Why Japan's Fragmented Grid Can't Cope

Editor's Note: This is part of IEEE Spectrum's ongoing coverage of Japan's earthquake and nuclear emergency. For more details on how Fukushima Dai-1's nuclear reactors work and what has gone wrong so far, see our explainer.

6 April 2011—The earthquake and tsunami that destabilized Japan’s Fukushima Dai-1 nuclear power plant last month also blew a large hole in the country’s power supply. Eleven nuclear reactors in eastern Japan shut down, including three that were running at Fukushima Dai-1 and four at the nearby Fukushima Dai-2 plant. In all, more than 27 gigawatts of power generation were out of commission, forcing Tokyo Electric Power Co. (TEPCO)—operator of the Fukushima reactors and power supplier to greater Tokyo—to ration power by instituting rolling blackouts.

TEPCO’s supply situation would look less grim were it not for a quirky split that divides Japan’s power grids in half: While Tokyo and the rest of eastern Japan run on 50-hertz electricity, the big cities southwest of Tokyo and the rest of the country run on alternating current that cycles at 60 Hz. It’s a historical accident from the 19th century, when Tokyo’s electrical entrepreneurs installed 50-Hz generators mainly from Germany, while their counterparts in Osaka selected 60-Hz equipment from the United States. The result is a national grid whose two halves cannot directly exchange AC power, which limits TEPCO’s ability to seek help from the 56 percent of Japan’s power-generating capacity that lies to the west.

"It’s a shame. The western grids can supply a lot. I think they could cover [TEPCO’s] peak demand," says Kent Hora, executive vice president for Mitsubishi Electric Power Products, the U.S. arm of Japanese power-engineering giant Mitsubishi Electric.

As it stands, just three small installations can squeeze power across Japan’s AC frequency frontier. These are converter stations that use high-voltage electronics to pull alternating current off one grid, convert the power to high-voltage direct current (HVDC), and then synthesize a novel AC wave to add the power to the other grid. Together these three facilities can push up to 1.2 GW of power east or west. TEPCO is using them at full capacity, says Junichi Ogasawara, a senior researcher at the Institute of Energy Economics, Japan (IEEJ), a Tokyo-based think tank.

Analysis by Ogasawara’s group, however, shows how short that leaves the utility. TEPCO has mapped out a plan to boost power output from less than its present level of 40 GW to at least 50 GW this summer, largely by reactivating idle coal-fired power plants, including roughly 900 megawatts of generating capacity at steel mills operated by Nippon Steel and Mittal. That leaves TEPCO projecting an 8 to 9 GW shortfall under a summer peak load of up to 60 GW.

The embattled utility hopes to make up some of the gap before the July-to-September peak season with the express installation of extra gas-fired turbines at existing TEPCO plant sites. But IEEJ is betting on more rolling blackouts. "TEPCO is making utmost efforts to expand its supply capacity. Still, a considerable power shortage is expected," according to a report issued last week by Ogasawara’s group [PDF].

The prospect of ongoing generation shortfalls has Japan’s Ministry of Economy, Trade and Industry and its grid managers hatching plans to beef up its west-to-east power flow capabilities. The government is looking to have additional capacity in place in two years, according to a ministry official quoted by Bloomberg last week.

Some independent experts are more bullish, arguing that new converters could be moving power in the summer of 2012. "Under normal conditions, these kinds of systems would take 18 to 24 months. Could we get one installed and in service in less than 12 [months] in an emergency situation like this? Absolutely," says Gregory Reed, a power engineering professor at the University of Pittsburgh and director of its Power & Energy Initiative.

One way to move faster is to use systems more advanced than the traditional HVDC technology employed in Japan’s three operating converter stations. (Some of these stations were leaders in their day. When it started up in 1965, the 300-MW Sakuma station was the first example of back-to-back use of HVDC converters to synchronize AC grids. And the 600-MW Shin-Shinano station pioneered the use of photo-triggered thyristors when its original power switches were replaced in 1992.)

The best way forward, according to Reed, is voltage source converter (VSC) technology. VSC-based HVDC uses relatively advanced switches, such as insulated-gate bipolar transistors, to simultaneously transmit DC power and regulate the voltage of neighboring AC lines. This flexibility has made VSC increasingly popular for use in merchant power lines and links to offshore wind farms, as well as in the advanced flexible AC transmission systems, or FACTS, that moderate power flows on AC networks.

