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

UNSTEADY FLOW MODELOF

LOWER SNAKE RIVERUSING

HEC-RAS v. 3.1.2

JOHN HEITSTUMAN, PE

CE 502 AND CE 504C

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

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

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

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….

If I have seen further it is by standing on the shoulders of giants.

Isaac Newton, Letter to Robert Hooke, February 5, 1675

Dr. Danny L. FreadDr. Danny L. Fread received his BS in Civil Engineering in 1961 from the Missouri School of Mines andMetallurgy where he ranked first in the graduating class. He worked six years for Texaco Inc. where hewas 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 receiveda PhD. His studies focused on hydraulics / hydrology / mathematics, and his research was centered onunsteady flow and numerical / experimental simulation of breached dams.Dr. Fread retired recently from the National Weather Service (NWS) after 29 years of service, where heprogressed during 1971-1978 from Research Hydrologist (GS-12 , 13, and 14) to Senior ResearchHydrologist (GS-15) to Director of Hydrologic Research (SES-4) from 1988-1994 to Director of theOffice of Hydrology (SES-4) from 1995-1999. During the 70’s and 80’s he personally formulated, codedand tested the following mathematical simulation programs (computer models): DWOPER, DAMBRK,SMPDBK, BREACH, and FLDWAV. The DWOPER model simulates unsteady flows from rainfallrunoffin river systems; the DAMBRK and SMPDBK models simulate unsteady flows from breacheddams in a single river; BREACH simulates the erosive formation of breaches in earthen dams; andFLDWAV is an improved simulation model of unsteady flows from rainfall-runoff and from breacheddams in a single river or network of rivers. Currently and for over 20 years, these models have beenextensively 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 anduniversities 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 24hours of training, to over 3000 engineers. He has authored 50 and co-authored 42 professional scientificpapers, and contributed chapters on modeling unsteady flows to four books including the Handbook ofHydrology. His last decade with the National Weather Service consisted of directing for six years ahighly trained staff of 35 who were engaged in hydrologic research and development of the NWS RiverForecasting 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 forecastingservices to the Nation.Dr. Fread received national awards for his work including the Department of Commerce Gold Medal, theAmerican Society of Civil Engineers (ASCE) Huber Research Prize and J.C. Stevens Award, the FederalLaboratory 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 consultingwork which includes: enhancing the FLDWAV model and extending it for unsteady flows in stormsewers, serving as an expert witness in dam breach arbitration cases, providing several dam breachanalyses for private firms and State Agencies, and providing training workshops on the FLDWAV,DAMBRK and BREACH models.

BREIF ORIENTATION TO RIVER REACHES MODELED

• LOCATION MAP

• SELECTED PHOTOS

• HEC-RAS SCHEMATIC

• INVERT PROFILES

54CLEARWATER3CLEARWATER21

1

GRANDE RONDE

1

1

1

1

1

SALMON

1

1

4

SN

AK

E

3

SN

AK

E

2

SN

AK

E

1SNAKE

CONF CLRWTR

CONF SALMON

CONF GR

LOC / SEL

SF CLR

NF CLRPOT

LAP

Partial GIS data

0 50 100 150 200 250

200

400

600

800

1000

1200

1400

1600

SNAKE RIVER INVERT PROFILE

Main Channel Distance (mi)

Ele

vatio

n (f

t)

Legend

Ground

SNAKE 1SNAKE 2

SNAKE 3SNAKE 4

SALMON 1GRANDE RONDE 1

CLEARWATER 1

0 20 40 60 80 100

600

800

1000

1200

1400

1600

CLEARWATER RIVER INVERT PROFILE

Main Channel Distance (mi)

Ele

vatio

n (f

t)

