When it comes to repairing the tallest dam in America, sometimes it helps to shrink the problem to a more manageable size. That’s why California water officials are relying on a scale model of the damaged spillway at Oroville Dam to plan their repairs.
The model, constructed by engineering professors and students at Utah State University in Logan, is built to a 1/50th scale out of wood, steel, concrete and acrylic plastic. At 120ft (36m) long and 30ft (9m) wide, it’s roughly the size of a tennis court.
About 50 tests have been run on the model since March to simulate the hydraulic forces acting on the structure. More are planned in the weeks ahead, said Michael Johnson, a professor of civil and environmental engineering at the university.
The model, Johnson said, helps fine-tune the rebuilt structure’s design before a single bucket of concrete is poured. It was built under a $277,000 contract with the California Department of Water Resources (DWR).
The goal is to help avoid a repeat of the disaster that unfolded in February, when massive water releases from the reservoir caused the spillway to break apart, prompting evacuation of nearly 200,000 people downstream as a precaution.
“No computer models can give all the answers,” said Johnson, chief engineer on the project, which is housed at the at the Utah Water Research Laboratory. “A physical model is really an inexpensive insurance measure that can be taken to resolve some of the unknowns. It takes some of the guesswork out of the end product.”
Such scale models are commonly used when building a new dam to help verify calculations about water flow, depth and pressure. In fact, a model was built of Oroville Dam prior to its original construction in the 1960s.
The model simulates Oroville’s entire main spillway, including the water-control gate structure at the top and a portion of the reservoir profile upstream. It does not include the adjacent emergency spillway.
The model uses clear acrylic plastic for the gates and the spillway chute surface, because this material effectively simulates the friction between water and concrete at full scale, Johnson said.
Water for the model is drawn from the Logan River into a giant cistern beneath the warehouse that holds the model. Pumps circulate water through the model and return it to the cistern, allowing it to be used repeatedly.
Johnson’s team also modeled damaged areas of the spillway and the eroded hillside beneath it. They replicated those conditions using measurements taken by DWR using laser imaging and computer-aided design programming.
The purpose of modeling the damaged areas is to simulate what might happen if a freak storm requires massive water releases while repairs are under way, or if DWR can’t complete the repairs in time for next winter.
“There’s a lot of unique features about this particular project that really make it interesting and exciting to dive into,” Johnson said. “It’s the tallest we’ve ever built in our lab, the most vertical elevation we’ve ever modeled, and it’s got the longest spillway chute. It’s a pretty substantial model.”
The model is now being converted to replicate the finished repairs, and a new round of tests will be conducted using that version.
Ted Craddock, DWR’s project manager for the Oroville emergency spillway recovery effort, said the physical model has helped verify computer and mathematical modeling already completed. It has shown, for instance, that the spillway’s existing side walls are large enough to contain maximum water releases from the reservoir.
The model will also be used to test a new design feature that may be added to the spillway during repairs: aeration slots. These are cutouts in the concrete surface designed to prevent cavitation, a process that can erode concrete during high water flows. Aeration slots have been added at numerous other large dams as a safety measure, including Glen Canyon Dam on the Colorado River.
“Ultimately what we’re using this information for is to … understand the flow characteristics,” said Craddock. “Sometimes you get eddies and things like that that you need to have some awareness of. It’s a way to increase the confidence in the design.”
The spillway chute will be rebuilt with thicker concrete and more steel reinforcing, a response to flaws identified by an expert panel that investigated the spillway after the disaster. The original design for the Oroville’s spillway specified concrete 18in (46cm) thick, but many areas were found to be only 12in (30cm) thick.
The upcoming repairs specify 30in (76cm)-thick concrete. In addition, many of the damaged areas will be built up with what Craddock called “leveling concrete,” instead of relying on bedrock, which the expert panel found to be suspect in some areas.
DWR’s contract for the work, valued at $275 million, is with Kiewit Corp. and runs through to January 2019. It includes rebuilding the spillway from its terminus at the Feather River channel to 600ft (183m) above the damaged areas. The remainder of the spillway up to the water-control gates will also be rebuilt, Craddock said, but that work is being delayed to the following year, and is not likely to be as extensive.
“As we’ve dug into the records, what we’ve found is that the bedrock from the gate structure downstream is better than the area where we had a problem this year,” Craddock said.
The current year’s contract also includes fortifying the emergency spillway. It was used for the first time in February after DWR officials shut down the main spillway when it began to break apart.
The emergency spillway is nothing but a massive concrete curb, with no water-control gates. When the reservoir reached maximum capacity, water simply spilled over the curb onto the hillside below, causing massive erosion that threatened to undermine the spillway and topple it over. This is ultimately what led to the evacuation order in February.
Environmental groups warned about that risk during Oroville Dam’s recent federal relicensing process, but were ignored.
Kiewit will also be working on this emergency spillway. Its contract for this year includes replacing the weakest portion of the emergency spillway on western end to make it “more robust,” Craddock said, including thicker concrete and more reinforcing steel.
Kiewit will also build a vertical concrete cutoff wall, parallel to the emergency spillway and about 600ft downstream, intended as an erosion-control measure.
In a followup contract next year, DWR plans to add a heavy concrete buttress against the downstream face of the emergency spillway, essentially giving it more thickness and a bigger foundation. It will also build a concrete “splash pad” between this buttress and the new cutoff wall as another erosion-control measure.
Craddock said he is confident the main spillway will be ready to perform at full capacity by November 1, the typical start of the flood-control season.
“We’re going to be able to pass the flow that the original design was for,” Craddock said. “We need to be able to use this spillway with the assumption we’re going to have record rain again next year. That’s what everyone’s working towards.”