Washington’s Columbia Generating Station Is a Seismic Timebomb
New studies reveal nuclear plant is near major faults
This story has been corrected and modified since its original posting.
The landscape of eastern Washington State is deceptively tranquil: a pastiche of vineyards, farms, scrub grass, ridges and windmills. But what appears peaceful and settled in the moment has proven restive and violent over geologic time. Beneath the glacial trough of the Puget Lowland, and extending east through the Cascades to the Columbia Basin, lies a hidden landscape of geomorphic rubble – broken basalt, vast shards of continental rock, volcanic ash, and layers of ancient sediment. Like a picnic blanket spread over a minefield, the Columbia Basin's flat meadows and rolling hills veil an oft-times explosive past.
Photo by the US Department of Energy
Not the best place, you might think, to build a nuclear power plant.
The Columbia Generation Station, Washington's only commercial reactor, sits inside the Department of Energy's Hanford Nuclear Reservation, a former nuclear weapons production site. Powered by a General Electric Mark II boiling water reactor, Columbia began operating in December 1984. In 2009, the industry-funded Institute of Nuclear Power Operations ranked Columbia as one of the country's two reactors "most in need of improvement." Of the 80 unplanned shutdowns (or "scrams") that hobbled the US commercial nuclear fleet that year, Columbia accounted for five.
Today, the reactor has become the focus of a growing debate over the safety of nuclear plants built in seismic trouble spots. The problem stems from the fact that the seismic studies available to the Washington State Public Power System engineers who designed the reactor only ran from 1974 to 1981, and new faults have been discovered since then.
When the atomic plant was still on the drawing boards, there were only two known historic earthquakes that drew concern. In 1872, a magnitude 6.5 to 7.4 quake rumbled through the Cascades, sending massive landslides tumbling into the Columbia River. In 1936, a window-cracking magnitude 5.7 to 6.1 quake opened 200-foot-long fissures in the Walla Walla Valley along the Washington-Oregon border.
After pro-reactor advocates conspired to "locate" the epicenter of the 1872 quake in the North Cascades (180 miles from the proposed Columbia site), the state's engineers only needed to focus on potential impacts of the smaller 1936 quake, whose epicenter was 55 miles southeast of the Hanford Site. This convenient relocation of risk enabled the Nuclear Regulatory Commission to green-light the reactor's construction.
In wasn't until after the 1,170MW reactor went operational in 1984 that scientists began to discover that Washington's seemingly placid landscape masked a troubling and rambunctious past. Initially, geologists thought the state's earthquakes were largely confined to the sea-facing portion of Washington, west of the Cascades. They believed the faults beneath the inland ridges of the Columbia Basin were "uncoupled" – short, shallow and unconnected fractures that posed little risk. We now know that much of the Hanford Reservation is transected by several significant faults.
Geologist Bill Bakun offered a dire assessment of Central Washington: "It’s all riddled with faults," he said. "It wouldn’t surprise me to have a magnitude 6.8 quake anywhere in that region, including near Hanford." In 2002, Bakun and several colleagues uncovered evidence that located the 1872 quake's epicenter at the southern end of Lake Chelan, a mere 99 miles from the Columbia plant. Bakun set the quake's magnitude at 6.8 – with a margin of error ranging from 6.5 to 7. (Other seismologists rate the quake at magnitude 7.4.)
The Lake Chelan quake rocked at least 151,000 square miles and may have been felt as far north as Alaska. Had the Columbia energy station existed when the quake occurred, it most likely would have sustained moderate to severe damage.
Energy Northwest (as Washington State Public Power System is now known) insists that its reactor – built to withstand a "very strong" to "severe" 6.5 magnitude quake – could handle a "violent" 6.9 magnitude event "based on conservative practices in design, manufacturing, fabrication and installation, plant structures, systems and components." But dealing with a magnitude 7.4 quake – nearly eight times more powerful than a 6.9 quake – would be a different matter.
In 2009, a swarm of more than 1,000 mini-quakes shook the Hanford Nuclear Reservation. While the quakes were no larger than magnitude 3.3, they struck close to the surface and produced a significant "peak ground motion." Casting a worried eye toward the Reservation's shuttered Cold War nuclear weapons facilities and its aging radioactive-waste-storage tanks, seismologist Annie Kammerer observed: "Frankly, it is not a good story for us. The plants were more vulnerable than they realized."
