Bacon's Rebellion

Towards a Smarter Grid

Dominion Virginia Power is using big data to increase the reliability of its electric distribution network. The result: Fewer disruptions and shorter outages for customers.

by James A. Bacon

One day earlier this month, 10,000 people living in Fairfax County lost their electric power around 7:30 a.m. Thanks to sensors and devices that Dominion Virginia Power had installed in its electric distribution system, company operators were able to quickly identify and isolate the problem. Fifteen minutes later, they had restored service to 9,000 residents; within half an hour, electric power was back online for everyone.

If you’re a Dominion customer and it seems as if you’re suffering fewer and shorter electrical outages, it’s not your imagination. Harnessing data to target maintenance spending with better precision, the power company has made a concerted effort over the past decade to improve the reliability of its electric service. Since 2008, Dominion’s 2.5 million customers have experienced a 26% decline in minutes lost to routine service disruptions (excluding major storms) when calculated on a three-year rolling average.

“We’re data driven. We’re a six-sigma company,” says Steven Chafin, director of reliability. Dominion, he says, has evolved from a company employing rough industry rules of thumb to one that collects data across the distribution system to drive continuous improvement.

System Average Interruption Duration Index (SAIDI): three-year rolling average of annual customer minutes without electric service, excluding major storms.

Comparing reliability performance to that of other power companies is difficult because each utility contends with different terrain, weather and settlement patterns. Dominion benchmarks against itself, tracking the performance of 35,000 miles of overhead electric lines, 22,000 miles of underground lines and thousands of miles of high-capacity transmission lines. The numbers exclude major storm events, which are so random and are of such a magnitude as to obscure trends in routine operations. Five different events since 1998 — the Christmas Eve ice storm, Hurricane Dennis, Hurricane Isabel, Hurricane Irene and the Derecho wind storm — caused massive outages that took eight to fifteen days to fully restore.

A 2015 J.D.Power survey of electric utility customers found that the quality and reliability of electric power is a major factor influencing customer satisfaction nationally. Dominion scored a customer satisfaction ranking of 684, above average for large utilities and an improvement from 661 in 2013.

The heart of Dominion’s reliability initiative is a portfolio of more than a dozen programs ranging from tree-and-brush clearance to the upgrading of neighborhood transformers. “We adjust our investment in these programs annually,” says Chafin. The company allocates capital to programs that offer the greatest bang for the buck.

In 2015 more than a quarter of Dominion’s reliability spending was dedicated to clearing trees and brush, the greatest source of downed power lines, Chafin says. The company once hewed to a regular, three-year cycle for pruning vegetation near overhead lines. It was a reasonable rule of thumb, but analysis of the data showed that tree-related outages could be improved — some places need trimming more frequently, others less often, depending upon how fast the trees grow and the voltage of the line, among other factors. (Higher-power lines increase the chances of electricity arcing from the line to a nearby tree.)

The second largest reliability initiative is the “capital asset rebuild” program. Much of Dominion’s capacity was installed in the 1930s, ’40s and ’50s, an era with less advanced technology and less rigorous performance standards. Improving the design of older distribution facilities and rebuilding them to current standards reduces the number of disturbances.

Under its circuit reconditioning initiative, Dominion began ranking the performance of each of its 1,800 circuits for preventable failures and prioritizing the worst performers for fixing. Then, diving deeper, the company started collecting data on breakers, reclosers and fuses in each circuit, some 180,000 devices in all.

The neighborhood transformers program addresses increasing demand in residential neighborhoods. In the 1930s, ’40s and ’50s, the company installed distribution lines and transformers to meet electricity demand of an era before air conditioning, big-screen televisions, computers, appliances and other energy-sucking devices. The same houses today draw more electricity and transformers can get overloaded. To deal with the problem, Dominion has replaced 4,000 transformers during periods of low demand. That proactive maintenance, says Chafin, is preferable to waiting until a transformer blows out on a hot summer afternoon.

As much as electric customers fume at routine disruptions, the lengthy storm-related outages are the ones they really remember, says Le-Ha Anderson, manager for media relations. “When electricity stops, life kind of stops.”

Last year Dominion submitted to the State Corporation Commission (SCC) a plan designed to reduce the length of major-storm outages by burying critical stretches of overline wire. High winds blow branches and other objects into power lines and knock them out. Undergrounding the electric distribution system statewide would be prohibitively expensive — on the order of $83 billion, according to an SCC report. That would amount to thousands of dollars per customer and cost generations to complete, Chafin says. The benefits don’t justify the cost.

After examining the data, however, Dominion discovered that 20% of overhead lines were responsible for a disproportionate percentage of outages and time lost. By burying just the most outage-prone tap lines (the small lines that stem from feeder lines to individual houses), the company calculated, it could cut average restoration times after major storms in half. Dominion said the program would cost only $2 billion up front. (An SCC report concluded that the program would cost customers $6 billion in capital costs, property and income taxes, and financing costs over the life of the assets.)

In a hearing before the SCC last year, Dominion asked the commission to approve the first stage of the project, covering 526 miles of tap lines at an initial investment of $263 million. The project would require a rate adjustment of $24.4 million in the first year, less than $1 per month on the average electric bill.

Consumer groups opposed the petition. Testified the Attorney General’s consumer counsel:

Despite the unprecedented size of the proposed [Strategic Underground Plan], the company has not conducted a cost-benefit analysis, has not provided any estimate regarding reliability improvements or economic benefits to customers, and has not considered any lower-cost alternatives.” Based on this record, we cannot conclude that it is reasonable, and in the public interest for Dominion to invest $263 million — and ultimately to charge customers over $700 million — for the first portion of the SUP…

The SCC proposed instead that Dominion conduct a pilot program targeting tap lines with the worst reliability record to gather data for a realistic cost-benefit analysis. Dominion has not publicly said whether it plans to submit a new proposal.

Meanwhile, the company is applying a cautious test-and-learn approach to integrating solar energy into its energy mix. Dominion worries that the intermittent shining of the sun creates fluctuations in voltage that could disrupt the transmission and electric systems. With new solar projects in its North Carolina service territory and the proposed Remington industrial-scale facility in Virginia, the company is building experience with solar that will enable it to model the impact of larger-scale projects in the future, says Anderson.

Whatever the future of solar facilities and underground lines, Dominion’s collection and analysis of data is likely to improve reliability in routine operations for years to come.

There are currently no comments highlighted.