5/5/2017

1

Jeff Lam, M.Eng., P.Eng.

Rocscience, Toronto, Ontario, Canada

ProbabilisticSlopeStabilityAnalysisofanEarth‐filledEmbankmentDam– CaseStudy

Geo3T2, Cary, North Carolina

April 12, 2017

Motivation

Background – Summary of Analysis Techniques

Case Study – Methodology

Case Study – Analysis Results

Outline

5/5/2017

2

•Determine the feasibility of implementing probabilistic based design with data from existing projects

•Validate results from probabilistic based design with observed performance

Motivation

Background

Deterministic Probabilistic

Traditional Standard Based

Non‐traditional probability based

5/5/2017

3

Background

Deterministic Probabilistic

Traditional Standard Based

Non‐traditional probability based

Background

Deterministic Probabilistic

5/5/2017

4

Background

Deterministic Probabilistic

Background

Deterministic Probabilistic

Traditional Standard Based

Non‐traditional probability based

5/5/2017

5

Background

Deterministic Probabilistic

Traditional Standard Based

Non‐traditional probability based

Background

Deterministic Probabilistic

Traditional Standard Based

1.2

Non‐traditional probability based

2.3

5/5/2017

6

Background

Deterministic Probabilistic

Uncertainty handled discretely through the use of “most likely scenarios”, “worst case scenarios”, etc.

Most Likely Case FoS = 1.5

Worst Case FoS = 1.2

The combined effects of various scenarios and their probability of occurrence considered holistically.

Pf = 0.5%

Background

Deterministic Probabilistic

Analytical

α

Iteration• Limit Equilibrium 

Approximation• Finite element analysis• Tables and charts

Analytical

ϕ ϕ

Simulation• Monte Carlo simulation

Approximation• Tables and charts• Uncertainty propagation• Other reliability methods

5/5/2017

7

Background

Deterministic Probabilistic

Analytical

α

Iteration• Limit Equilibrium 

Approximation• Finite element analysis• Tables and charts

Analytical

ϕ ϕ

Simulation• Monte Carlo simulation

Approximation• Tables and charts• Uncertainty propagation• Other reliability methods

Background

Deterministic Probabilistic

Analytical

α

Iteration• Limit Equilibrium 

Approximation• Tables and charts• Finite element analysis

Analytical

ϕ ϕ

Simulation• Monte Carlo simulation

Approximation• Tables and charts• Uncertainty propagation• Other reliability methods

5/5/2017

8

Background

Background

Deterministic Probabilistic

Analytical

α

Iteration• Limit Equilibrium 

Approximation• Tables and charts• Finite element analysis

Analytical

ϕ ϕ

Simulation• Monte Carlo simulation

Approximation• Tables and charts• Uncertainty propagation• Other reliability methods

5/5/2017

9

Background

Deterministic Probabilistic

Analytical

α

Iteration• Limit Equilibrium 

Approximation• Finite element analysis• Tables and charts

Analytical

ϕ ϕ

Simulation• Monte Carlo simulation

Approximation• Tables and charts• Uncertainty propagation• Other reliability methods

μ= 1.00σ= 0.10Pf = 0.48

Case Study

5/5/2017

10

Methodology

Methodology

5/5/2017

11

Methodology

Methodology

5/5/2017

12

Analytical Model:Morgenstern‐Price 2D Limit Equilibrium

Reliability Analysis Method:Monte Carlo Simulation

Defining the Failure: Factor of Safety <1

Defining Material Properties: Literature review (Phoon & Kulhawy 1999), tailings data from similar mines, calculated some based on lab test data.

Analysis Tool: Slide by Rocscience

Methodology

Defining Material Properties: 

Methodology

5/5/2017

13

Defining Material Properties: 

Methodology

Model Calibration: 

Methodology

5/5/2017

14

Model Calibration: 

Methodology

μ= 1.00σ= 0.10Pf = 0.48

Model Calibration: 

Methodology

5/5/2017

15

Defining Geological Model: 

Methodology

10’

650’

Defining Geological Model: 

Methodology

5/5/2017

16

Defining Geological Model: 

Methodology GW96-1A

5/5/2017

17

10’ thick - 650' diameter

Defining Geological Model: 

Methodology

5/5/2017

18

Methodology

Defining Geological Model: 

Methodology

10’ – 20’

10’

650’

5/5/2017

19

Defining Geological Model: 

Methodology

20’ – 30’

10’

650’

Defining Geological Model: 

Methodology

30’ – 40’

10’

650’

5/5/2017

20

Results

•Risk = Probability x Consequence

•Discount cash flow valuations when valuing projects for business cases

•Justifying site investigation costs – “You pay for a site investigation program whether you have one or not.” So how many boreholes do we need? 

Conclusion

5/5/2017

21

Questions?

Results