Report: Flaws in Design, Building and Upkeep Led to Oroville Spillway Failure
An independent analysis of the spillway failure at Oroville Dam concludes that extensive sections of the concrete structure disintegrated due to a series of design, construction and maintenance flaws stretching back more than half a century.
The root-cause analysis, by Robert Bea, a retired professor of civil engineering at UC Berkeley and co-founder of the university's Center for Catastrophic Risk Management, contends that "the gated spillway failures are deeply rooted in pervasive design defects and flaws developed by the California Department of Water Resources" -- the agency that built the dam in the 1960s and now owns it.
Bea has been called on to investigate high-profile engineering failures, including the loss of the space shuttle Columbia, the levee failure in New Orleans after Hurricane Katrina and the Deepwater Horizon oil-rig disaster in the Gulf of Mexico.
In a report for the UC Berkeley center circulated Monday, Bea contends that "the gated spillway failures are deeply rooted in pervasive design defects and flaws developed by the California Department of Water Resources."
Those defects include slabs that were much too thin -- an estimated 4 to 6 inches in critical spots -- and lacking in the reinforcement and anchoring they needed to withstand the destructive power of repeated water releases.
Report: Flaws in Design, Building and Upkeep Led to Oroville Spillway Failure
DWR compounded those initial flaws, Bea's report says, by failing to ensure that the giant concrete chute was built on competent rock. Contractors also used "native soils" -- mostly clay -- to grade sections of the spillway foundation, allowing extensive erosion and displacement of gravel intended to support a crucial drainage system, Bea concludes.
The spillway's built-in problems led to extensive cracking of its concrete surface and openings in the joints between adjacent slabs of concrete. Bea says repeated repairs to the concrete were ineffective in preventing large volumes of water from getting under the concrete and further eroding the surface below.
Bea also faults DWR for allowing large trees to grow adjacent to the spillway walls. Their roots could block pipes meant to drain water from beneath the concrete.
The dam is the key facility of the State Water Project, which conveys agricultural and urban water supplies from the massive reservoir of Lake Oroville to farms in the San Joaquin Valley and agencies serving 26 million people in the San Francisco Bay Area and Southern California.
The Department of Water Resources had no immediate response to a request for comment on Bea's analysis. In past statements, the agency has said the Oroville spillway met the design and construction standards in place during the 1960s and that repeated inspections found it to be well-maintained and approved for continued use.
Bea's analysis challenges both of those DWR assertions. He concludes "the wrong standards and guidelines are being used to requalify many critical infrastructure systems for continued service."
Spillway Breach Led to February Crisis
The dam became the focus of worldwide attention in early February when a large breach appeared in its main spillway just as a surge of storm runoff flowed into Lake Oroville, the state's second-largest reservoir.
With spillway releases reduced, a rapid rise in the lake sent water flowing over an untested emergency spillway. Severe erosion on the slope beneath that massive overflow structure prompted fears that it would collapse, triggering evacuation orders affecting about 180,000 people from communities along the Feather River.
As that near-catastrophe unfolded, Bea said in interviews, he began assembling materials to try to explain it. He says the Department of Water Resources rebuffed a request for original design documents of the spillway and adjacent emergency weir.
DWR representatives told him that agency guidelines barred the release of information on critical infrastructure systems -- data protected by a post-9/11 law designed to prevent such resources from falling into the hands of potential attackers.
"They looked at me and smiled and said, 'You might be a terrorist,' " Bea recalls.
Bea then assembled on his own a collection of publicly available information -- including partial design information released by DWR in the 1960s and '70s, a selection of the department's dam inspection reports, and photographs from the agency, media sources and even Google Earth -- to try to analyze the causes of the spillway failure.
While acknowledging that he was working with what he called "incomplete information," Bea says the materials pointed to the defects he outlined in his analysis.
A Drainage System and Thin Concrete
One of the crucial flaws in the construction, Bea says, concerns the complex drainage system installed beneath the slab.
