Both terms get used as though they meant the same thing, usually by people commissioning one and receiving the other. They sit on different standards, belong at different points in a project, and produce different documents. Getting the choice wrong wastes the study.
Two Questions, Not One
The distinction is easiest to hold if you treat each method as asking a different question. A HAZID asks what hazards exist here at all. A HAZOP asks how this particular process can deviate from what it was designed to do. The first is wide and early. The second is narrow and late.
A HAZID applies guideword-based brainstorming under ISO 17776 to surface the broad range of hazards a facility introduces, typically at concept, pre-FEED or FEED stage. It is deliberately qualitative and wide-angle, and it is usually conducted against layout drawings or a three-dimensional model rather than detailed engineering. Its natural territory is hazards arising from the external environment, from planned activities, and from the inherent properties of what will be handled: things a layout can address while a layout is still changeable.
A HAZOP is a different exercise entirely. Developed by ICI in the 1960s and now codified in IEC 61882, it examines a defined process unit node by node against piping and instrumentation diagrams. The team establishes the design intent for each node, then applies guide words to each relevant parameter to generate deviations from that intent, recording causes, consequences, existing safeguards and recommendations for each.
A HAZID tells you what you are dealing with. A HAZOP tells you how the thing you built can misbehave. Neither answers the other’s question.
What a HAZOP Actually Does
The mechanical core of a HAZOP is the guide word set, and its systematic application is what distinguishes the method from a discussion. IEC 61882 defines seven: none, more, less, as well as, part of, reverse, and other than. Each is applied to each relevant parameter at each node, with parameters typically including flow, pressure, temperature, level and composition.
The point of the discipline is that it removes reliance on imagination. A team asked to think of what might go wrong will produce the failures it has seen before. A team required to consider “reverse flow” at every node where flow is a design intention will reach scenarios nobody in the room had thought about, which is precisely where the value sits. Reverse flow and low temperature are the deviations most often underweighted in upstream scopes, and reverse flow through a single check valve has contributed to serious hydrocarbon releases offshore.
HAZID: Concept, pre-FEED or FEED. Layout drawings or 3D model. Wide coverage of hazard categories. Output informs layout, siting, escape routes, and which detailed studies will be needed.
HAZOP: Detailed design, on P&IDs that are stable. Node by node against the seven guide words. Output feeds safety integrity level determination, alarm rationalisation, operating procedures and permit scope.
Both, in sequence: Normal on larger projects. The HAZID identifies what needs studying; the HAZOP studies it properly once there is enough engineering to study.
Management of change: Most jurisdictions expect a documented HAZOP or equivalent before commissioning a hazardous facility, and again before any major modification. A change that alters design intent invalidates the study that assessed the old intent.
The Timing Error That Wastes the Money
The most common and most expensive mistake is running a HAZOP too early. Under schedule pressure, teams convene one at FEED stage while the P&IDs are still preliminary. The workshop happens, people attend, a report is produced.
What comes out is an action register full of items reading “to be confirmed in detailed design.” The study consumed senior engineering time and delivered almost no risk reduction, because there was not enough engineering definition to assess deviations properly. A HAZOP validates a design; it cannot validate a design that does not yet exist.
An action register where half the entries say “to be confirmed later” is not a study finding. It is a record that the study was held too soon.
The converse error is running a HAZID too late, once the layout is fixed. A HAZID’s value lies in surfacing hazards while siting, separation distances and escape routes can still be changed cheaply. Held after those decisions are locked, it produces hazards nobody can now design out, and the organisation is left managing operationally what it could have engineered away.
What the Output Should Look Like
Here is the test a client should apply. The deliverable of a HAZOP is not a report. It is a worksheet capturing, for every node examined, the deviations considered, the guide words applied, identified causes, consequences, existing safeguards, risk ranking, and recommended actions with assigned owners.
