You are on page 1of 38

Introduction to Quantitative Risk

Assessment
Webinar June 2017

14 June 2017

1 DNV GL © 2015 14 June 2017 SAFER, SMARTER, GREENER


Speakers Speakers

Kenny Shaba Mark Hunter


Phast and Safeti Product Manager Chartered Engineer
with extensive experience in in DNV GL’s Advisory business
Technical Safety projects 10 years’ experience of safety
(especially QRA) for both onshore related studies in the oil & gas
and offshore assets sector

2 DNV GL © 2015 14 June 2017


Notes Notes

Any questions during the webinar?


Please use the chat window on Go to Webinar control panel
to send your questions.

The questions will be collected and handled during Q&A session


at the end.

Please accept our apologies if we do not have time to answer


all your questions. In that case we will follow up with you
after the webinar.

For further inquiry, demo, test, training or quote,


please contact us:
software@dnvgl.com

3 DNV GL © 2015 14 June 2017


Agenda Agenda

What is QRA? Drivers for QRA


01 02 why QRA?

Challenges that QRA Pros/Cons of


can help 03 04 QRA
address/Applications
of QRA

Doing a QRA – Conclusion


Practical 05 06
considerations, best
practices, pitfalls

4 DNV GL © 2015 14 June 2017


What is QRA?

5 DNV GL © 2015 14 June 2017


What is Risk?

Commonly used definition …

The LIKELIHOOD and consequence of a specified


undesired event occurring within
a specified period or under
specified circumstances

R = f(L,C)

6 DNV GL © 2015 14 June 2017


What is Quantitative Risk Assessment?

is a risk assessment methodology that allows

for numerical estimates of the level of risk


associated with a certain activity or series of activities to
be estimated and then assessed

7 DNV GL © 2015 14 June 2017


What is Quantitative Risk Assessment?

What Can Go Wrong?


Hazard Identification

How Often? How Big?


Frequency Analysis Consequence Analysis

So What?
Risk Assessment

What Do I Do?
Risk Mitigation

8 DNV GL © 2015 14 June 2017


Typical QRA Outputs

 Individual Risk  Societal (Group) Risk


– “Relationship between frequency
and the number of people
Solo rock UNACCEPTABLE suffering from a specified level of
climbing RISK Riskiest
(5 hr per week) 1 in 1,000 Industry harm from the realisation of
specific hazards”

1 in 10,000 Traffic
Accident
(driving 10h
Heart per week)
Disease
1 in 100,000

Smoking
(10 cigarettes
per day)
Safest
1 in 1,000,000 Industry
BROADLY
ACCEPTABLE
RISK
Accident 1 in 10,000,000 Struck
at home by lightning

9 DNV GL © 2015 14 June 2017


Typical QRA Outputs – Individual Risk

LSIR IRPA

Location Specific Individual Risk Individual Risk Per Annum

0.00 0.20 0.40


km

“1 in 10 000/Yr” or “10-4/Yr”

“1 in 100 000/Yr” or “10-5/Yr”

“1 in 1 000 000/Yr” or “10-6/Yr”

“1 in 10 000 000/Yr” or “10-7/Yr”

10-8/Yr

10-9/Yr

10 DNV GL © 2015 14 June 2017


Individual Risk – ALARP Principle

ALARP – As Low as Reasonably Practicable

HIGH RISK

Risk cannot be justified


Unacceptable region save in extraordinary
Maximum
Tolerable circumstances
Criterion

Tolerable only if risk reduction is


The ALARP or Tolerability impracticable or if the cost is grossly
region disproportionate to the improvement
(Risk is undertaken only if gained
a benefit is desired) Tolerable if cost of reduction would
exceed the improvement
Desirable
Target

Broadly acceptable region Necessary to maintain assurance


(No need for detailed working to that risk remains at this level
demonstrate ALARP)

