Risk Modeling 101 AWS
Mason Matthews Athens, AL Gas Department NOVEMBER 14, 2017
Risk Modeling 101
Disclaimer
Question… What does “Safe” mean? “Freedom from unacceptable risk.” (ISO Guide 51:1999E) Are we “safe” here today? What is “Risk?” Severity x Probability How severe is your severity? What is the higher risk today….?
Sources
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/index.htm
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/index.htm
PHMSA Risk Modeling Work Group (RMWG) Participants TransCanada Kinder Morgan Columbia Gas B31.8S committee representative INGAA (2) NAPSR (2) AGA APGA
CenterPoint Energy Consumers Energy Idaho National Laboratory API Phillips 66 Marathon Colonial PHMSA (4)
PHMSA Risk Modeling Work Group (RMWG)
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
“Risk informed decision making” https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
Back to career plan #1….
Back to career plan #1….
Back to career plan #1….
Risk Modeling 101 Risk is defined operationally as a set of triplets: – The scenario(s) leading to degraded performance with respect to one or more performance measures. – The likelihood(s) (qualitative or quantitative) of those scenarios. – The consequences (qualitative or quantitative severity of the performance degradation) that would result if those scenarios were to occur. Source: NASA, NPR 8000.4A, Agency Risk Management Procedural Requirements, p. 22
Risk Modeling 101 Risk assessments, which are the very foundation of an integrity management program, can vary in scope or complexity and use different methods or techniques. The ultimate goal of assessing risks is to identify the most significant risks so that an operator can develop an effective and prioritized prevention/ detection/mitigation plan to address the risks. Source: ASME B318.S
Risk Modeling 101 B31.8S Risk Model Approaches Subject Matter Expert (SME) Relative Assessment Model Scenario Based Model
Event Trees, Decision Trees, Fault Trees
Probabilistic Model Source: ASME B318.S
Risk Modeling 101 Risk analysis can be fairly simple with values ranging from 1 to 3 (to reflect high, medium, and low likelihood and consequences)... Multiplying the relative likelihood and consequence numbers together provides the operator with a relative risk for the segment and a relative priority for its assessment. Source: ASME B318.S
Risk Modeling 101
1
2
3
3
3
6
9
2
2
4
6
1
1
2
3
Risk Modeling 101
1
2
3
3
3
6
9
2
2
4
6
1
1
2
3
Risk Modeling 101 Segment 15
Segment 15 currently has a low risk attributed to third party damage This segment is not located near a populated area (consequence = low) and there has been little to no excavation history in the area (likelihood/frequency = low)
Risk Modeling 101
Segment 4 currently has a medium risk attributed to third party damage This segment is located away from populated areas but in a cultivated field (consequence = medium) and the farmer has been plowing the area (likelihood/frequency = medium)
Segment 4
Risk Modeling 101
Segment 9 currently has the highest risk of third party damage in the system This segment is located near homes, businesses and a US Highway (consequence = high). This is an area of active construction with an increasing number of locate tickets (likelihood/frequency = high)
Segment 9
Risk Modeling 101 Segment 15
Annual Probability of Excavation Damage Segment 15 2.10E-04 Segment 4 1.50E-03 Segment 9 7.20E-02
Segment 9 is 342.9 times more likely to experience excavation damage than Segment 15 Segment 9 is 48 times more likely to experience excavation damage than Segment 4
Segment 9
Segment 4
Segment 4 is 7.1 times more likely to experience excavation damage than Segment 15
Risk Modeling 101
Tools for Quantitative Analysis Decision Trees Influence Diagrams Event Trees Fault Trees Monte Carlo Simulation
“Scenario-Based Models. This risk assessment approach creates models that generate a description of an event or series of events leading to a level of risk, and includes both the likelihood and consequences of such events. This method usually includes construction of event trees, decision trees, and fault trees. From these constructs, risk values are determined.” Source: ASME B318.S
Risk Modeling 101
Fault Tree Logic and Probability Example
Risk Modeling 101
Fault Tree Logic and Probability Example
Fault Tree Handbook with Aerospace Applications, NASA Office of Safety and Mission Assurance, Volume 1.1, August, 2002
Risk Modeling 101
Example of Fault Tree “Or” Gate The output event occurs only if one or more of the input events occur ADD the probabilities
.01 + .01 = .02 .01
.01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
.01
Single Fuel Cell .01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
Single Fuel Cell .01
.01
Redundant Fuel Cell .01 x .01 = .0001 .01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
.01
.01
Single Fuel Cell .01 Redundant Fuel Cell .01 x .01 = .0001
.01
Redundant Fuel Cell plus Battery .01 x .01 x .01 = .000001
Risk Modeling 101
Fault Tree Analysis
Risk Modeling 101
Fault Tree Analysis
Risk Modeling 101
Fault Tree Analysis
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Work Group Meeting Topics 1. RMWG Overview 2. Likelihood Modeling 3. Consequence Modeling (and likelihood) 4. Facilities Modeling 5. Data and Risk Tolerance 6. Index Model Applications and Evolving to more Quantitative Models
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Concerns for IOUs and public gas systems...
