For three weeks, Oroville Dam’s fractured main spillway and the surrounding hillsides have taken a nearly nonstop pounding. The stunning waterfall crashing down what’s left of the 3,000-foot concrete span has split the spillway in two and carved massive canyons on either side.
The Department of Water Resources, which operates the dam, has had little choice. The state agency had to force water down the damaged spillway to avert a potentially catastrophic “wall of water” from pouring out of Lake Oroville when the reservoir filled to the brim Feb. 12 – and dam operators watched in horror as the hillside below the never-before-used emergency spillway started to wash away. That sparked the frantic two-day evacuation of 188,000 downstream residents.
In the weeks since, the state has continued to blast water onto the battered main spillway, albeit in reduced volumes, to reduce water levels in the immense reservoir that sits behind the earthen dam. Lake Oroville – the state’s second-largest reservoir – has receded to the levels required this time of year to absorb a huge rainstorm.
So, with at least five more weeks to go in California’s rainy season, and billions of gallons of water waiting to rush into the reservoir when the snowpack blanketing the Sierra Nevada melts this spring, the question on the minds of many living in the Sacramento Valley is this:
Just how long can the badly damaged spillway hold up?
In practical terms, DWR officials are hoping the spillway can operate in its hobbled state through the rainy season, which this year is on pace to break records. Typically, the season stretches into April. According to several independent experts contacted by The Sacramento Bee, that appears to be a realistic goal.
Even with a fissure as long as a football field cutting across the midsection of the structure, they say the damaged chute appears strong enough to make it through spring. The top two-thirds of the span, including its critical release gates, appears to be undamaged. And the bottom third, while broken up, appears to sit on stable rock.
The jagged tear in between has misdirected a portion of the powerful flows cascading down the spillway onto the adjoining hillsides, carving deep ruts on either side. But that erosion, while dramatic, isn’t undermining the top of the spillway or posing a threat to the main earthen dam itself, according to geologist Jeffrey Mount of the Public Policy Institute of California. The spillway and dam are separated by a large triangle of hillside sitting on bedrock.
The experts said the bedrock underneath the main spillway has proved more durable than the crumbly material that washed away so dramatically below the adjacent emergency spillway Feb. 12.
But they also said it’s clear the state is in uncharted territory at Lake Oroville: DWR has ad-libbed its response from the start of the crisis, and can’t be sure exactly what will evolve as the weeks progress.
“That’s the No. 1 question,” said Joe Countryman, a flood expert and former engineer with the U.S. Army Corps of Engineers in Sacramento. “They’ve got 24-hour surveillance on how it’s operating, and so far, so good.”
For his part, DWR acting director Bill Croyle now speaks with confidence about the ability of the main spillway to last through spring. But the agency wasn’t always so sure.
After the crater formed in the lower third of the main spillway on Feb. 7, DWR closed the spillway gates to check the damage. A day later, engineers resumed the flow of water, but gingerly, conducting test runs at low volumes of water to see how the spillway would stand up. The tests sent tons of concrete and soil down the chute in a gush of water the color of chocolate milk. Water also flowed over the sides, eroding the spillway sidewall and carving into the hillsides.
Within a day or so, DWR officials said the water was pouring out clear and that the erosion on the spillway itself seemed to be in check. But they resisted ramping up the outflows and at one point dialed back slightly, to make sure the hillside erosion wouldn’t topple an electrical transmission tower they deemed vital to dam operations.
Meanwhile, with runoff from a wet, warm Sierra storm pouring into the reservoir, lake levels continued to climb. And DWR, in effect, was forced to choose between two bad, untested options:
Speed up releases over the damaged main spillway and possibly sacrifice the transmission tower. Or, let lake levels rise to the point that they crested the lip of an adjacent emergency spillway that had never been used in the dam’s 48-year history. Officials wanted to avoid the emergency spillway. They knew the spills would be uncontrolled and send mud and timber rolling down the unlined hillside beneath the structure into the Feather River below.
On Feb. 11, the water from the swollen reservoir rolled over the emergency spillway lip. Although the flows peaked at just 12,000 cubic feet of water per second – a fraction of the 250,000 cfs the structure was supposed to handle – the effect was far worse than imagined. On the afternoon of Feb. 12, the hillside directly beneath the spillway was eroding fast and on the verge of collapse.
The hillside looked “like flour being blown off by a fire hose,” J. David Rogers, a dam expert with the Missouri University of Science and Technology, recalls being told by an engineer at the scene.
In one area, the erosion came dangerously close to the emergency spillway’s concrete apron. If the water were allowed to eat away those remaining yards, officials feared it would undercut the emergency structure. That would have sent an uncontrolled blast of deadly, fast-moving water down the Feather River, flooding a vast swath of the eastern Sacramento Valley from Oroville to Yuba City.
