The Oroville Incident – an opportunity for dam safety

The incident at Oroville Dam in California captured the world’s attention in February as pictures and video showed flows over the main and auxiliary spillways and the damage to both structures was revealed. Equally compelling was the sight of residents evacuating the downstream area following the evacuation order and the announcement of ‘Imminent Failure of Auxiliary Spillway on Oroville Dam!’ (NWS, 2017).

Now that repairs are being made and the clamor for immediate answers has subsided, the difficult work of carrying out a root cause analysis has started. At the direction of the Federal Energy Regulatory Commission (FERC), the California Department of Water Resources (CADWR) has appointed a forensics team to identify the root causes of the Oroville event as well as any contributing factors. While the work of the forensic team will take time, it is reasonable to anticipate they will unravel the puzzle, answering most if not all of the questions as to what happened and why. The process will take months, and probably longer for lessons to be identified, learned and transformed into improved dam safety practices.

Historical incidents

Events like the incident at Oroville Dam, while seemingly a surprise to many, are not particularly rare. There have been incidents involving major spillway damage in the U.S. in the past; Hoover Dam in 1941, Yellowtail Dam in 1967, Glen Canyon Dam in 1983, and Fort Peck Dam in 2011, among others here and around the world. While each case is unique in its own way, they are fundamentally similar; flows below the original design basis led to significant damage. Issues with spillways are not unique areas of surprise. In 1995, the failure of a spillway gate at Folsom Dam during normal operations motivated significant changes in the way gates are inspected, periodically tested and maintained. In 2005, the failure of the Taum Salk Pumped Storage project brought to light concerns about the operation and reliability of pumped storage projects and reservoir control systems. In 2014, a major incident (in terms of costs to the owner) occurred at Wanapum Dam on the Columbia River in Washington state. A crack in a spillway monolith was discovered following a visual observation by a utility employee of a misalignment of a bridge railing. The failure mode that occurred was a result of a number of factors, including an error in the original design, that over a period of 50 years, evolved and manifested itself into a misalignment that was readily apparent in February 2014. Noteworthy is the fact the failure mode had not been identified in the potential failure mode analysis process. Further, monitoring data that showed this monolith was performing quite differently than the others which were of similar size and design was misinterpreted, leaving the failure mode undetected. 

Forty-one years ago, the failure of the U.S. Bureau of Reclamation’s (USBR) Teton Dam on first filling proved to be a watershed event for dam safety. It precipitated changes in the culture of the USBR, and it captured the attention of a new President who directed the U.S. federal agencies involved in the design, operation and regulation of dams to review their dam safety practices and to work together to establish consistent dam safety guidelines across the federal government. President Carter’s memorandum on April 23, 1977 made dam safety a national concern. Stepping back from the details of the Oroville incident and the specific lessons it will teach us; it seems this event and others that preceded it, may be defining another watershed moment. A moment where we should be asking the substantive questions about the manner in which we carry out the work of dam engineering, safety, and management.

Dams are significant assets that society has invested in to provide a range of benefits; water supply, electrical power, recreation, flood protection, irrigation, etc. When the operation of these assets is compromised due to extreme events, mis-operation, or incidents that are the result of ageing (i.e., corrosion, etc.), services and project benefits are lost and repairs are often costly. In the case of Oroville these costs continue to move upward and are now reported to be $275 million (Sacramento Bee, 2017).  In the worst case of course, a catastrophic failure of dam can lead to permanent loss of the project assets (e.g., Teton Dam), and significant downstream consequences (i.e., fatalities, injury, damage to public and private property, business impacts, environmental damage, etc.). In view of the societal investment in dams and the potential losses that could be associated with the uncontrolled release of the reservoir, or the downtime due to mis-operation or other incidents, are we investing the time and resources in processes (e.g., dam inspection, maintenance, analysis of risks and remediation) for maintaining reliable and safe dam systems? If the American Society of Civil Engineers (ASCE) infrastructure report card is any indication, the answer is clearly no; leading one to conclude that dams are being significantly under-valued as critical assets.