In 2013, the Washington and Oregon chapters of Physicians for Social Responsibility hired geologist Terry L. Tolan to review the available seismic research. Tolan's study confirmed that when Columbia was designed, geologists were only aware of six faults. We now know the region is crossed by 12 major faults. Furthermore, these faults are more numerous, much longer, far deeper, and potentially more destructive than previously believed, with the potential to rattle the reactor with forces double those the plant was designed to survive. USGS studies published in 2009 and 2011 revealed that two active faults actually bracket the reactor site, with one running approximately 6.5 miles to the north and another just 2.3 miles south of the nuclear core.
Infographic courtesy of Seattle Times
The Yakima Fold and Thrust Belt, which extends east of the Cascade Range to the Hanford site, is now known to be far more seismically active and interconnected than once believed. In the mid-2000s, the USGS discovered that the belt had produced at least seven magnitude-7 earthquakes with ground motions exceeding the Columbia reactor's design limits.
The region's most dangerous surface fault is believed to be the South Whidbey Island Fault. Unlike most faults, which parallel coastlines and mountain ranges, the Whidbey fault crosses through the Cascade Range, reaching as far as the Tri-Cities in southeast Washington. The South Whidbey Island Fault is composed of a complex band of fractures, with a greater fault track running 200 miles from Vancouver Island to the Cascade foothills. Geologists have found evidence of four major quakes along the fault during the past 16,000 years.
The Seattle Fault – a 44-mile fracture that underlies metropolitan Seattle – now is understood to be part of the South Whidbey Island Fault. Together, they form a system that extends across the Cascade Range to the Hanford Reservation. A 2011 USGS report traced the Umtanum Ridge fault through the Cascades and linked it with the Seattle and South Whidbey Island Fault fracture zones in the Puget Sound area, nearly doubling its length – from around 77 miles to 124 miles.
"The faults don't just end in Puget Sound," USGS research geophysicist Richard Blakely notes. "Our hypothesis is that many big faults in eastern Washington go through the Cascades." Blakely's research suggests that the active faults west of the Cascades actually extend 250 to 300 miles from the Olympic Peninsula and through the Cascade Range, where they merge with the basalt formations of Eastern Washington, at least as far as Pasco – a town located about 20 miles southeast of the Columbia reactor.
As the Pacific Northwest National Laboratory noted in a 2012 report, larger faults can produce more slippage, which can generate larger quakes and more intense ground motion. "If you have a fault system that's 300 kilometers long and you rupture half or a third of it, that's a big earthquake," says USGS geologist Brian Sherrod. "That's a magnitude 7.5."
Contrary to long-held opinion, the "shallow" faults beneath central Washington were found to extend more than 12 miles below the surface. "The faults that formed the ridges are much more dangerous than anyone realized," Sherrod says. "It's a fundamental rethinking of the seismic risk."
A Geological "Train Wreck"
Even without an earthquake, the Pacific Northwest is in constant motion, moving about a half-inch per year. And, with every creeping millimeter of movement, the pressures continue to mount. It is estimated that since 1700, the Northwest coast has moved more than 25 feet closer to Japan. As USGS scientist Ray Wells puts it: "It's a train wreck on a geological scale."
If it is a train wreck, then the "locomotive" is the Pacific Plate, which continues to chug implacably northward at a rate of two inches per year, pulling much of California along for the ride. Rotating under strain and pushed northward, Oregon presses into Washington. But Washington's northward progress is blocked by the unyielding bedrock beneath British Columbia. Pushed from the south and blocked by the north, Wells explains, the Evergreen State "crumples like a line of box cars slamming into a mountain."
"The Puget Lowlands are being compressed by about a quarter of an inch a year," Wells says. "That adds up to more than 20 feet of crunch since the last time the Seattle Fault fired off…. Inexorably, the pressure is accumulating, loading the Seattle Fault and its associates like springs. The squeeze on the Puget Sound region is enough to produce a magnitude-7 quake every 500 years."
Planning for the "Expected" – Not the "Unexpected"
The Columbia nuclear energy plant was not designed to survive a specific magnitude earthquake. Instead, the facility was designed to shrug off a hypothetical "Safe Shutdown Earthquake" with a ground motion of 0.25g (i.e., one-fourth the force of gravity).
As the NRC explains, when a Safe Shutdown Earthquake strikes, "all structures, systems, and components important to safety are designed to remain functional." (This standard may seem a bit wishful since it seems to presume there will never be such a thing as an "Unsafe Shutdown Earthquake.")
In 2011, during the Columbia’s relicensing process, the NRC expressed concern that Energy Northwest was still relying on seismic studies from 1994. Despite protests from citizen's watchdog groups (understandably alarmed by the spectacle of three similar GE reactors reduced to radioactive rubble by the Fukushima quake and tsunami), the license-renewal process remained on track.
