Did Pacific Gas & Electric sidestep what some people consider to be critical seismic requirements for Diablo Canyon nuclear power plant following the discovery of the Shoreline Fault? Whether that’s even a valid question to raise depends on to whom you’re talking.
The U.S. Geological Survey’s 2008 Shoreline discovery prompted PG&E to quickly assess whether the fault—a roughly 12-kilometer-deep, 24-kilometer-long vertical strike-slip that intersects with the nearby Hosgri fault to the north, and lies some 600 meters from the plant’s power block—is capable of producing an earthquake large enough to damage vital plant components.
A subsequent PG&E report found that the Shoreline is capable of producing up to a magnitude 6.1 earthquake. The plant was built to withstand a 7.5-magnitude event.
But practical application of knowledge of the fault, as well as a hazy grasp on how it could interact with neighboring faults, turned into a years-long internal Nuclear Regulatory Commission debate focusing on the plant’s seismic safety design basis.
An operable plant’s safety components must be theoretically capable of functioning even after the impact from seismic events likely to occur at a facility given its known neighborhood hazards. Most plants have two design basis testing requirements to meet. Diablo Canyon is the only plant in the country required to meet three requirements: the Design Earthquake (DE), the Double Design Earthquake (DDE), and the Hosgri Event (HE). They differ by something called dampening assumptions, which are standards for how the plant’s components react to different ground motions. The DDE is considered the most conservative criteria, in that it’s the hardest to pass.
The worst-case scenario that a plant’s shutdown components could withstand is considered a “safe shutdown earthquake.” So which of the three applies to Diablo?
“One of the requirements we look at for it to be operable is that it has to be able to work after the safe shutdown earthquake,” Michael Peck, the NRC’s senior resident inspector for Diablo Canyon until 2012, told New Times from his new office at the NRC’s technical training center in Chattanooga, Tenn.
Following the release of the final Shoreline report, Peck, who’d been the plant’s senior inspector since 2007, questioned the way PG&E was plugging the new seismic data from the Shoreline into those three criteria. In short, he found that they weren’t—and he began to ask questions.
“That struck me as odd,” he said. “Then I realized that they were looking at it probabilistically, but we’re licensed for deterministic.”
What he means is that PG&E seismologists were looking at the probability of a worst-case earthquake, which essentially allowed them to weight the worst seismic hazard on how likely it is to occur—as opposed to how damaging it would be should it occur.
“When I asked them about it, there was no answer,” Peck said.
More than 100 internal PG&E e-mails and call logs obtained by New Times confirm that Peck’s questions caught PG&E staffers off-guard. Immediately after his first inquisition, Diablo chief Loren Sharp began rallying his staff to create an argument countering Peck’s concerns.
“Be aware [Peck] will start asking about how we know we can safely shut down with the new spectra (ie, do we meet DDE),” a plant manager—whose name was redacted—advised Sharp in a Jan. 6, 2011, e-mail. “I suggest we’ll have to keep him focused on addressing safety and capability vs. licensing compliance.”
In another incident, when Peck was again asking questions, one PG&E supervisor suggested having an NRC seismologist—who was “on our side”—on the phone during a meeting should Peck “continue to imply he will make statements relative to [Diablo Canyon Power Plant] safety.”
“My management then asked PG&E to evaluate the new info so we know the emergency cooling system can still operate. ... PG&E said, ‘We’re not going to do it.’ My management said, ‘OK, no problem. If you [Peck] have a problem, write a white paper.’ So I did,” Peck told New Times. “That got some traction.”
In September 2011, Peck recommended formally stating that the plant was in violation of its license agreement.
On Jan. 7, 2011, PG&E submitted its final report on the Shoreline to the NRC. It included the determination that the three local earthquake faults—Shoreline, San Luis Bay, and Los Osos—could produce about 70 percent greater ground motion than the Double Design Earthquake. However, the utility argued that the ground motions extrapolated from the new information met the requirements when only applied to the Hosgri Event, which they argued NRC guidelines allowed them to do.
The NRC said no, and originally concluded that the new information needed to be evaluated against all three design basis earthquakes. But rather than do the evaluation, in October 2011—seven months after Fukushima—PG&E instead applied for a license amendment to lower the plant’s licensing requirements and leave the two more difficult requirements (the DE and DDE) out of the licensing criteria.
