RAC. Southeast Louisiana Flood Protection Authority - East Baseline Risk Assessment Study Results. July 20, Engineers & Economists, LLC

Transcription

1 RAC Southeast Louisiana Flood Protection Authority - East Baseline Risk Assessment Study Results July 20, 2017

2 Probabilistic Performance Analysis for all Major Structures in the HSDRRS and Redevelopment of Operations and Maintenance Manuals and Emergency Action Plans RAC Project Background/History Review of Overall Project Objectives Phase 1 and Phase 2 Risk Study Framework HAZOP Block Diagrams for Each Structure FTA FMEA Risk-Informed Decision Support Tool Results of Analysis Risk Reduction Measures Future Applicability

3 RAC Project Background/History How the Study was Formulated HUD CDBG-DR Funded Study Components - Nine Complex Closure Structures Goals of the Study Further Reduce Risk Through: Additional Spare Parts On-Hand Additional Operation and Maintenance Procedures Needed Emergency Contracts Written Procedures Maintenance Activities Proposed Repairs 3

4

5 RAC Project Objectives Manage risks associated with closure of nine gates during a hurricane or tropical storm warning (HTSW). Identify the main risk contributors Identify an initial set of potential risk reduction measures Provide a foundation for developing a risk-informed maintenance program Risk-Informed Decision Support Tool It should be noted that this study was not a system-wide risk assessment or a condition assessment. It addressed only the performance of the nine complex gate structures and did not include the performance of the levees, floodwalls or other HSDRRS features. The study included the normal components of the risk assessment process, including risk identification, risk estimation, risk analysis, risk evaluation and risk assessment, as described in the final report. The study did not identify any immediate cause for concern for public safety. 5

6 RAC Risk Management Definition Phase 2 Phase 1 6

7 Phase 1 Phase 2 RAC Reviews & Inspections Failure Modes & Effects Analysis (FMEA) Gate System Components Block Diagrams Initial Failure Modes Ranked Failure Modes HAZOP Study Workshop Potential Risk Reductions Fault Tree Analysis (FTA): Per Demand Fail to Close Probs. Initial Fault Trees Initial Cut Sets Event Probabilities Existing Fail to Close Probs. Risk Reduction Fail to Close Probs. Project Outcomes System Consequences & Ann. Probs. Existing System Conseqs. & Ann. Probs. Risk Reduction System Conseqs. & Ann. Probs. Abbreviations: Ann. Annual Conseqs. Consequences HAZOP Hazards and Operability Probs. Probabilities Risk Reduction Recommendations Spare Parts Inventory Risk Reduction Measures Workshop Recommended Risk Reductions

8 RAC Risk Framework - HAZOP Overview The HAZOP (Hazard and Operability Study) Workshop took the participants through an analysis of risks for each of the nine gated risk structures in the HSDRRS. Potential risks were categorized into four separate risk areas that were further broken down into elements and components. Risk Areas Four RISK Focus Areas Personnel, Structure, Operational Support (Power and Communications), Procedures Elements Twelve ELEMENTS - Staff, Contractors, Controls, Hydraulics, Mechanical Systems, Structural, Power, Communications, SOPs (Operating, Maintenance, Communication, Water Control) Components Sixty-one COMPONENTS Communications: Power/Primary/Secondary/Tertiary/Generator 9 Flood Control Systems 549 COMBINATIONS

9 Risk Framework - Block Diagram Example BAYOU DUPRE: Review and Inspections Identify Gate Systems Components RAC 9

10 Manual Generator 1 Generator 2 Diesel Tank Transfer Switch Power Supply Subpanel Board 1 Electric Motor 1 Manual Transfer Switch Leaf 1 Leaf 2 Panel Board Manual/Auto Transfer Switch Main HPU Portable Generator* Subpanel Board 2 Electric Motor 2 Diesel Same as Leaf 1 Gate Group A Sector (Gear): Bayou Dupre Risk Framework - Functional Block Diagram * Kept at yard and shared with Caernarvon Hagglund Gear Motor Portable HPU Gasoline Debris or Silt Rack and Pinion Gate Winch (No procedure) Impact Damage Legend Normal Operation Backup Hazard 10

11 RAC Risk Framework - Fault Tree Analysis Overview What combination of events will cause another (top) event to occur (e.g., gates do not fully close)? Also provides an estimate of the per demand probability that the top event will occur Utilize deductive logic, where the causative conditions and factors are deductively identified, organized in a logical manner. Logic represented in a fault tree diagram FERC Risk Training Course Phase 1 Logan, Utah

