The hazard and operability study remains one of the most powerful tools in process safety, yet many studies fall short of their promise. A HAZOP is only as good as the discipline behind it, and the guide word method described in IEC 61882 rewards teams that stay patient and structured. When a study drifts, the causes are usually the same handful of avoidable mistakes. Understanding these failure modes helps a team run a review that actually finds the hazards worth finding rather than producing a thick report that gathers dust.
Treating the study as a paperwork exercise
The first and most damaging mistake is approaching the HAZOP as a box to tick before startup or before a management of change closes. When the team believes the outcome is predetermined, the questioning becomes shallow and the deviations get waved through. A genuine study assumes nothing and lets the guide words drive an honest exploration of what could go wrong. If the room feels like everyone wants to finish early, the leader has already lost the plot.
Choosing the wrong nodes and boundaries
Poor nodalisation quietly undermines many studies. Nodes that are too large blur the intention of a line or vessel so that deviations lose meaning, while nodes that are too small drown the team in repetition and fatigue. A good node captures a clear design intent such as pressure, temperature, flow, level and composition for a defined section of the process. Boundaries should follow the process logic rather than the drawing sheet edges, and every node should have a stated intention before the guide words begin.
A team without the right people
A HAZOP is a team activity and the quality of the outcome depends on who is in the room. Missing disciplines create blind spots that surface later as incidents. The core participants usually include the following.
- A trained and independent leader who owns the method and keeps the pace
- A scribe who captures findings in real time so the team can focus on thinking
- Process and operations knowledge from people who actually run the plant
- Instrumentation, control and mechanical input for protective systems
- Design or project engineering to explain the original intent
Letting one voice dominate
Even with the right people present, a study can be captured by the most senior or most talkative person in the room. When that happens the quieter operators and younger engineers stop contributing, and their front line knowledge is exactly what catches the subtle deviations. A skilled leader manages the group dynamics, invites every discipline to speak and protects the space for a junior engineer to ask the awkward question that turns out to matter.
Rushing the schedule
Fatigue is the enemy of a good HAZOP. Sessions that run for eight or nine hours a day for a week produce a sharp decline in the quality of thinking, and the deviations examined late on a Friday rarely get the attention they deserve. Sessions of four to six focused hours with real breaks keep the team alert. It is far better to add a day to the plan than to sign off a node that nobody was really thinking about.
Confusing causes, consequences and safeguards
A common recording error is to muddle the three columns that give a HAZOP its structure. A cause is an initiating event, a consequence is what results if nothing intervenes, and a safeguard is a measure that reduces the likelihood or severity. When a team records a safeguard as though it removes the hazard entirely, the residual risk gets understated. Consequences should be evaluated as if the safeguards were absent, so the team can judge honestly whether the protection in place is enough.
Overcrediting safeguards
Related to the previous point is the habit of listing every conceivable safeguard and assuming the sum makes the risk acceptable. A pressure gauge that nobody watches is not a safeguard, and an alarm that shares a sensor with the control loop is not independent of it. The team should ask whether each safeguard is genuinely independent, whether it is tested, and whether an operator has time to act. Where the honest answer falls short, the deviation deserves a recommendation rather than a comfortable note.
Vague and unactionable recommendations
A recommendation such as review the design later is worthless. Good actions are specific, assigned to an owner and framed so that someone can decide when they are complete. The study loses its value if the recommendations are not tracked to closure through the management of change process, and many organisations find that the follow up is where their diligence quietly collapses.
Ignoring human factors and non routine operations
Studies that consider only the steady state normal running of the plant miss the periods when incidents most often happen. Startup, shutdown, maintenance, catalyst change and abnormal operation all deserve their own attention because the protective systems and operator workload differ from normal running. Layering human factors into the discussion, such as the time available to respond and the clarity of the alarm, turns a mechanical study into a realistic one.
Failing to revalidate the study
The final mistake is treating a HAZOP as a one time event. Plants change through modifications, new feedstocks and evolving operating envelopes, and a study from ten years ago may no longer describe the plant that exists today. Many frameworks expect a periodic revalidation so the hazard understanding stays current. A study kept alive is an asset, while a study left to age becomes a false comfort.
Avoiding these mistakes is less about clever technique and more about discipline, the right people and honest questioning. If you are planning a new study or wondering whether an older one still reflects your plant, a process safety specialist can help you scope the work and keep the method rigorous.