VSC was introduced commercially in the early 1990s, and Mitsubishi Electric demonstrated its use for frequency conversion in 1999, when the company installed a 37.5-MW system at Shin-Shinano. Nevertheless, Japanese utility Chubu Electric Power went back to traditional HVDC for the country’s third converter, the 300-MW Higashi Shimizu station that powered up in 2006.

VSC technology costs about 25 percent more on average than traditional HVDC—a premium worth paying in Japan’s situation in exchange for what Reed and others predict would be faster installation.

Photo: TMT&D

Click on the image for a larger view.

One reason VSC might get up and running more quickly in Japan is that it requires less space than HVDC systems. That’s because VSC produces a cleaner synthetic AC wave and therefore requires less filtering equipment. "You can take about 35 percent out of the footprint of a conventional installation," says Reed. That smaller footprint would help squeeze converters into sites adjacent to the three existing converter stations. It would also simplify any required expansion of the AC lines feeding the stations; new lines could run along the existing transmission rights-of-way.

Jan Johansson, an HVDC expert with European engineering firm ABB, stresses another VSC advantage—the fact that VSC-based converters do not require the extensive grid modeling studies that go into siting and integrating traditional HVDC stations. Japan’s present HVDC needs the AC grid to be stabilized in order to avoid voltage fluctuations, but VSC systems actively stabilize the grid on their own. "It is very much less sensitive to different properties of the AC networks," says Johansson.

How much converter capacity should TEPCO install? Mitsubishi’s Hora advocates a major expansion to make Japan’s grids more flexible to respond to future emergencies. These potential calamities might otherwise be out of reach of the added generating capacity TEPCO is installing on the eastern grid. "It may happen on the western side in the future. We have to prepare for that," says Hora.

In contrast, power systems expert Akihiko Yokoyama, an electrical engineering professor at the University of Tokyo’s department of advanced energy, says Japan must take into account cost-effectiveness—even in the current urgent situation. Yokoyama endorsed the suggestion last week by officials at the Electric Power System Council of Japan, a grid-management body, to add a modest 300 MW of additional east-west exchange capacity. More-ambitious proposals to double the exchange capacity, he says, are both excessively expensive and impractical.

Yokoyama says adding 1 GW of capacity could require an additional high-voltage line on a new right-of-way, regardless of the HVDC technology employed, thus pushing the total price tag to US $2 billion or more (roughly $600 million for the converter station and $1.4 billion to $2 billion for 200 kilometers of 500 000-volt transmission). That may be more expensive than called for if, as he suggests, such a large loss of power generation "occurs once [every] several hundred years."

The question may be moot, anyway, bets Yokoyama. While Japanese citizens suffering in the wake of last month’s natural disasters appear stoic, Yokoyama nevertheless expects any new transmission lines to confront "NIMBY" (not-in-my-backyard) opposition. If a new line is required and protests erupt, Yokoyama says the project could take a decade to complete. "It is very difficult to find the right-of-way for transmission lines and to get agreement from the residents near the lines," says Yokoyama.

Johansson says HVDC technology offers an answer to that, albeit one that will further increase cost. "It doesn’t have to be through a back-to-back station," he says. "There can be a transmission link in between." Imagine HVDC converter stations at shore points on the east and west grids, pumping DC power through a well-hidden subsea line that snakes its way around the island. According to Johansson, ABB delivered one such subsea line using VSC technology, linking Estonia and Finland. The installation took less than 20 months—including laying the cables.

At that pace, two summers would pass in Japan before the extra capacity kicked in. But eastern Japan may still need the power then. TEPCO has already announced that it will scrap at least four of the six reactors at Fukushima Dai-1. It took two years for TEPCO to gain approval to restart its Kashiwazaki-Kariwa nuclear plant, which suffered comparatively minor damage in a 2007 earthquake. In fact, given rising protests against nuclear power, and the contamination of surrounding communities, it is an open question whether any of Fukushima’s reactors will ever be allowed to run again.

This article was modified on 15 April 2011.