Asst. Prof. of Geotechnical, Earthquake, and Risk Engineering

Transcription

1 RISK ANALYSIS OF THE CALIFORNIA BAY DELTA LEVEE SYSTEM Robb Eric S. Moss, Ph.D., P.E. Asst. Prof. of Geotechnical, Earthquake, and Risk Engineering

2 Interstate and State Highways Rail Corridors and Infrastructure Deep Water Port Facilities High Dollar Agricultural Land Residential and Municipal Land Power Transmission and Storage Aquaducts and Canals (>20 million users) ~2000 km of levees after DWR 1993

3 Risk=(probability of failure)(consequences) Component Failure single levee failure System Failure failure of weak link results in overall failure Consequences property loss, crop loss, infrastructure and building loss, and life loss.

4 Loading Function: Ground Shaking High Water Static Failure Mechanism: Bearing Sliding Slump/ Spread Seepage Erosion Over- Topping from Moss and Eller 2007?Mechanisms are often compounded and cascading in nature??causative Loading mechanism can be difficult to pinpoint?

5 Bearing Failure => weak foundation + static or seismic load Sliding => weak foundation + static or seismic load Slump/Spread => weak foundation or weak levee + seismic load Seepage => porous levee or foundation + high water Erosion => weak levee material + wave or river action Overtop => insufficient levee elevation + high water Bearing Sliding Slump/Spread Erosion Seepage Over-topping from Moss and Eller 2007

6 Novel Research Concepts being Implemented 1) Using data-based semivariogram/autocorrelation functions which quantify the spatial variability to statistically define the levee reach for probability of failure analysis. 2) Application of extreme value distributions for evaluating most likely failure modes and failure locations. 3) A time-stepping degradation/aggradation of resistance and load variables to model creep, strength loss, deferred maintenance, sea level rise, and other... 4) Use of reverse risk modeling, setting the consequences first and backing through all possible contributing failure modes. 5) Application of FORM and MC-based reliability in component and system analysis. 6) Method calibration/validation against previously documented failures case histories.

7 Resistance parameters as random variables: 1. Weak Foundation -sandy soil with high k -weak clay or organics with low su/p and low qc -liquefiable material with low qc or N and low PI -seepage paths (integrity/maintenance problems) 2. Weak Levee -sandy soil with high k -soil with low φ and c -clayey soil with low qc -rodent burrows -mobile or unarmored surface soil 3. Flawed Design -low crest elevation -penetrations, connections, intersections, ect freq All measured resistance parameters can be modeled as lognormally distributed random variables. Erosion and liquefaction failures are related as density and cohesion resist both. su/p, qc, k, c&φ, integrity/maintenance

8 Loading parameters as random variables: 1. Static total and effective stress 2. Seismic PGA or some other intensity measure 3. Water Level freeboard height (H) Subsidence, earthquake, flood, climate change, sea level rise, land use change are all considered. Elevation is the key variable in all these analyses. freq σ, σ, PGA, H

9 Extreme Values Lower bound resistance values tend to be the culprit in stress/strain failures. These need to be modeled as extreme values: Type III Smallest (Weibull) sample population (n) located at 1 or 2 sigma from a running/bracketed mean from Seed et al th St. Canal Failure (NOLA)

10 Spatial Variability and Measurement Uncertainty General Relative Variogram (Issaks and Srivastava 1989) has much utility and is compatible with the meaning and definition of point estimated coefficient of variation. Relative Semivariance γ ( x) 2 2 δ Squared COV h (m) Distance Geomorphically constrained variograms with explicit measurement of spatial variability and measurement uncertainty. The levee reach distance can then be statistically defined and cross-correlation of failure modes can be taken into account.

11 Create GIS (Google Earth) map of documented failures Define consequences geographically Generate geomorphically constrained variograms from DWR data Overlay geomorphic boundaries with vulnerable infrastructure and parse delta accordingly Evaluate loading and resistance per reach and per mode Carry out FORM and MC-based reliability to rank the different failure modes per reach Compounded or cascading failure modes considered in a second-phase correlated FORM analysis Time-stepped degraded/aggraded resistance distributions, time based loading distributions, seasonal or quarterly peak water levels, etc. Static loading results calibrated/validated against documented failures

12 Research Goals Provide the most rational and comprehensive basis for a risk analysis as possible utilizing novel higher order probabilistic tools. Identify the most critical levees and failure sequences based on vulnerability and consequences. Identify non-essential levees that can be considered functionally sacrificial. Quantify consequences in terms of life loss and/or dollar figures to facilitate communication with planners, policy makers, decision makers, and other stake holders. Couch the geographic representation of the data/results in an open source format to facilitate collaboration and for simplified updating. Provide a independent assessment of risk in the Bay Delta to compare with the state DRMS project. Provide guidance on targeted mitigation measures to reduce risk.

13 References DWR (1993). Sacramento San Joaquin Delta Atlas. Department of Water Resources, Sacramento. Issaks, E.H. and Srivastava, R.M. (1998). An Introduction to Applied Geostatistics. Oxford University Press. Moss, R. E. S., and Eller, J. M. (2007). "Estimating the Probability of Failure and Associated Risk of the California Bay Delta Levee System." GeoDenver, ASCE conf. proc.. Seed, R. B., Bea, R. G., Athanasopoulos, A. G., Boutwell, G. P., Bray, J. D., Cheung, C., Cobos-Roa, D., Harder, L. F., Moss, R. E. S., Pestana, J. M., Riemer, M. F., Rogers, J. D., Storesund, R., Vera-Grunauer, X., and Wartman, J. (2008). "New Orleans and Hurricane Katrina. III: The 17th Street Drainage Canal." Journal of Geotechnical and Geoenvironmental Engineering, 134(5),