1 USACE DAM RISK ASSESSMENT PROGRAM OVERVIEW AND HOW ENGINEERING GEOLOGY CONTRIBUTES TO THE LEVEL OF CONFIDENCE AND RESULTS 255 255 255 237 237 237 217 217 217 200 200 200 0 0 0 163 163 163 131 132 122 Todd N. Loar, PG, CEG Senior Geological Engineer Risk Management Center Lakewood, CO 14 September 2017 Association of Engineering Geologists Colorado Springs, CO September, 2017 File Name 80 119 27 239 65 53 Hurricane Katrina, 2005 110 inundation map 135 120 252 174.59 112 92 56 62 102 130 102 56 48 130 120 111
USACE INVENTORY OVERVIEW Dams: 710 structures (at 556 different flood management projects) Levees: 15,000 mi of federally authorized levees (USACE and other stakeholders) Navigation Systems
USACE INVENTORY OVERVIEW
RISK MANAGEMENT CENTER (RMC) RMC Offices IWR Offices RA Team: Part of USACE HQ (Institute for Water Resources, IWR). Lead risk management for centrally-located dam/levee safety program: Program management Risk analysis and assessments Review and support Establish project prioritization, plan studies, and modification activities Data management / compilation Develop technical competency (training) Risk Cadre District Team Multi-disciplinary Team: Civil, geotech, geologic, H&H, structural, mechanical, economists, environmental
RISK MANAGEMENT CENTER (RMC)
We live in a world filled with risk and its in our nature to continually evaluate our actions, environment & decisions relative to the potential consequences and/or benefits.
RISK AND RISK ASSESSMENTS Probability of Failure Incremental Risk= Probability Consequences of the Loading Given the Loading Given Failure X X Failure Likelihood (loading & system response) Includes statistical & subjective probability Consequence Level The risk assessment process attempts to answer the following questions: Consequences include: life, economic, and environment and other non-monetary impacts. Life safety is paramount.
TOLERABLE RISK GUIDELINES (TRG S) Risks that society is willing to live with so as to secure certain benefits. Risks society does not regard as negligible. Decisions are risk INFORMED: Emphasis is on the decision, NOT the number or the model Numbers are NOT decisions Risks that society is confident are being properly managed by the owner. Risks that the owner keeps under review and reduces still further as practicable.
LEVELS OF A RISK ASSESSMENT 1. Periodic Assessments (PA) and Semi- Quantitative Risk Assessment (SQRA) 2. Issue Evaluation Studies (IES) 3. Dam Safety Modification Study (DSMS) 4. Post-Implementation Evaluation (PIE) Each Level involves an increasing amount of data compilation, review, evaluation, site characterization, and engineering analysis
Geologic Contributions to Risk Assessment - Loading Seismic Loading: Probability Probability of Failure Consequences of the Loading Given the Loading Given Failure Deterministic & probabilistic earthquake loadings for dam stability analyses Potential for and geometry of surface fault rupture % Contribution to Hazard Incremental Risk= Crustal Faults Deep Intraslab Source Cascadia Megathrust 9,950 yr
Geologic Contributions to Risk Assessment - Loading Incremental Risk= Hydrologic Loading: Probability Probability of Failure Consequences of the Loading Given the Loading Given Failure Constrain hydro-loading using long-term flood records: Geomorphic indicators and investigations of paleoflood stage Geomorphic indicators of stage non-exceedance (upper bound) Age-dating to define frequency of extreme loading House et al. (2002) AGU Paleoflood Paleo-flood point
Geologic Contributions to Risk Assessments Failure / Consequences Probability of Failure Incremental Risk= Probability Consequences of the Loading Given the Loading Given Failure Consequence analysis for various loading, breach & non-breach scenarios are developed by the MMC This is where engineering geology contributes the most to the RA Process
DAM FAILURES IN THE 20TH CENTURY Fontenelle St. Francis Bayless Quail Creek ~1/3 of all dams failed by overtopping ~1/3 failed by seepage and piping through the dam or into the foundation ~1/3 failed due to foundation and miscellaneous causes ~55% of all concrete dam failures attributed to foundation issues (ICOLD, 1995) Approx. 70% of failures can be attributed to geologic and/or geotechnical issues