12 Fault Tree Diagrams SERIES System: OR-gate (BUC) Failure of system A occurs if one or both of the input events B and C occur PARALLEL System: AND-gate (B C) Failure of system A occurs if both of the input events B and C occur RAC System A Input Event B Occurs System A Event C Occurs Output Input Event B Occurs Event C Occurs Output System A Fails Top Event System A Fails OR-gate AND-gate Event B Occurs Event C Occurs Fault Events Event B Occurs Event C Occurs

13 RAC Fault Tree Diagram Illustration Top Event Emergency generator fails to start OR Gate AND Gate Or-gate Fault Events Fault in start-up signal for diesel generator Fault in diesel generator Or-gate Or-gate Fault Events Fault in transmission of signal Fault in sending the signal Fault in reception of of signal Fuel supply problem Mechanical fault in diesel generator Or-gate 2 Transfer Event Page 3 Blocked fuel intake 3 No fuel Basic Event

14 RAC OR Gate Fault Tree Illustration: First Order Cut Sets AND Gate Top Event True Or-gate Emergency generator fails to start True Fault Events Fault in start-up signal for diesel generator Or-gate True Or-gate Fault in diesel generator True Fault Events Fault in transmission of signal Fault in sending the signal Fault in reception of of signal Fuel supply problem Mechanical fault in diesel generator Cut set - the event(s) that must occur simultaneously for a system to fail Or-gate Blocked fuel intake True No fuel 2 Transfer Event Page 3 3 Basic Event

15 Bayou Dupre Sector Gate Fault Tree - Index to Fault Tree Page Numbers RAC 15

16 Risk Framework - Failure Modes and Effects Analysis (FMEA) Overview RAC A systematic procedure for the analysis of a system to identify the potential failure modes, their causes and effects on system performance The system is represented as a functional block diagram where performance of its elements can be defined The analysis proceeds in a bottom-up fashion until the end effect on the system is identified A thorough FMEA is a result of team work that stimulates the thought process, and ensures necessary expertise FMEA generally deals with individual failure modes and the effect of these failure modes on the system. Each failure mode is treated as independent. 16

17 RAC FMEA Example Bayou Dupre PFM # Failure Progression Resulting Impacts BD1* - BD125 BD126 BD174 BD174 BD177 The Rack and Pinion fails. The gate is then, unable to be closed manually via Uncontrolled flooding of the Central Wetlands and possible failure of the Adverse Factors (more likely) Operators or equipment fail to reach the site in time to close the gates during Uncontrolled flooding of the Central Wetlands and possible failure of the From an event. Initial FTA secondary protection system. The Main Hydraulic Power Unit (HPU) fails due to either; failure of shared Uncontrolled flooding of the Central Wetlands and possible failure of the Cut Set Risk Elements to electric cables, panel board failure, or failure of electric motor. In addition, the secondary protection system. Additional damage to the gates due to Verbal Description of First Existing Discussion Risk Reduction of Portable HPU fails due to; bad fuel, mechanical issues, blown hose, operator vibration/surging Possibly Mitigate injury, lack of knowledge, or the Portable The HPU Failure is unavailable. The gate Mode is Consequences Measures then, unable to be closed manually via tugboat due to; lack of knowledge, lack Based on Block Diagram of written procedure, the tug boat breaks down or is unavailable. Outside the HAZOP Favorable Factors (less likely) Fault Tree Analysis Order A1, A2, A3 F1, F2 Third A3, A5, A8 F4, F12, F13 Third A3, A4 F5, F6 Second tugboat due to; lack of knowledge, lack of written procedure, the tug boat secondary protection system. Additional damage to the gates due to breaks down or is unavailable. vibration/surging BD178 BD181 A hydraulic hose on one of the Hagglund Gear Motor bursts. The gate is then, Uncontrolled flooding of the Central Wetlands and possible failure of the A3, A4, A5 F4, F5 Second unable to be closed manually via tugboat due to; lack of knowledge, lack of secondary protection system. Additional damage to the gates due to written procedure, tug boat breaks down or is unavailable. vibration/surging BD182 Debris accumulates on the sill that keep the gates from completely Uncontrolled flooding of the Central Wetlands and possible failure of the A3, A6, A7 F6, F7, F11 First closing/sealing during a storm surge or extratropical storm event. The gate is secondary protection system. Additional damage to the gates due to BD183 then, unable to be closed manually via tugboat due to; lack of knowledge, lack of written procedure, the tug boat breaks down or is unavailable. The gate is severely damaged prior to the event by vessel or other impact and is unable to close or seat properly. The gate is then, unable to be closed manually via tugboat due to; lack of knowledge, lack of written procedure, the tug boat breaks down or is unavailable. vibration/surging. F6 Regular Exercising, Operation and Maintenance of Gates Uncontrolled flooding of Central Wetlands and possible failure of the secondary protection system. Additional damage to the gates due to vibration/surging. A3, A7 F8, F9, F10 First 17