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Publicly owned transmission statistics Nearly 1,000 municipally owned gas systems in the U.S. 730 APGA members 62 municipal gas systems operate a total of about 2,700 miles of transmission More than half of total mileage is 8” or smaller
*** 40 of the 62 currently do not contain a High Consequence Area (HCA)
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom
Active vs. Passive Components
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Calibrating the model… Bounding uncertainties….
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.”
Is the process you choose as important as the output?
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
“As a risk person I am a generalist, not an expert-expert.”
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
“As a risk person I am a generalist, not an expert-expert.” “Risk modeling is a process not an event.” i.e., “if I can only get through this!”
PHMSA Risk Model Work Group (RMWG)
Conclusion?
13
Questions? Mason Matthews Athens, AL Gas Department 256-232-1440 [email protected]
Risk Modeling 101
Disclaimer
Question… What does “Safe” mean? “Freedom from unacceptable risk.” (ISO Guide 51:1999E) Are we “safe” here today? What is “Risk?” Severity x Probability How severe is your severity? What is the higher risk today….?
Sources
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/index.htm
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/index.htm
PHMSA Risk Modeling Work Group (RMWG) Participants TransCanada Kinder Morgan Columbia Gas B31.8S committee representative INGAA (2) NAPSR (2) AGA APGA
CenterPoint Energy Consumers Energy Idaho National Laboratory API Phillips 66 Marathon Colonial PHMSA (4)
PHMSA Risk Modeling Work Group (RMWG)
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
“Risk informed decision making” https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
PHMSA Risk Modeling Work Group (RMWG)
https://primis.phmsa.dot.gov/rmwg/docs/RMWG%20Template%20Presentation_10.07.2016.pptx
Back to career plan #1….
Back to career plan #1….
Back to career plan #1….
Risk Modeling 101 Risk is defined operationally as a set of triplets: – The scenario(s) leading to degraded performance with respect to one or more performance measures. – The likelihood(s) (qualitative or quantitative) of those scenarios. – The consequences (qualitative or quantitative severity of the performance degradation) that would result if those scenarios were to occur. Source: NASA, NPR 8000.4A, Agency Risk Management Procedural Requirements, p. 22
Risk Modeling 101 Risk assessments, which are the very foundation of an integrity management program, can vary in scope or complexity and use different methods or techniques. The ultimate goal of assessing risks is to identify the most significant risks so that an operator can develop an effective and prioritized prevention/ detection/mitigation plan to address the risks. Source: ASME B318.S
Risk Modeling 101 B31.8S Risk Model Approaches Subject Matter Expert (SME) Relative Assessment Model Scenario Based Model
Event Trees, Decision Trees, Fault Trees
Probabilistic Model Source: ASME B318.S
Risk Modeling 101 Risk analysis can be fairly simple with values ranging from 1 to 3 (to reflect high, medium, and low likelihood and consequences)... Multiplying the relative likelihood and consequence numbers together provides the operator with a relative risk for the segment and a relative priority for its assessment. Source: ASME B318.S
Risk Modeling 101
1
2
3
3
3
6
9
2
2
4
6
1
1
2
3
Risk Modeling 101
1
2
3
3
3
6
9
2
2
4
6
1
1
2
3
Risk Modeling 101 Segment 15
Segment 15 currently has a low risk attributed to third party damage This segment is not located near a populated area (consequence = low) and there has been little to no excavation history in the area (likelihood/frequency = low)
Risk Modeling 101
