Model Development to Support Assessment of Flood Risk for the Columbia River Treaty Review

U . S . A r m y C o r p s o f E n g i n e e r s

2012 AWRA Washington State Conference Sara Marxen, P.E.

U.S. Army Corps of Engineers Seattle District

September, 2012

Slide 2

U . S . A r m y C o r p s o f E n g i n e e r s

Presentation Outline I. Modeling Goals and Approach II. Data Collected III. Data Developed IV. Models built V. Use in the Treaty Studies

Slide 3

U . S . A r m y C o r p s o f E n g i n e e r s

The Columbia River System

Originates in Canada, Canada has 15% of the basin area, 30% of average annual flow is from Canada, 50% of worst Columbia flood flows (1894) came from Canada.

Flows 1,200 miles through 4 U.S. States, drainage area of 259,000 square miles

Over 60 large dams and reservoirs owned and operated by many different entities

Mica

Keenleyside

Duncan

Libby

Slide 4

U . S . A r m y C o r p s o f E n g i n e e r s

Flood Risk Management Task: Evaluate

Flood Risk Under Current

and Future Operating Scenarios

Goal: Provide Information to

Support a Decision on Future of the

Treaty

EM 1105-2-101: Requires Risk Analysis • Expected Annual Damage • Annual Exceedance Probability • Long Term Risk • Conditional Non-Exceedance • Residual Risk

Slide 5

U . S . A r m y C o r p s o f E n g i n e e r s

Modeling Goals Implement Flood Risk Management

Models should be flexible, maintainable, adaptable

Accurate geospatial economic data

Integrate hydrologic/reservoir/river/economic

Increase automation in the approach

Slide 6

U . S . A r m y C o r p s o f E n g i n e e r s

What are we modeling?

There are 1,000 of dams in the basin

Really only 8 of them have specific operations that are regularly used and manageable

Historically there are 22 reservoirs that are modeled as part of “System” Flood Risk Management

More than 60 dams are modeled for hydropower

Slide 7

U . S . A r m y C o r p s o f E n g i n e e r s

Data Collected 1. LIDAR – 1600 river miles, 3000 square miles 2. Use of existing and collection of new river cross-

sections/bathymetry 3. 3-D levee centerlines 4. structure inventory

Slide 8

U . S . A r m y C o r p s o f E n g i n e e r s

Data Developed 1. Hydrology

• 2000 level modified flows dataset

• “Synthetic” Low Probability Events

• Climate Change

2. Standardized run-off volume forecasts

3. Levee fragility

Slide 9

U . S . A r m y C o r p s o f E n g i n e e r s

Models Developed 1. HEC-Ras

2. HEC-ResSim

3. HEC-FIA

4. HEC-WAT with FRA

Slide 10

U . S . A r m y C o r p s o f E n g i n e e r s

Hydraulic Modeling

US

US

US

US

US

US

US

Libby

Hungry Horse

Dworshak

McNary

The Dalles

Grand Coulee Albeni Falls

Slide 11

U . S . A r m y C o r p s o f E n g i n e e r s

CRT System HEC-ResSim

Model

67 projects in the

model

36 with storage that can affect flood operations

8 which operate for “the system”

Slide 12

U . S . A r m y C o r p s o f E n g i n e e r s

CRT System HEC-ResSim Model

Simulates reservoir operations at a planning level on a daily time-step. Reservoirs were initially chosen for inclusion based primarily on

volume, and whether they were included hydropower or flood studies.

Does not model ecosystem function (biop) operations Simulates the majority of CRT Flood and Hydropower operations Includes - flow routing, automated refill, continuous or refill based

mode

Slide 13

U . S . A r m y C o r p s o f E n g i n e e r s

CRT System HEC-ResSim Model

Approach to developing post-2024 operations. What will effective use and called upon look like? How can they be designed to provide the same level of flood risk

but meet the treaty description? What does that mean for other users of the system?

Slide 14

U . S . A r m y C o r p s o f E n g i n e e r s

HEC-WAT Structure

HEC-FIA Model

Simulation.dss

WAT Simulation

With Default Program

Order

HMS Model HEC-RAS

Model

HEC-WAT

WAT Simulation

With Default Program

Order

2000 Modified Flows/

Synthetic Events

HEC-ResSim Model

Flow

Hydrologic Data

HEC-ResSim Plug-In

HEC-RAS Plug-In

HEC-FIA Plug-In

Hydrographs

Hydrographs Hydrographs

Damages

Slide 15

U . S . A r m y C o r p s o f E n g i n e e r s

Flood Risk System In

flow

Qi

Time

Inflo

w Q

i

Time

Inflo

w Q

i

Time

Qi Qb

Qo

Out

flow

Qo

Time

Reservoir 1

Reservoir 2

.

Slide 16

U . S . A r m y C o r p s o f E n g i n e e r s

HEC-FIA Model

HEC-RAS Model

Regulated Hydrographs

Breach Hydrographs

Damages

HEC-WAT with FRA option

Model Convergence

?

Freq Sampler

Spreading Model

Spreading Model

Frequency

Hydrograph Sets

Damage Sets

No

Fragility Sampler

Simulation.dss

Depths, Velocities per Grid

CSRA

Benefits

Net Benefit

Sampler

Expected Net Benefit,

CDF

Yes

HEC- ResSim

RAS Mapper

Forecast Sampler

Unregulated Hydrographs

Mean Forecast

Depth Grids

HEC-WAT Structure - FRA

Slide 17

U . S . A r m y C o r p s o f E n g i n e e r s

Flood Risk System In

flow

Qi

Time

Inflo

w Q

i

Time

Inflo

w Q

i

Time

Li Sta

ge (

ft)

Probability of Levee Failure

Qi Qb

Qo

Out

flow

Qo

Time

Reservoir 1

Reservoir 2

.

