Resilience and reliability concepts applied to Oroville Dam
Feb 15th, 2017
So, we are in a conceptual exercise today. We discuss how to apply Resilience and reliability concepts to Oroville Dam.
We will stay away from numbers, as we do not know them.
Oroville Dam system
A catchment area of 3,607 sq mi (9,340 km2) brings water to the Feather River Valley upstream of the dam location.
The Oroville dam bars the Feather River Valley mainly for water supply, hydroelectricity generation and flood control. The Dam’s design and building procedure complied with codes. The embankment was completed on October 6, 1967. That was after a series of mishaps, including a railroad accident and a catastrophic flood (1964, peak flow of 250,000 cubic feet per second (7,100 m3/s) above the Oroville Dam site) generated delays.
The Oroville Dam design comlied with seismic codes. A complex monitoring network reportedly similar to those in hydro dams in Europe (water pressure and settlements, deformations) completed the system.
Two safety features were built-in:
- a service spill-way (with manoeuvrable gates) followed by a concrete channel designed to lead escaping waters to the original valley-bottom,
- an emergency spill-way, a reinforced low point threshold, at higher elevation that the service spill-way, but lower than the dam crest. The emergency spill-way is a “last ditch” protection against over-top. Over-top is a potential cause for dams’ breaches. It leads escaping waters to the natural slope below the dam toe, where natural soil, vegetation is present, in such a way that escaping waters will find their way back to the river.
An inundation study and evacuation plans considered the population at the time. They covered the case the dam would breach, or some other catastrophic mishap may occur.
What occurred over the last decades
Reportedly three environmental groups filed a motion in 2005 with the Federal Energy Regulatory Commission. They requested a concrete protection for the dam’s earthen emergency spillway as “in the event of extreme rain and flooding… heavy erosion” would occur. Regulators considered the request unnecessary or excessive. In 2006, a senior civil engineer sent a memo to his managers stating that the emergency spillway met FERC’s guidelines for an emergency spillway, and that these specify that during a rare flood event, it is acceptable for the emergency spillway to sustain significant damage.
The question is, like usual, what is “rare”? Various media noted that the auxiliary spillway (not emergency anymore) began carrying water for the first time since the dam’s construction in 1968. People seem to believe that one time in forty years is “rare”. But that’s ridiculous, especially in view of the fast changing climate conditions in this world and considering the criticality of the system (exposed population, infrastructure, environment).
We note that over time the spill-way changed denomination from “emergency” to “auxiliary” without any corresponding physical upgrade.
A famous James Bond theme song says “diamonds are forever”, but concrete has a very definite life duration. Concrete can crack:
- on its own due to shrinkage,
- chemical reactions exacerbated by air pollution,
- because of subtle (or not so subtle) changes in the foundation conditions, such as settlement, deformation, seismic events,
- water table raises and lowering (like in a severe meteorological cycle after a long drought, for example).
Thus, the spillway started cracking in 2013 leading to interviews of a Senior Civil Engineer with the Department of Water Resources by the Sacramento Bee who promptly explained the phenomenon is common and “There were some patches needed and so we made repairs and everything checked out.”
What has occurred to the Oroville Dam
After the Californian drought, relentless storms brought very strong continuous precipitations. Water has flown into the Oroville Dam reservoir. At a point in time (on February 7, 2017) action of the service spill-way gates occurred. The goal was a flood control release of about 50,000 cubic feet per second (1,400 m3/s).
Was it too late? Answering that question is not within the scope of this discussion. However we note that many historic floods around the world were allegedly the result of dam operators having “sticky finger” in the gate opening decision. The reason for the “sticky finger” syndrome is oftentimes simple. Indeed, water is a precious resource (whether used for drinking, energy). “Wasting it”, especially in drought stricken regions is a tough decision to make. It is loaded with responsibility and potential consequences for the decision-maker.
Concrete started to erode
When the gates were opened and the service spill-way flow reached a high service capacity, concrete started being eroded (for any of the reasons cited above or due to concrete cavitation . At that point water got “under and around” the original concrete channel spiralling toward a self-boosting chain reaction of failure. Thus the hope that using the damaged spillway could drain the lake fast enough to avoid use of the auxiliary spillway proved false. It became necessary to lower the discharge. It went from 65,000 cfs (1,800 m3/s) to 55,000 cu ft/s (1,600 m3/s) due to potential damage to downstream infrastructures.
Water in the lake rose and at a certain point in time, the emergency spill-way level was reached. From the picture above we can see flooding traces in the parking at the end of the emergency spillway. Indeed, water escaped from the right end of the weir.
Water escaping the emergency spill-way started eroding the sloped downstream. That lead to very serious concern, triggered evacuation plans, panic, etc. Media as widely reported these events.
So, what about Resilience and reliability concepts applied to Oroville Dam ?
Thus it seems that Resilience and reliability concepts supporting the Oroville Dam were sound and bullet-proof as well as code compliant. Were they?
We will close this post with a series of questions and ask for your opinion:
- Do you think “unheard of”, “rare” adjectives are reasonable for events occurring once every 40-60 years? Do not forget, for example, the catastrophic flood of 1964 with 250,000 cubic feet per second (7,100 m3/s)).
- Shouldn’t the probability of this type of mishap be way lower, considering that such an event can alter the life of nearly 200 thousand people?
- Is it reasonable to hide behind codes that state that an emergency spillway could sustain heavy erosion and damages? That is without wondering what that means (multidimensional consequences) from a technical and social point of view?
- Do you think it is reasonable to have one officer state “There were some patches needed and so we made repairs and everything checked out”. In fact without a serious risk analysis of the damaged/repaired system and its implications for the public?
- Did the criticality of the structure change due to the population density increase? If positive why not plan additional mitigations?
Once again the idea a myth is destroyed. If one designs a system to withstand all the worst-case credible accidents, the system is NOT “by definition” safe against any credible accident.
Tagged with: emergency spill-way, Oroville Dam, Reliability, Résilience, service spill-way
Category: Consequences, Crisis management, Risk analysis, Risk management
A very well presented and thought analysis. “designed to withstand all the worst-case credible accidents (issues)” .. this is where we need to alter practice…once the design is rated as such — it has to be monitored for the life of the structure, all aspects – internal, site specific, environmental, at depth, surficial ..if so you have an arsenal of data telling you the pulse and life of the structure .. to think this is not “real” — that these man-made assets somehow stay in stasis once built — that is perhaps economical, but foolish.
In my short 4 years of experience as an engineer in Northern California (1991-1994) we had two 100 year rain events and a 250 year event that nearly exceeded the capacity of our water diversion systems. Only through expeditious water management practices (two big pumps, lots of sand bags, and a large hole in the ground) did we avoid violating our water discharge permit. There is nothing “rare” about these events in that area. I would certainly not rely on these probabilistic approaches to design of water embankments and diversions in this area given my experience.
Was the design and oversight, state or federal (or both?). Who had the ultimate authority ?