As state officials clamp down on records at Oroville Dam, one of the country’s foremost experts on catastrophic engineering failures has used state inspection reports, photographs and historical design specifications to piece together an autopsy detailing why the spillway at the country’s tallest dam failed so spectacularly this winter.
The independent analysis by Robert Bea, of the Center for Catastrophic Risk Management at UC Berkeley, points to design and construction flaws dating back to the spillway’s construction in the 1960s. Bea said the gaping crater that formed in the spillway on Feb. 7 was all but inevitable given that the design problems were compounded by inadequate upkeep and maintenance.
Bea’s 78-page report, which he has shared with The Sacramento Bee and other media outlets, says the spillway was undermined by a variety of factors, including thin concrete, the presence of “soils and incompetent rock” below the concrete and evidence of water undermining that material. Bea’s findings dovetailed with the conclusions made last month by four consultants advising the state on Oroville’s repairs.
Subsequent reports by those consultants have been sealed, along with several other documents connected to the Oroville recovery effort. Bea said he’s troubled that federal and state officials are citing terrorism concerns to block access to these reports. Greater third-party scrutiny could help guide the $275 million repair job at Oroville and point to flaws in other dams, he said.
“In essence, their fear for security is something that was largely built up in their own minds as a defensive measure,” said Bea, a retired engineer whose credentials include conducting an independent investigation into why the levees around New Orleans failed in 2005 during Hurricane Katrina.
State officials say the Department of Water Resources is trying to balance the public’s right to information while also preventing critical design elements from falling into the wrong hands. Federal regulators have mandated that independent consultants guide repairs and do a formal forensic analysis that will go into greater depth than Bea’s, said Erin Mellon, a spokeswoman for the California Natural Resources Agency.
“The forensics analysis team is reviewing thousands of documents and recent geotechnical information taken from the spillway,” Mellon said in an email. “They will base their analysis not on original design drawings but on actual construction – there is a difference – and all of the facts, which are being collected and evaluated. … We look forward to the team’s final report, which will be made public.”
After receiving a flurry of complaints from state and federal lawmakers outraged that the DWR blocked public access to nearly every document the agency files with federal regulators detailing ongoing construction planning, bids and safety information, acting DWR director Bill Croyle told reporters last week that he’s weighing whether to release redacted copies of those reports.
Robb Moss, an engineering professor at Cal Poly San Luis Obispo who worked with Bea on Katrina, said independent analyses using those types of documents are critical to a better understanding not just of what went wrong but also how to prevent future problems caused by hidden defects in the nation’s aging dams and infrastructure.
“You have to have as many eyes on these things as possible so that you don’t miss things,” he said, after reviewing Bea’s Oroville report. “If the Governor’s Office or DWR is going to take the approach of ‘We know what’s best and we’re going to lock things down,’ that’s not how you solve problems like this, where you’re talking about thousands to tens of thousands of people that are in harm’s way. There’s needs to be a lot of light that’s shone on this.”
Bea’s analysis, which relied on inspection records and design specifications obtained before the state clamped down on releasing documents, found that the spillway’s concrete was too thin – as little as 4 to 6 inches in some places. While concrete-strengthening steel rebar was embedded throughout the structure, Bea concludes there wasn’t nearly enough steel bridging the heavy concrete slabs to keep them from pulling apart from the force of gravity and powerful water flows gushing down the spillway. He argues that the structure also was inadequately anchored to the hillside. Compounding those risks, he points to inadequate earthen material and a poorly designed drainage system underneath the spillway’s concrete – failings undermined further by “repeated ineffective repairs” on cracks.
The state also failed to remove trees growing along the sides of the spillway whose roots likely plugged drains, he said.
The spillway cracked Feb. 7, prompting a temporary shutdown of the 3,000-foot-long structure as a major rainstorm rolled through the region. The lake filled so high that water flowed over the adjacent emergency spillway for the first time ever. When it appeared the emergency spillway would fail, officials ordered 188,000 downstream residents to evacuate for two days. Disaster was averted when dam engineers dramatically ramped up water releases from the damaged main spillway, arresting the flow of water over the emergency structure.
The chasm in the main spillway is so extensive that repairs will take two years, although DWR officials say the spillways will be functional next winter.
Robert Bea’s Preliminary Root Causes Analysis