Guiding principles

Following Hurricane Katrina and the flooding and levee failures that occurred, along with other infrastructure system failures, the ASCE prepared “Guiding Principles for the Nation’s Critical Infrastructure” (ASCE, 2009). To move forward in managing critical infrastructure, ASCE identified four guiding principles:

• Quantify, communicate, and manage risk.
• Employ an integrated systems approach.
• Exercise sound leadership, management, and stewardship in decision-making processes.
• Adapt critical infrastructure in response to dynamic conditions and practice.

The authors note “these guiding principles are fully interrelated. No one principle is more important than the others and all are required to protect the public’s safety, health, and welfare.”

With the ASCE principles as a guide, it would seem the Oroville incident (and others in recent years that have preceded it) provides an opportunity to closely and critically examine how the work of dam safety is being carried out; to ask the difficult questions and find solutions that will improve the management and safety of a critical part of society’s infrastructure system. To be clear, the improvements in dam safety in the U.S. since the Teton Dam failure (1976) and the levee failures in New Orleans during Hurricane Katrina (2005), have been positive, significant and effective. Since these events, the USBR and the U.S. Army Corps of Engineers (USACE) have successfully transitioned from being traditional, standards-based dam engineering agencies where the potential for unsatisfactory performance of dams was not discussed, let alone analyzed; to being risk-informed managers of a sizable portfolio of large dams in the U.S. While the USBR and USACE dam safety programs continue to improve, and evolve (each goes through periodic independent peer review), their ability to transition and improve is not as widely shared in dam safety.

The transition of the USACE following Hurricane Katrina and the failure of levees that were part of the New Orleans Hurricane Protection System (HPS) is particularly noteworthy. This event and the post-mortem evaluation that led to the recognition the HPS was ‘a system in name only’ (USACE, 2009), was a major impetus for the USACE to fully transition to become a risk-informed dam safety program. As the owner and operator of over 700 dams and the largest engineering organization in the country, this was a significant and successful undertaking that was accomplished in a period of about a decade.

Areas to examine

In the spirit of looking at the Oroville incident as an opportunity to examine the state of dam practices, areas we might examine include:

• Dam Safety Governance
• Dam Engineering and Safety Basics
• Learning from Experience
• Understanding and Managing Risks

Dam Safety Governance – Do dam owners and regulators have organizational philosophies, organizational structures, processes (policies and practices), and staffing to effectively evaluate and manage a critical societal asset which also poses a range of often significant organizational and public risks (economic, public safety and environmental)? 

Dam Engineering and Safety Basics – The day-to-day business of dam engineering and safety is multi-disciplinary and dynamic; periodic dam safety inspections, review and evaluation of monitoring data, period evaluation of project safety assessments, emergency action plan updating and exercises, project operations during normal and high inflow periods, maintaining project documentation, etc. For some, there is concern that we are not taking care of the basics – inspections are not thorough enough, monitoring data is not properly reviewed and interpreted, project records are scattered, etc. These concerns are exacerbated by the aging cadre of professionals with design, construction and operations experience who are moving on and the difficulty in attracting enough talented, young professionals to dam engineering. These issues pose a diversity of challenges to the future management of an aging infrastructure.

Learning from Experience – In a number of industries (transportation, nuclear power, public health, among others) and even in individual companies, there is a recognition of the importance to invest in a process of monitoring and evaluation of system performance data as a means to; identify new issues or trends, educate the next generation of professionals, and pro-actively leverage experience data into improved practices. Do we do enough in dam engineering to learn from the operating experience of dams such that positive changes to engineering management practices, design, operations and maintenance are realized? While it is common to read in the literature about ‘lessons learned’, this may too often be a misnomer. Lessons learned often refers to lessons identified, not lessons that have been broadly learned and integrated into the knowledge base of practitioners and transformed into improved practices.