LOW RISK NEGLIGIBLE RISK

11 DNV GL © 2015 14 June 2017


Individual Risk Criteria – Global Examples
3 bands
Country 2 Bands 3 Bands
UK: Safety Cases/Reports Intolerable
UK: Land Use Planning
Netherlands
Belgium: Flanders ALARP
Hong Kong
Brazil (3 states)
Singapore Tolerable
Malaysia
Australia: Western Australia
Australia: New South Wales
2 bands
Australia: Queensland
Australia: Victoria
Intolerable
Canada
France
Switzerland ALARP
USA: Federal Agencies
USA: Santa Barbara County, CA
USA: New Jersey
IMO
Abu Dhabi

12 DNV GL © 2015 14 June 2017


Typical QRA Outputs – Societal (Group) Risk

FN Curves PLL

Frequency/Number of fatalities Potential Loss of Life

13 DNV GL © 2015 14 June 2017


Sample Societal Risk Criteria (Upper Limit)
10-2
Frequency (/year) of N or more fatalities

10-3

10-4

10-5
UK HSE
10-6 RdJ & RGdS
Victoria
Hong Kong
10-7 São P & WA existing
Flanders
Santa Barbara
10-8 Netherlands & CH
WA new
NSW
10-9
1 10 100 1000 10000
Number of fatalities, N

14 DNV GL © 2015 14 June 2017


Drivers for QRA

15 DNV GL © 2015 14 June 2017


Drivers for QRA

NRC WASH-1400 Reactor Safety Study

16 DNV GL © 2015 14 June 2017


Drivers for QRA

Today, QRA is a …

Decision
support tool

Core process
safety methodology

Key to
Risk Regulation

17 DNV GL © 2015 14 June 2017


Challenges that QRA can help
address? /Applications of QRA

18 DNV GL © 2015 14 June 2017


Examples of QRA Applications

Operational
Life Cycle Asset Risk Land Use
Risk
Management Planning
Safety Case Management
Development
• Design
• Construction
• Operations
• Decommissioning
Cost Benefit
Decision Analysis
Compliance
Support
(Legal
Requirement)
Equipment
Design
Escalation
Assessment
Quantify the
Technology benefit of
Option mitigation Insurance
Layout Selection measures
Optimization
• Site
• Equipment

19 DNV GL © 2015 14 June 2017


Example of Cost Benefit Analysis

Risk 1 Cost a lot


Little reduction in risk

2 Cost a little
Little reduction in risk

Base 1 2 3 4 3 Cost is moderate


Moderate reduction in risk

Cost 4 Cost is relatively little


Large reduction in risk

20 DNV GL © 2015 14 June 2017


Pros/Cons of QRA

21 DNV GL © 2015 14 June 2017


Pros of QRA

Enables risk (various) Offers an Understanding an


to be quantified objective/rational Engineered System
approach to risk
assessment

Visual Depiction of Identify Risk Drivers


Risk

22 DNV GL © 2015 14 June 2017


Strengths of QRA - Understanding an Engineered System
• Flammable
Hazard
material
• Toxic
material

• Weather
Environment
• Ignition
• Explosion
sources

• People
Receptor • Assets

“The key benefit of estimating risk lies in the achievement of detailed understanding
of the engineered system and the implications of various siting and technical
options”. (Royal Society, 1999)

23 DNV GL © 2015 14 June 2017


Strengths of QRA – Visual Depiction of Risk

0.00 0.20 0.40


km

“1 in 10,000/Yr” or “10-4/Yr”

“1 in 100,000/Yr” or “10-5/Yr”

“1 in 1 000,000/Yr” or “10-6/Yr”

“1 in 10 000,000/Yr” or “10-7/Yr”

10-8/Yr

10-9/Yr

24 DNV GL © 2015 14 June 2017


Strengths of QRA – Identify Risk Drivers

Risk at Control Room Risk at Control Room


Breakdown by Scenario Breakdown by Hazard Type

Scenario 1 Scenario 2
Scenario 3 Scenario 4 Toxic Explosion Jet Fire
Pool Fire Flashfire Fireball

“The process of undertaking a QRA can lead to a better understanding of the


important features contributing to risk and weaknesses in the systems as
well as allowing a numerical estimate of the residual risk to be derived” (HSE,
1999)

