DWR_2717593 - Laserfiche WebLink

CARRIAGE HILLS NO. 2 DAM, Dam Failure Forensic Investigation Report Page 39 February 27, 2015 9. CONCLUSIONS The Carriage Hills No. 2 Dam failed during the catastrophic September 2013 flood event. Both the Carriage Hills No. 1 and No. 2 Dam spillways were inadequate to pass inflows and both dams overtopped. Failure of the No. 2 dam may have been caused by a combination of overtopping erosion and internal erosion along an old CMP conduit. A low area of the crest around the assumed outlet location likely contributed to the failure by concentrating overtopping flows. Historically the Dam Safety Branch cited spillway obstruction by willows and inadequate spillway size and freeboard as safety problems. In 2002 the DSB required the owner to restore the spillway to the size on the approved C‐1195 construction plans. The DSB also cited the improperly abandoned CMP outlet conduit as a safety problem and required the owner to rehabilitate the outlet works. Neither of these required repairs was apparently done. Clearly the Sept. 2013 flood was extreme, especially for high elevations above 7500 ft (Carriage Hills No. 2 Dam crest elevation was around 7645 ft). Our analysis shows the 48‐hr rainfall for the Carriage Hills basin had less than a 1/1000 Annual Exceedance Probability based on NOAA Atlas 14. We cannot say for sure if the No. 2 dam would have survived had it been properly maintained and repaired, but it certainly would have stood a better chance. The causes of the dam failure can be attributed to poor maintenance and repair of the dam as well as the infrequent, large magnitude of the flood. It seems likely that the dam began to fail on Sept. 12, 2013, around 22:00 hours with a failure time of approximately 14 minutes, causing the surge flow observed ½ mile downstream by Mr. Bonza at around 22:20 hrs. The timing corresponds to the third rain spike of the storm, which may have dealt the final blow to the dam. We estimate the floodwave duration was approximately 36 minutes, and it briefly added between 677‐ 801 cfs on top of the natural flood flow in Fish Creek. The dam breach floodwave likely contributed to the peak flood stage observed at the Bonza house around 22:20 hrs on Sept. 12th while Country Club Road was still intact; however, we are not convinced that dam breach caused the peak discharge on Fish Creek, due to the timing of runoff and channel erosion associated with subsequent failure of Country Club Road. High water mark estimates, Lake Estes hydrographs, the timing of debris entering the Fish Creek Arm of Lake Estes, and the eyewitness account of peak flows and debris higher up on Fish Creek after midnight on Sept. 13th all suggest that a larger natural flood peak occurred overnight than what can be explained by the dam breach hydrograph. Although it seems likely that the dam failure contributed to the observed surge at 22:20 hrs, we have to acknowledge the large uncertainty associated with the Sept. 2013 Fish Creek flood event despite our efforts to synthesize the best available information. A number of other factors could account for the surge and peak stage observed at the Bonza house: (1) the surge and peak stage could have resulted solely from rain‐driven runoff from the third and final spike in rain intensity; the timing and peak flows may not be reflected in the Jacobs’ HEC‐HMS model and our mass‐balance hydrograph due to the potential sources of error described above, or (2) the surge and peak stage could have been caused by the failure of any number of debris dams, upstream road crossings, landslide debris, etc. as explained in the Jacobs (August 2014) report. If the dam failure flood contributed to the peak flood stage at the Bonza house, it may have caused incremental damage at other low‐lying structures along Fish Creek, particularly upstream of Country Club Road. The flood