Legend

Ground

CLEARWATER 1CLEARWATER 2

CLEARWATER 3CLEARWATER 4

CLEARWATER 5SF CLRWTR 1

SELWAY 1POTLATCH 1

NF CLRWTR 1

LOCHSA 1

LAPWAI 1

GAGE DATA

USED FOR MODEL INPUT, CALIBRATION,

AND VALIDATION

/SNAKE/HCDIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

20000

40000

60000

80000

100000

HELLS CANYON OUTFLOW

/SALMON/WHBIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

20000

40000

60000

80000

100000

120000

SALMON RIVER AT WHITE BIRD, ID

/GRANDE RONDE/TRYOQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

5000

10000

15000

20000

25000

30000

35000

40000

GRANDE RONDE RIVER AT TROY, OR

/SNAKE/ANAWQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

SNAKE RIVER NEAR ANATONE, WA

/LOCHSA/LOCIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

5000

10000

15000

20000

25000

30000

35000

LOCHSA RIVER NEAR LOWELL, ID

/SELWAY/SELIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

SELWAY RIVER NEAR LOWELL, ID

/CLEARWATER/ORFIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

CLEARWATER RIVER AT OROFINO, ID

/NF CLEARWATER/DWRQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

5000

10000

15000

20000

25000

30000

DWORSHAK OUTFLOW

/CLEARWATER/PEKIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

CLEARWATER RIVER NEAR PECK, ID

/CLEARWATER/SPDIQR/FLOW/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

CLEARWATER RIVER AT SPALDING, ID

/SNAKE/SRCIHG/STAGE/01SEP1995/1HOUR/OBS/

1996 1997 1998 1999 2000 2001 2002

1995 1996 1997 1998 1999 2000 2001 2002

730

732

734

736

738

740

SNAKE RIVER AT LEWISTON, ID CONTROL POINT(FEET NGVD 29)

/SNAKE/LWGFB/STAGE/01SEP1995/1HOUR/OBS/

729

730

731

732

733

734

735

736

737

738

739

1996 1997 1998 1999 2000 2001 20021995 1996 1997 1998 1999 2000 2001 2002

0

50000

100000

150000

200000

250000

LOWER GRANITE FOREBAY ELEVATIONLOWER GRANITE DISCHARGE

MODEL CALIBRATIONINCLUDING STATISTICAL ANALYSIS

STATISTICS COMPUTED WITH HEC-DSSCORRELATION AND LAG COMPUTED USING

ITSM 2000 PROFESSIONALBY PETER J. BROCKWELL

ANDRICHARD A. DAVIS

COLORADO STATE UNIVERSITY

/SNAKE 2/167.206*/FLOW/01APR1996/1HOUR/PLAN WY97/

21 5 19 2 16 30 14 28

Apr1996 May1996 Jun1996 Jul1996

0

20000

40000

60000

80000

100000

120000

140000

160000

SNAKE RIVER NEAR ANATONE, WAAPRIL THRU JULY 1996

MODEL VS. GAGE

2.00E +04

4.00E +04

6.00E +04

8.00E +04

1.00E +05

1.20E +05

1.40E +05

1.60E +05

0 500 1000 1500 2000 2500

S eries 1

2.00E +04

4.00E +04

6.00E +04

8.00E +04

1.00E +05

1.20E +05

1.40E +05

0 500 1000 1500 2000 2500

S eries 2

==================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

/CLEARWATER 4/2.68421*/FLOW/01APR1996/1HOUR/PLAN WY97/

21 5 19 2 16 30 14 28

Apr1996 May1996 Jun1996 Jul1996

0

10000

20000

30000

40000

50000

60000

70000

80000

CLEARWATER RIVER AT OROFINO, IDAPRIL 1996 THRU JULY 1996

MODEL VS GAGE(24% OF DRAINAGE UNGAGED)

/CLEARWATER 3/2.81308*/FLOW/01APR1996/1HOUR/PLAN WY97/

21 5 19 2 16 30 14 28

Apr1996 May1996 Jun1996 Jul1996

0

10000

20000

30000

40000

50000

60000

70000

CLEARWATER RIVER NEAR PECK, IDAPRIL 1996 THRU JULY 1996

MODEL VS GAGE

/CLEARWATER 1/9.80866*/FLOW/01APR1996/1HOUR/PLAN WY97/

21 5 19 2 16 30 14 28

Apr1996 May1996 Jun1996 Jul1996

0

10000

20000

30000

40000

50000

60000

70000

80000

CLEARWATER RIVER AT SPALDING, IDAPRIL 1996 THRU JULY 1996

MODEL VS GAGE(UNGAGED PORTION OF DRAINAGE CAN PRODUCE MAJORITY OF FLOW

GENERALLY DUE TO RAIN AT LOWER ELEVATIONS)

/SNAKE 1/139.3/STAGE/01APR1996/1HOUR/PLAN WY97/

21 5 19 2 16 30 14 28

Apr1996 May1996 Jun1996 Jul1996

733.0

733.5

734.0

734.5

735.0

735.5

736.0

736.5

737.0

737.5

SNAKE RIVER AT LEWISTON, ID CONTROL POINT(NOTE NOISE IN GAGE DATA)

MODEL RESULTS FOR RAPIDLY VARYING

DISCHARGE AND STAGEVALIDATION

JANUARY 1997

/SNAKE 2/167.206*/FLOW/01JAN1997/1HOUR/PLAN WY97/

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Jan1997

40000

60000

80000

100000

120000

140000

160000

180000

SNAKE RIVER NEAR ANATONE, WAJANUARY 1997

MODEL VS GAGE

/CLEARWATER 4/2.68421*/FLOW/01JAN1997/1HOUR/PLAN WY97/

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Jan1997

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

22000

CLEARWATER RIVER AT OROFINO, IDJANUARY 1997

(24% OF DRAINAGE UNGAGED)

