-hoc” meteorologists with weather impacts based on regional weather forecasts and past experiences.

The control center then must juggle multiple inputs to create and deploy response plans making it difficult to aggregate the information and develop a proactive response. Using one source of truth for weather intelligence, correlated with risks and risk tolerance in each area of interest, improves internal communication of weather risk, operational and response decisions, and initiate emergency protocols to engage and optimize resources.

Use the Power of Data Analytics

The ability to forecast and manage weather events has benefited from digital transformation in recent years, giving utilities access to sophisticated, predictive weather analytics. This includes ensemble models that consider numerous atmospheric conditions and data from multiple sources to suggest a realm of scenarios or probabilities. These powerful datasets can help prepare for the efficient, effective coordination of weather-related power outages.

The most advanced tool for helping energy companies plan and respond to outages uses artificial intelligence (AI). Advanced modeling shortens the time between data collection, analysis, and action. It also provides actionable insights expected to provide the best outcomes overall. Machine learning, a subset of artificial intelligence, automatically learns from historical data and adapts to new intel by absorbing massive amounts of unstructured data. The information is then organized and integrated with other relevant data streams to deliver a more informed and up-to-date analysis of the dynamic environment.

But until a decade ago, it was assumed that only a few large utilities had access to this type of data with in-house teams running custom models. The power companies without the budget or resources—yet still held to the same regulations, outage standard, and public sentiment—were forced to use the data at hand to make outage mitigation decisions. But robust computing technology, the development of open-source AI models, and enterprise solutions with integrated AI technology have accelerated access to utilities of all sizes. One of the biggest advantages of using AI with weather modeling is the ability to confidently plan outage response teams, even as extreme weather events are evolving.

A recent example of this is DTN Storm Risk Analytics. The outage-prediction solution combines seven years of verified, historical outage and weather data with advanced weather and machine learning models that can be tailored to a utility’s operating region and topography without the need for custom modeling.

Consider the advantage of having these insights when planning and positioning response crews when large-scale weather events like Hurricane Ian occur (Figure 2). Not only did the large Category 4 hurricane trigger requests for mutual aid from multiple states to compete for outside assistance, but the predicted track also fluctuated by hundreds of miles as it developed, making it more challenging for businesses, emergency crews, service providers, and residents to prepare.

2. The day before Hurricane Ian made landfall, DTN Storm Risk Analytics predicted 4.42 million customers would experience outages (left) between Sept. 27 and Oct. 4, 2022. That was within 3% of the actual outage count of 4.29 million customers (right). Courtesy: DTN

As the hurricane intensity and track evolved, new real-time information was modeled with updated predictions delivered every six hours to predict customer outages. Utilities using this and other outage solutions driven by AI have insights to better determine likely damage locations and the scope of future outages. It is estimated that by using DTN Storm Risk Analytics, utilities could reduce outage durations by as much as 50%. Ultimately, AI and advanced weather modeling will help a utility prior to a storm’s arrival, rather than reacting in the aftermath.

Keeping a Human in the Loop

While the most advanced tool for helping energy companies plan and respond to outages uses artificial intelligence, keeping a human—or in this case a risk communicator—in the loop is still critical. AI and machine learning are learning from the past, but as energy companies have experienced, the past doesn’t dictate the future. It is a valuable guide that, with evolving data, will become a powerful predictor of extreme weather events. But for today, organizations planning or responding to extreme weather need a meteorologist or risk communicator to add experience and adaptability to the forecast.

A risk communicator is a meteorologist with deep industry expertise (Figure 3), who works with the decision-maker from the planning stage, through the weather event, and during post-storm analysis. A risk communicator can also help the utility assess weather risks for service areas and trouble spots, as well as determine the risk tolerance levels for each identified risk. Increasingly utilities are collaborating with meteorologists to create operational response workshops that include mock extreme weather conditions with action planning, trigger chart development/review, and build out of a communication structure.

3. Meteorologists Brad Nelson and Andrew Polk evaluate a potential tropical storm development to keep a customer forewarned early in formation. Courtesy: DTN

Prepare Now for Extreme Weather of Tomorrow

The triad of robust technology, experienced meteorologists, and the evolution of the science of weather and the climate are the powerful tools that utilities can access and use to better prepare for extreme weather events. As the threat for future unprecedented and record-breaking events inch closer to today, utilities can be more weather resilient by taking the steps necessary and making the agile response decisions needed for the next extreme weather event.

Renny Vandewege is general manager of Weather and Climate Intelligence with DTN.

SHARE this article