Segment 4 currently has a medium risk attributed to third party damage This segment is located away from populated areas but in a cultivated field (consequence = medium) and the farmer has been plowing the area (likelihood/frequency = medium)
Segment 4
Risk Modeling 101
Segment 9 currently has the highest risk of third party damage in the system This segment is located near homes, businesses and a US Highway (consequence = high). This is an area of active construction with an increasing number of locate tickets (likelihood/frequency = high)
Segment 9
Risk Modeling 101 Segment 15
Annual Probability of Excavation Damage Segment 15 2.10E-04 Segment 4 1.50E-03 Segment 9 7.20E-02
Segment 9 is 342.9 times more likely to experience excavation damage than Segment 15 Segment 9 is 48 times more likely to experience excavation damage than Segment 4
Segment 9
Segment 4
Segment 4 is 7.1 times more likely to experience excavation damage than Segment 15
Risk Modeling 101
Tools for Quantitative Analysis Decision Trees Influence Diagrams Event Trees Fault Trees Monte Carlo Simulation
“Scenario-Based Models. This risk assessment approach creates models that generate a description of an event or series of events leading to a level of risk, and includes both the likelihood and consequences of such events. This method usually includes construction of event trees, decision trees, and fault trees. From these constructs, risk values are determined.” Source: ASME B318.S
Risk Modeling 101
Fault Tree Logic and Probability Example
Risk Modeling 101
Fault Tree Logic and Probability Example
Fault Tree Handbook with Aerospace Applications, NASA Office of Safety and Mission Assurance, Volume 1.1, August, 2002
Risk Modeling 101
Example of Fault Tree “Or” Gate The output event occurs only if one or more of the input events occur ADD the probabilities
.01 + .01 = .02 .01
.01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
.01
Single Fuel Cell .01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
Single Fuel Cell .01
.01
Redundant Fuel Cell .01 x .01 = .0001 .01
Risk Modeling 101
Example of Fault Tree “And” Gate The output event occurs only if ALL of the input events occur MULTIPLY the probabilities
.01
.01
Single Fuel Cell .01 Redundant Fuel Cell .01 x .01 = .0001
.01
Redundant Fuel Cell plus Battery .01 x .01 x .01 = .000001
Risk Modeling 101
Fault Tree Analysis
Risk Modeling 101
Fault Tree Analysis
Risk Modeling 101
Fault Tree Analysis
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Work Group Meeting Topics 1. RMWG Overview 2. Likelihood Modeling 3. Consequence Modeling (and likelihood) 4. Facilities Modeling 5. Data and Risk Tolerance 6. Index Model Applications and Evolving to more Quantitative Models
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Concerns for IOUs and public gas systems...
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Publicly owned transmission statistics Nearly 1,000 municipally owned gas systems in the U.S. 730 APGA members 62 municipal gas systems operate a total of about 2,700 miles of transmission More than half of total mileage is 8” or smaller
*** 40 of the 62 currently do not contain a High Consequence Area (HCA)
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom
Active vs. Passive Components
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom Calibrating the model… Bounding uncertainties….
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.”
Is the process you choose as important as the output?
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
“As a risk person I am a generalist, not an expert-expert.”
PHMSA Risk Model Work Group (RMWG)
Comments, concerns and nuggets of wisdom “We are engineers for the most part, and engineers communicate differently.” “Don’t throw away good knowledge for bad data.”
“As a risk person I am a generalist, not an expert-expert.” “Risk modeling is a process not an event.” i.e., “if I can only get through this!”
PHMSA Risk Model Work Group (RMWG)
Conclusion?
13
Questions? Mason Matthews Athens, AL Gas Department 256-232-1440 [email protected]
