Year Completed:
Estimated 2001

Construction Cost:
$ 200,000.00


Halff Associates, Inc.
4000 Fossil Creek Boulevard
Fort Worth, TX 76137


Lynn Lovell, P.E. (Halff)
(817) 847-1422
John McCloud (LCRA)
(512) 473-2573

Nature of the Work:
  • Hydrologic Modeling
  • Hydraulic Modeling
  • Flood Frequency Analysis

Colorado River Flood Damage Evaluation

Project Description

Espey Consultants, Inc. (EC) participated as a member of the Halff Associates consulting team involved in the award winning Colorado River Flood Damage Evaluation Project for the Lower Colorado River Authority (LCRA). The project includes the use of four primary hydraulic and hydrologic models: 1) HEC-HMS, 2) HEC-RAS (unsteady), 3) HEC-FFA and 4) HEC-5.

The basic hydrologic runoff model used an hourly time step HEC-HMS model of the entire Colorado River basin, utilizing the UT-CRWR GIS-generated dendritic model. Using area-adjusted frequency rainfall depths from NWS TP-40, the HEC-HMS model was calibrated to frequency peak discharge curves. Design storm rainfall from TP-40 with appropriate areal reduction or adjustment and storm centering was also utilized. Verification and additional calibration of the model has been accomplished using the December 1991, May 1995, and June 1997 floods. The initial verification was based on the unregulated flow and volume data (i.e., simulate the floods without the reservoirs). The data output was stored in the HEC-DSS database. The calibrated HMS model of the entire basin for unregulated flow provides an updated look at the overall benefits of the Highland Lakes system on flooding.

A HEC-RAS hydraulic model was used to develop modified Puls routing data for river reaches. Cross sections for these models were taken electronically from recent LCRA aerial mapping (GIS/DTM) information using Geo-RAS. Roughness coefficients and other parameters were estimated from observations in the field, aerial photographs and previous studies. The HEC-RAS model was calibrated to best available information, including high water marks, gages and previous studies. In the lower basin, near Wharton and below, additional care was required for the hydraulic modeling of the overflow into the San Bernard River. For the preliminary analyses, the steady-state HEC-RAS model was used. The final HEC-RAS modeled for delineating floodplain limits were performed using the unsteady flow option available in version 3.0.

The project team developed HEC-5 models of the reservoir operations system, including all significant reservoirs in the basin. The model is based on current documented operation policies for each reservoir. Inflow data was taken from the series of HEC-HMS watershed models. Output from the HEC-5 model included computed inflows, releases, and pool elevation hydrographs for each reservoir, flows and pool elevations with observed data. Results showed flow hydrographs for each control point.

The HEC-5 model was calibrated by comparison of the historical December 1991, May 1995, and June 1997 floods (and possibly other storm events) and comparing the computed from the execution of the model, which established computed 100-year peak flood discharges for each control point and computed 100-year peak flood elevations at each reservoir, including Lake Travis.

Services Provided

Espey Consultants, Inc. took a leadership role in the development of the “unregulated” flood frequency curves for 3 main stem gages and 13 tributary gages located throughout the lower Colorado River basin. Based on the methods described in Bulletin 17B, EC developed flood frequency curves using HEC-FFA. The analysis included both recorded and historical data.

Initial hydrologic model calibration is required in order to analyze “critical design storm centers.” EC performed calibration runs on Pecan Bayou (2,199 square miles), Middle Colorado River (1,649 square miles) and San Saba River (3,146 square miles) using HEC-HMS. NEXRAD rainfall data and observed hydrographs from the October 1998 flood event are read from HEC-DSS for storm simulation and comparison of hydrograph response. Model calibration included timing (time to peak), runoff volume and peak flow rate. Calibration was performed by adjusting the initial and uniform loss rates as well as Snyder unit hydrograph parameters (Cp and Ct). Design storm centers were determined and the HEC-HMS models were calibrated to the flood frequency curves developed by EC.


©2001-2011 Espey Consultants, Inc. All Rights Reserved.
Site designed by Winchell's Web Works

DHTML Web Menu by OpenCube