Introduction to Risk Engineering

Nazmul Rahmani

Mary Kay O’Connor Process Safety Center

Chemical Engineering Department
Texas A&M University System
Safety?

• What is Health? •What is Safety?

• Absence of illness/sickness •Absence of incidents or losses
• How to measure and •How to measure and
maximize? maximize?
•Check for any incident -
• Check for any illness or
causing situation –RISK
sickness –RISK
•Minimize Risk – Safer design
• Minimize Risk- Avoid and operation
unhealthy conditions •Monitor and Manage Risk
• Do regular checkups

Safety to the System is Health to the human

2
Safety vs Risk

Risk
Assessment of the presence and impact
of unwanted situation at time t
Risk (t)= Scenarios {occurrence of
unwanted situations & its impact}

𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 𝑡𝑡 = 𝑆𝑆 𝑡𝑡 {𝐹𝐹 𝑡𝑡 . 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 𝑡𝑡 }

Safety
Absence of unwanted situation in the system
at time t
1
𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑡𝑡 ⍺
𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 (𝑡𝑡)

3
 Safety & Security
Safety Security
•The aim here is to protect the people, assets •The aim here is to protect a system from
and environment from the unintended event unwanted access and damage from the
or malfunction of a system external environment.
•It preserves the integrity of the people and •It preserves the integrity of the system.
environment •The threat (potential event) is from the
•The threat (potential event) is from the intended situation.
unintended situation

Buncefield fire

Colonial Pipline

4
WHAT IS PROCESS SAFETY?

Chemical Process Safety is a blend of engineering

and management skills focused on preventing
catastrophic accidents, particularly explosions,
fires, and toxic releases, associated with the use of
chemicals and petroleum products.

Center for Chemical Process Safety

American Institute of Chemical Engineers

5 5
PROCESS SAFETY VS PERSONAL SAFETY

What Is the difference and why might that be important ??

Process Safety addresses high consequence, low frequency events.

Major Offsite
Incident
Consequence

Process Safety
Serious Onsite Incident

Slips, Trips and Falls

Personal Safety

Frequency

6 6
PROCESS SAFETY VS PERSONAL SAFETY

Personal Safety
Process Safety
(Occupational Safety)
•Focuses on: Prevention of
incidents involving leaks, •Personal safety focuses on
spills, fires or explosions by elements that may cause injury
ensuring well-designed, or harm to an individual.
safely operated and properly •Related risks usually result in
maintained facilities. high frequency-low consequence
events (e.g. slips, trips and falls).
•Involves: •Note that a human fatality could
•Installation of systems in also be a consequence
place to monitor and control •Over reliance and a focus on
hazards. good occupational illness/injury
•Technical, management and rate performance can lead to a
operational systems to false sense of confidence about
achieve desired outcomes. Process Safety.

7 7
 Personal Safety vs. Process Safety

Workers  Follow the rules & Ensuring facilities are properly

work safely designed and safely operated
• Authorization/work permits and maintained
(entry to confined spaces, • Identify, assess, eliminate or
ensuring gas testing, etc.). reduce risk.
• Ensure isolation before • Prevent fires, explosions,
electrical work. chemical releases.
• Do not smoke/use drugs or
alcohol.

8
Historical Events that Define Process Safety

9 9
FATAL OCCUPATIONAL INJURIES
10

Ref: US Bureau of Labor Statistics

What is Risk?
11

 The New World Dictionary defines risk as the chance of injury,

damage, or loss.
 To put oneself "at risk" means to participate voluntarily or be
involuntarily involved in an activity that could lead to injury,
damages, or other losses.
 This definition of risk does not explicitly state the likelihood or
chances of gain and success, which we can achieve by accepting
levels of risk, such as monetary risk
 Our preference is always that risk for gain be as low as feasible or
As Low As Reasonably Practicable (ALARP), as practiced in Industry.
 Hazard & Risk
12

A hazard is a chemical or physical condition that has the potential for

causing damage to people, property, or the environment. (i.e. a risk
receptor)

Risk is a measure of damage to hazards taking into account both the

likelihood and magnitude of the damage (consequence).
The expected damage can be in terms of human injury, environmental
impact or economic loss.

