Now, Exelon is now getting a bigger net. The power company’s utilities division just signed a deal with GE Power to sift the data with apps built on Predix, GE’s software platform for the industrial internet. “Predix is going to dip into the data lake and make the most use of it by pulling out the most helpful data,” says Andy Gay, the Exelon program manager for GE Power.
A good place to start is with slashing maintenance-related costs and downtime. By crunching current performance and historical data, Predix apps can suggest when to repair or replace specific parts. This way, Exelon can avoid costly downtime and reduce maintenance costs by taking equipment out of commission when necessary, rather than on a fixed maintenance schedule.
Exelon also can use Predix software to better prepare for storms and other forces of nature that lead to power outages. Using data from previous tempests, such as wind speeds, information about trees and other vegetation in the area and maintenance records, Predix apps can identify probable trouble spots and what Exelon will need to do to restore customers’ electricity faster. “Instead of waiting to see where the outages are, they’ll be able to get crews and equipment in place before a storm hits,” Gay says.
What’s more, Predix software will also help Exelon balance electricity generation as the share of renewable energy grows. Energy grids need to maintain a delicate balance of supply and demand or risk shutting down. Renewables make it especially hard to maintain that balance because the wind doesn’t always blow and the sun doesn’t always shine. Predix apps will help Exelon predict demand and know when and to what extent to engage the gas-fired power plants that must pick up the slack when renewables waver.
The seven-year program will start by looking at five major categories: network connectivity, asset health, outage prediction, storm response and outage history. Exelon has the opportunity to add 13 other electricity grid analytics over time. “At the beginning, there’s going to be a massive training push,” Gay says. “But once they get used to the system, they’re going to be able to develop their own (algorithms) to perform tasks that haven’t been thought of yet.”