One Lake Oroville disaster threat was enough. Record heat and wildfires held our attention the past few months while a winter 2017 disaster inched towards restoration. As we begin this year’s rainy season, construction should be winding down on the Lake Oroville spillway that fell apart last winter. For $250 million, Kiewit Corporation promised to fix it by November 1st.
Nearly nine months ago, Lake Oroville held so much water that, for the first time since the lake filled in 1968, water flowed over the emergency spillway. Apparently, the disaster began last February when the main spillway started to fall apart at the same time inflows to the lake increased. But the emergency spillway use didn’t work out well and one day later an emergency evacuation order was issued due to a potential failure of the structure. Water was eroding the slope. If it eroded under the emergency spillway’s concrete lip, an uncontrolled release of flood waters from Lake Oroville would dump downstream. The evacuation order sent 180,000 people rushing to shelters and the safety of friends.
But perhaps the disaster began decades ago in the 1950’s when the gated spillway was designed. Was it managed and regulated to failure? And what does the persistent green spot of vegetation on the dam slope indicate? Locals have been pointing to it with alarm for decades while officials claim it’s spring water flowing uphill or the result of rainfall. One unmistakeable disaster threatening 180,000 people brings up all the questions, including how safe is it to live downstream of the nation’s tallest earthen dam (770 feet)? Will it continue to hold back 3.5 million acre feet of water from a 3,607 square mile watershed?
Here’s the site before it fell apart.
After rain in early winter 2017 refilled the lake, water was released into the primary spillway as intended. The first sign of trouble was big concrete hunks bouncing down the Feather River near the fish hatchery about a mile below the spillway (baby salmon were evacuated from the hatchery!). The concrete originated at the spillway but it was impossible to see what was happening due to the water. California Department of Water Resources reduced flows to the spillway to asses the damage and found a crater in the spillway that caused the water to jump out of the spillway chute.
But warm rain that also melted some snow caused heavy inflows and forced DWR to resume higher water releases.
The problem worsened. Here are the rough details at 9:30am on February 11th 2017.
The lake level was 101% full and water flowed over the emergency spillway on February 11th.
The emergency spillway became an emergency situation but DWR wasn’t worried. They’d had workers clear the slope of trees and de-energize the transmission lines on tall towers that crossed the slope from the power plant at the base of the dam. (My bold added.)
This aerial view from a California Department of Water Resources drone shows water flowing over the emergency spillway at Oroville Dam on Saturday, February 11, 2017, after the lake level exceeded 901 feet elevation above sea level.....The volume of water is expected to pose no flood threat downstream and should remain within the capacity of the Feather River and other channels to handle. Oroville Dam in Butte County itself remains safe with no imminent threat to the public.
As DWR asserted, the dam was safe — but the slope below the emergency spillway wasn’t. At 4:11 pm the following day (Feb 12th), due to flash flood potential, an evacuation order (GET OUT right now don’t wait!) was issued for the cities of Oroville, Marysville, Yuba City and everyone downstream. Failure of the facility was anticipated to occur within an hour when erosion would undercut the emergency spillway concrete lip and a wall of water gush forth. Residential areas begin only a few thousand feet downflow of the spot expected to fail.
To reduce flow over the emergency spillway and potentially halt erosion, water releases were increased to 100,000 cubic feet per second in the primary spillway.
The evacuation order was lifted two days later, but erosion of the primary spillway and adjacent slope continued.
I reported on the situation in early March: Man proposes but nature disposes.
Amidst inspections, reviews, frantic funding requests, community alarms, and DWR protestations that this was something no one could have predicted, a $275 million dollar contract was issued for repairs to the eroded area and the destroyed spillway. The contract called for completion of a spillway capable of handling water flows of 100,000 cubic feet per second by November 1st, 2017.
Area residents and environmental groups have been saying the emergency spillway disaster potential exists and needs attention for years but no one in authority from the DWR to the federal government (FERC) paid attention to their warnings.
The bid was awarded mid-April.
The plan calls for filling in the giant craters in the main spillway, which fractured Feb. 7, with fast-drying concrete but leaving the massive chasm in the adjoining hillside untouched. The chasm could be used for handling excessive outflows of water during the next rainy season, but would be filled in during summer 2018.
The contractor would also partially line the nearby emergency spillway with concrete; the structure currently consists of a concrete lip perched atop an unlined hillside.
Putting it all back together required blowing up pieces.
In Repair is too small a word, I gave a July 1st update on the work.
