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

DAM,BREAK MODELING

51

difficult. Field selected n-values (see Supplemental Cross-
Section Data section at the end of the report) ranged from
an estimated 0.15 on the Roaring River to 0.035 along
the lower reach of the Fall River, and the Big Thompson
River. Composite n-values used in the model are listed
by reach in table 8. Initially, expansion coefficients were
-0.5. and the contraction coefficients were 0.5. Cross sec- .
tions were assumed to contain only active-flow area based
on flood and postflood field observations.
The first model runs wonld not compute, as the flow
regime alternated between subcritical and supercritical
flow. depending on the flow rate. Therefore, runs were
made by increasing or decreasing n-values to maintain
subcritical flow or supercritical flow in various segments,
based on slopes (as previously discussed). When flow was
assumed to be supercritical, that is lower n-values, the
model peak discharges were too large; water depths were
too shallow; and traveltimes were too fast. Therefore, n-
values were increased to 0.2 along the Roaring River and
to 0.1 on the lower reaches of the Fall River and the Big
Thompson River to maintain subcritical flow in all seg-
ments. which provided the best results. This substanti-
ates that subcritical flow predominated in the river
reaches studied These modified or calibrated model n-
values averaged 78 percent larger than field-selected n-
values; they are listed by segment in table K These n-
values were reasonable since Leutheusser and Chisholm
(1973) measured an n-value of 0.225 in a heavily vegetated
channel; flow also was subcritical. The use of the inac-
tive flow area did not improve model results. Results of
the model calibration of observed and computed peak dis-
charges, flow depths. trave1times of the leading edge of
the flood wave, and percent differences for selected loca-
tions are shown in table 9. Peak discharges also are shown
as a peak-discharge profile in figure 11. The peak-
discharge profile was based on indirect peak-discharge
measurements and model results. Selected modeled flood
hydrographs and the observed hydrograph at Lake Estes
are shown in figure 46. Flood profiles are not shown
because of the extremely high-gradient slopes and
shallow depths of flow.
Results of the calibration phase indicated that the
model has the potential to simulate dam-break floods in
high-gradient stream channels. The range in difference
of observed and modeled peak discharges varied from
-3,200 ft'ls to 600 ft'/s. At worst. the model under-
predicted peak discharge by 27 percent 3.6 mi
downstream from Cascade Lake dam (Site 3). Results
were quite reasonable, considering the dynamics of the
breach development. The range of difference of observed
and modeled maximum flood depth was -1.3 to 2.6 ft;
the range averaged 1.0 ft. The range of difference of
observed and modeled leading edge of traveltime was
-0.4 to 0.05 hour. Considering the complex problems

TABLE B.-Field selected- and calibrated model-composite Manning's
o.values

Modo!
lIegmentEl.

Distance downstream
from Lawn Lake dam,
in miles

Field selected Calibrated model

n-values

1
2
3
4
5

0.00- 4.74
4,74- 6.50
6.67- 7.78
7.78-11.46
11.46-12.52

0.125 0.20
.075 .11
.055 .10
.055 .10
.035 .10

asee figure 45 for segment location.

previously discussed, the model reasonably simulated
this dam-break flood and demonstrated that the model
could be used to provide supplementary hydrologic in-
formation that could not be obtained otherwise. How-
ever. it must be emphasized that model results would
have been significantly different without the extensive
calibration and the assumptions made to allow the
model to operate.

MODEL SIMULATIONS

Selected hypothetical scenarios. of dam-breach
development and the probable impact of the failure of
Cascade Lake dam were made. Analyses of the
scenarios were made to indicate the range in possible
flood conditions resulting from the failure of Lawn Lake
dam. Factors that could result in the greatest range of
peak discharge from the failure of a dam appeared to
be the time and width of breach development. Shorter
time and greater widths of breach development resulted
in larger peak discharges for a given volume of stored
water. This particularly applied to Lawn Lake dam
where the embankment was narrow and long. Converse-
ly, the embankment of Cascade Lake dam was narrow
and short and probably would have failed in the same
manner, regardless of the magnitude of the inflow-flood
hydrograph from the failure of Lawn Lake dam. How-
ever. it is important to understand what the flood
hydrology would have been if Cascade Lake dam had
not been present. or if Cascade Lake dam had not failed.
These two general scenarios are discussed separately
to provide information on the rang&of flood conditions
that could have occurred.

VARIATIONS OF LAWN LAKE DAM BREACH
WIDTH AND TIMING

Accuracy of the modeled peak discharges was lowest
immediately downstream from Lawn Lake and Cascade
Lake dams, primarily because of the uncertain aspects
of length of time of full breach development, as well as