Risk = Probability of Failure (Likelihood) x Consequences (Severity)

Voluntary Risks
13

Voluntary risks result from hazards of activities we choose to engage

in, such as:
• driving a car or riding a motorcycle
• climbing a ladder
• entering a laboratory
• skydiving
• working on a farm
• working in a plant
Involuntary Risks
14

Involuntary risks result from hazards of activities or events without

prior consent or choice of persons. Examples include:
• Acts of nature such as lightning, fires, floods, tornados, etc.
• Exposure to environmental contaminants

Which of these two types of activities (Voluntary or Involuntary),

are people more willing to tolerate to a higher level? Why?
Probability or Frequency
15

Risk = Probability of Failure (Likelihood) x Consequences (Severity)

 Probability or likelihood - A quantitative measure of the uncertainty
in the occurrence of a value, value range, event, or an event
sequence.
• Probability (Pr) is defined as a unitless number within [0,1].

 Frequency - Number of occurrences of an event per unit time, often

hour or year, or per distance, area, or volume.
• Frequency of occurrence is used to estimate probability of
occurrence.
• In its simplest form, Relative Frequency= Point value of Probability
of Occurrence = [(# of specific events, n)/(# of all events, N)
Consequences
16

 The various outcomes of taking risk (losses/gain)

 Fatalities, monetary losses, property/asset damage,
Environmental impact
 Various commercial softwares for consequence analysis
(e.g. PHAST, ALOHA, FLACS-CFD etc.)
 Risk estimation example
 Hazards

Which tool when used presents the greatest hazard to user?

Why?

17
 Risks

Which tool when used presents the Highest risk to user?

Why?

18
How to do Risk Assessment:
(the five questions of a robust risk assessment)
19

 What can go wrong?

 Hazard Identification
 What are the adverse impacts or consequences ?
 Consequence Risk Analysis
analysis
(qualitative or quantitative)
 How likely is it to happen?
 Frequency Analysis
 Do I need to do anything about it?
 Risk Evaluation / Risk Assessment
 What should I do about it?
 Risk Control
If you planning a job in an operating plant – any other question?
Basic Definitions
20

 Risk Analysis: process of assessing, managing, decision making, and

informing others about existence, nature, magnitude, probability,
contributing factors, and uncertainties of potential losses (or gains).

 Risk level: size or magnitude range due to hazard exposure: Estimates

of both the probability (or, frequency) of upset scenarios and
outcome consequences
 Scenario: sequence of specific events beginning with an initiation
event, intermediate events, and terminating in an outcome event.
Various inter-related events can influence a scenario.
Risk Tolerance
21

 Industrial risks are now less acceptable to industrial employees,

company stakeholders, and to the public than they were previously

 A generally acceptable guideline for acceptable voluntary risk is that

industrial employees should be exposed to no more risk on the
(voluntary) professional work job than they are exposed to involuntary
activities outside of work.
Risk Tolerance or Acceptability Levels
22

 Risk tolerability ranges today are based on events in human life:

 An involuntary risk (fatality risk) is generally considered tolerably low
by the public if it is similar to or lower than the risk of fatality per
person-year due to natural hazards
• Historically ~10-6 per year.
 A voluntary risk is unacceptably high to the public if it is near to or
above the risk of fatality per person due to disease
• Historically ~10-3 per year.
 Risk acceptance criteria for events that could lead to fatalities were
founded on the range between the two frequency values: [10-6, 10-3]

(Shortreed, et al, 1995)

Risk Tolerance Criteria
23
Types of ALARP Demonstration (HSE a)

As
Low
As
Reasonably
Practicable

HSE a, “Guidance on ALARP Decision in COMAH”, accessed February 19, 2021, https://www.hse.gov.uk/foi/internalops/hid_circs/permissioning/spc_perm_37/
Traditional Measures of Risk Acceptability
24

 Regulations (such as for toxic exposures, e.g., OSHA, EPA)

 Industry sector practice
 Company practice
 Engineering judgment, expert judgment based on experience
 Risk communication

OSHA: Occupational Safety & Health Administration

EPA: Environmental Protection Agency
Engineering Risk Analysis Methods and Critical Role of Decisions Under
Uncertainty During System Life Cycle
Conceptual design Decisions about alternative design options