25 DNV GL © 2015 14 June 2017


Cons of QRA

Uncertainty in the risk Cost/time intensive Susceptible to “GIGO”


estimates  Addressed by software tools  Expertise is key
 Addressed by adopting a such as Safeti reduce the
conservative approach analytical burden approach

Highly specialised
activity
 Expertise e.g. DNV GL is
available

26 DNV GL © 2015 14 June 2017


Doing a QRA – Practical
considerations, best practices,
pitfalls

27 DNV GL © 2015 14 June 2017


QRA is akin to predicting the weather…

 Goal is to predict the weather based on


available data, models

 Accuracy is dependent on a range of


factors:

28 DNV GL © 2015 14 June 2017


Key Considerations for QRA

Uncertainty Quality Management

Choice of Software application Competency of resources

29 DNV GL © 2015 14 June 2017


Practicalities

 Do I need a QRA?
– Is it appropriate for my site?
– How will I use the results?
– Regulatory requirements?

 Information needed?

 Can I do the QRA myself or do I need someone else to do it?


– Capabilities/competencies
– Software
– 3rd party assurance (may be regulatory requirement)
– Maybe mix both in-house and external

 Documentation – the QRA should be well documented (all assumptions etc.)

30 DNV GL © 2015 14 June 2017


Software Applications for QRA

 Many solutions exist, most are in-


house proprietary and Microsoft
Excel® based (i.e. spreadsheet
models)

 Commercial quality QRA software is


available e.g. Safeti/Safeti Offshore for
on/offshore QRA respectively
– High quality, integrates codified best
practice and industry standards

31 DNV GL © 2015 14 June 2017


How to check a QRA Study/Report for quality etc.

 Independent checks - verification

 Validation versus comparable assets

 Specifically, check:
– Main risk contributors – are they as expected and consistent with experience?
– Consistency of:
– Individual versus Societal risk
– FN Curve and PLL
– FN Curve and individual risks

32 DNV GL © 2015 14 June 2017


The need to dig deeper: Hazards With Same PLL

No of
Frequency PLL
fatalities

1 per 10 years 1 0.1

1.0E-01 1 per 100 years 10 0.1

1 per 1000 years 100 0.1


1.0E-02
1 per 10000 years 1000 0.1
Frequency of Exceedance

1.0E-03

1.0E-04

1.0E-05

1.0E-06
1 10 100 1000
Number of Fatalities

33 DNV GL © 2015 14 June 2017 Slide 33


Conclusion

34 DNV GL © 2015 14 June 2017


Conclusions

 What is QRA?

 Why/ how is it done?

 What is it used for?

35 DNV GL © 2015 14 June 2017


Useful resources on QRA and Risk Analysis

 American Institute of Chemical Engineers. Center for Chemical Process Safety,


2000. Guidelines for chemical process quantitative risk analysis (Vol. 1). Wiley-
AIChE.
 CPR (1999) CPR18E “Purple book”. Sdu Uitgevers, Den Haag, 1999
 HSE (1999) Reducing Risks, Protecting People
 McQuaid, J (2007) “A Historical Perspective on the Tolerability of Risk” in “Bouder,
Frédéric, and David Slavin. The tolerability of risk: a new framework for risk
management. Earthscan, 2007”.
 Royal Society (1992) Royal Society Study Group on Risk “Risk: analysis,
perception and management” - report of a Royal Society Study Group. London:
Royal Society , 1992.)
 OECD (2003) “OECD Guiding principles for Chemical Accident Prevention,
Preparedness and Response”. OECD

36 DNV GL © 2015 14 June 2017


Q&A

37 DNV GL © 2015 14 June 2017


Thank you for attending Notes

Webinar recording, presentation slides, together with a Q&A


document will be sent to all registered delegates in ~ 2 wks

Any further questions can be sent to software@dnvgl.com

For further inquiry, demo, test, training or quote,


please contact us:
software@dnvgl.com

38 DNV GL © 2015 14 June 2017