Cumulative Risk Assessment Model to Analyze Increased Risk due to Impaired Barriers in Offshore Drilling Rigs

Transcription

1 OESI Advisory Committee Meeting Cumulative Risk Assessment Model to Analyze Increased Risk due to Impaired Barriers in Offshore Drilling Rigs Syeda Zohra Halim

2 A Holistic Look into Risk 2

3 Deviations can interact and sum up Unacceptable Risk Deviations plus a dominating unacceptable risk Deviations leading to unacceptable risk Normal situation where cumulative risk from deviations is understood and reduced Figure: Cumulative risk and tolerability [1] 3

4 Past Incidents [3,4 ] Macondo Disaster April, killed 4.9 million barrels of oil spilled Fine: $18.7 billion + others U.S. Chemical Safety and Hazard Investigation Board, Investigation Report: Refinery Explosion and Fire, 2007 Texas City Refinery Explosion March, killed, 180+ injured Fine: $87 million 4

5 Cumulative Risk Assessment: The Challenge Integration of human and organizational factors with technical and operational factors Complexity and size of system Dependencies of components and events Temporal aspects (Dynamic) Uncertainties of parameter estimation 5

6 Objective Develop a framework for merging all factors together and build a model based on this framework that will enable determination of cumulative risk Identify causes behind offshore incidents and understand the effect of impaired barriers Conduct literature review for developing required framework Find applicable tool for developing a model that can handle required criteria 6

7 Previous Work Accident Evolution and Barrier Analysis (AEB): Svensson (2001) Barrier Analysis(BA): Dianous and Fievez (2006) Management Oversight and Risk Tree (MORT): Johnson (1980) Events and Causal Factor Charting and Analysis (ECFCA):DOE (1999) Swiss Cheese Model : James Reason (1990, 1997) The Shortcut Risk Analysis Method (SCRAM): Davis et. Al (2011) Functional Resonance Accident Model (FRAM): Hollnagel (2004) PyraMAP: Bellamy et. Al (2008) Human Factors Analysis and Classification System (HFACS): FAA/NTIS (2000) System Theoretic Process Analysis (STPA): Leveson (2009) Semi-Quantitative Fault Tree Analysis (SQUAFTA): Hauptmanns (2004) Barrier and Operational Risk Analysis (BORA): Aven et.al (2006) Bowtie: ICI (1979) Acci-Map: Rasmussen et. al (2000) Tripod BetaGroeneweg (2008) I-Risk: Bellamy et. Al (2003) and many more 7

8 Previous Work (2) APPROACH ALL FACTORS QUANTI- TATIVE DEPEN- DENCY DYNAMIC OTHERS Swiss Cheese Model: James Reason MORT: Johnson STAMP/ STPA: Nancy Leveson FT becomes too large No software to identify all the loops, depends on expertise and control structure diagram FRAM: Erik Hollnagel Requires iteration, difficult and time consuming BORA, Risk-OMT: Aven et. al. Scoring depends on RIFs and hence on the expert team 8

9 Objective Develop a framework for merging all factors together and build a model based on this framework that will enable determination of cumulative risk Identify causes behind offshore incidents and understand the effect of impaired barriers Conduct literature review for developing required framework Find applicable tool for developing a model that can handle required criteria 9

10 Available Tools Event-Cause Trees (Includes Fault Tree, Event Tree, Bow-Tie) Markov Chains Bayesian Network Petri Nets 10

11 Bayesian Networks It is a directed graph consisting of a set of nodes and arcs. Handles dynamic systems, distributions and dependencies and allows probability updating Based on Bayes Theorem: P(p i e) = P(e p i )P(p i ) k å i=1 P(e p i )P(p i ) 11

12 Petri Net Basics Petri Nets is a directed bipartite graph where system is analyzed by movement of tokens from one place to another via firing of transitions. 12

13 Case Study T s V 1 V1 V1 C1 C2 Consequence C 1 C1 Open Delayed Rupture Open Close Success Open Open Early Rupture Close Close Delayed Rupture Open Delayed Rupture Open C2 C 2 Close Success Close Open Early Rupture Close Close Rupture 13

14 Case Study: Event Tree 14

15 Case Study: Bayesian Network 15 Bobbio et al. (2003) Bearfield et al. (2005) Weber et al. (2012) Khakzad et al. (2013)

16 Case Study: Petri Nets 16 Liu et al. (1997) Labeau et al. (2000) Nyvlt et al. (2012) Pasman (2015)

17 Results ET BN PN Successful Mitigation 36.90% 36.90% 36.89% Early Rupture 5.90% 5.90% 5.89% Rupture 3.69% 3.69% 3.67% Delayed Rupture 53.51% 53.51% 53.55% 17

18 Conclusion An integrated model may be required to analyze cumulative risk A framework is to be used to merge technical, operational, human and organizational factors together BN and PN achieve their common goal through different approaches. Selection should depend on intended application Further work on development of the model is underway 18