The work of restoring a safe functional spillway for Lake Oroville is a massive undertaking estimated to cost nearly $300 million dollars. The lower 1,400 feet of the spillway is being demolished as part of the reconstruction. [...]
The work won’t be finished this year. The goal is to have the lower spillway area that was damaged last winter rebuilt this summer and then next summer (dry season) demolish and rebuild the upper 1,000 feet. By time the project is completed, workers will have placed approximately 800,000 cubic yards of roller-compacted concrete, 146,000 cubic yards of concrete, and installed 8.5 million pounds of reinforcing steel. DWR is aiming to have a spillway capable of releasing 100,000 cubic-feet-per-second by November 1, 2017.
Work continued day and night through summer wildfire evacuation warnings as some of the people who evacuated due to flood threat in February evacuated due to wildfire threat in summer.
Here is the scene on September 5, 2017 — less than two months ago.
It looks somewhat more encouraging a month later, October 6, 2017.
Close up view of the intensive work on October 9, 2017.
And — ta dah — on October 30th, 3,000 feet of primary spillway has concrete the entire length. The top 730 feet below the spillway gates were repaired but not replaced this summer. (This segment will be destroyed and rebuilt next summer.) We hope this repair work was handled differently than the original construction that didn’t excavate the “native soils and incompetent rock overlying the competent rock foundation.” The forensic evaluation of last winter’s disaster also identified the failure to properly address flaws in the spillway channel. Was the “patch and pray” form of maintenance used? The spillway gates have leaked since the late 1980’s and are part of the “regulated to failure” problem. They were repaired this summer, but was it done to higher standards? When will we know?
It looks like the November 1st deadline has been met but the total cost has nearly doubled and now is estimated at $500 million (or more). The scale of the project becomes dramatically evident at the 0:48 mark in this video when you see the large equipment and again at the end near 2:12 when you see the trucks driving the road atop the spillway gates.
So, yeah, fixing that huge eroded chasm in the slope and replacing 2,270 linear feet of spillway atop the fill can be done in seven months. The emergency spillway erosion has been filled in with boulders and concrete slurry, but otherwise the slope is still au naturale. Lake capacity is 900 feet and right now the water level is 695 feet. Water outflows through the base of the dam (not over the spillway) are increasing each day and now at 4,448 cubic feet per second. Inflows are negative now that outflows have been increased and the lake level has dropped 0.71 feet in the past day.
But if this season’s rains are ample (or if they are warm and don’t fall as snow in the lake’s upper watershed), the spillway will be used. Will it need to handle 100,000 cfs flow? When rainfall inflows to the lake require water releases down the spillway, we will begin to see the practical results of the repair work. That will be the bigger test of construction objectives. The deadline for completion is being met, but what about the spillway’s ability to carry 20,000 cfs of water, let alone 100,000 cfs?
The weather forecast predicts rain for Northern California beginning later this week and continuing through the weekend. This won’t be enough to raise the lake level to the spillway gates, but it indicates the beginning of wet season (officially the Water Year began October 1st).
Meeting the spillway repair deadline is a good start. Gates that hold and release water properly and a spillway with integrity that doesn’t crumble under use will help repair the integrity of DWR. But this summer, as the spillway repair proceeded, the integrity of the dam itself was questioned (local residents have discussed this for years).
After studying the structures, in July Robert Bea, forensic engineer and founder of Berkeley’s Center for Catastrophic Risk, asked “is Oroville Dam leaking?”
One of the country’s foremost experts on catastrophic engineering failures released a new report Thursday on the troubled Oroville Dam that asks a disturbing question: Is the country’s tallest dam leaking?
State dam managers have insisted for months that there’s no problem, and that persistent green wet spots near the top left abutment of the nearly 770-foot-tall earthen dam are nothing more than natural vegetation growth caused by rainfall.
DWR has an entire section on their website about these questions and the dam persistent green wet spots. Bea isn’t convinced. He says the DWR assurances based on measurements in seepage pools do not provide direct evidence of dam integrity. Direct evidence requires data from the internal sensors (piezometers) in the dam and they’ve been broken for years.
“FERC (the Federal Energy Regulatory Commission) has been asking DWR for years to be able to measure the internal water condition in the dam, yet DWR doesn’t mind that 100 percent of the piezometers are now non-functional,” they wrote in response to DWR’s report . . . “That is what is so concerning about the green wet area.”
It will take more than one season’s successful flow of water down the spillway for downstream residents to feel the sense of security they had before the 2017 disaster. O’dam.
Additional Resources
For more photos of the disaster and repairs, visit the CA DWR photo gallery.
These articles have information on the disaster’s forensic assessment and repairs.