Cognitive Computing

How Analytics is Helping Utilities Weather Storms

Share this post:

While an aging infrastructure and growing populations makes the electrical grid susceptible to failure, the majority (70 percent) of power outages are caused by the strains of wind, rain, sleet, snow and ice. The negative impact of storm-related outages to the U.S. economy is estimated between $20 billion and $55 billion annually. In a world in which the climate is changing, and major weather ‘events’ are becoming more frequent and stronger, the data suggests the number of weather-related outages will continue to rise.

As an American Meteorological Society Certified Consulting Meteorologist who has worked for a major utility company in the New York metropolitan area, I have witnessed the direct impact weather can have on a utility infrastructure and its customers.

What keeps most utility operations managers up at night is the fear of the unknown. The storm that no one forecasted. The storm that everyone expected but showed up stronger and more ferocious than anticipated.

A good example of the latter was the rare autumn snowstorm in the Northeast five years ago. Due to a warmer than normal month, trees still had significant leaf cover when the snow hit, and the increased surface area allowed it to accumulate more so than on a bare branch. The added weight caused branches to break and even whole trees to topple over and fall onto utility power lines and other infrastructure. So, while the heavy wet snow was forecast, the potential negative impact was underestimated.

That’s just one example of how a complex combination of weather and other factors came together to cause significant utility problems. Utility managers had to examine multiple sources of diverse and mostly unstructured data to figure out the weather forecast. Then, in a separate analysis, figure out what the impact of that weather would be. If they missed one seemingly innocent feature or data point, like leaves still on trees, the whole scenario changes.

Our understanding of weather and weather prediction is continuing to benefit from rapid advancements in computing power to combat risk and uncertainty. This has advanced our insights into the chaotic nature of the atmosphere and our ability to forecast potential outcomes.

For utilities, a cognitive business places a premium on making fact-based decisions using data rather than reacting after the fact. A cognitive business uses every opportunity to interact with data to reason, adapt and continuously learn. Core to this shift is the capability to tap into and integrate countless data sets, including weather data, and combining it with business and external data to add context, depth and confidence to every business decision.

Uncertainty exposes a utility company to risk — the risk of being unprepared to respond with the right amount of trucks and lineman standing by to respond to an event. Being unprepared is costly. Utility companies must decide if and when to call for mutual assistance from other companies.

Other utilities must decide if they are out of harm’s way so they can allow their crews to help others. Restoration crews may have to travel for days before they can help the affected utility.

Enough supplies must be on hand to support the restoration effort. It’s not only utility infrastructure like poles, wires, and transformers, but also supplies such as food and safety gear as well as staging areas to set up equipment. The inverse is true too – preparing for a storm that never happens is just as costly. No one wants to spend money unnecessarily.

Imagine if you could use high-resolution weather forecasts over your specific service territory that is connected to the very data that is necessary to calculate the estimates of impact on your utility system.

IBM’s advanced analytics can help combine the weather, infrastructure and historical impact information saving time and allowing you to focus on proactively responding to the storm.

Lead-time will be in days, not hours, allowing you to make the appropriate calls for restoration resources. Notifications can go out to employees so staffing plans can be developed. Improved customer communications will prepare them for potential outages. This will improve their experience with the utility and reduce regulatory scrutiny. If the weather forecast changes or you want to run scenarios on different storm strength, path, or timing it’s all possible with our cloud based solution. You can even perform signature analyses and compare current weather to past events.

Weather will always be unpredictable, but the power of cognitive insight, combine with analytics can help utilities better prepare, and create new systems that reason and learn over time. Combining weather data with traditional business data from an unprecedented number of Internet of Things enabled systems and devices has the potential to significantly impact decision-making.

This week IBM is at DistribuTECH 2016 at the Orange County Convention Center, Orlando, Fla., to demonstrate how its partnering with utilities to deliver advanced solutions that mitigate uncertainty and risk brought on by inclement weather.

Offering Manager IBM Analytics; AMS Certified Consulting Meteorologist

More Cognitive Computing stories

The Next Enterprise Platform

“What ultimately makes a platform worth using in the long run are the applications that run on it.” – Ben Thompson, Stratechery Platforms drive commerce. Whether in technology or other industries, the creation, acceptance and adoption of platforms spur innovation, efficiency, and productivity. Consider the U.S. Interstate Highway System, which dates back to the 1950’s, […]

Continue reading

University of Oklahoma Taps AI to Strengthen Student Retention Rates

Graduating from a four-year college in four years should be an achievable goal, but only just over 40 percent of students are able to reach this milestone. A critical driver for achieving on-time graduation is first-year retention. But for public institutions, like the University of Oklahoma (OU), the national average of full-time, first-time students who started […]

Continue reading

Bringing the Power of Deep Learning to More Data Scientists

New AI technologies like machine learning and deep learning are fitting ever more snugly into the shifting enterprise landscape. Deep learning in particular is being adopted by an increasing number of enterprises for expanded insights and with the aim to better serving their clients. Thanks to more powerful systems and graphics processing units (GPUs), we […]

Continue reading