Lessons learned |
failed suddenly. The failure section was 12-24m wide and located in the left embankment. An upstream remnant remained standing for 10 to 20 seconds, before it gave way and the embankment was fully breached. The breach enlarged over the next few hours, releasing the water stored in Wixom Lake. The IFT is confident that the embankment did not
overtop. Internal erosion was also judged to not be plausible as the primary mechanism of failure because the observed physical characteristics of failure are not consistent with this: ● No seepage exiting the ground surface was detected.
● No turbid water discharge was detected. ● No evidence of a developing open pipe, sinkhole, or progressive sloughing that might indicate global backward erosion piping was observed.
Above: Flooding in Michigan in May 2020 after failure of the Edenville and Sanford Dams. The independent forensic team has shared its findings regarding the physical mechanisms of dam failure © Tyler Dittenbir
j “It’s absolutely critical that there’s buy-in from the very top. We need to move from awareness and shared understanding to actual commitment,” Branigan continues. “I really believe that we have an ethical obligation
to ensure that anybody in our society feels capable of pursuing a career regardless of their gender, their background, or any other parts of their identity. As a sector, we are better when everybody can participate. And that’s why the diversity and inclusion work is so important.”
Michigan failures
On 19 May 2020, the Edenville and Sanford Dams, located in central Michigan, US, failed. To investigate the failures and the physical and human factors that contributed to them, the Federal Energy Regulatory Commission (FERC) engaged a five-member independent forensic team (IFT) in August 2020. In September 2021, the IFT announced it had essentially completed its evaluation of the physical mechanisms of the dam failures. Although its work and final report will not be completed for several months, the IFT wanted to share its findings with the dams profession and the public. The Edenville and Sanford Dams were two of four dams in Michigan owned at the time of the failures by Boyce Hydro Power and located in series along the Tittabawassee River; the other two Boyce Hydro dams are Secord Dam and Smallwood Dam. All four dams were built in the 1920s. At the time of the failures, Secord, Smallwood, and Sanford were active hydroelectric facilities. Edenville’s powerhouse was inactive because its FERC license had been revoked in September 2018. The watersheds upstream of the four dams received significant, but not extreme, rainfall from 17-19 May 2020. Despite all spillway gates being open, lake levels continued to rise. Wixom Lake which is impounded by the Edenville Dam, continued to rise until the time of the failure when the lake level is estimated to have been about 1.6m above normal pool level and 0.3- 0.45m below the nominal embankment crest. Based on a video recorded by a local resident,
photographs, and eyewitness accounts, it became clear that a downstream section of the Edenville dam
28 | December 2021 |
www.waterpowermagazine.com
● The kinetics of the failure are not consistent with historical observations of internal erosion failures. In the IFT’s opinion, internal erosion may have contributed to the depression in the crest that was observed about 35 minutes before the failure and may have affected the phreatic surface and pore water pressures within the embankment, but it does not explain the primary physics of the failure. In the IFT’s opinion, the most plausible principal mechanism for the failure of Edenville Dam is static liquefaction (flow) instability of saturated, loose sands in the downstream section of the embankment. Static liquefaction has been receiving increasing
attention in recent years in the tailings dam arena because of several recent tailings dam failures. This failure mechanism has been rare, but not unprecedented, for water storage dams, and water storage dam engineers have not typically considered it. The conclusion regarding static liquefaction at Edenville Dam is supported by: ● The accelerations and velocities of the failing soil mass evident in the dam failure video.
● Strong evidence of loose, uniform fine sand in the embankment.
● Strength loss behaviour exhibited in laboratory tests on loose specimens of uniform sand collected from the breach remnant.
● A reasonably close match of a simplified kinetic analysis with the characteristics of the failure shown in the dam failure video. Although there is uncertainty concerning the exact trigger or triggers that led to the static liquefaction failure, there are several phenomena that are plausible triggers, either individually or in some combination, as explained in the report. The physics of the Sanford Dam failure are very clear and was the result of embankment overtopping. The breach outflows from Wixom Lake after the failure of Edenville Dam caused the water level in Sanford Lake to rise more quickly than could be accommodated by the spillways at the Dam. Given the failure of Edenville Dam, the failure of Sanford Dam was not unexpected. Regulators and engineers understood that should a breach occur at Edenville, Sanford would almost certainly be overtopped and fail.
South Carolina reports Since 2015, a series of historic weather events have
caused more than 80 state-regulated dams across the US state of South Carolina to fail, driving a renewed
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37
