1. UNSTEADY FLOW MODEL OF LOWER SNAKE RIVER USING HEC-RAS v. 3.1.2 JOHN HEITSTUMAN, PE CE 502 AND CE 504C

2. PROJECT PURPOSE LEARN TO USE HEC-RAS UNSTEADY FLOW MODEL DEVELOP UNSTEADY FLOW MODEL OF LOWER SNAKE RIVER AND TRIBUTARIES FOR REAL TIME MODELING AND PLANNING STUDIES

3. DATA COLLECTION AND ANALYSIS LATEST AVAILABLE GEOMETRIC DATA FOR APPROXIMATELY 257 MILES OF THE SNAKE RIVER COLLECTED, MODIFIED, AND INTERPOLATED. GEOMETRIC DATA FOR TRIBUTARIES, OTHER THAN LOWER CLEARWATER RIVER ADAPTED FROM FLDWAV MODEL DEVELOPED BY GENE R. SPANGUDE, PE GEOMETRIC DATA FOR SNAKE RIVER UPSTREAM OF ASOTIN, WA, PROVIDED BY IDAHO POWER DISTANCE DISCRETIZATION OR CROSS-SECTION SPACING OF APPROXIMATELY 500 FEET USED FOR ALL REACHES DUE TO STEEP SLOPES AND VARIABLE CHANNEL GEOMETRY STAGE AND DISCHARGE DATA COLLECTED, FILTERED, AND FORMATTED FOR FIFTEEN SITES FOR WATER YEARS 1996 THROUGH 2002 ALL STAGE AND DISCHARGE DATA CONVERTED TO HOURLY DATA LOWER SNAKE RIVER DAMS’ GATE AND TURBINE DISCHARGES APPROXIMATED BY TAINTER GATE AND SLUICE GATE EQUATIONS AVAILABLE IN HEC-RAS TIME DISCRETIZATION FOR ALL MODELING RANGED FROM ONE TO FIVE MINUTES; THREE MINUTES WAS FOUND TO BE OPTIMUM MODEL CALIBRATED FOR THE PERIOD APRIL THROUGH JULY 1996 MODEL VALIDATED FOR JANUARY 1997, FEBRUARY THROUGH JULY 1997, AND MAY THROUGH JULY 2002

4. PENDING ANALYSIS TIME SERIES GATE OPENINGS FOR LOWER SNAKE RIVER DAMS BECAUSE FULL BOUNDARY CONDITIONS ARE NOT SUPPORTED IN HEC-RAS v. 3.1.2 REAL TIME VALIDATION OF MODEL

5. HISTORICAL PERSPECTIVE HEC-RAS DEVELOPED TO REPLACE HEC-2, HEC-6, AND HEC-UNET HEC-RAS IS PART OF THE CORPS WATER MANAGEMENT SYSTEM (CWMS) SUITE OF PROGRAMS CWMS DEVELOPED TO UNITE CORPS DATA BASES AND STANDARDIZE MODELS AS A RESULT OF LESSONS LEARNED FROM 1993 MIDWEST FLOODS VERY ABREVIATED HISTORY OF UNSTEADY FLOW MODELING LEADING TO HEC-RAS….

7. If I have seen further it is by standing on the shoulders of giants. Isaac Newton, Letter to Robert Hooke, February 5, 1675

