FLOOD06904

• County: Larimer
• Basin: South Platte
• Title: Hydrology, Geomorphology, and Dam-Break Modeling of the July 15, 1982 Lawn Lake Dam and Cascade Lake Dam Failures, Larimer County
• Date: 1/1/1986
• Prepared For: Larimer County
• Prepared By: USGS
• Floodplain - Doc Type: Flood Documentation Report

HYDROLOGY, GEOMORPHOLOGY, AND DAM-BREAK MODELING OF THE
JULY 15, 1982, LAWN LAKE DAM AND CASCADE LAKE DAM
FAILURES, LARIMER COUNTY, COLORADO

By ROBERT D. JARRETT and JOHN E. COSTA

ABSTRACT

At approximately 0530 Mountain Daylight Time on the morning
of July 15, 1982, Lawn Lake dam, $ 26-foot-high earthen dam located
in Rocky Mountain National Park, Colorado, failed. The dam released
674 acre-feet of water and an estimated peak: discharge of 18.000 cubic
feet per second down the Roaring River valley. Three people were killed
and damages totaled $31 million. The Colorado State Engineer deter,
mined. that the probable cause of failure was deterioration of lead caulk-
ing used for the connection between the outlet pipe and the gate valve.
The resulting leak eroded the eartbfill. and progressive piping led to
failure of the embankment.
Floodwaters from Lawn Lake dam overtopped a second dam,
Cascade Lake dam, located 6.7 miles downstream, which also failed.
Cascade Lake dam, a 17-foot high concrete gravity dam, 12.1 acre-
foot capacity dam, failed by topp!iJ1g with 4.2 feet of water flowing
over its crest. The flood continued down the Fall River into the town
of Estes Park, which received extensive damage from the overbank
flow.
This report presents the setting, a summary of the causes of the
dam failures, the hydrologic data, aIld geomorphic effects of the flood.
Data on dam-breach floods on high-gradient streams are limited. A
dam-break computer model was used to evalute the model's capabil-
ities on high-gradient streams, to enhance and provide supplemental
hydrologic information, and to evaluate various hypothetical scenarios
of dam-breach development and probaWe impact of the failure of
Cascade Lake dam.
Flood data were obtained at two gaging stations (06732500 Fall
River at Estes Park and 06733000 Big Thompson River at Estes Park)
and five miscellaneous sites downstream from the Lawn Lake dam.
Peak discharges were determined using a variety of indirect methods.
Because of extensive scour and erosion along the Roaring River, peak
discharges were estimated from the dam-break model. Calculated peak
discharges for the flood were 18,000 cubic feet per second from Lawn
Lake dam, 12,000 cubic feet per secoJ1d at Horseshoe Falls where R0ar-
ing River joins the Fall River, 7,210 cubic feet per second into Cascade
Lake dam at the east end of Horseshoe Park, 16,000 cubic feet per
second from the failure of Cascade Lake dam, 13,100 cubic feet per
second about I mile downstream from Cascade Lake dam, 8,520 cubic
feet per second just upstream from Estes Park, 6,550 cubic feet per
second for gaging station 06732500 Fall River at Estes Park, and 5,500
cubic feet per second for gaging station 06733000 Big Thompson River
at Estes Park. Maximum d$pths ranged from 6.4 to 23.8 feet; max-
imwn widths ranged from 97 to 1,112 feet; and mean velocities ranged
from 3.3 to 12.6 feet per second. TraveJtimes of the flood were deter-
mined from eyewitness accounts. The leading flood wave took 3.28
hours to travel 12.5 miles (average 3.8 miles per hour). Flood peaks
were 2.1 to 30 times the 50o-year flood. for selected locations along
.the flood path. Geomorphic and sedimentologic evidence suggest that

it probably was the largest flood in these basins, at least since the
retreat of the glaciers several thousands of years ago.
Geomorphic effects of the flood resulting from the dam failures were
profound. Channels were widened tens of feet and scoured from 5 to
50 feet locally. In the Roaring River valley. alternate river reaches
were either scoured or filled, depending on valley slope. Generally,
reaches steeper than 7 percent were scoured, and reaches less than
7 percent were filled In the Roaring River, &6 percent of the channel
was scoured, some by as much as 50 feet, and 44 percent was filled
with coarse sediments, 2 to 8 feet thick.
An alluvial fan of 42.3 acres, containing 364,600 cubic yards of
material, was deposited at the mouth of the Roaring River. The fan
has a maximum thickness of 44 feet and an average thickness of 5.3
feet. The largest boulder thought to have moved in the flood,
14X17.5X21 feet and weighing an estimated 452 tons, was located
on the alluvial fan. Down the flow axis, averflge particle size changes
fro'" 7 .5-foot boulders to fine sand and silt in a distance of 1,900 feet.
The alluvial fan dammed the Fall River, forming a lake of 17 acres
upstream. from the fan.
Satisfactory results were obtained from the dam-break model, but
not without significant difficulties in proper operation of the model.
To calibrate the model, Manning n-values between 0.1 and 0.2, or an
average of 78 percent greater than field.selected values, were required;
subcritica1 flow was verified. The occurrence of numerous debris dams
caused localized backwater, resulting in predominantly subcritical
flow. However, when these debris dams broke, flow probably was
supercritical for a short distance until another debris dam formed.
Without the extensive calibration of the model and the assumption
of subcritical flow, results would have been significantly different.
Peak discharges from dam-break modeling reflect water-only
discharges; total discharge may have been considerably higher on the
Roaring River and on the Fall River immediately downstream from
Cascade Lake dam from sediment and debris. At Horseshoe Falls and
for tl. short reach downstream from Cascade Lake dam, geomorphic
and sedimentolOgic evidence indicates the flow temporarily became
a turbulent, high-concentration, cohesionless sediment-gravity flow.
The sediment and debris may have bulked the peak water flow by 50
to 60 percent. The range of difference of observed and modeled peak
discharges varied from -3,200 cubic feet per second to 600 cubic feet
per second. The range of difference of observed and modeled maximwn
flood depth was -1.3 to 2.6 feet and averaged 1.0 foot. The range of
difference of observed and modeled leading edge of traveltime was
-0.4 and 0.15 hour,
Comparisons were made for hypothetical breach widths of (1) 25
feet and (2) 200 feet. They were comparad with model results of the
$ctual breach width of 55 feet:
1. For a breach width of 25 feet, the peak discharge would have been
7,000 cubic feet per second less downstream from Lawn Lake dam
to 1,300 cubic feet per second less at Estes Park. Maximum flood.
1