Example Geological PFM: Description Loading -> Flaw exists > Initiation > Continuation > Progression > Intervention > Breach
Example Geological PFM: Description Loading -> Flaw exists > Initiation > Continuation > Progression > Intervention > Breach
Example Geological PFM: Event Tree Breach Unsuccessful Intervention Progression: Clogging 6 Continuation Initiation Flaw: Ineffective Treatment Flaw: Open Rock Discontinuities 1) Reservoir Loading 4 3 Probability that the surface treatment / 2 grouting in the core Probability that rock trench fails to cut off the seepage path defects exist & are 1 open/continuous enough to transport embankment materials Probability that gradients and velocity are sufficient to initiate scour? Probability an unfiltered exit exists downstream? Probability that intervention is unsuccessful? Probability that the embankment materials are unable 5 to self-heal or clog the failure paths and a stope forms? Geologic / Geotech / Hydrogeologic Uncertainties: Open/continuous network of discontinuities downstream of core/curtain within the SS/LS unit allowing embankment material to migrate/erode Evaluate foundation treatment and effectiveness. Obtain information to understand the gradients across the core trench and between the core material and downstream rock Potential Targets for Field Studies (mapping, drilling, testing, instrumentation) Probability the stope enlarges to collapse 7 the crest of the dam leading to breach by overtopping?
Example Geological PFM: Characterizing the Uncertainty at Each PFM Node MORE LIKELY / LESS LIKELY FACTORS Geomechanical mapping, drilling angled borings, packer testing, down-hole televiewer survey, additional piezometers might improve our confidence and reduce uncertainty with this node of the PFM High confidence we are unlikely to revise our estimate with more information. Moderate confidence we are unsure about the potential to change the estimate with more information Low confidence likely to revise estimate with more information.
SUCCESSFUL ENG. GEO. IN RIDM Has experience / background in dam & levee design and construction Basic understanding of geotechnical, civil, hydrologic, and structural engineering disciplines Knowledge of different PFM s and how geologic conditions influence the PMF mechanics. Understands probability estimates/analysis for various nodes in the event tree. Familiar with past precedents, incidents, and failure case histories. Ability to sort, query, analyze, compile, and portray different types of geologic/geotechnical spatial data and create plots, figures and drawings that communicate the spatial geological conditions related to the PFM s. Must know what data is IMPORTANT and NECESSARY to evaluate foundations and solve engineering problems related to the PFM.
SUCCESSFUL ENG. GEO. IN RIDM Reviews, understands, summarizes, & prioritizes existing data prior to the RA Understands what the instrumentation data is telling the RA team in the context of the geology and PFM s (data tabulation is not interpretation) Thinks about PFM s and uses all geologic tools to most accurately reflect the subsurface and make realistic estimate of depositional environments, spatial extents and infer the most realistic behavior given the PFM Creates detailed plans and section drawings to display, summarize, and communicate material properties and instrumentation data Has genuine enthusiasm to solve specific technical geologic problems (be a problem-solving detective).
LESSONS LEARNED PFMA is starting point for RA Training of facilitators, technical team members, and participants is essential Engineering geologists are fully capable to lead / facilitate risk assessment teams Select realistic engineering properties and ranges (avoid conservatism in RA) PFM must be well developed and presented to characterize the failure modes and answer specific questions associated with the event tree nodes There is no clear line between geotechnical engineering and engineering geology Archived data or previous findings must be identified early as it can significantly alter the risk estimates Reports and geologic evaluation must be focused on the PFM and point to specific evidence supporting & justifying the conclusions and interpretations
Life Safety is Paramount Protecting People, Not Infrastructure