Survey: U.S. Utilities Still Feel Underprepared for Weather-Related Outages

8/24/2017

The Weather Company, an IBM Business, and Zpryme have released the results of a survey that found that U.S. utility companies still feel underprepared for weather-related outages and that their reactive approach to outage prediction leads to lost time and resources. Key findings include:

The Challenges of Being Reactive to Adverse Weather and Outage Prediction
According to a survey of 150 energy providers across the U.S., 93% report the impact of severe weather on their operational decisions has stayed the same or grown over the past three to five years. However, due to a lack of solutions in market and the lack of resources to manage and build their own; utilities say they have struggled to add a prediction model to their storm preparation toolbox. As a result, many utilities are forced to be reactive rather than proactive in their storm preparation activities.

Many utilities surveyed (65%) still rely heavily on generic weather forecasting services, which usually are not tailored to their specific weather impact thresholds. This means that advance knowledge of how the weather specifically will affect the utility infrastructure and how many outages will result is not known with much confidence in advance. This hinders preparation efforts that could help utility crews restore customers faster and more efficiently after a storm hits. 81% of the survey respondents say managing an adverse weather response and restoring outages is a top operational cost. Utilities can spend millions of dollars per year² replacing and repairing damaged assets (71%), as well as mobilizing crews (56%).

“Through this research, we found it is critical for utilities to determine the level of impact that weather can have on their system and take the appropriate actions in advance of both major storms and everyday changes in weather patterns,” said H. Christine Richards, vice president of research, Zpryme. “To do this, utilities need to better invest in the right predictive analytical tools that will provide operational and financial benefits for their organizations.”

Utilities Need Advanced Analytics and Machine-Learning Solutions
According to the survey, 57% of utilities feel that they need to improve their data analytics and develop predictive solutions. Most respondents believe that data-driven outage prediction tools will reduce outage duration (70%) and improve customer satisfaction (61%). Additionally, they believe that predictive tools would improve emergency management and preparedness (34%) and help them better pre-position crews (30%).

The Weather Company recently introduced a new outage prediction solution that uses a machine-learning model to combine historical weather data and the most up-to-date, hyperlocal weather forecasts with past outage and/or infrastructure damage information from a utility’s service area. By combining this information, the utility can see within 72 hours of a predicted storm what areas are predicted to be hit the hardest. This is critical in deciding how a utility will mobilize their storm response and where and when to pre-stage restoration crews and equipment if required.

“Using more cognitive, cloud-based solutions – such as outage prediction models – utilities have the opportunity to take a proactive stance against impactful weather,” said Maia Sisk, director of offerings, location and new markets for The Weather Company. “These machine-learning predictive models enable a utility’s operations team to confidently make critical decisions ahead of an anticipated weather event – helping them to control costs and improve restoration times.”

For more results from this survey, read Zpryme’s white paper, Utility Outage Prediction: Embracing Advanced Analytics and Machine Learning Solutions.