Decisions to help prevent, reduce or eliminate hazards (inherently

Design safer design),
minimize life-cycle cost, maximize performance

Identify & test systems that contribute most to risk,

Development
quality assurance, warranty development

Regulation (Risk Decisions concerning system elements that contribute most to risk,
Management) set monitoring and performance criteria, perform inspections

Decisions to optimize cost of operation, tests, and maintenance,

Operation (Risk
define surveillance requirements, schedules, replacement policies &
Management)
decisions, aging estimation & management, security measures

Decisions for safe decommissioning alternatives,

Decommissioning
select disposal methods, assess long term liability
Decision Making Under Uncertainty
26

 Decisions about designs, engineering projects, test and maintenance

strategies, or Emergency Planning:
• Assume only worst-case conditions?: overly costly
• Assume only low case conditions?: poor safety
 Engineering variables are random variables
 The inherent variability can be modeled and communicated in
histograms, which are frequency diagrams displaying observed
values
How to measure uncertainty
27

 Sample characteristics include:

• Sample mean or average μ, median, mode
• Sample variance: average of the squared deviations from the sample mean
due to randomness
• Sample standard deviation 𝜎𝜎 = (𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 0.5 ); it is a measure of dispersion of
the annual rainfall intensity due to randomness of the rainfall.
𝜎𝜎
• Coefficient of Variability (COV) = ; it is an unitless metric for the uncertainty in
𝜇𝜇
the mean as a measure of data dispersion.
Risk Analysis Approaches
28
 Traditional Risk Analysis
• Manage technical systems, consider accidents due to operator error and
equipment error
• Decisions based on point value alone
• Assumed independence of components
• Easy to do, but not effective
 Systems Approach to Risk Analysis
• Manages Socio-Technical Systems (STS) and their organizations
• Investigates accidents or near misses to determine underlying causes
• Makes decisions based on ranges and distributions of information
• Considers significant interdependencies
• Requires more effort to learn and practice but more realistic
System Risk Analysis Steps
29

 Identify and characterize hazards

 Develop scenarios of upset events, conditions
 Assess risk and represent in risk profiles
 Categorize, prioritize contributions to the overall risk
 Reduce risk, Manage risk within acceptable ranges
 Communicate risk with personnel and stakeholders
 Train/retrain personnel, educate/update stakeholders
 Monitor, measure system & organization, find trends
 Continually update system data and parameters
 Update behavior predictions of components and system
Risk Assessment, Management, Communication Process Map
System Data &
Complementary Info

Hazards
Continue Updating Characterization & Risk Re-Evaluation
Scenario Development

Probability/Freq. Consequence
Analysis & Distribution Analysis & Distribution

Risk
Distribution

Risk Criteria Risk Evaluation Risk Reduction

System Monitoring Risk Governance

Accepted Risk Level Risk Management30
& Measurement
Risk Communication
Types of Risk Assessment
31

 Quantitative RA (QRA): calculate risk in form of a numerical

probability (or frequency) distribution of an event and the event
consequence magnitude distribution. Data-centric, model-based.
Generally needed for most critical parts of a system.
 Semi-Quantitative RA: use scores or order of magnitude for
frequency and outcome magnitude for non-critical parts of a
system.
 Qualitative RA: use of linguistic or ordinal scales (e.g., low,
medium, high; 1, 2, 3) for probability or the outcome magnitude.
Often used for screening, but the scales are ambiguous and are
not reliable. Subjective
Semi-Quantitative & Quantitative RA
32
 Semi-Quantitative: Rank-ordered approximation of probabilities (or
frequencies) and consequences to rank scenario outcome events with
an example:
 Frequency: Freq/yr  Consequences:
• Frequent >1 – Catastrophic: deaths
• Probable 1-10-1
– Critical: 10 major injuries
• Occasional 10-1-10-2
• Remote 10-2-10-4 – Marginal: 15 minor injuries
• Improbable 10-4-10-6 – Minor: 5 hours down time
• Incredible < 10-6

Linguistic scales are ambiguous and should be defined by semi-

quantitative ranges as shown. A suitable RA must be semi-quantitative
as above or fully quantitative!