10. Dr. Danny L. Fread Dr. Danny L. Fread received his BS in Civil Engineering in 1961 from the Missouri School of Mines and Metallurgy where he ranked first in the graduating class. He worked six years for Texaco Inc. where he was promoted to Senior Engineer and specialized in the design of gravity and pressurized piping systems. In 1971 after completing four years of graduate studies at the University of Missouri - Rolla, he received a PhD. His studies focused on hydraulics / hydrology / mathematics, and his research was centered on unsteady flow and numerical / experimental simulation of breached dams. Dr. Fread retired recently from the National Weather Service (NWS) after 29 years of service, where he progressed during 1971-1978 from Research Hydrologist (GS-12, 13, and 14) to Senior Research Hydrologist (GS-15) to Director of Hydrologic Research (SES-4) from 1988-1994 to Director of the Office of Hydrology (SES-4) from 1995-1999. During the 70’s and 80’s he personally formulated, coded and tested the following mathematical simulation programs (computer models): DWOPER, DAMBRK, SMPDBK, BREACH, and FLDWAV. The DWOPER model simulates unsteady flows from rainfallrunoff in river systems; the DAMBRK and SMPDBK models simulate unsteady flows from breached dams in a single river; BREACH simulates the erosive formation of breaches in earthen dams; and FLDWAV is an improved simulation model of unsteady flows from rainfall-runoff and from breached dams in a single river or network of rivers. Currently and for over 20 years, these models have been extensively used for unsteady river flow modeling by Federal and State Agencies, consulting firms, mining and hydropower firms, and for educational purposes in universities; these agencies, firms and universities are located in the United States and Canada and in more than 20 other nations worldwide. Since 1976, Dr. Fread has provided over 120 training workshops on the models, each consisting of 24 hours of training, to over 3000 engineers. He has authored 50 and co-authored 42 professional scientific papers, and contributed chapters on modeling unsteady flows to four books including the Handbook of Hydrology. His last decade with the National Weather Service consisted of directing for six years a highly trained staff of 35 who were engaged in hydrologic research and development of the NWS River Forecasting System, and for four years a staff of 85 who were involved in hydrologic research / development and policy / support services for the NWS mission of providing real-time river forecasting services to the Nation. Dr. Fread received national awards for his work including the Department of Commerce Gold Medal, the American Society of Civil Engineers (ASCE) Huber Research Prize and J.C. Stevens Award, the Federal Laboratory Consortium Award for Technology Transfer, the Association of State Dam Safety Officials (ASDSO) National Award of Merit, and the American Meteorological Society Award of Fellow. Since his retirement from the National Weather Service in 2000, Dr. Fread has engaged in consulting work which includes: enhancing the FLDWAV model and extending it for unsteady flows in storm sewers, serving as an expert witness in dam breach arbitration cases, providing several dam breach analyses for private firms and State Agencies, and providing training workshops on the FLDWAV, DAMBRK and BREACH models.

12. BREIF ORIENTATION TO RIVER REACHES MODELED LOCATION MAP SELECTED PHOTOS HEC-RAS SCHEMATIC INVERT PROFILES

25. GAGE DATA USED FOR MODEL INPUT, CALIBRATION, AND VALIDATION

38. MODEL CALIBRATION INCLUDING STATISTICAL ANALYSIS STATISTICS COMPUTED WITH HEC-DSS CORRELATION AND LAG COMPUTED USING ITSM 2000 PROFESSIONAL BY PETER J. BROCKWELL AND RICHARD A. DAVIS COLORADO STATE UNIVERSITY

42. ================== ITSM::Multivariate(CCF) ================== # of Lags = 3 Sample Cross-correlations: Cor[X_1(t+h),X_1(t)] 1.0000.9991.9975.9953 Cor[X_1(t+h),X_2(t)].9934.9945.9949.9946 Cor[X_2(t+h),X_1(t)].9934.9910.9883.9854 Cor[X_2(t+h),X_2(t)] 1.0000.9991.9975.9952

