What is the What-if method?
What-if analysis is a qualitative technique for examining situations that could disrupt the intended functioning of a process. Through hypothetical questions, a team identifies scenarios, plausible causes, consequences, and existing measures. The question initiates the analysis; the useful outcome is a justified conclusion regarding the necessary actions.
Its flexibility allows it to be applied to focused reviews and changes, among other uses. This same flexibility requires preparation to avoid relying solely on someone’s recall during a meeting. The method must have a defined scope, participants, and recording criteria, along with sufficient information about the work being reviewed.
Prepare the scope and information
The team determines which installation, operation, or modification to study and under what conditions. Relevant states, such as startup, normal operation, shutdown, and maintenance, must be included. Drawings, procedures, materials, and background information allow for the formulation of specific questions and the avoidance of scenarios unrelated to the system.
It is advisable to review the information before the session and identify any gaps. If a piece of information influences a conclusion, it should be noted as pending and its verification assigned. The absence of documentation should not be interpreted as a consensus-based approval. Time should also be allocated to review interfaces with other activities or resources.
Formulate questions that allow for analysis
A useful question describes a specific deviation: a loss of supply, an incorrect sequence, unforeseen material, or a function unavailability. It should be precise enough for the team to explain what happens next. Simply asking if everything is safe doesn’t allow for scenario building or evaluating controls.
The session can be organized by phases, equipment, or deviation categories, using adapted support checklists. These checklists help verify coverage but do not replace process knowledge. It is important to include infrequent conditions and auxiliary tasks, where people may be exposed in ways that differ from their usual routine.
Review causes, consequences, and controls
For each scenario, plausible causes and relevant consequences for individuals are documented. Then, measures that prevent its progression or limit its effects are examined. A distinction must be made between existing and proven protections and proposed improvements that are not yet implemented.
The response also needs to consider control failures and dependencies. Two measures can share a power source, sensor, or human intervention. Listing them as if they were independent barriers can exaggerate the level of protection. When the relationship is complex, it’s advisable to move the issue to a fault tree analysis or another study with the necessary detail. A worksheet can separate the question, cause, consequence, existing control, evidence, and pending action. This separation prevents assuming a measure is implemented when it has only been proposed. It also allows for reviewing which conclusion changed upon obtaining new data, without losing the traceability of the initial analysis.
Participation and quality of the session
The analysis brings together individuals with complementary knowledge of design, operation, maintenance, and prevention. The experience of those performing the task allows for the identification of discrepancies between the written procedure and the actual work. Proper coordination facilitates the expression of doubts without turning them into personal blame.
The session should distinguish between facts, hypotheses, and opinions awaiting verification. If technical disagreements arise, it’s advisable to document what evidence will help resolve them. The goal is not to accumulate questions or close every worksheet entry during the meeting, but rather to produce well-founded decisions and identify areas where further analysis is needed.
Differences with HAZOP and other methods
HAZOP uses a systematic examination of deviations using parameters and guide words within a defined structure. What-if analysis offers greater flexibility in scenario formulation. Choosing one or the other requires considering complexity, available information, and purpose; a brief meeting does not automatically replace a necessary detailed study .
FMEA focuses on functions and failure modes. Bow-tie analysis helps represent threats, barriers and consequences. What-if results can inform these methods or benefit from them, while keeping their differences clear and identifying the questions that each analysis has actually covered.
Practical example
Before modifying a transfer operation, a team asks what would happen if the receiving vessel did not have the expected capacity. They review how the vessel is identified, what information the operator uses, and what controls would prevent incompatible filling. They also assess the actual availability of those controls during the operation.
The review reveals that a check relies on information that does not always reach the workstation. A process improvement and verification procedure are defined before implementing the change. The scenario is documented, including the responsible party and the required evidence. The example illustrates preventive reasoning, not a design instruction applicable to any installation.
Actions, review and limits
Each action should clearly state what will be done, who is responsible, when it will be verified, and what evidence will allow it to be closed. Change management should incorporate relevant results before implementing modifications. If a decision is postponed, the provisional conditions and their corresponding evaluation should be made explicit.
Common mistakes include copying generic questions, accepting answers without evidence, and considering a long list of safeguards sufficient. What-if analysis doesn’t calculate a numerical probability on its own. Its value lies in exploring scenarios in a disciplined manner, identifying areas for improvement, and maintaining a reviewable record of preventive decisions.