/CLEARWATER 3/2.81308*/FLOW/01JAN1997/1HOUR/PLAN WY97/

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Jan1997

8000

10000

12000

14000

16000

18000

20000

22000

24000

CLEARWATER RIVER NEAR PECK, IDJANUARY 1997

MODEL VS GAGE

/SNAKE 1/139.3/STAGE/01JAN1997/1HOUR/PLAN WY97/

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Jan1997

732.5

733.0

733.5

734.0

734.5

735.0

735.5

736.0

736.5

737.0

737.5

SNAKE RIVER AT LEWISTON, ID CONTROL POINTJANUARY 1997

MODEL VS GAGE WITH OBSERVED FOREBAY ELEVATION(NOTE RESERVOIR LOWERED THRU 6 JANUARY DUE TO HIGH INFLOW)

MODEL RESULTSVALIDATION

FEBUARY THROUGH JULY 1997

/SNAKE 2/167.206*/FLOW/01JAN1997/1HOUR/PLAN WY97/

Mar Apr May Jun Jul

1997

20000

40000

60000

80000

100000

120000

140000

160000

SNAKE RIVER NEAR ANATONE, WAFEBRUARY THRU JULY 1997

MODEL VS GAGE

/CLEARWATER 4/2.68421*/FLOW/01JAN1997/1HOUR/PLAN WY97/

Mar Apr May Jun Jul

1997

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

CLEARWATER RIVER AT OROFINO, IDFEBRUARY THRU JULY 1997

MODEL VS GAGE(24% OF DRAINAGE UNGAGED)

/CLEARWATER 1/9.80866*/FLOW/01JAN1997/1HOUR/PLAN WY97/

Mar Apr May Jun Jul

1997

10000

20000

30000

40000

50000

60000

70000

80000

90000

CLEARWATER RIVER AT SPALDING, IDFEBRUARY THRU JULY 1997

MODEL VS GAGE(UNGAGED PORTION OF DRAINAGE CAN PRODUCE MAJORITY OF FLOW

GENERALLY DUE TO RAIN AT LOWER ELEVATIONS)

/SNAKE 1/139.3/STAGE/01JAN1997/1HOUR/PLAN WY97/

Mar Apr May Jun Jul

1997

732.5

733.0

733.5

734.0

734.5

735.0

735.5

736.0

736.5

737.0

737.5

SNAKE RIVER AT LEWISTON, ID CONTROL POINTFEBRUARY 1997 THRU JULY 1997

MODEL VS GAGE WITH FOREBAY ELEVATION

MODEL RESULTSVALIDATION

MAY THROUGH JULY 2002

(AVERAGE TO LOW FLOW YEAR)

/SNAKE 2/167.206*/FLOW/01MAY2002/1HOUR/PLAN WY97/

12 26 9 23 7 21

May2002 Jun2002 Jul2002

10000

20000

30000

40000

50000

60000

70000

80000

90000

SNAKE RIVER NEAR ANATONE, WAMAY THRU JULY 2002

MODEL VS GAGE

/CLEARWATER 4/2.68421*/FLOW/01MAY2002/1HOUR/PLAN WY97/

12 26 9 23 7 21

May2002 Jun2002 Jul2002

0

10000

20000

30000

40000

50000

60000

CLEARWATER RIVER AT OROFINO, IDMAY THRU JULY 2002

MODEL VS GAGE(24% OF DRAINAGE UNGAGED)

/CLEARWATER 1/9.80866*/FLOW/01MAY2002/1HOUR/PLAN WY97/

12 26 9 23 7 21

May2002 Jun2002 Jul2002

10000

15000

20000

25000

30000

35000

40000

45000

50000

55000

60000

CLEAR WATER RIVER AT SPALDING, IDMAY THRU JULY 2002

MODEL VS GAGE

/SNAKE 1/139.3/STAGE/01MAY2002/1HOUR/PLAN WY97/

12 26 9 23 7 21

May2002 Jun2002 Jul2002

733.0

733.5

734.0

734.5

735.0

735.5

736.0

736.5

737.0

737.5

SNAKE RIVER AT LEWISTON, ID CONTROL POINTMAY THRU JULY 2002

MODEL VS GAGE WITH LOWER GRANITE FOREBAY ELEVATION

HEC-RAS VISUALIZATION TOOLS

• PROFILE ANIMATION

• PSEUDO 3D VIEWER

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

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

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)

HEC-RAS DATA INPUT FORMATS

• TEXT FILES – GEOMETRY = .G01– UNSTEADY FLOW = .U01– PROJECT FILE = .PRJ– PLAN FILE = .P01

• DSS FILES– TIME SERIES DATA

VIEW TEXT FILES WITH ULTRAEDIT

DEMONSTRATE HEC-DSS VUE

DEMONSTRATE HEC-RAS INTERFACE