19 Acknowledgements Dr. M. Sam Mannan Dr. Hans Pasman Dr. Yogesh Koirala Jim Pettigrew All members of OESI Advisory Committee All members of MKOPSC 19

20 References 1. Halim S. Z., Koirala, Y., Mannan, S. M., Probabilistic Methods of Quantitative Risk Analysis: A Case Study with Bayesian Network and Petri Net Approaches, MKOPSC International Symposium 2016, College Station, Alsyouf, I., "The role of maintenance in improving companies productivity and profitability." International Journal of Production Economics (2007): Neill, M., Improving Risk-based Decision Making by Connecting PSM Systems to Day-to-Day Plant Operations, Global Congress on Process Safety, , AIChE Spring Meeting, (2015). 4. U.S. Chemical Safety and Hazard Investigation Board, Investigation Report: Refinery Explosion and Fire, Accessed 10/21/ U.S. Chemical Safety and Hazard Investigation Board, Final Report: Macondo Blowout and Explosion, Accessed 10/21/ Weber, P., Medina-Oliva, G., Simon, C., and Lung, B., "Overview on Bayesian networks applications for dependability, risk analysis and maintenance areas." Engineering Applications of Artificial Intelligence 25.4 (2012): Labeau, P.E., Smidts, C., and Swaminathan, S., "Dynamic reliability: towards an integrated platform for probabilistic risk assessment." Reliability Engineering &System Safety 68.3 (2000): Blacklaw, A., Ward, A., and Cassidy, K., The Cumulative Risk Assessment Barrier Model, SPE , (2011) 9. Dugan, J.C., Bavuso, S.J., and Boyd, M.A., "Dynamic fault-tree models for fault-tolerant computer systems." IEEE Transactions on Reliability 41.3 (1992): Khakzad, N., Khan, F., and Amyotte, P., "Risk-based design of process systems using discrete-time Bayesian networks." Reliability Engineering & System Safety 109 (2013): Hosseini, S.M.H., and Takahashi, M., "Combining static/dynamic fault trees and event trees using Bayesian networks." International Conference on Computer Safety, Reliability, and Security. Springer Berlin Heidelberg, (2007). 20

21 References (Contd.) 12. Nývlt, O., and Rausand, M., "Dependencies in event trees analyzed by Petri nets." Reliability Engineering & System Safety 104 (2012): Kulkarni, V.G., Modeling and Analysis of Stochastic Systems. CRC Press, (1996). 14. Siu, N., "Risk assessment for dynamic systems: an overview." Reliability Engineering & System Safety 43.1 (1994): Bobbio, A., Ciancamerla, E., and Franceschinis, G., "Sequential application of heterogeneous models for the safety analysis of a control system: a case study." Reliability Engineering & System Safety 81.3 (2003): Hu, J., Zhang, L., Cai, Z., Wang, Y., and Wang, A., "Fault propagation behavior study and root cause reasoning with dynamic Bayesian network based framework." Process Safety and Environmental Protection 97 (2015): Bobbio, A., Portinale, L., Minichino, M., and Ciancamerla, E., "Improving the analysis of dependable systems by mapping fault trees into Bayesian networks." Reliability Engineering & System Safety 71.3 (2001): Boudali, H., and Dugan, J.B., "A discrete-time Bayesian network reliability modeling and analysis framework." Reliability Engineering & System Safety 87.3 (2005): Bearfield, G., and Marsh, W., "Generalising event trees using Bayesian networks with a case study of train derailment." International Conference on Computer Safety, Reliability, and Security. Springer Berlin Heidelberg, (2005). 20. Khakzad, N., Khan, F., and Amyotte, P., "Dynamic safety analysis of process systems by mapping bow-tie into Bayesian network." Process Safety and Environmental Protection 91.1 (2013): Liu, T.S., and Chiou, S.B., "The application of Petri nets to failure analysis." Reliability Engineering & System Safety 57.2 (1997): Nývlt, O., Ferkl, L., and Haugen, S., "Stochastic Coloured Petri Nets as a modelling language for complex Event Trees." Nutritional Care of the Patient with Gastrointestinal Disease (2015): Pasman, H., Risk Analysis and Control for Industrial Processes - Gas, Oil and Chemicals: A System Perspective for Assessing and Avoiding Low-Probability, High-Consequence Events, Butterworth-Heinemann, (2015) 21

22 Questions? Comments? If history repeats itself, and the unexpected always happens, how incapable must Man be of learning from experience. George Bernard Shaw 22

23 A Simple Petri Net Example P1: Operating (ON) µ T Repair Reachability Graph: T Failure P2: Failed (OFF) 2,0 1,1 0,2 Pr(1) Pr(2) Pr(3) µ µ 23

24 A New Approach Use of two colored tokens to represent probability of failure and success Straight forward and easy approach to obtain system s failure probabilities Reduced simulation time compared to previous approaches 24