A Smarter Power Grid for U.S. Utilities
On a recent muggy morning, Jeff Myerson, director of business transformation for Houston-based utility CenterPoint Energy, is pointing to mundane-looking gray metal boxes high on a wooden utility pole. They mask state-of-the-art wireless relays that collect data on customers’ power use every 15 minutes. Myerson is showing a visitor around Braeswood, a tree-lined neighborhood of several hundred 1950s-era ranch homes about 7 miles southwest of downtown. It’s being refitted for CenterPoint’s “intelligent grid.” This first phase is scheduled to be completed by yearend at a cost of $138 million, with the help of $50 million from the federal government.
When fully deployed, the grid will encompass a network of sensors, switches, smart meters, and data analysis software that will give CenterPoint more control over and insight into its almost 50,000 miles of power lines. Breaks in the lines typically take hours, even days, to find, particularly in the wake of bad storms. Starting in Braeswood, CenterPoint, which powers more than 2 million homes and businesses, is gaining the ability to almost instantly pinpoint a problem to a single block. It plans to use algorithms that can analyze weather patterns and spot weaknesses in the lines to predict failures. The smarter equipment can also reduce the risk of overload by telling customers when energy costs them the least. “We’re bringing a mechanical system into the digital age,” says Myerson.
The U.S. electrical grid, once one of the world’s great marvels, is crumbling after decades of underinvestment. Valued at $876 billion by the Edison Electric Institute, an industry group, the grid is an amalgam of almost 7,000 power plants that send electricity over 450,000 miles of high-voltage transmission lines and 2.5 million miles of feeder lines. All this is managed by 3,300 utilities serving 150 million customers. The grid is arguably the biggest machine on earth. It’s also something of a relic, largely built after World War II from designs that date to Thomas Edison. Half of U.S. homes still have mechanical meters that require workers to show up and read them. Most utilities can’t analyze their customers’ energy use beyond the monthly bill.
CenterPoint is among a growing number of utilities remodeling their pieces of the grid with the kinds of data analysis tools that hardware makers have been hawking to consumers. While talk of a smart grid has been around for years, the utilities’ efforts mark the start of a huge, albeit decentralized, infrastructure makeover that could total almost half a trillion dollars in retrofitting costs before it’s completed, according to the Electric Power Research Institute, a research arm of the utility industry.
The deteriorating grid is forcing some utilities to act quickly. Power outages are up 285 percent since 1984, and the U.S. ranks last among the top nine Western industrialized nations in the average length of outages, which the federal U.S. Energy Information Administration says cost businesses as much as $150 billion a year. Hundreds of thousands of miles of power lines can’t be monitored from a central location, so repairers spend 60 percent of their time searching for breaks. “Many utilities are blind beyond the substation,” says Edward Kennedy, chief executive officer of Tollgrade Communications, which makes grid sensors and consults with utilities on smart-grid projects.
Besides CenterPoint, Columbus (Ohio)-based American Electric Power, Sempra Energy’s San Diego Gas & Electric, and New York Independent System Operator, which runs New York State’s grid, are smartening their systems. Giant PG&E has established an R&D facility with more than 120 engineers and technicians testing ways to speed smart-grid adoption. In Detroit, DTE Energy has begun a five-year, $250 million grid upgrade. Working with Cranberry Township (Pa.)-based Tollgrade, the utility is installing outage-predicting sensors and software that it estimates will eliminate about 500,000 outage minutes for its 2.1 million customers over the next three years. In Austin, Texas, smart-grid researcher Pecan Street is using 300 test homes to show how a utility that tracks power flow can change how and when to generate electricity as well as how and when consumers use it. Ultimately, the goal of these projects is to digitize energy-use data so that power generators and customers can exchange information about demand to eliminate overbuilding and waste.
The smart grid may be a necessity if the utility system is to absorb and manage changes in U.S. power generation. A surge in rooftop solar has turned tens of thousands of consumers into generators, and regulators and green groups are pushing utilities to incorporate more of that power. “There is a lot of technology coming, and if you don’t integrate it successfully, you’re going to have one hell of a mess on your hands,” says Roger Duncan, chairman and one of the founders of the Pecan Street project. “The current grid is just not designed to handle this.”