By Candace Stuart
Small Times Senior Writer
Electricity disruptions like those that plague California could be averted using microsensors and other small tech devices implanted in transformers and power lines, researchers and utility companies say.
Like good scouts, microsystems will search for signs of deterioration and immediately alert utilities of irregularities. With sophisticated software, they could also reroute power to avoid blackouts.
“You take a big dumb system and make it smarter,” said Harry Roman, a technology specialist at the Public Service Electric and Gas Co. in New Jersey who is involved in a pioneering project to integrate MEMS into the aging power grid. He envisions inexpensive devices that work non-stop to make the grid more reliable and efficient.
A MEMS acoustic sensor will eavesdrop on transformers, reporting if it detects signs of deterioration in oil that insulates electrical parts. The plug-size device will fit into an existing hole in the transformer, listening for a characteristic signal from arcing, which is the flash emitted when current jumps from its course. Arcs occur when the oil breaks down or is contaminated.
Intermittent arcing can go on for years, sapping the system of energy. Intense arcing can cause outages and possibly widespread damage to the grid infrastructure.
“By being proactive, you avoid equipment failures,” Roman said. Damages could cost utility companies and consumers several million dollars if the transformer goes, and up to $20 million if other grid structures are affected, he said.
There are also other repercussions. “It’s an ugly mess that inconveniences people for a long time,” he said.
The New Jersey Institute of Technology is designing the acoustic sensor using its MEMS expertise. NJIT and PSEG helped form a consortium with the state of New Jersey and other utilities and universities to promote energy research and development. PSEG is funding the initial work, but hopes the consortium will participate in building and testing of the sensor. Roman expects to complete the design stage within nine months and begin testing the sensor on a transformer next year.
The Department of Energy is acutely aware that the grid, with some parts almost a century old, is at risk. In a March speech at the national energy summit, Secretary of Energy Spencer Abraham called it “woefully antiquated and inadequate to meet our future needs.” California was but a harbinger for the rest of the nation, he said, if fixes aren’t made.
Sandia National Laboratory in Albuquerque, one of 16 DOE labs, is applying microsystems for another grid component: transmission lines. Its microsensors will measure strain in the grid, for instance on hot days when everyone cranks on their air conditioners. The sensor will identify the stressed points and route the energy path around them, averting power dips or outages.
“This will definitely be part of people’s lives in five years,” said Al Sylwester, a microsensor inventor and manager at Sandia.
The technology can’t come too soon for California, where the region’s network is pushed to its limits — and sometimes beyond — by energy-hungry computer-based companies and the burdens of deregulation.
“The system is being used in new ways, ways it wasn’t designed for,” said Massoud Amin, a researcher with the Electric Power Research Institute, an organization in Palo Alto, Calif. that encourages energy-related technology development.
Nationally, demand for electricity has grown 35 percent in the past decade, and will jump another 45 percent in the next 20 years, according the DOE. But complicating factors ranging from permits to purse strings limit similar growth in energy hardware.
Deregulation adds another layer of stress on the system. Before deregulation, utilities generated their own power relatively near the point of use, drawing on outside supplies in emergencies. But deregulation allows electric suppliers from increasingly wider realms to transmit their wares through the grid, adding yet more traffic.
Working with a team of 28 universities and two utility companies, Amin is creating software that will give the system not just brains but near omniscience.
At the ground level, the program lets communities of sensors share information and collaborate to help the grid run optimally. At the next level, software digests information from the entire network of sensors to ensure each community’s actions benefit the grid as a whole, and corrects them when they don’t. A top level oversees the entire system, reacts to real-world threats like storms or deliberate attacks and adjusts the grid accordingly.
Amin’s goal is to demonstrate the management system in three years, and have a real-world program ready within 10 years. He estimates it will cost a couple hundred of millions of dollars.
“But if it can prevent one big failure a decade,” he says, “it is worth it.”
The last major blackout to hit the U.S. started on August 10, 1996 in Portland, Ore., and spread to 14 states and Canada, affecting 7 million people. The cost was placed between $1.5 billion and $2 billion, Amin said.
The elaborate and ambitious program, which Amin calls a “self-healing energy infrastructure,” relies on its foot soldiers, the sensors. “They need to sense and react to changes in condition, in quality, in demand, and do it in a way that is proactive,” he said. “They’re monitoring and data-mining at the smallest scale.”
For his purposes, they needn’t be microsystems. But industry insiders such as Beau Farmer, industry consultant and director of the Microelectronics Research Center at New Jersey Institute of Technology, consider microsystems a good candidate for any grid because of their size and low cost.
Understanding in real time what is happening inside the grid — and why — is critical, Farmer said. “The future of the power industry rests on knowing this information,” he said. “They need an intelligent system, and a critical component of sensing and knowing will be microsystems.”