Like everyone else, Robert Bea was appalled when almost 200,000 Californians living below Oroville Dam were ordered to flee for their lives on February 12. The evacuation was necessitated by severe erosion of the dam’s primary and emergency spillways caused by massive releases of water following torrential winter rains. But unlike most citizens, Bea knew the incident wasn’t engendered strictly by the vagaries of nature or an act of God. Human error was at play.
A UC Berkeley Engineering professor emeritus, one of the nation’s foremost forensic engineers, and the founder of Berkeley’s Center for Catastrophic Risk Management (CCRM), Bea had long been worried about the design and management of the dam. He expressed those concerns to a state Department of Water Resources (DWR) engineer while visiting the Oroville site in April. The engineer opined that fixes DWR was making to the spillways were sufficient to stabilize the structure. Bea vigorously disagreed, countering DWR was downplaying the risks.
Bea and CCRM colleague Tony Johnson then launched an exhaustive study of the Oroville incident. The document resulting from that inquiry—Root Causes Analyses of the Oroville Dam Gated Spillway Failures and Other Developments—has just been released, and it makes for a riveting read. The paper not only dissects Oroville’s dire design and management problems in detail, but it also obliquely indicts America’s engineering culture as insensitive to risk, championing a certain cowboy élan and “get ‘er done” attitude over public safety.
We recently spoke to Bea about his findings. The conversation has been edited and condensed for clarity.
CALIFORNIA: You investigated some of the nation’s most spectacular disasters, including Hurricane Katrina, Deepwater Horizon, and the Challenger space shuttle. How would a major breach at Oroville compare to those incidents? And is such a failure even possible?
Bea: Yes, it is possible. And it would be worse than any of them. A breach at Oroville would send a wall of water down the Feather River, through the Sacramento Valley and ultimately into the Sacramento/San Joaquin Delta. It would destroy towns along the Feather and Sacramento Rivers, flood major portions of Sacramento, and blow out levees throughout the Delta, permanently flooding much of the region. The huge government pumps near Tracy that send water to Southern California cities and farms would be incapacitated. There would be tremendous loss of life and property, and it would be years before a permanent water delivery system to the south state could be reestablished.
Furthermore, it wouldn’t necessarily take a tremendous amount of rain and uncontrolled releases as we saw in February to trigger such a failure. It could happen on a bright, sunny summer’s day. The situation is that serious.
You had expressed concern about Oroville’s status prior to the spillway issues and evacuation. Is there anything you discovered in your most recent research that surprised you?
We were able to obtain some important documents and photographs we hadn’t seen before, and we learned a few things that, frankly, shocked me. One thing I had always wondered about was the construction of the primary spillway. The original design specified construction on rock and concrete, but the documents confirmed it was built on graded fill. I’d always suspected something like that, but had never found any corroboration. Graded fill simply doesn’t provide the strength and stability that rock and concrete provide, as we saw in February, when huge portions of the spillway were destroyed and tremendous erosion occurred.
The report also concludes that DWR has been aware of Oroville’s problems for many years, correct?
The main portion of the spillway that blew out is a section known as Station 3300. We found documents and photographs dating all the way back to a 1998 inspection report that confirmed there were major problems with this section. Essentially, they were screams for help that went largely ignored. DWR tried patching cracks, even filling up voids, but on that structure, finding the hollow areas is like trying to find a stud behind a wall by tapping it with a hammer. There’s a lot of room for error, and the repairs were far from adequate. Also, some of the patches actually trapped water so it couldn’t drain out, and that further compromised structural integrity. All this has been going on since at least 1998, so it’s no surprise that things went wrong at Station 3300.
What else did you find that concerns you?
There are very serious problems with the headgates [the structures that control water releases from the dam]. There is a lot of corrosion, and there are large cracks in two gate control tendons. Some of the gates don’t even open and close all the way. They’re like refrigerator doors that just stay ajar.
