Do You Really Know Your Surveillance and Monitoring Program?

FERC's Engineering Guidelines address Surveillance and Monitoring Programs (SMP) and other tools such as the Potential Failure Mode Analysis can enhance dam safety. With the trend to replace personnel with equipment monitoring, the potential for a "weak link" failure increases. The questions are: 1) How do you know a significant "change" occurs and can it be mitigated?; 2)Are there redundant or parallel SMP mechanisms?; and 3) Does SMP reflect the intent of the project's original design and subsequent modifications? These questions may seem simple to answer, but when interacted with the human, electrical, mechanical and telemetry factors, the probability for a potential failure increases.
When a potential failure mode is identified, the questions are "what are the consequences and the best way to mitigate it?" Subsequent investigations often include more detailed analysis and evaluations such as a fault tree analysis. A fault tree analysis looks at failure mechanisms that could occur from the point of origin to it's completion. For example, if the remote monitoring equipment shows the reservoir levels rising and gates open, yet nearby projects show that the water is receding, there may be a problem? The response could range from no problem to failed telemetry or sensors. Again, the consequences of the failure coupled with experience needs to be considered and thence mitigation measures may go beyond just a visual SMP.

Learning from incidents should not be the rationale for developing an SMP. Experience and investigations to determine weak links in electrical, mechanical and other component monitoring can and should be a proactive approach to dam safety. The projects experiences range in age from the early1900s, hydraulic heads of 50 to 1000 feet and flows from a few to thousands of cfs.

As a minimum, with examples, the following will be addressed: 1) human error; 2) reliability of equipment, including telemetry; 3) rapid shutdown that can rupture a conduit; 4) multiple and single alarms; 5) remote and on-site monitoring; 6) routine testing and maintenance of equipment; 7) operator training and 8) usage of a fault tree analysis. 13 pp.