Res

ervo

ir

Elev

atio

n

Time

Fore

cast

V

olum

e

Observed Volume

Slide 18

U . S . A r m y C o r p s o f E n g i n e e r s

What does this mean for the Flood Risk Analysis

Year x has a peak of 435,000 cfs at The Dalles with no Called Upon

Rather than this…..

Slide 19

U . S . A r m y C o r p s o f E n g i n e e r s

What does this mean for Flood Risk Analysis

Forecast Volume

Peak

Flo

w

Example: Each color represents a single year and the effect of forecast uncertainty on the resulting peak flow

Slide 20

U . S . A r m y C o r p s o f E n g i n e e r s

Uncertainties Incorporated in FRA (short-term)

Forecast uncertainty

Starting reservoir pool elevation

Stage/flow rating curves

Roughness coefficients

Levee breaching parameters/fragility

Slide 21

U . S . A r m y C o r p s o f E n g i n e e r s

HEC-ResSim Starting Storage/Elevation Stream routing coefficients Stage/flow Rating Curves Reservoir physical data:

storage/elevation, release capacity, etc. Demands (water, power)

Current Power Capacity (outages) Sedimentation changes HEC-RAS

Roughness Coefficients Weir Coefficients Bridge Debris Gate Coefficients Ice thickness Bridge/culvert coefficients Dam/levee breeching parameters Contraction/Expansion coefficients

Boundary Conditions (normal depth slope, etc.)

Terrain Data

HEC-FIA Foundation Height Ground Elevation Structure value Content Value Car Value Other Value Depth/Damage functions Population at Risk

Uncertainties Incorporated in FRA (long-term)

Slide 22

U . S . A r m y C o r p s o f E n g i n e e r s

Many Other Uncertainties What will be an agreed upon called upon operation? What will be an agreed upon effective use operation? What could a CanadianTreaty Terminates operation look like? What Treaty operation will be agreed upon vs. what we assume? What biological operations will be desirable in U.S. and Canada? How will the basin change over time (people, places, demands)? How will the climate change over time? How will an operation change in real-time implementation, is it

significant?……….

Slide 23

U . S . A r m y C o r p s o f E n g i n e e r s

Example of using FRA to address this

Arrow Elevation: 70 Yr Avg.

1426.0

1428.0

1430.0

1432.0

1434.0

1436.0

1438.0

1440.0

1442.0

1444.0

1-Oct OCT NOV DEC JAN FEB MAR APR I APR II MAY JUNE JULY AUG I AUG II SEP

Elev

atio

n (ft

.)

Case 1 Case 2: 165 Case 2: 225 Case 3Case 4 Case 4b Case 7 Case 10Case 14 Case 8 Case 9 Case 3bCase 15 Case 4c Case 4q Case 4FB

Arrow Elevations: 70 Yr. Avg.

1395.0

1400.0

1405.0

1410.0

1415.0

1420.0

1425.0

1430.0

1435.0

1440.0

1445.010

/1/2

010

10/1

5/20

10

10/ 2

9/20

10

11/1

2/20

10

11/2

6/20

10

12/1

0/20

10

12/2

4/20

10

1/7/

2011

1/21

/201

1

2/4/

2011

2/18

/201

1

3/4/

2011

3/18

/201

1

4/1/

2011

4/15

/201

1

4/29

/201

1

5/13

/201

1

5/27

/201

1

6/10

/201

1

6/24

/201

1

7/8/

2011

7/22

/201

1

8/5/

2011

8/19

/201

1

9/2/

2011

9/16

/201

1

9/30

/201

1

Out

flow

(kcf

s)

Case 1 (BCH) Case 4C (Recom.)

Case 4FB (Alt. Arrow) Avg 1995-2011 (17 yrs)

Apply probability to the operations cases based on expert knowledge. Incorporate that into flood risk. Case 1 – 50% chance of occurring Case 4FB – 25% chance of occurring Case 4C – 20% chance of occurring Case NCS – 5% chance of occurring

Slide 24

U . S . A r m y C o r p s o f E n g i n e e r s

What does this mean for the Treaty Studies

We now have a tool in which operational changes in Canada or the U.S. can, relatively quickly, be evaluated for flood risk.

This is only one piece of the puzzle that will be use to make a decision. Multiple studies through STT/SRT are ongoing: – How should called upon and effective use work operationally? – What is the impact of a less conservative flood risk operation? – What is the impact of trying to meet more normative flow levels in the

mainstem? – What is the impact of more normative reservoir levels? – Can levees be improved o reduce flood risk? – Can levees be removed to reduce flood risk? – What is an appropriate system-based dry year operation?

Slide 25

U . S . A r m y C o r p s o f E n g i n e e r s

Questions

Acknowledgement goes to multiple engineers in Seattle, Portland, Walla Walla, Division Columbia Basin Water Management, Hydrologic Engineering Center, BPA and West Consultants Pete Dickerson, Ryan Cahill, Ron Malmgren, Geoffrey Walters, Kristian Mickelson,Tracy Schwarz, Jeremy Giovando, Travis Ball, Margaret McGill, Joan Klipsch, Matt Fleming, Chris Nygaard, and more!