FLOOD06904 - Laserfiche WebLink

Home Browse Search

48 LAWN LAKE DAM AND CASCADE LAKE DAM FAILURES. COWRADO protected sites, such as inside sharp channel meanders, and in wide valley cross sections. Maximum thickness measured was a deposit of coarse sand and gravel about 2 ft thick. Where the Fall River entered the west end of Estes Park at river mile 11.5, at an elevation of 7,550 ft, the streiun passed under a narrow bridge opening and followed ,R channelized path along the south end of town, before join- ing the Big Thompson River near the eastern end of Estes Park. The majority of the flood water left the channel of the Fall River at the narrow bridge crossing and over- flowed down Elkhorn Avenue, the main street in Estes Park (figs. 23, 24). No significant amounts of scour or deposition occurred in this short reach between Estes Park and the streamflow-gaging station (06733000 Big Thomp- son River at Estes Park). SEDIMENT IN LAKE ESTES There were no accurate measurements taken of the sediment and debris deposited in Lake Estes by the flood However, the US. Bureau of Reclamation spent approx- imately $80,000 for heavy debris and sediment removal following the flood. Periodic sediment surveys have not been performed at Lake Estes because sedimentation was not considered a problem prior to this flood. A ma- jor volume of sediment was deposited in Lake Estes dur- ing the spring of 1983, 1 year after the flood. This accumulation was moved into the reservoir by the sus- tained high spring snowmelt flood. The source of sedi- ment was primarily from erosion of exposed river banks and channels that had lost their protective covers of vegetation and rock armor during the big flood. Sub- sequent work in and near the channel by adjacent prop- erty owners also created some streambed instability. Zenas Blevins (US. Bureau of Reclamation, oral com- mun., 1983) estimated that there Was at least 10 times more sediment deposited in Lake Estes during the spring runoff of 1983 than was deposited by the Lawn Lake flood. The unstable sand and gravel deposits along the Roaring River, Fall River, and Big Thompson River indicates that the river is not yet stabilized and will continue to transport large amounts of sediment into Lake Estes for several years. Water quality problems are expected each spring during the next several years. The river banks across the glacial moraine deposits are still very steep and unstable and will continue to slough for many years. DAM.BREAK MODELING Hazard mitigation of floods resulting from dam fail- ure requires documentation of historic dam failures to understand the processes involved Future hazard mitiga- tion also requires analysis of past dam-failure floods to estimate potential flood discharges, depths, boundaries, and traveltime. These analyses of dam-break floods are often made with deterministic digital-computer models. However, most of the documentation and analyses of dam-failure floods generally have been made on relative- ly low-gradient streams (Land, 1980; Chen and Arm- bruster, 1979); whereas, the Lawn Lake dam and Cascade Lake dam failures flood occurred on very high-gradient streams (fig. 2). Some aspects of the documentation of this flood could not be made because of hydraulic pro- blems previously discussed. Questions also need to be answered concerning various scenarios of possible dam- failure floods of this type. Therefore, dam-break computer modeling was undertaken to evaluate the model's capabilities on high-gradient streams, to enhance and pro- vide supplemental hydrologic information, to evaluate various hypothetical scenarios of dam-break develop- ment, and to assess probable impacts of the failure of Cascade Lake dam, as well as impacts of other situations. These analyses provide the range of possible flow characteristics from the actual conditions as well as other possible conditions. The following section of this report provides a summary of the dam-break model analysis. DAM-BREAK MODEL The selected dam-break flood model was formulated, developed, and documented by Fread (1977) and modified and documented to meet Us. Geological Survey needs by Land (1981). The model was selected because of its general purpose formulation. Studies by Land (1980) in- dicate that it is the most accurate, economical, flexible, numerically stable, and easiest to apply of four models tested on three documented dam-failure floods. The model formulation is based on the two one-dimensional open- channel flow equations (Saint Venant) of continuity and momentum for shallow depth and unsteady conditions. The numerical analysis technique used is a nonlinear im- plicit finite-difference method. Minor modifications to the dam.break model by Land (1981) were made to meet the objectives of a general purpose model for the needs of the Us. Geological Survey. These included eliminating the multiple dam-failure option, altering the computation of the time step, reorganizing data entry and printout formats, preventing the model from switching from one state of flow to the other (subcritical to supercritical) in a given subreach, and allowing for tributary inflow. The model is designed to simulate a dam-break flood in one computer run; however, the simulation is done in two distinct parts. The first part is to route an incoming flood through the reservoir and compute an outflow