Every element in that list is load-bearing, and the last one decides whether anything happens. An action without a named owner is a suggestion. An action without a date is a suggestion with an aspiration attached. Studies that change nothing almost always fail at exactly this point: the technical work was competent, and then the recommendations went into a document that circulated and settled.
Why Most Studies Go Nowhere
Two failure modes account for nearly all of it. The first is generic output. A study that produces hazards applicable to any facility of that type has not examined this one. Generic findings are a symptom of insufficient design information, a team without the operational knowledge to challenge the drawings, or a facilitator working through a template rather than the plant.
The second is the follow-through gap. The register is issued, the project moves on, and nobody tracks closure. Six months later the actions remain open, the design has changed around them, and the study describes a facility that no longer exists. This is the same pattern that produces safety management systems full of procedures nobody follows and maintenance plans that miss the fire damper: the analysis was done, the discipline of acting on it was not built.
Both are visible in advance to a client who knows what to ask for.
What to Demand from Whoever Runs It
Commissioning a study is a procurement decision with a technical outcome, and a handful of questions separate a useful engagement from an expensive workshop.
Ask which standard the study will follow and expect ISO 17776 or IEC 61882 to be named rather than “industry practice.” Ask who facilitates, and whether they are independent of the design team, since a facilitator reviewing their own colleagues’ work is poorly placed to challenge the design intent. Ask what engineering deliverables will be available on the day, because that determines whether a HAZOP is feasible at all. Ask who will be in the room, and insist on operations and maintenance representation rather than designers alone, since the people who will run the plant know how it will actually be operated. And ask what the deliverable is, expecting a worksheet and an action register with owners rather than a narrative report.
Finally, ask what happens after issue. A study that ends at the report is a study that ends. The engagement worth paying for is the one that hands over a register someone is accountable for closing, because the analysis was never the point. Changing what gets built, and what gets done, was.
Frequently Asked Questions
What is the difference between HAZID and HAZOP?
A HAZID identifies what hazards exist, using guideword-based brainstorming under ISO 17776 at concept, pre-FEED or FEED stage, typically against layout drawings or a 3D model. A HAZOP examines how a specific process can deviate from its design intent, applying IEC 61882 guide words node by node against P&IDs during detailed design. They answer different questions and belong at different points in a project.
What are the HAZOP guide words?
IEC 61882 defines seven: none, more, less, as well as, part of, reverse, and other than. Each is applied systematically to every relevant parameter at every node, with parameters typically including flow, pressure, temperature, level and composition. The systematic application is what makes the method reliable, because it produces deviations the team would not have thought of unprompted.
Can a HAZOP be run during FEED?
It can be convened, but it rarely produces value. Where P&IDs are still preliminary there is insufficient engineering definition to assess deviations properly, and the result is an action register dominated by items to be confirmed in detailed design. A HAZOP validates a design rather than developing one, so it belongs once the design is stable. A HAZID is the appropriate study earlier.
What should the deliverable be?
A worksheet recording, for every node, the deviations examined, guide words applied, causes, consequences, existing safeguards, risk ranking and recommended actions with assigned owners, rather than a narrative report. Each action needs a named individual, a date and a closure criterion, and the register needs reviewing until every action is closed with evidence. Without that, the technical work is sound and nothing changes.
Sources: IEC 61882, Hazard and operability studies (HAZOP studies), Application guide, edition 2.0 (2016), including the guide word set and the node-by-node examination of deviations from design intent · ISO 17776, Petroleum and natural gas industries, Offshore production installations, Major accident hazard management during the design of new installations, guideword-based hazard identification · US OSHA 29 CFR 1910.119, Process Safety Management of Highly Hazardous Chemicals, acceptable process hazard analysis methodologies · CCPS, Guidelines for Hazard Evaluation Procedures
MarineCraft Journal
Get the next briefing in your inbox
Safety protocols, offshore training, regulation and market analysis, written for Malaysia’s maritime and offshore industry. No sales pitches.
We use your address only to send MarineCraft Journal. Unsubscribe any time. See our Privacy Policy.