Also, we were alarmed by green grass growing on the abutments. That’s because the grass shouldn’t be green in that part of California during the summer and fall. It should be brown, so when you see lush green grass, it indicates there’s seepage through the dam face. What’s particularly worrisome is evidence we found indicating this seepage has been going on for about 50 years. DWR says that it’s just some ‘natural springs,’ and that’s it’s no cause for worry. But a ‘natural spring’ in that area would have to flow uphill, which runs counter to the laws of physics. And understand this: Any seepage through an earth-fill dam is extremely worrisome.
You’ve often talked about a certain laissez faire attitude among engineers and bureaucrats regarding the risks of large infrastructure projects. How does that apply to Oroville?
First, rigorous state and federal guidelines exist for the risk-based management of dams. Some of them were issued by President Carter following the Teton Dam failure in Idaho in 1976 [which killed eleven people and caused massive property damage]. But they’re simply being ignored. Dams should be viewed like nuclear power plants or jet airliners. With both nuclear plants and passenger jets, there is tremendous emphasis on managing and minimizing risks, and that’s appropriate. But dams are treated like inert piles of dirt or concrete that have no potential for catastrophic failure, and that’s simply not the case. Oroville isn’t just a big dirt plug with a bunch of cows and corn downstream. If it failed, it would be the worst disaster in the history of the United States. The people at DWR are not bad human beings, but it’s apparent to me that they’re not up to the risk management challenges they’re facing.
Governor Jerry Brown (B.A., Cal Classics) has just appointed Grant Davis (B.S., Cal Political Science) as the new director for DWR. He was formerly the director of the Bay Institute and the Sonoma County Water Agency. In both positions, he was a strong advocate for environmental protections and sustainable water use. What do you think of him?
He’s certainly an improvement over recent past directors, but he’s no [Admiral] Hiram Rickover [widely regarded as the father of the U.S. nuclear submarine fleet and a visionary engineer who promoted safety and risk aversion in both marine and terrestrial nuclear power plants]. Davis has good environmental credentials, but I don’t think he knows much about managing risk, and I’m not sure he has the guts to say what needs to be said to the people above him. I just don’t think he has the capabilities we need at this point.
So how should we address Oroville? Can it be repaired to the point that the risk it represents is acceptable?
There are certainly things we should—must—do. We did it with Folsom Dam [on the American River], which was also showing danger signs, including on its spillway. It cost a billion dollars, but we did it, and it is now stable for the foreseeable future. So we need to make the same serious commitment to Oroville, and that doesn’t mean a quick patch job like we’re now seeing, followed by an announcement that everything is okay. We need to convene independent experts and follow their recommendations. Right now, the response is ad hoc, and that’s both inadequate and extremely dangerous.
From the larger perspective, the problem with Oroville, with our entire water delivery infrastructure, really, where there are multiple danger points, isn’t concrete, steel, dirt, or water. It’s people. Many of the people who are in decision-making positions don’t have the expertise they need to make those decisions. Water is our most essential natural resource, we need it, but if you turn your back on it, it can kill you. We have to have people in power who understand that, who are willing to take the actions necessary to protect the public.
So what’s next for you? Are you going to continue your work on Oroville?
I don’t think so. Both Tony and I are getting up in years, and candidly, I’m getting tired, a bit worn out. I’m running out of teachable moments. I’m usually an optimistic guy, but I’ll be honest, what we’ve found with Oroville has been deeply discouraging. Since word of our work has gotten out, we’ve had engineers and water managers contacting us from all around the country, and the basic message is ‘Holy Crap! We’re having the same problems with our dams.’ Oroville is just the most obvious signal of a national, systemic problem.
Posted on July 27, 2017 - 1:07pm
Filed under: Law + Policy
Science + Health
Related topics: UC Berkeley
University of California
UC Berkeley engineering
Root Causes Analyses of the Oroville Dam Gated Spillway Failures and Other Developments
Berkeley’s Center for Catastrophic Risk Management
Department of Water Resources
Department of Water and Power
Admiral Hiram Rickover