abhishek singh, phd, pe · abhishek singh, phd, pe incorporating uncertainty into integrated...

1

Abhishek Singh, PhD, PE

Incorporating Uncertainty into Integrated Regional Water Resources Planning

Oct 1, 2019

2

Planning under Uncertainty

• Planning with certainty is a rare luxury

• Planning under uncertainty is the norm

3

Outline

• Planning under uncertainty• 5-steps for uncertainty assessment• Case-Study – Texas water planning• Lessons learned

4

The Planning Process

Now

How we get there

Decisio

ns, Acti

ons

Future

Where w

e want

 to be 

Success

5

Planning under Uncertainty

Now

How we get there

Decisio

ns, Acti

onsWhere w

e want

 to be 

Future?

Success

6

• Account for uncertainty

Alternatives?

• Ignore uncertainty• Deterministic• Reactive• Near-Term

• Probabilistic• Scenarios• Adaptive• Long-Term

7

Planning under Uncertainty• How do we maximize our chance of success given

everything we do not know?

8

5-step Program

1. Identify Uncertainties • Known Unknowns/Unknown Unknowns

2. Characterize Uncertainties

3. Relate Uncertainty to Key Decisions

4. Assess Sensitivity to/Importance of Uncertainties

5. Manage Uncertainty• Reduce Uncertainty

• Increase Reliability/Reduce Risk

• Increase Resilience

• Monitor, Measure, and Adapt

9

Case Study

10

Texas Water Planning Framework• Regional water planning

process for resilient water supply– Drought of record– Deterministic

• Stake-holder driven process– Regional water planning

groups• 50-year planning horizon

– Updated every 5 years

•

11

Objective

• Methodology to inform decision-makers how to characterize and account for uncertainty in regional water resources planning– Build on the current (deterministic) water planning

framework

12

Methodology

13

Uncertainty Characterization

• Uncertainty in demand– Population projections– Water usage rates

• Uncertainty in supplies– Water supplies in future droughts – Climate-change impacts

• Create multiple demand and supply scenarios

14

Demand Scenarios

• 6 Pop. Proj. x 3 usage rate = 18 demand scenarios

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

500,000 1,000,000 1,500,000 2,000,000 2,500,000

2050 Population

Rela

tive

Lik

elih

ood

0.12

0.14

0.16

0.18

0.2

0.22

0.24

Year

Hig

her

than

Ave

rage

Per

-Cap

ita

Usa

ge R

ate

(ac-

ft/Y

) High Usage Rate

Low Usage Rate

Avg. Usage Rate

0%

5%

10%

15%

20%

25%

30%

150,000 200,000 250,000 300,000 350,000 400,000 450,000

2050 Demand (ac-ft/yr)

Rela

tive

Lik

elih

ood

High UsageAvg. Usage

Low Usage

15

Supply Scenarios

• LCRA/SAWS Climate Change Study used as basis for modeling uncertainty in climate

• 2 GCMs x 2 Future Emission Scenarios + 1 Baseline = 5 Supply Scenarios

16

Supply Scenarios

17

Water Needs Scenarios

• Water Need = Water Demand – Water Supply

18

Water Needs Scenarios

19

Evaluating Projects

• Baseline strategy = Conservation and Reuse (C&R)– Meets (deterministic) projected water needs (10,000 AFY)– Only 22% reliable

20

Evaluating Projects

• 6 potential strategies to meet the deficit

21

Project Portfolios

22

Cost-Reliability Trade-Off

23

Sensitivity to Uncertainty

24

Sensitivity to Uncertainty

25

Summary

• Framework to plan under uncertainty based on existing planning framework

• Identified and characterized key uncertainties in demands and supplies

• Developed project portfolios to improve reliability of water plan

• Evaluated trade-offs in cost and reliability to rank and select project portfolios

•

Other Planning Studies…

Other Planning Studies…

28

Lessons Learned

• Demonstrate importance of considering uncertainty– ‘Baseline’ solution not reliable

• Having a well-defined deterministic planning framework is key

• Start simple – easier to communicate ideas to stakeholders– Sequentially add ‘layers’ of uncertainty– Scenarios keep things ‘real’– Sensitivity analysis shows importance of different uncertainties

• Enables more robust decision making•

29

?