Dam Safety Evaluation and Management – As the dam safety profession has moved away from traditional standards-based practices toward a risk-informed approach to dam safety management, methods such as potential failure modes assessment (PFMA) and quantitative risk analysis are being advanced. These advances have been positive, though not universally adopted. Oroville Dam, like many hydropower projects regulated by the FERC, had been the subject of a PFMA. With the backdrop of the Oroville incident and the structural cracking at Wanapum Dam, among others, we need to ask ourselves are PFMAs and risk analyses (if performed at all) carried out to the level of detail that provides a complete understanding of the vulnerability of dam systems that support effective decision making, informed project management, emergency planning, and community resilience? The experience of Grant County Public Utility District suggests maybe not. A lesson they note from the Wanapum Dam incident states, “If you think you may have all the potential failure modes covered for your structures, think again, the answer is ‘probably not’. For example, the effects of seasonal temperature variations causing stress reversals within the body of this concrete gravity dam was not previously understood.” (Mishalanie, et al., 2016). They go on to state, “Learning from exposure to other dam safety incidents can be invaluable. Be open-minded when revisiting failure mode scenarios…...”

In closing, there are certainly positives to be found from the events that played out at Oroville Dam in February. There was no uncontrolled release of the reservoir.  The evacuation of nearly 188,000 people went well given the threat of an ‘Imminent Failure of the Auxiliary Spillway’. More importantly are the benefits/advances in the management, engineering, operation and maintenance of dam systems that can be derived in the months and years ahead. 

Martin McCann is an Adjunct Professor in the Department of Civil and Environmental Engineering at Stanford University and the Director of the National Performance of Dams Programme (NPDP) (http://npdp.stanford.edu). The programme seeks to gather and evaluate information on the performance of dam systems to support that advancement of dam safety.

Citations

1. American Society of Civil Engineers, (2009). Guiding Principles for the Nation’s Critical Infrastructure, prepared by the ASCE Critical Infrastructure Guidance Task Committee.
2. Mishalanie, D., et al. (2016). “The Wanapum Dam Spillway Incident: Response, Remediation, Post Repair Performance and Lessons Learned,” Proceedings of the Association of State Dam Safety Officials, 2016 Annual Conference, Lexington, Ky.
3. Sacramento Bee, April 21, 2017, http://www.sacbee.com/news/state/california/water-and-drought/article145980829.html, accessed 4/21/2017.
4. National Weather Service, (2017), ‘Imminent Failure of Auxiliary Spillway on Oroville Dam!,” February 12, 2017, Sacramento, CA.
5. U.S. Army Corps of Engineers, (2009). “Performance Evaluation of the New Orleans and Southeast Louisiana Hurricane Protection System Final Report of the Interagency Performance Evaluation Task Force, Volume I – Executive Summary and Overview, June.

A Formal Start to Dam Safety in the US – A Memorandum From the President

The 23 April 23 2017marked the fortieth anniversary of US President Jimmy Carter’s memorandum to federal agencies with an interest in dams. The memorandum marks the start of the nation’s interest in and need to focus on dam safety. While the memorandum was certainly motivated in part by the Teton Dam failure, a federally owned dam, that occurred the year before on June 5, 1976 in Idaho, it took a much broader view of the state of dam safety in the Federal government at the time. It directed the Federal agencies to take a number of specific actions, one of which led to the development and publication in 1979 of the Federal Guidelines for Dam Safety. However, the memorandum was also rather visionary in that it identified specific areas that required ‘special attention’, including:

• The use of risk-based analysis in all aspects of dam engineering and operation.
• The use by the Federal agencies of external experts, evaluating the effect of earthquake and other geologic hazards on dams.
• The effects of cost-saving incentives on decision making.
• Reviewing practices for identifying and resolving dam safety problems.
• The involvement of communities in addressing dam safety issues.

In many ways, the President not only made dam safety a national priority, he also laid out a roadmap for its advancement that is applicable today.

Dr. Martin McCann is currently working on a paper related to the President’s memorandum and its impact on dam safety.