Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Learning from Failure | Oroville Dam spillway

Dk oro spillway damage 5318 02 27 2017

On 7 February this year, engineers spotted a hole in the concrete spillway of the tallest dam in the United States.

Over the next few days Oroville Dam in California made international headlines as the damage spiralled out of control.

Its two spillways — overflows for excess water — had suffered erosion damage as a result of the heavy rainfall which had spread through the state, after several years of drought. That month, more than half of the 3,000 people working for the California Department for Water Resources (DWR) were dedicated to emergency response work.

Operational since 1968, Oroville Dam is 235m tall and its crest is 275m above sea level. During heavy rain, excess water first flows down the primary, or main, spillway. An auxiliary spillway provides extra capacity in an emergency.

The discovery of damage to the main spillway alarmed officials in February, and prompted the slipway’s immediate closure to allow for inspections. It was then gradually reopened to controlled water flows.

But rainwater continued to flood into the reservoir, and with nowhere to go, the 275m crest was breached and water started to flow down the auxiliary (emergency) spillway for the first time in the dam’s history. As the auxiliary spillway was just a channel cut into the hillside, erosion damage soon appeared at its head, undermining the concrete weir.

Oroville

Oroville

Concerns about the resulting high volumes of water downstream of the dam — by this time water was crashing down the main spillway at 2,831m³/s — led to evacuation orders being issued for 180,000 residents living in the danger zone, downstream.

“It was very crazy, very hectic, very stressful,” says DWR chief engineer Jeanne Kuttle, who was working on site on 7 February when the initial spillway damage was discovered.

“We’ve had floods a number of times here in California, and I’ve worked those a number of different times — 1997, 1998, 2005 — but none of those were nearly as intense as this past 2017 February and March.”

In the immediate aftermath there was no time to reflect on what had gone wrong. Engineers raced to repair the damage between storms, plugging the hole in the auxiliary spillway with rocks and gravel dropped from helicopters, and starting to clear away 764,555m³ of debris at the bottom of the main spillway.

Kg oroville repair 16066

Kg oroville repair 16066

Repairs are underway

On 19 May, the last water for the season was released from the dam and work started on interim repairs to both spillways in a race to get them ready for the next rainy season.

“Everything has been a challenge,” says Kuttle.  “The biggest unknown, of course, is what’s going to happen this winter. Is it going to be a dry season? Is it going to be a wet season? How wet would it be?

 

While water was still running

While water was still running

High velocity water flows triggered the spillway’s failure

“If someone could say, ‘hey, it’s not going to rain until December’ then we have the ability to breathe a little bit, more a sigh of relief, but no-one can make that prediction so we have to be prepared for everything. And we are; we intend to be.”

DWR’s contractor, Kiewit Infrastructure, has worked up to seven days a week and in double shifts to carry out the vital repairs by 1 November. Parts of the concrete chute of the main spillway have been demolished. The slope has been stabilised, foundations laid and roller compacted concrete added.

Now water is being released to lower water levels in the reservoir before the rainy season hits. Through a series of controlled releases, water levels in the reservoir will be brought down to 213m above sea level; significantly lower than the usual 238m expected at this time of year.

Geologists survey damage at the bottom of the main spillway crop

Geologists survey damage at the bottom of the main spillway crop

Investigators study the damage

“That gives us a bit of a buffer to be able to take any sort of inflows into the lake without having to encroach on the spillway until much later in the season,” says Kuttle.

Once winter arrives, construction work will slow and there will be more time to reflect on what happened and what the next steps are. But questions are already being asked about what caused the incident, who is at fault, and what lessons can be learned to make sure this year’s situation is not repeated.

An Independent Forensic Team (IFT) was set up in the wake of the emergency to provide the answers. Made up mostly of engineers, it has already released its interim findings into what went wrong.

Firstly, it addresses what happened to the main spillway. The IFT believes a small section of concrete on the main spillway chute was loosened and removed by the high velocity water flows mid-morning on 7 February. The foundations in the rock and soil-like material under the chute quickly eroded and more concrete slabs were smashed away by the water, creating the hole first seen by engineers.

According to the IFT report, the spillway had endured much greater water flows in the past. So what went wrong this time?

Dk oro spillway damage 5657 03 06 2017

Dk oro spillway damage 5657 03 06 2017

Investigators examine voids beneath the chute

Corrosion of the reinforcement across joints in the concrete had occurred, and new damage to previous repairs had gone unnoticed. In addition, anchors beneath the concrete slabs had not been properly encased in grout, leading to corrosion. Although the IFT does not believe a deep void had formed beneath the spillway chute, it is examining the possibility that existing shallow voids formed by erosion could have expanded, increasing the possibility of spillway collapse.

It would be easy to blame engineers and other officials for failing to pick up on the warning signs. But Kuttle does not believe more inspections would have made a difference in identifying the underlying causes of the spillway failure.

“Any maintenance or inspections that produced anything that required attention or repairs was done, and usually done pretty efficiently and expeditiously,” she says.

“In this case, I think the factors that led up to the spillway failure on the seventh [of February] were undetectable from a visual standpoint or from a maintenance standpoint.”

The IFT came to a similar conclusion in its report.

“Physical inspections, while necessary, are not sufficient to identify risks and manage safety. At Oroville Dam, more frequent physical inspections would not likely have uncovered the issues which led to the spillway incident,” says the IFT report.

Dam inspection overhaul

Instead, it recommends an overhaul of the dam inspections process. While the dam itself was regularly inspected, the appurtenant structures — such as the spillways — were not included as part of the routine.

It says comprehensive reviews of the original design and construction details, conducted by professionals with a high level of technical expertise, should be carried out regularly to check existing structures against current best practice.

This is an issue in California, where much of the state’s key infrastructure is 50 years old or more. At Oroville Dam, the IFT found that design and construction flaws made chute vulnerable to erosion damage.

A relatively large aggregate size had been used in the concrete, making it prone to cracking and spalling problems. Cracks above many of the herringbone drains running under the chute allowed water to pass through the slab, amplifying the spalling issues.

Below spec foundation preparation and treatment

The IFT discovered that during construction of the chute slab, the foundation preparation and treatment failed to measure up to what was specified. In some areas, more than 50% of the foundations had not been properly treated. Poorly designed drains added to the troubles with insufficient collector drain capacity for the high flows and no filtering system.

Despite these failings, Kuttle thinks it is unfair to blame the designers.

 “It would be the same as driving a 1950s’ car and calling the designers of that vehicle flawed for not having seatbelts or airbags,” she says.

“It’s things that are developed with state of practice, so I think it’s important to look at where we are today in history, and how things are done today independent of how things were done then.”

In fact, the redesign of the main spillway – which will be fully constructed by April 2019 – does not deviate far from the original plans.

“The general concept of the design is very similar, so that means that the folks back then didn’t get it completely wrong. We all make improvements in everything we do,” says Kuttle.

No evidence of critical examination of spillway

According to the IFT, there is no evidence that a critical examination of the main spillway has ever been carried out. It believes a review would have most likely connected the dots and found risks including initial construction and design shortcomings, drain flows which were far higher than the intended design capacity and chute repairs which were not designed to withstand high velocity water flows.

Kuttle believes there was no fault or negligence on the part of the engineers currently working on the dam because the high level reviews recommended by the IFT are not common practice.

But she is clear  that the underlying causes of the spillway failure came to a head because of intense pressure from the water released due to unusually heavy rain. Around the world, extreme weather events are set to become more common: engineers have to be ready to react quickly in a crisis. She adds that more training in how to respond during an emergency would ensure that engineering professionals are sufficiently prepared. 

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.