In April 2011, a coalition of Northwest public interest groups petitioned the NRC to put the Columbia Generating Station's relicensing application on hold pending an assessment of the new seismic findings. The NRC rejected the petition, claiming that it raised "issues that are outside the narrow scope of the NRC's safety review for license renewal." According to the NRC, the only issues of concern during the relicensing process are those "limited to managing the effects of aging on certain passive structures, systems and components." Seismic reviews are part of "the ongoing regulatory oversight process."
Energy Northwest continues to insist there is no significant danger associated with operating a power plant that has undertaken no structural seismic safety improvements for nearly 30 years. A former Columbia employee (who requested anonymity) believes it would be "virtually impossible to upgrade the foundation to meet the standards that we now know the plant should have." Moreover, the former employee confided: "it would be impossible to upgrade the piping."
In 2012, the NRC announced that, as part of the NRC's "Post-Fukushima Daiichi Lessons Learned" response, a nationwide "update to the seismic hazards assessment is in progress." The review and recommendations are to be provided by the plant operators themselves and are not due until March 2015. When nuclear watchdogs asked why the NRC was prepared to wait four years before assessing the Columbia plant’s potential Fukushima problem, the NRC's Lara Uselding explained there was no cause for public concern since "the NRC knows of no significant changes to possible seismic hazards of the region."
It's not as if the NRC doesn't realized the dangers. In an email to colleagues four days after the Fukushima quake, Brian Sheron, head of the NRC's Office of Nuclear Regulatory Research, referenced some of the alarming findings about new fault hazards in Central and Eastern Washington. This data, Sheron wrote, demonstrated the NRC "didn't know everything about seismicity…. And isn't there a prediction threat the West Coast is likely to get hit with some huge earthquake in the next 30 years or so? Yet we relicense their [nuclear] plants."
These concerns notwithstanding, on May 22, 2012, the NRC relicensed the nearly 30-year-old reactor to continue operating for another 30 years. The decision came at the same time TV screens were broadcasting images of Fukushima's three smoldering GE reactors.
Fukushima's "Lessons Learned" but Not Applied
Anti-nuclear activists in Japan had repeatedly warned of the specific dangers posed by Fukushima's GE-built reactors. Their complaints eerily foreshadowed the very problems that prompted concerns about the Columbia plant. AiIeen Mioko-Smith, director of Green Action, Japan's leading anti-nuclear organization, noted that the Tokyo Electric Power Company (TEPCO) was "operating on 1978 earthquake-resistant guidelines" and company officials had ignored recent studies revealing greater-than-imagined seismic dangers. TEPCO's analysis of the earthquake risk was "unscientific and grossly underestimated," Mioko-Smith charged. "The study's main technique considered fault lines as short and separate threats when they were clearly parts of a much larger system."
Unfortunately, the risks posed by the radiation that continues to spill from the damaged Fukushima reactors appears not to have prompted an "excess of caution" from officials at the NRC or Energy Northwest. In October 2013, Dave Swank, Energy Northwest's vice president of engineering, assured the media: "I don't have any concerns." Echoing the NRC, Swank explained there was no cause for alarm because the "odds" of a major quake were low. Swank also quoted from a letter NRC chief Allison Macfarlane's sent to Physicians for Social Responsibility. In her letter, Macfarlane maintained: "the NRC continues to conclude that CGS has been designed, built and operated to safely withstand earthquakes likely to occur in its region."
Macfarlane's stance on Columbia seems out-of-sync with the NRC's concerns for the rest of the country. As The New York Times recently reported, operators at 24 nuclear power sites in the central and eastern US have informed the NRC they could no longer guarantee their facilities could withstand the greater quake risks predicted by newly updated seismic studies: "The new earthquake threat was larger than what they were designed to face."
After a six-month investigation, Japan's Fukushima Nuclear Accident Independent Investigation Commission concluded the disaster "was the result of collusion between the government, the regulators and TEPCO." The official report also confirmed that at least one of the reactors had succumbed to the forces of the earthquake, not to the floodwaters.
Despite the warnings from Japan, the NRC continues to insist "the newest seismic data suggest that, although the potential seismic hazard at some nuclear power plants… may have increased beyond previous estimates, all operating nuclear plants remain safe with no need for immediate action." Japan's 660-page Fukushima investigation concluded that the crisis was a "profoundly man-made disaster that could and should have been foreseen and prevented."
Let's hope this is never said about the Columbia Generating Station.