Peck filed a formal report, essentially disagreeing. Despite this, in October of that year, the NRC signed off on the deal, releasing what’s called a Research Information Letter agreeing that the ground motion from the Shoreline has already been shown safe for the plant.
Peck doesn’t disagree with the findings in the letter, however, which focused on the seismology surrounding the plant; rather, he questions how that information would affect operability of the plant’s safety components. He also took issue with the agency’s decision that the DDE issue would best be determined in the license amendment process.
“You have to understand that [the NRC geologist whose findings are the basis for the letter] is a geologist, and she’s looking at the seismology. … I don’t disagree with any of that,” Peck said. “I have a different question: Given that it’s going to move the ground that much, how does that compare with design and licensing basis? That’s what we look at [to determine] operability. The RIL doesn’t do that.
“[NRC branch chief, Peck’s former boss] Neil O’Keefe said, ‘Thank you very much but we disagree.’ At that point, the issue is dead,” Peck said. “But I disagreed—and still do.”
Then, in March 2012, the NRC—as part of its Fukushima “Lessons Learned” follow-up—mandated a nationwide Senior Seismic Hazard Assessment process for all of the nation’s nuclear plants, creating a committee to evaluate seismic safety design. First priority plants—like Diablo—are required to submit their committee report to the NRC by 2015.
PG&E argued that the new review provided a different mechanism to address the issues raised in the license amendment—i.e., that the DE and the DDE criteria should be excluded from consideration for all new data—and that the committee process would allow them to extend the actual evaluation to at least 2015.
They then withdrew their amendment request, and the issue remains unresolved—for now.
O’Keefe, speaking to New Times from his office in Arlington, Texas, said the issue boiled down to PG&E’s ability to demonstrate that the plant’s strong enough to handle the dominant seismic hazard. (That’s an earthquake from the Hosgri Fault.)
“This plant is the most analyzed plant in terms of seismic in the country,” O’Keefe said.
O’Keefe said that PG&E and the NRC’s assessment of the plant’s operability in the face of what’s considered the most severe seismic hazard is solid, and that the committee process is unlikely to change the plant’s licensing basis.
“These [components] are very sturdy,” he said. “Even if these events happened a couple minutes apart, they’re still separate events.”
O’Keefe added that the agency determined that applying anything but the Hosgri Event was like “pulling the wrong tool out of the toolbox,” and would give analysts “calculational garbage,” or extreme data.
“PG&E chose not to do all three [tests] because it’s a waste of money,” he said. “They’d already met our requirements.”
When asked if Diablo Canyon, with its unusual licensing requirements, is considered by the regulatory agency to be the troubled child of nuclear plants, and whether the utility is just kicking the licensing requirements down the road through further evaluation, the line went silent for a few seconds. Then came a chuckle and a light-hearted “no comment” from O’Keefe.
“As an inspector, I can only make recommendations,” Peck said. “My boss signs my report out. This is a documented case where I’m differing from my boss.”
NRC Region IV Spokesperson Lara Uselding pointed out that all this is somewhat old news; the NRC held a public workshop on the topic in San Luis Obispo in fall 2012.
PG&E is moving ahead with the Senior Seismic Hazard Assessment process. Their analysis is due in late 2015, when the NRC will determine if more analysis is needed. However, should the utility not satisfy the design basis for the plant—and if they have to apply the DE and DDE requirements, it’s documented they won’t—the process would extend through at least three more years of evaluation. The utility needs to complete its application for relicensing the plant’s two reactors by 2019. Those are currently set to expire in 2024 and 2025, respectively.
“Most plants know that their re-evaluated ground motions are going to exceed design basis and are developing risk assessments,” an NRC scientist, Clifford Munson, reported during a June 19 California Energy Commission workshop on California nuclear issues. “PG&E is in that group.”
Following the difference of professional opinion, Peck remained at Diablo for nine more months before applying for, and receiving, the instructor gig in Chattanooga. He told New Times there wasn’t any nefarious attempt to push him out, but rather that inspectors are only permitted to stay at one plant for seven years to avoid conflict.
All told, where Peck and the higher-ups in the NRC seem to differ—when all technical jargon is set aside—is whether or not applying the license requirements boils down to a question of safety.
“Dr. Peck himself realizes that this is not a safety issue at the plant. The plant is safe,” NRC Spokesperson Uselding told New Times. “It’s an administrative issue.”
“I wouldn’t have pursued it if it wasn’t a safety issue,” Peck said.
News Editor Matt Fountain can be reached at firstname.lastname@example.org.