By Rich Heidorn Jr.

CARY, N.C. — Thanks to smart meters, phasor measurement units (PMUs) and the forecasting challenges of renewable generation, utilities and RTOs are becoming increasingly voracious consumers of data.

“Instead of getting SCADA data every four seconds,” notes Stephen Rourke, vice president of system planning at ISO-NE, “we’re getting PMU data every two cycles.”

But some are not using the information as well as they could, speakers said at the RTO Insider/SAS ISO Summit at SAS headquarters last week.

“The amount of operations data that’s being created is just incredible,” said Bill McEvoy, industry principal for OSIsoft. “And a lot of it is still being created in silos.”

That’s a mistake, said Jill Dyché, vice president of best practices at SAS. While many companies perform only “random acts of analytics,” forward-thinking organizations have created analytics “Centers of Excellence,” she said.

“Analysis is a collection of very distinguished skill sets that may not exist elsewhere in the organization, so there’s an argument for leveraging those in a sustained way through some kind of COE or marketplace,” she said. “For energy companies, the potential is not just in optimizing our business but also becoming data businesses. There’s a huge potential in using our data in fresh new ways.”

Data Quality

Most companies fail to measure the cost of poor, missing or inaccurate data. “Typically when I get a ‘yes’ on that question, it’s after the fact,” Dyché said. “We realize that a business initiative failed because we didn’t have the data, or the data was wrong,” she said. “Data quality can make or break some pretty serious decisions these days.”

“I think it’s wise to watch where your numbers come from,” said Brad Lawson, a SAS industry consultant. “Working in the utility industry, we could never get customer numbers to match. You went to one group and there was a customer count of whatever. The next group may be 10,000, 15,000 more. So we finally got a group together to talk about, what is a customer? What we found within this utility was that we had about seven different definitions of ‘customer.’”

Similar disparities exist in data on solar capacity and generation, he said, complicating utilities’ forecasting challenges.

Forecasting EV Charging

Electric vehicles are a growing concern for utility forecasters, particularly in California, which has more than 40% of the EVs in the U.S.

“The utilities would like to be able to manage that charging,” said Ralph Masiello, senior vice president of Quanta Technology. “In other words, if everyone comes home from work [and] plugs the car in at 6 p.m., that’s right when that duck curve is ramping the worst. So it’s the last time that you want another 1,000 MW of EV charging.

“The big data need for the utilities is how do they know when those customers are going to plug in? And the answer is they’ve got to monitor [the California Department of Transportation] for traffic conditions. Because a traffic accident on Interstate 405 can mean 1,000 Teslas plugging in a half-hour later than they normally would.”

Data Volumes Taxing Hardware

As one of the biggest users of data analytics at ERCOT, Manager of Demand Response Carl Raish is looking forward to a refresh of the ISO’s hardware platform. “I’m hopeful that this hardware change is going to make my life better. I’ve done what I could with software in terms of trying to leverage capabilities that are in SAS to make code run better. But it’s really the volumes of data [that are challenging] at this point.”

Helping to provide answers to RTOs’ challenges is the Electric Power Research Institute. EPRI’s Market Design research group has created webcasts and an ISO/RTO Market Design Tech Forum, in which technical market designers discuss the challenges of changing markets. Its Market Design and Operations Research Program provides analytical support for research projects.

Erik Ela, senior technical leader of EPRI, said the organization is tackling the industry’s biggest research and development challenges, including providing adequate compensation to prevent retirement of resources that are not used often; incentives to encourage system flexibility; pricing schemes; and incorporating policies that favor technologies for reasons other than cost.

ISO markets currently dispatch resources based on reliability and cost, Ela noted.

“But there are a lot of other things that we don’t do at the ISO. We don’t have an environmental [input in market clearing engines]. We don’t care about job preservation. … Fuel diversity is very hard to quantify. So there’s a lot of these other aspects out there that a lot of the states have a lot of incentive to try to keep …  [But] that’s not built into the way we clear our markets.

“So how do we interact with these policies in the way that we are running the markets? That’s a big area and [one] I think that we’ll see more and more questions about.” (See related story, Ott Seeks ‘Resilience’; Clark Handicaps ZECs.)