The crisis was averted when dam operators gambled on the broken main spillway. They sent an explosion of water, at 100,000 cfs, nearly twice as strong as what they had been dumping onto the fractured chute just hours before. The move paid off. The main spillway held. The transmission tower stood. And lake levels dropped below the emergency spillway. The mass evacuations in Oroville and other downstream communities were lifted two days later.
How could the main spillway handle so much water in its dilapidated state? Experts say the bedrock below the remaining concrete chute appears to be nothing like the crumbly rock that eroded in the early days of the crisis. It’s certainly much stronger than the weathered material on the upper portion of the hillside below the emergency spillway.
“They’re like two different geologies,” said Rogers. “One’s severely weathered, and the other one isn’t. The other one is youthful and fresh and hard.”
Mount, with the Public Policy Institute of California, said he initially worried the erosion would eat its way up the spillway, possibly endangering its gates and rendering the structure inoperable. At that point, the spillway is also much closer to the earthen dam that holds back the reservoir. But that upward erosion hasn’t happened.
The whole structure “seems to be surprisingly stable,” Mount said.
Paul Tullis, an engineering consultant in Logan, Utah, who has studied spillway designs, said a kind of self-healing mechanism appears to have kicked in: The crater in the chute became large enough, and filled up with so much water, it’s acted something like a trampoline. The biggest gushes bounce away without beating too much into the chute.
“It has the ability to absorb and dissipate the energy,” Tullis said.
In the meantime, helicopters and dump trucks have hauled in thousands of tons of aggregate and concrete slurry to patch up the hillside below the emergency spillway. With the repairs taking hold, Croyle, acting director of DWR, has said “I would not hesitate to use the emergency spillway” if necessary.
Countryman, a member of the state’s Central Valley Flood Protection Board, agreed that it looks as though the emergency spillway “would be operable” given the bolstering. Nevertheless, he and others say it’s preferable to avoid letting the reservoir fill completely again, sending water down the emergency structure uncontrolled.
Over the past week, dam operators have throttled back flows from the main spillway nearly in half, to 50,000 cfs. “We haven’t seen any further damage up and down the spillway,” said Chris Orrock, a DWR spokesman. “We have seen erosion ... carve a new path down the hillside, but we haven’t seen any additional erosion up and down the cement structure.”
Eventually DWR wants to reduce outflows from the main spillway to zero to do a more thorough inspection of the damage and figure out repair options. “We have a lot to learn on the geology,” Croyle said.
Cutting back the releases also will enable crews to start dredging and hauling away the 150,000 cubic yards of rocky debris piled up in the Feather River channel beneath the concrete chute. The piles have raised channel levels to the point that dam operators can’t run Oroville Dam’s hydroelectric power plant, normally its main source of water releases.
“We are going to come in with barges, with heavy equipment on them and just create a channel that will allow the normal flow of water to return,” Orrock said. “We’re days away from that.”
Resuming operations at the hydro plant would play a small but pivotal role in the dam’s overall recovery.
The power plant can only release about 14,000 cfs, a fraction of what the main spillway can discharge. But that could handle a major portion of the Sierra snowmelt that will flow into the reservoir this spring and summer, Countryman said.
The hope, eventually, is that DWR can shut down the main spillway and start a permanent fix. Croyle has said he wants the spillway repaired – or maybe rebuilt in another location – in time for the next rainy season, which begins in October.
Croyle has declined to speculate on what caused the initial fracture in the main spillway. The Federal Energy Regulatory Commission, which licenses the dam, has ordered the state to open an investigation into the cause “as soon as feasible,” in part because the findings could influence how the repairs proceed.
Some outside experts already have weighed in with theories as to why the chute, rated to handle 250,000 cfs, broke open Feb. 7 with a relatively modest 55,000 cfs pouring down.
Some, including Tullis, point to a phenomenon known as “cavitation,” in which the blast of tiny water bubbles gushing down the chute at 50 mph effectively jack-hammered holes in weakened sections of concrete. Others, such as Robb Moss, a professor of geotechnical engineering at Cal Poly San Luis Obispo, speculate that roots from trees growing along the chute expanded weaknesses in the concrete.
Rogers, the dam failure expert from Missouri, said the spillway may have had cracks that weren’t properly patched. He also theorized the failure may have been tied to California’s five-year drought: The aging spillway could have weakened as it underwent contractions due to the sudden heavy soaking following years of dry weather.
Whatever the cause, the experts say, problems at Oroville should serve as a wake-up call on the state of America’s aging dams. At 48 years old, Oroville is actually slightly younger than the average American dam. According to the American Society of Civil Engineers, the average age of the 84,000 dams in the country is 52 years.
“The aging issues are without precedent,” Rogers said. “You’re going to see failure modes that you haven’t commonly seen before.”