Meter data, applied
- Posted on March 1, 2012
- 19 views
Two utility data analytics case studies, one from Florida and one from Wisconsin, handily illustrate the ways in which meter data analytics (MDA) can be applied to other aspects of the utility's day-to-day processes.
Mark Schweiger, a senior business analyst at Florida Power & Light Company (FPL), told Utility Analytics Institute Summit attendees last month that his utility is using meter data from the approximately 2.9 million smart meters already deployed to assist in its revenue protection. (Florida Power & Light is set to complete its 4.4 million meter deployment in 2013.)
"We are taking advantage of the deployment to look at the smart meter data and analyze patterns that may indicate theft," Schweiger said. "We anticipate that revenue protection processes will need to change."
Prior to the installation of smart meters, the utility's theft detection was strictly visual in nature, he said. Now, it's network-reported.
In analyzing the meter data, FPL found three types of electricity theft patterns. The first, Schweiger said, includes those who "pulled" or bypassed their meter for specific periods of time. The second type includes the "weekend people" who pull their meters out on Friday evening, and reconnect them on Monday morning. And the third type includes those who pull their meters nightly.
In the case of a meter bypass, analytics performed on the meter data is able to identify meters where there is an immediate and sustained drop in consumption. "We looked for an isolated outage followed by a drop in usage," Schweiger explained.
But in order to ensure that the analysis was truly effective, front-end work is required. "We have a seasonal demographic, with a lot of variability in type of customer," he said. "So we had to look at 'What is a normal customer? What does a normal customer look like? What is a normal usage pattern?'"
Test effectiveness rates for the analytics solution FPL is using are quite high, Schweiger said, with the utility experiencing an approximate 80 percent effectiveness in detecting diversion conditions. "Our long-term test development strategy will address all major categories of theft," he added, including unauthorized meters, meter bypass and internal tampering, magnet use, self-reconnects and inactive premise consumption. "We will be working with our meter technology center to recreate certain conditions in a lab environment," he said.
FPL has some early lessons learned, too:
- Event notifications (tamper flags) from the meter on their own are not good indicators of theft.
- Usage analysis is necessary to have effective tests.
- Effective and useful tests are proving to be elusive, especially considering FPL's seasonal customers.
We Energies is applying MDA to its meter-to-bill procedures, Peggy Clippert, We Energies' manager of customer research, told attendees. The utility began a mass rollout of automated meter reading (AMR) in 2002 and completed it at the end of 2010. Daily reads of electric meters (residential and small commercial) now total 910,000, and daily reads of gas meters (all segments) total 750,000. We Energies also still does a monthly drive-by read on 560,000 gas and electric meters.
"We were at a crossroads when the AMR deployment matured," Clippert said. "Should we turn around and invest in costly new two-way technology to capture more benefits, or should we deploy another meter data management solution to manage the AMR data?"
Instead, We Energies chose a different approach, deploying an affordable and high-value MDA solution to enhance existing business processes, rather than replacing them and, in Clippert's words, "exploit the value found in all the meter data."
We Energies' MDA solution uses data from metering systems, the CIS, and third party data. Internally, billing analysts, field technicians and revenue protection, customer service representatives, finance and some third-party users are all using the system to access the necessary data.
The UAI Summit case study workshop was sponsored by DataRaker, with presenters including DataRaker president and CEO Rick Brakken and DataRaker director of analytics services Mike Blumberg.
For other UAI Summit coverage, see also Utilities using data analytics to leverage the "data tsunami", A utility challenge: enterprise information management and 'Big Data'" not necessarily a tsunami.
Editor-in-chief, Intelligent Utility magazine
No discussions yet. Start a discussion below.