|Shortly before a catastrophic failure. Image: NPR.|
Every dam has a spillway at some level below the top of the dam -- usually at least a few feet lower. This spillway defines what should be the maximum level of water behind the dam. It must be designed with a capacity to allow for ANY possible flow of water around the dam. If water is allowed to flow over the wall of the dam ("topping" the dam) failure is quite likely. This is especially true of earthen dams, but is also a danger with concrete or stone dams. Dave Petley's American Geophysical Union blog post on which I found this video includes more technical details, including some that are in the form of questions, as this system is much more complicated than it appears to non-specialists.
The failure of this dam brings immediately to mind the 1889 Johnstown Flood. In both cases, innocent downstream communities were devastated by a failure of both engineering and conscience. Wealthy owners had been warned that their structures were unlikely to survive expected rain events, but chose to risk not only their investments but also the properties -- and the very lives -- of downstream neighbors. Lee Mueller, the owner of the Edenville Dam, is doubly culpable because he has actively ignored federal and state safety regulators and he must know the history of Johnstown. He is a perfect illustration of the need for environmental regulations, and the need to empower those agencies charged with enforcing them. Sadly, at least 2,500 dams in the United States are considered dangerous, meaning that they are both in a position to cause deaths if they fail and in a state of repair that makes such failures more likely than they ought to be.
When I commented about the good luck that nobody was killed by the Edenville collapse (though thousands were evacuated and millions of dollars of damage was done), a friend and former professor of mine quickly corrected me: nobody died as an immediate result of the collapse, but the flood damage included the inundation of an inadequately prepared chemical plant downstream.
As I explain in my 2016 Houston, Too Close to New Orleans post, tanks storing hazardous chemicals are required to be surrounded by secondary containment, meaning a wall or berm that will contain the chemical in the event of flooding or failure of the tank. I know this because I have done the calculations for such containment in Puerto Rico. The inside dimensions must be adequate to contain the entire contents of the dam, plus a small freeboard capacity. Similarly, the outside dimensions (height above ground) must be high enough to withstand a 100-year flood. Tanks on Dow properties downstream from Edenville were not adequately contained, so my friend is correct that increased casualties are likely. But they will be the result of long-term, low-dose chemical exposure that will be impossible to document on an individual basis.
In the case of the Dow facility, it is not yet clear what the nature of the miscalculation was; it could very well be that the berms met legal requirements and failed anyway. This is because the regulations worst-case conditions on which regulations are based might assume only meteorological flooding, not that caused by independent failures upstream. Moreover, the 100-year flood level is no longer an adequate way to estimate flood risk. It describes a flood that would have a 1/100 probability of being exceeded in any given year in the past. As I explain in my 2018 post Not in the Cards, even in the hands of statisticians, hydrologists, and engineers who do understand precisely what the term does and does not mean (and these people are rare), existing records are not adequate to estimate flood probabilities, because climate change means that today's conditions are statistically not related to those of the past.
Ellicott City Comparison
|Image: Scott Weaver|
We are not fans of reality television -- even if it pertains to food -- but of course we watched this show with great attention (and through torrents of tears, to be honest). PLEASE watch it if you can, because ultimately it is a hopeful story. Because it is not a show about hydrology, though, it is understandable that a mistaken comment about the floods was included without correction. Someone in the program said that the town had been struck by two 1,000-year floods. They did not go on to say that the odds of this would be close to one in a million, but it is implied. The reality, unfortunately, is that the floods in Ellicott City were inevitable given. climate change and recklessly inadequate regulation of upstream land use. The three blog posts above describe the details.
Incidentally, Ellicott City is still vulnerable, but engineering remedies -- including those described in Ramsay's show -- have likely reduced the risk in general terms and have removed some of the most vulnerable victims completely.
Strictly speaking, my master's thesis was not about dams: it was about soil erosion. But the unexpectedly rapid sedimentation of a reservoir behind a medium-scale dam was the motive for our study, and we examined several dozen small-scale reservoirs to study the problem. When we started the project, I could barely discern small reservoirs on aerial photos; after a full year of study, I felt as though I knew some of them personally.
I am not certain to what extent dam safety has been a factor in the decision, but I recently learned that a small dam about a mile from my house in Bridgewater is slated for removal. This will create some environmental benefits and some environmental hazards (principally from the draining of artificial wetlands and the possible liberation of long-dormant toxic sediments). I will be following that with nerdy interest!