18 RAC Basic Event Probability Estimation Initial Probability Estimates Expert elicitation Probability mapping table Snap Shot in Time Resources: Project information Off Shore (OREDA) USACE: Age and frequency of use-testing for flood control projects Industry failure rate data for mechanical hardware, electrical equipment Human reliability available for individual actions considered critical actions Life Safety and Operator Injuries are not accounted for Event Reliability (per demand) Descriptor in 10 Critical in 100 Low-Poor in 1,000 Moderate in 10,000 Moderate-High in 100,000 High in 1,000,000 Very High 18

19 RAC Repair/Recovery Time Estimated for each event as: Zero hours for Quick Fix = Diagnosis plus repair is less than 1 hour Hours for diagnosis, decision and resolution No for times exceeding time available during gate closing assigned a nominal value of 1,000 hours Estimates considered: Diagnosis of problem Decision to repair Resources and Available Spare Parts Getting parts to site Repair Feasibility during emergency within total 96 hr. window Procedure Available/in-place Assume very low nominal probability for quick fixes under current situation = 1 in 10,000 Sum Repair Times for each cut set and reduce gate unavailability to account for repairs Are repair times reasonable? Are spares or other resources (equipment, special skills) currently available or do procedures/contracts need to be developed? Are there opportunities for faster diagnosis and repair (based on importance) as risk reduction measures (RRMs) 19

20 Repair/Recovery Time Impacts on Failure to Close Probability RAC Gate Total Unavailability (per demand) Reduction in Unavailability Due to Repair (per demand) 1 in 1 in Percent Reduction Net Unavailability (per demand) Time (hrs) Bayou Bienvenue Sector 2.91E 04 3, E 04 6,897 50% 1.46E 04 6, Bayou Bienvenue VLG 2.60E E % 1.23E Bayou Dupre Sector 8.74E 04 1, E 04 4,082 28% 6.29E 04 1, Bayou St John 4.12E 04 2, E 04 6,246 39% 2.52E 04 3, Caernarvon Sector 8.74E 04 1, E 04 4,082 28% 6.29E 04 1, GIWW Barge Gate 3.32E E % 3.15E GIWW Sector Gate 3.18E E % 9.96E 04 1, Seabrook Sector Gate 1.93E E % 7.69E 04 1, Seabrook VLG #1 2.50E E % 7.33E 04 1, Seabrook VLG #2 2.50E E % 7.33E 04 1, All Gates 9.33E E % 3.76E in Expected Value of Cut Set Repair 20

21 Per Demand Failure to Close Probabilities - Unavailabilities RAC Gate System Unavailabilities Linear Scale 1.E 1 8.E 2 6.E 2 4.E 2 2.E 2 0.E+0 Total Unavailability (per demand) Net Unavailability (per demand) 21

22 Per Demand Failure to Close Probabilities - Unavailabilities RAC Gate System Unavailabilities Log Scale 1.E+0 1.E 1 1.E 2 1.E 3 1.E 4 Total Unavailability (per demand) Net Unavailability (per demand) 22

23 Phase 1 Phase 2 RAC Reviews & Inspections Failure Modes & Effects Analysis (FMEA) Gate System Components Block Diagrams Initial Failure Modes Ranked Failure Modes HAZOP Study Workshop Potential Risk Reductions Fault Tree Analysis (FTA): Per Demand Fail to Close Probs. Initial Fault Trees Initial Cut Sets Event Probabilities Existing Fail to Close Probs. Risk Reduction Fail to Close Probs. Project Outcomes System Consequences & Ann. Probs. Existing System Conseqs. & Ann. Probs. Risk Reduction System Conseqs. & Ann. Probs. Abbreviations: Ann. Annual Conseqs. Consequences HAZOP Hazards and Operability Probs. Probabilities Risk Reduction Recommendations Spare Parts Inventory Risk Reduction Measures Workshop Recommended Risk Reductions

24 System Consequences & Annual Probabilities RAC Objective: Quantify flooding and economic impacts due to failure of gate to close under various loading conditions (50- year, 100-year, 500-year events) Consider event duration to estimate inflows Estimated Values Frequency of surge levels in IHNC basin and on exterior gates Overtopping / breach assumptions of IHNC floodwalls Flood depths within polders Converting flood depths to damage estimates 24

