I sat in on a briefing yesterday in which understanding outages and improving outage and restoration notification came up, so I thought I'd share. This topic has figured prominently in some of our coverage.
Connecticut, hit badly by storms last fall, has actively pursued reliability solutions in the form of microgrids. See "Connecticut: In Search of Microgrids."
Dayton Power & Light got dinged by local media for not implementing interval meters and advanced metering infrastructure (AMI). See "Dayton Power & Light: An Exception Worth Considering."
And while we lauded Baltimore Gas & Electric for its live blogging and responses to customers last fall, BG&E subsequently got an earful from both public and regulators during hearings on the utility's storm-related response. See "BGE Handles Howling, and Hurricane."
Stakeholder expectations in general—the public and regulators in particular—have raised the bar for utility handling of outages, an ongoing trend propelled by last fall's hurricanes, tropical storms and early winter weather, particularly along the Eastern Seaboard and across the Northeast, Accenture executives noted.
The advent of advanced metering infrastructure, meter data management and analytics were supposed to be a boon for outage management, right? Well, while that statement is correct, the Accenture team pointed out that a utility's understanding of an outage—and thus its restoration work and public notification—isn't as simple as it sounds.
"With the storms of the past year or two, folks want more precise information—they expect that when an 'estimated time to restoration' is communicated that the utility either delivers on that promise or quickly manages the expectations throughout the process," said Wade Malcolm, Accenture's senior director for Smart Grid Operational Technology.
"Originally, when AMI systems started going in, the engineering side of the house on the transmission and distribution side, said, 'Hey, if we manage our systems properly, the data from the AMI system really shouldn't yield any substantial changes to knowledge or awareness of an outage," Malcolm said.
Some Accenture clients have begun working with data warehousing and gone back to valid interval data and correlated that with other knowledge they have about an outage as recorded in their outage management system and they're seeing differences, the executive said.
That means there's an opportunity to improve the resolution of knowledge about an outage by applying data recorded by the AMI system. There's been a lot of industry learning around how to integrate that meter data management system into an outage management system as well.
"In the 'old days,' people tried to bolt the head-end of the MDM right into the outage management system," Malcolm said. "Say a lightning storm went through and you'd have a momentary outage. Depending on the types of smart meters and communications systems you're using, it's conceivable you could have thousands of 'I'm-out-of-power' alarms immediately followed by thousands of 'I'm back,' based on momentary interruptions due to the lightning impact. That was enough to cripple some of the early integrations and bring that outage management system down. So there's been a lot of work to filter some of those alarms."
Other utilities have gone in a different direction, opting to turn off the 'last gasp' notification and approach the situation with different logic. Some utilities struggle with being able to discern whether they've actually got nested outages or whether they've simply lost communications with end points. Systems whose sophistication increases over time due to correlating events and outcomes can narrow the cause of ostensible outages as well.
"We're still relatively early in the learning process," Malcolm concluded. "What's happened, more broadly, is that you have new tools, new streams of data, just on the outage management side. We're seeing new applications, new dashboards being made available that allow information to be integrated more effectively. We're also seeing folks doing detailed analytics to understand the particular issues they have. We're just at the early stages of benefits from bringing meter data in and taking advantage of it."
These questions are important, of course, because in many cases utilities have sold smart grid-related capital investment and cost recovery to regulators and the public based on purported benefits. In the case of AMI, those business cases include improved outage management and restoration notification. That interval meters and foundational IT systems will improve outage management has become axiomatic in the industry, as the Dayton Power & Light case illustrates. (DP&L missed out on a stimulus grant and couldn't make the business case work, so it postponed advanced metering, only to be dinged by local media for not pursuing technology that could improve its response to storms and resulting outages.)
Perhaps until the integration and business intelligence issues are ironed out, this area will remain, as mentioned, an exercise in managing expectations. That cannot go on indefinitely.
Intelligent Utility Daily