The system consists of a network of clay pipes with an inside diameter of 6 inches and outside diameter ranging from about 8 inches to 10.5 inches at their connecting ends. The pipes are laid out in a herringbone pattern beneath the concrete. They are designed to collect water that seeps through cracks in the slab and route it to drains in the spillway's sidewalls.
Bea says that since the slab is generally just 15 inches thick, the concrete over the drain pipes is much thinner -- between 4 and 6 inches thick, or about the same as a standard sidewalk.
The thinness of the concrete over the pipes showed up as cracks on the surface of the spillway -- a fact Bea says was noted after extensive repairs to the concrete were performed in 2009.
"The concrete along the spillway chute has been repaired," a dam inspector wrote in June 2010. "The repaired herringbone crack pattern is said to reflect the underlying drain system."
Agency Consultants Share Similar Concerns
In some respects Bea's conclusions echo concerns raised in a report last month by a board of engineering consultants DWR retained, at the direction of federal regulators, to oversee the agency's plans for rebuilding the spillway.
The board expressed concern about the structure's repair history. Members noted that the slab was just 12 inches thick and thinner in some areas -- much thinner than current design standards -- and had been cut open repeatedly in the past to fill voids that had opened beneath the spillway. The engineers also noted that compacted clay had been used to fill spaces that have opened under the slab.
The use of the clay "calls into question whether the portions of the slab that appear undamaged by the failure should be replaced during the restoration," the board wrote.
The consultants also observed that a drainage system installed beneath the spillway slab was discharging "extraordinarily large" amounts of water even days after flows were halted down the concrete chute.
"It seems likely that piping [erosion] of foundation material beneath the chute slab may be responsible for the voids that have been found and repaired in the past," the consultants' report said. The board called for an investigation of that possibility, a task that will likely be taken up by a separate panel appointed to conduct a forensic investigation of the spillway failure.
DWR says the forensic panel, which was also created at the direction of federal overseers, began reviewing documents late last month and visited the spillway for the first time last week. There is no timeline yet for its report on the causes of the structure's failure, though it has been asked to submit preliminary conclusions to the spillway replacement design team by the end of the first week of May.
'A Regulated Failure'
Bea calls the breach and destruction of the spillway "a regulated failure."
"The regulatory processes are not sufficient to put a check and balance into the system to say, 'Hey, what a minute, what you're thinking here about design is not appropriate,' or 'What you're concluding from your inspection and repair operations is not appropriate,' " he said.
Why deficiencies in the spillway were allowed to occur and grow into a crisis, a process that took more than five decades, is an issue he says he's still wrestling with.
"It's obvious these things were accumulating and persisting. Why?" Bea says he asked himself.
"I looked back on my previous experience with our Department of Water Resources," Bea says. "I worked with them back in the 2005 and 2010 period and had gotten to know [former DWR director] Lester Snow pretty well, and a couple of engineers. They're not really what I would call the evil empire."
But Bea says it's clear to him that the agency and other oversight bodies, such as the Federal Energy Regulatory Commission, are using standards that don't account for the deterioration of infrastructure over time or outdated technology. He points to the fact that the Bureau of Reclamation's current standard for spillway construction calls for 36 to 42 inches of concrete and other measures to reduce the destructive effects of high water flows.
"Much as I learned from working on Katrina," Bea said, "DWR, FERC and our state regulatory system are using guidelines and standards that may have been appropriate for the 1960s, and were appropriate for the design and initial construction, but not appropriate for a 50-year-old system that's experienced aging and technological obsolescence. We're using the wrong standards to requalify these aged infrastructure systems."
Bea's analysis of the failure comes as DWR rushes forward with plans to rebuild the spillway. The department announced Monday night that it had awarded a $275 million contract for the project to Kiewit Infrastructure West, the Northern California branch of an Omaha-based construction conglomerate.
DWR says it estimates the project's cost at $231 million. Kiewit's was the lowest of three bids, which ranged up to $344 million.
The department says preparatory work will begin immediately. Acting Director Bill Croyle said as recently as last Thursday that it's still working to determine the project's scope and final design.