Big data is becoming more useful every day as we learn to collect an abundance of information and analyze it more efficiently. There are things we can learn from data collection that we would never have expected years ago. One of the most recent uses for big data is to help predict power outages in the event of a disaster.
Its not about whether or not power outages will happen. In many cases, they are inevitable and customers understand as much, especially during big storms like Hurricane Sandy, where affected areas were left without power for weeks. What its actually about is the timely restoration of said power.
How Big Data Can Accelerate Power Outage Recovery
When Hurricane Sandy hit in 2012, it devastated the northeastern part of the U.S. and left millions without power for days, some for weeks.
Worse yet, it revealed several vulnerabilities in the power systems used in affected regions and offered a real-world example of how slow these power companies can take to react after a disaster. The Kinetic Analysis Corp. conducted a pilot study in 2006 for the Long Island Power Authority, which pointed out some of the vulnerabilities and inconsistencies in its system. One thing the study revealed was that there would be a delayed, maybe even prolonged recovery period in the event of a failure.
Of course, LIPA claimed the information was inaccurate and it would only take a maximum of 10 days to restore power to its coverage area. As we now know, the study was much more accurate.
Naturally, companies are now working on big data systems that can help alleviate some of these issues in the future. GE was one of the first companies to introduce an analytics solution designed specifically to help power companies manage power outages and recover faster.
Introducing GEs PowerOn Response System
Colorado Springs Utilities in Colorado will be GEs first customer to implement the new analytics system, as its currently rolling out the first phase of the new software which handles most of the damage and outage reporting. The next phase will handle predictive analytics to allow officials to react accordingly in the event they need to plan for an outage or system failure.
GEs software is called PowerOn Response, and its designed to collect a variety of information from field and network service crews. It can gather and analyze data about facility damage, circuit failures and conditions, devices and equipment, and restoration periods for customers. The data can also be shared with several sources, including management teams, utility crews and even customers.
The predictive side of the software can obviously help teams prepare for future scenarios and identify weak points in the system. Field crews can react by replacing old equipment, retrofitting old circuit breakers and repairing damaged devices. The goal is to help these teams and companies avoid power failures and outages as much as possible, or at the very least recover faster when something does go wrong.
All of this software and collected data should help prevent wide-scale outages like the Northeast blackout of 2003. It left many customers in the U.S. and Canada an estimated 55 million people without electricity for nearly 16 hours in some locations. The worst part was that it was easily preventable, and was augmented by several factors, the two most relevant of which were miscommunication and poor coordination between service teams.
If a big data system was available, the scenario might have gone a bit differently. That is exactly what GE and its power-related customers are betting on. If it can quell and prepare for failures before they happen, it can ensure that people arent left in the dark for long, if at all.
Image by Iwan Baan
