After the devastating tsunami damage to the Fukushima Daiichi nuclear plant in Japan, seismologists in the United States are focusing on a potentially dangerous fault system near the Diablo Canyon nuclear power plant in California.
Perched over Point Buchon on the Central Coast, the PG&E plant was built four decades ago near two active faults: the Hosgri and the Shoreline. The electricity-generating facility on a barren stretch of coastline is about 12 miles from the college town of San Luis Obispo.
Even though PG&E has said the Hosgri and Shoreline are too small to threaten the aging plant, some government scientists suspect the faults - acting with others in the region - could produce an earthquake much more powerful than the plant was built to withstand.
For now, the scientists are being cautious. Sam Johnson, a U.S. Geological Survey scientist, would not speculate in an interview about whether the area's fault system is actually much longer, and therefore more powerful, than now predicted.
But during a presentation earlier this spring at the USGS headquarters in Menlo Park, Johnson showed how the Hosgri Fault could stretch 250 miles from Point Conception all the way to Bolinas just north of San Francisco - far longer than its current official length of 105 miles.
The longer the fault, the more powerful the rupture. Johnson showed the possibility of the Hosgri Fault connecting in a system with the San Simeon and San Gregorio faults north of the nuclear plant.
"If the fault were to run from Point Conception to Bolinas, that would be close to an 8.0," said Johnson. "That would be a big concern."
Johnson was quick to point out that the USGS has not said the fault runs that far. He and a team of researchers from the USGS, State Coastal Conservancy, National Oceanic and Atmospheric Administration and others, including PG&E, are collecting this information through the California Seafloor Mapping Program.
Documents show that Diablo Canyon was constructed to withstand ground shaking from a magnitude 7.5 earthquake 3 miles offshore, on the Hosgri.
In 1927, a magnitude 7.1 earthquake struck offshore near Lompoc, possibly on the Hosgri. Nobody was killed, but a concrete highway was cracked and a railroad bridge was thrown out of line, according to the Southern California Earthquake Data Center.
PG&E would not respond directly to questions about the structural integrity of the plant, or fault length and size. The utility's spokesman, Blair Jones, said in a statement: "Diablo Canyon was designed and constructed with seismic safety in mind and components of the facility were tested to withstand probable ground motions resulting from nearby faults."
And so far, scientists and top administrators for the Nuclear Regulatory Commission said that the plant is safe and that further studies are unwarranted.
"I would offer the seismology around Diablo Canyon has been thoroughly studied," said Elmo Collins, the commission's regional administrator for Region IV, which includes California, Oregon, Washington and Alaska. "We have to have a solid technical and scientific conclusion about safe operation of the facility today. We believe we have that."
Even so, seismologists have been paying close attending to one newly discovered fault near Diablo Canyon: the Shoreline. Located in 2008, the fault runs along the tip of Point Buchon, just a few hundred feet from the power plant.
In a 500-page report on the Shoreline Fault for the Nuclear Regulatory Commission, the federal overseer of nuclear power plants, PG&E suggested the fault was just 15 miles long, with the potential for a magnitude 6.5 earthquake.
The utility also denied the possibility that the Shoreline and Hosgri faults could act in concert, saying: "Rupture is inhibited from branching from the Hosgri Fault Zone to the Shoreline fault zone. North Segment dies out before, or terminates at, the Hosgri Fault Zone."
But a federal seismologist who discovered the Shoreline Fault said it's logical to assume the faults are connected. If the two faults did connect, that could put a magnitude 7.2 earthquake almost immediately beneath the plant.
"With the data I'm looking at, it actually doesn't make sense to think of these faults as not connecting to each other," said Jeanne Hardebeck, the USGS seismologist who discovered the fault, using PG&E and USGS data on recorded earthquakes. "An interpretation that says they don't connect doesn't seem to fit with the observations that we have."
Hardebeck points out that a magnitude 7.2 quake on the combined faults would cause greater ground shaking at the nuclear plant than would a magnitude 7.5 quake 3 miles offshore on the Hosgri Fault acting alone.
If the Hosgri Fault runs up to Bolinas, and if it's connected to the Shoreline Fault, there is the possibility of a magnitude 7.7 earthquake under the plant, Hardebeck contends.
"Our current thinking about how faults might work together has been really influenced by some earthquakes that have occurred over the last couple of decades," she said, "where earthquakes have started out at one fault have actually jumped to another fault, or gone around bends in a fault and jumped gaps up to about 3 miles."
Hardebeck pointed to the recent magnitude 9.0 earthquake and tsunami in Japan that derailed the Fukushima Daiichi nuclear power plant. She said in that case, scientists thought there were several unconnected faults offshore, "but in fact, when the earthquake did happen ... it ruptured through all of those fault segments."
3-D mapping planned
PG&E has received authorization from the California Public Utilities Commission to spend more than $16 million on 3-D seismic studies to better evaluate earthquake threats near the plant. The 3-D technology works like a CT scan on geological surfaces, showing subsurface layers, cracks and faults.
PG&E is in the process of trying to relicense the power plant, which is set to expire in 2024. The Atomic Safety and Licensing Board Panel of the Nuclear Regulatory Commission - made up of judges and scientific experts - told PG&E in June that it must complete these 3-D seismic studies by the end of 2015.
Relicensing is suspended until then.
Johnson, with the USGS, said the 3-D seismic mapping is "likely to yield some interesting, unexpected results - answers to questions we haven't even asked. We won't know what it will mean until we see the data."
Still, the 3-D mapping won't uncover critical information about potential dangers to Diablo Canyon. While geologists and seismologists think it may provide spectacular images of the Hosgri, it may not determine its length and the rate the fault is slipping.
The $16 million 3-D analysis PG&E will use to explore the Shoreline and Hosgri offshore faults is unlikely to resolve the relationship between the two faults. That's because the fault lies in a kind of muddy swirl of bedrock that is not conducive to 3-D analysis, said Hardebeck.
And if the studies come back with inconclusive information about the faults, said state Sen. Sam Blakeslee, a Republican from the Central Coast who has a doctorate in seismic studies, "that raises a really interesting question for the NRC."
"If it's reasonable to infer that a much larger earthquake could occur but the studies come back inconclusive," he said, "the NRC has one of two choices: to plan as though it could or to ignore that possibility because it couldn't be proven."
James Boyd, the California energy commissioner, thinks even the best-laid plans, predictions and models might fail in the face of unpredictability and nature.
"We don't know enough about the magnitude of a potential earthquake to say with absolute assurance that this site, based on its location on a bluff and all the other safety measures, makes us totally secure from any possible events in the future," he said. "Mother Nature has certainly demonstrated this year, particularly in Japan, what can happen."
This story was reported in conjunction with producer Joe Rubin, PBS' "Need To Know" and Capital Public Radio.