51. MODEL RESULTS FOR RAPIDLY VARYING DISCHARGE AND STAGE VALIDATION JANUARY 1997

56. MODEL RESULTS VALIDATION FEBUARY THROUGH JULY 1997

61. MODEL RESULTS VALIDATION MAY THROUGH JULY 2002 (AVERAGE TO LOW FLOW YEAR)

66. HEC-RAS VISUALIZATION TOOLS PROFILE ANIMATION PSEUDO 3D VIEWER

69. LESSONS LEARNED GEOMETRIC DATA –CROSS-SECTION EFFECTIVE FLOW AREAS MUST BE WELL DEFINED TO AVOID TRANSCRITICAL FLOW –USE ONLY ENOUGH POINTS TO DEFINE CROSS SECTION; EXCESS POINTS USE MORE MEMORY AND SLOW PROGRAM EXECUTION –FIND OPTIMUM CROSS-SECTION SPACING BY TRIAL AND ERROR MODIFYING ONLY ONE RIVER REACH AT A TIME –AT JUNCTIONS CROSS-SECTION INVERTS SHOULD BE MADE TO BE EQUAL CROSS-SECTIONS SHOULD BE REGULAR IN SHAPE AND CONSITST OF ONLY A FEW POINTS –CHECK ALL REACH LENGTHS RELATIVE TO STRUCTURES, GAGES, AND JUNCTIONS CAREFULLY –THERE IS A PRACTICAL LIMIT (MEMORY) TO THE NUMBER OF CROSS-SECTIONS IN A MODEL; ABOUT 3,500 –SPECIFY ENOUGH VERTICAL ‘SLICES’ FOR CROSS-SECTIONS TO DEFINE CONVEYANCE OVER ENTIRE RANGE OF FLOWS –CHECK THE SHAPE OF ALL CONVEYANCE CURVES FOR DISCONTINUITIES

70. LESSONS LEARNED (CONTINUED) TIME SERIES DATA –INPUT HYDROGRAPHS ALL STAGES AND FLOWS MUST BE SCREENED CAREFULLY FOR VERY LARGE OR SMALL VALUES (UNREASONABLE OR OUT OF POSSIBLE RANGE) MISSING DATA CAN BE LINEARLY INTERPOLATED USING HEC DSS VUE, HOWEVER MINIMUM FLOWS MUST BE SPECIFIED WELL ABOVE ZERO FLOW OR STAGE. –SCREENING OF HYDROGRAPHS IS CRITICAL FOR REAL-TIME OPERATIONS, AND DIFFICULT TO IMPLEMENT CORRECTLY –MOST OF THE TIME, WHEN UNET MATRIX FAILS TO CONVERGE; IT IS DUE TO UNREASONABLE HYDROGRAPHS (PROVIDING TIME AND DISTANCE DISCRETIZATION IS OK) –GATE OPENINGS FOR NOW, GATE OPENINGS MUST BE SPECIFIED BY FITTING RATING CURVES TO EQUATIONS FOR NOW TURBINE FLOW IS BEST MODELED BY SLUICE GATE CURRENTLY STAGE AND BOUNDARY CONDITIONS CANNOT BE SPECIFIED IN CONJUCTION WITH IN LINE STRUCTURES (HEC IS WORKING ON THIS NOW) –TIME DISCRETIZATION TIME DISCRETIZATION IS CRITICAL TO MODEL OPERATION AND ACCURACY OF RESULTS –REFER TO THE HEC-RAS FOR UNSTEADY FLOW FOR FIRST TRY –USUALLY SUGGESTED VALUE IS TOO SMALL AND A SLIGHTLY LARGER VALUE MAY BE USED –MONITOR OUTFLOW HYDROGRAPHS CLOSELY FOR RESOLUTION WHEN INCREASING TIME STEP –THE TIME STEP THAT IS ‘JUST RIGHT’ USUALLY CAUSES THE MODEL TO RUN WITH FEWEST ITERATIONS

71. LESSONS LEARNED (CONTINUED) RECOMMENDATIONS FOR FURTHER DEVELOPMENT OF HEC-RAS –MODEL SHOULD SUPPORT AUTOMATIC VARIABLE TIME STEPS AS DOES FLDWAV –RATING CURVES FOR TAINTER GATES AND TURBINES, AND HYDROPOWER OUTPUT SHOULD BE ADDED (SCHEDULED FOR RELEASE THIS FALL?) –ONCE PROGRAM IS ‘MATURE’ RELEASE THE SOURCE CODE TO FACILITATE DEBUGGING (THE SOONER THE BETTER)

72. HEC-RAS DATA INPUT FORMATS TEXT FILES –GEOMETRY =.G01 –UNSTEADY FLOW =.U01 –PROJECT FILE =.PRJ –PLAN FILE =.P01 DSS FILES –TIME SERIES DATA

73. VIEW TEXT FILES WITH ULTRAEDIT

74. DEMONSTRATE HEC-DSS VUE

75. DEMONSTRATE HEC-RAS INTERFACE