SAFETY INSIGHTS

Brainstorming for Hazard Identification and Risk Assessment

Brainstorming can reveal hazards and new control options. Learn how to encourage diverse ideas, manage group dynamics and turn discussion into safety actions.

Brainstorming helps a group generate ideas about what could go wrong, what matters to a decision and how risks could be controlled. In safety engineering, it can reveal possible failure modes, hazards, accident sequences and treatment options. Its value comes from combining relevant knowledge with deliberate encouragement of different ideas.

A useful session separates idea generation from evaluation. Participants first contribute possibilities without having to defend a complete analysis. The team then clarifies, tests and develops those ideas. The initial list is a starting point for assessment, not evidence that the system is safe.

Choose a clear purpose

Brainstorming can support work at any life cycle stage. Early sessions can identify concerns about a proposed concept. Later sessions can examine installation, operation, maintenance or disposal. A focused session can also explore a design change, an incident or options for risk reduction.

State the question that the session must address. “What could harm people during maintenance of this equipment?” gives the team a clearer task than “What are the risks?” Define the system boundary, operating conditions and people affected.

The method is useful when experience is limited or new technology creates unfamiliar interactions. Imagination can suggest scenarios for investigation. It cannot replace missing evidence about their likelihood, consequences or controls.

Prepare the people and information

Invite people who understand different parts of the problem. Include operational and maintenance experience alongside design knowledge. Add relevant specialists and people who work at important interfaces. A group with similar backgrounds can share the same blind spots.

Provide a concise description of the system and its intended use. Supply drawings, task descriptions, known limits and relevant incident information. Distinguish confirmed facts from assumptions. Too little context produces vague suggestions; excessive detail can distract from the question.

Appoint a facilitator and a recorder. Prepare prompts about operating modes, human actions, equipment failures, external events and shared services. Use these prompts to open new lines of thought when discussion slows.

Generate ideas before judging them

Explain the session rules. Invite a wide range of relevant ideas, allow people to build on earlier contributions and defer criticism during the initial generation phase. Capture concerns in the contributor’s intended meaning. Do not dismiss an idea because it is incomplete or difficult to describe.

The facilitator should keep the discussion within scope without forcing it towards a preferred answer. If one topic is exhausted, introduce another condition or perspective. Ask what changes during start-up, a power failure, maintenance or an emergency.

Brief factual clarification can prevent confusion, but avoid lengthy debates while ideas are still emerging. Record matters that need investigation and return to them during evaluation. Handle any immediate safety concern through the organisation’s established arrangements rather than waiting for the workshop to finish.

Give each participant a way to contribute

A lively discussion can still exclude valuable knowledge. Seniority, confidence or fear of criticism can affect who speaks. Invite contributions without treating silence as agreement. A short period of individual writing before group discussion can give people time to form their own ideas.

Electronic submission can allow simultaneous contributions. Anonymous input may help where people are reluctant to raise concerns, but anonymity depends on the tool and arrangements. Explain what is anonymous and how questions will be clarified.

Nominal group technique provides a more structured alternative. Health Canada describes silent idea generation, round-robin sharing, clarification and prioritisation. It is not defined by anonymous submission alone. [1] Voting can help organise further work, but an unpopular safety concern still needs a reasoned assessment.

Clarify and evaluate the results

After idea generation, make each concern specific enough to investigate. Separate hazards from causes, harmful outcomes and possible controls. Combine duplicates without losing differences in operating conditions or consequences. Keep a record of concerns set aside and the reasons.

Use evidence to assess credible scenarios. Identify existing controls and check their effectiveness. Mark proposed controls separately. Where knowledge is insufficient, assign an action to obtain the required information.

Use other techniques to examine gaps. Task analysis can explore human actions. SWIFT can add structured questions. HAZOP, FMEA or Fault Tree Analysis may be appropriate for particular technical issues. Brainstorming alone cannot demonstrate complete hazard coverage.

Practical example maintenance in an equipment compartment

A team reviews maintenance of a pump in a machinery compartment. Individual notes raise water leakage, unexpected start-up and restricted access. Group discussion adds a less obvious concern: removing the pump could damage a neighbouring cable tray.

The team develops this into a scenario. Lifting equipment or the suspended pump could strike cables that supply a protective system. The review must establish the lifting path, isolation boundaries and consequences of cable damage.

The output is an action to examine the installation and maintenance method, with a named owner and completion criterion. A possible lifting restriction is recorded as a proposal. This illustrative example shows how an idea becomes a defined issue for analysis.

Case study Columbia and the treatment of dissent

Space Shuttle Columbia was lost during re-entry on 1 February 2003, killing its seven crew members. The Columbia Accident Investigation Board identified organisational barriers that obstructed safety information and suppressed professional differences of opinion. It also criticised reliance on previous success in place of sound engineering evidence. [2]

For a brainstorming facilitator, the lesson is to preserve questions that challenge the prevailing view. Record uncertainty, allow alternative interpretations and ensure that unresolved safety concerns reach the people responsible for decisions. A meeting can generate ideas and still fail if those ideas have no effective route to assessment.

This is a lesson drawn from the investigation. It is not a claim that a brainstorming technique alone would have prevented the accident.

Turn the discussion into accountable work

Issue a record of the scope, participants, assumptions, findings and unresolved concerns. Assign actions and link significant issues to the Hazard Log or relevant assessment. Explain which decisions require further evidence and who can approve them.

Check action completion and communicate the outcome to contributors. The quality of the session depends on what happens after the discussion: credible concerns must receive a justified response.

Hashtags

#Brainstorming, #HazardIdentification, #RiskAssessment, #SystemSafety, #SafetyEngineering, #SafetyCulture, #RiskManagement

Sources

[1] Health Canada. Policy Toolkit for Public Involvement in Decision Making. 2000. Nominal Group Process section. Read source

[2] Columbia Accident Investigation Board. Report Volume I. August 2003. Executive summary and organisational findings. Official copy hosted by the US Department of Energy. Read source

Simon Di Nucci
Simon Di Nucci

Simon Di Nucci, CPEng, FIEAust, NER, brings 30 years of experience in safety engineering and consultancy. His work spans aircraft, ships, submarines, air-traffic management systems, trains and software in the UK and Australia.

About the author

THE SAFETY ARTISAN NEWSLETTER

Keep learning.
Stay curious.

Practical safety ideas and training updates from Simon, delivered to your inbox.

You can unsubscribe at any time. If the form does not load, open it directly.