25 RAC Event Tree - Bayou Dupre Sector 8 SB5 Surge SWL (ft NAVD88) 24 Bayou St. John Sector Gate 22 Seabrook Complex Structure GIWW East Closure Sector Gate 20 Bayou Bienvenue Vertical Lift Gate Bayou Dupre Sector Gate 18 Caervarvon Sector Gate Return Period (years) Adjusted Exterior Surge Hazards General curve shape taken from 2007 JPM Study; a log-linear adustment was added to make the 100- and 500-yr values equal to USACE's 2014 HSDRRS Hydraulic Design for 2007 Base Case; a 10-yr value was also added. Probability (Unavailability) from Fault Tree Polder Volume (acre ft) 350, , , , , , Gate Fails to Close 2 Gate Closes SB5 Economic Damages ($m) Innudation Elevation (ft. NAVD88) 50, Innudation Elevation (ft. NAVD88)

26 RAC Economic Consequences Average Annual Minimum Weighted Maximum Annual Probability Figure 12 Incremental Incremental Average Incremental (/yr) Gate Abbreviated Economic Economic Incremental Economic Gate Name Consequences Consequences Economic Consequences 1 in (AAEC) ($/yr) (Min EC) ($) Consequenecs ($) (Max EC) ($) Bayou Bienvenue Sector BB Sec 1.46E 05 68,617 $2 $66,666 $ 137,234 $215,472 Bayou Bienvenue VLG BB VLG 1.23E 04 8,157 $126 $288,166 $ 1,027,732 $1,256,240 Bayou Dupre Sector BD Sec 6.29E 05 15,898 $18 $205,924 $ 286,169 $307,069 Bayou St John Sector BSJ Sec 1.26E ,227 $2,233 $1,201,620,841 $ 1,771,666,667 $1,938,080,748 Caernarvon Sector C Sec 6.29E 05 15,898 $1,234 $3,859,879 $ 19,618,442 $51,863,347 GIWW Barge Gate GIWW Barge 3.15E $695,410 $1,536,002 $ 220,889,138 $808,387,498 GIWW Sector Gate GIWW Sec 9.96E 05 10,045 $21,991 $1,536,002 $ 220,889,138 $808,387,498 Seabrook Sector Gate SB Sec 7.69E 05 13,009 $25 $224,100 $ 330,950 $2,254,937 Seabrook VLG #1 SB VLG 7.33E 05 13,639 $19 $224,100 $ 259,138 $1,146,104 Seabrook VLG #2 SB VLG 7.33E 05 13,639 $19 $224,100 $ 259,138 $1,146,104 26

27 Annual Probability vs. Weighted Average Incremental Economic Consequences RAC 1.E 2 Probability GIWW Barge 1.E 3 Annual Probability SB VLG BB VLG 1.E 4 SB Sec SB VLG BD Sec C Sec Av Ann Econ Risk = Prob * Conseq GIWW Sec 1.E 5 BB Sec Consequences BSJ Sec 1.E , ,000.0 Weighted Average Incremental Economic Consequences ($million)

28 Phase 1 Phase 2 RAC Reviews & Inspections Failure Modes & Effects Analysis (FMEA) Gate System Components Block Diagrams Initial Failure Modes Ranked Failure Modes HAZOP Study Workshop Potential Risk Reductions Fault Tree Analysis (FTA): Per Demand Fail to Close Probs. Initial Fault Trees Initial Cut Sets Event Probabilities Existing Fail to Close Probs. Risk Reduction Fail to Close Probs. Project Outcomes System Consequences & Ann. Probs. Existing System Conseqs. & Ann. Probs. Risk Reduction System Conseqs. & Ann. Probs. Abbreviations: Ann. Annual Conseqs. Consequences HAZOP Hazards and Operability Probs. Probabilities Risk Reduction Recommendations Spare Parts Inventory Risk Reduction Measures Workshop Recommended Risk Reductions

29 RAC Risk Reduction Measures

30 RAC Risk Reduction Measures

31 Risk-Informed Decision Support Tool Realizing Value Into the Future RAC Potential Future Benefits Periodic Update and Assessment Modular Maintaining and Improving the Risk-Informed Decision Support Tool/Approach Maintenance Database Feedback System Performance Expand to System-Wide Application Updated Information Economics, Fragility, CLARA Model

32 RAC Questions? Robert Turner, Executive Director, Southeast Louisiana Flood Protection Authority East David Moore, PE (Tetra Tech)