A Microgrid That Wouldn’t Quit
How one experiment kept the lights on after Japan’s earthquake
Editor's Note: This is part of the IEEE Spectrum special report: Fukushima and the Future of Nuclear Power.
26 October 2011—On 11 March, when the 9.0-magnitude earthquake struck off the northeast coast of Japan and triggered one of the deadliest tsunamis the world has ever seen, the bustling port city of Sendai was directly in harm’s way. The port was destroyed, the airport was swamped, and waves reportedly rolled 8 kilometers inland, killing hundreds of people. While downtown Sendai escaped heavy structural damage, activity in the city ground to a halt. Every traffic light and office lamp went dark after the wall of water knocked out the electricity grid for the entire city. In some areas, the outages would last for weeks.
But in one small section of the city, the lights stayed on. At Tohoku Fukushi University, in the northwest part of town, the laboratories’ servers kept on humming, the clinic’s MRI machines didn’t lose a tesla, and the hospital’s lights and equipment operated without a hitch. These facilities were the beneficiaries of an experimental microgrid project fed by three types of energy generators—fuel cells, solar panels, and natural gas microturbines. Because the project also uses the thermal exhaust from the gas turbines and fuel cells to heat the buildings, the hospital’s patients were kept warm through northern Japan’s cold March nights.
The project was intended to demonstrate a microgrid’s potential to improve power-supply reliability. It seems safe to say that its performance in the wake of the Great East Japan Earthquake provided ample proof.
The Sendai microgrid is a small distributed energy system with a total output of only 1 megawatt, but its setup yields certain advantages that make up for its diminutive size. For example, the microgrid’s power sources are close to its customers; this arrangement decreases energy losses during transmission and allows for the distribution of heat as well. And although the microgrid is connected to the larger "macrogrid," its independent energy sources make it less vulnerable to problems in the larger system.
The resulting reliability makes microgrids attractive to customers such as hospitals—which need a guaranteed, uninterrupted flow of power—while allowing utilities to avoid costly across-the-board improvements.
"Today, people have no options," says Keiichi Hirose, the head of the Sendai microgrid project. "The idea is to provide some options for electricity." In the Sendai system, customers pay different rates depending on the level of reliability they need.
The Sendai project was proposed and carried out by NTT, Japan’s largest telecom company, with initial funding from a government R&D agency. But even after the government’s four-year funding commitment came to an end in 2008, NTT decided to keep the microgrid alive and operational. Hirose, a senior research engineer with NTT Facilities, says the company is experimenting with distributed energy because it is looking for new ways to power its massive data centers.
The power center that forms the heart of the microgrid is noisy and industrial but oddly homey. The tops of the natural gas turbines’ cooling towers are covered with green nets to keep out cherry blossom petals from a row of trees that flower spectacularly each spring. The control room, packed with computers and ranks of batteries for energy storage, is housed in an undistinguished metal building. Still, workers trade shoes for slippers when they go inside.
Under normal circumstances, the Sendai microgrid is a bit player in the larger story of energy distribution throughout the city. It’s connected to the area’s macrogrid, quietly contributing its solar, fuel-cell, and natural-gas power. But when an outage occurs in the macrogrid, the Sendai project goes into "island mode." Its connection to the macrogrid is temporarily severed, and the energy produced onsite is routed directly to its customers. Says Hirose: "After March 11, the whole city of Sendai, including the downtown, had two days of outage. So for about two days, our microgrid system operated in island mode."
Sendai’s macrogrid was brought down by a combination of generation and transmission problems on 11 March, explains Alexis Kwasinski, an expert on disaster forensics at the University of Texas at Austin’s department of electrical and computer engineering. The earthquake triggered automated shutdowns at nearby nuclear power plants; the seismic forces also wrecked a substation and four transmission towers. The tsunami followed hard on its heels, wiping out a coal-fired generating facility and another substation.
"The microgrid is a local concept, and that’s what makes it important," says Kwasinski, who has visited the Sendai site twice since the earthquake. "Even if you lose all the infrastructure outside your immediate area, the microgrid can keep operating."
But for every silver lining, there is a cloud. And for microgrids, the downside is their initial costs. The Sendai project was blessed with government funding to finance the purchase of power sources and their installation. But start-up costs will pose a problem for commercial entities that try to follow NTT’s example without government aid. "Maybe we could reduce initial costs if we develop a standard or a cookie-cutter system," says Hirose.
Kwasinski agrees that adoption of microgrid technology will be slow until capital costs come down. "But I see value in specific applications, where you have initial costs that can be offset with gains in reliability," he says. He notes that in the United States, the Department of Defense is working with companies such as Boeing, Lockheed Martin, and Siemens to create microgrids for military field stations and tactical operations centers. When people place a premium on reliability, he says, the benefits of these miniature grids can far outweigh the costs.