ModulesProcess Hazard IdentificationProcess Safety
Find the hazards while the design can still change.Guide word by guide word, one record per node.
For project and process safety engineers: screen each node with your company’s guide words, answer present, not present or not applicable, and rate every consequence on your published matrix.
| Guide word · event | Cause → consequence | Initial | Safeguards | Residual | Recommendation |
|---|---|---|---|---|---|
OverpressureLoss of containment from V-201 | Steam valve TV-201 fails open, outlet blockedOverpressure of V-201, rupture, propane fire | S5·L4High | S1 PSV-201 relief to flareExisting S2 PAHH-201, operator isolates steamExisting S3 Independent high-pressure trip on the steam supplyRecommended | S5·L2Medium | R4 |
One node, three views of the same study. An example study; the fields are the product’s, the values illustrative.
Where it connects
On the same loop: HAZOP & LOPA, Emergency Preparedness & Response, Hazard & Risk Register and 38 more modules in one workspace.
Method note
A checklist, or an open room? Both schools are right.
Prompts make sure nothing obvious is skipped. Open discussion finds what no list predicted.
Leandro · Founder, UProcess
One school runs HAZID from a fixed list of prompts, so every node is asked the same questions and nothing depends on who turned up. The other keeps the list short and lets experienced people talk, because the hazard that hurts is often the one nobody wrote a prompt for.
So the guide word list is yours, and the room can add to it.
- Company listYour guide words on every node, each answered present, not present or not applicable.
- Added in the roomA prompt typed during the session joins your company’s list for the next study.
- Free scenariosWhere a prompt is present, the team writes the event, causes and consequences in its own words.
Where Process Hazard Identification usually breaks
Four places an early study goes wrong, and what the record does about each.
HAZOP is asked to do the screening too.
Practitioners report HAZOP used as the only hazard identification method, and HAZOPs run on concepts or incomplete designs, where intent is unclear and the actions come out trivial.
Source: Whitty and Foord, IChemE Hazards XXI
In UProcess HAZID screens each node first, against your own guide words, before anyone works deviation by deviation. A HAZOP study that continues a HAZID shows where it came from.
HAZOP-0007 · from the HAZID
- Study
- HAZOP-0007 continues HAZID-0003
The HAZID is not changed. Deviations, causes and safeguards are worked in the HAZOP.
Two people, two ratings for the same risk.
Risk matrices can give very different risks the same rating, and with subjective severity categories, different users can rate the same risk in opposite ways.
Source: Cox, Risk Analysis 28(2), 2008
In UProcess Each study is rated on one published matrix version, chosen for the study. Every consequence carries its initial and residual risk on that version.
Corporate 5×5 · published 12 Aug 2026
Initial · S5·L4 · HighResidual · S5·L2 · Medium
A safeguard on the list is taken as working.
OSHA staff warn that a listed safeguard counts only if it is tested and the test is effective, citing a release where valves were tested daily but never confirmed to close.
Source: Lay et al., AIChE Global Congress on Process Safety 2012
In UProcess HAZID records safeguards per consequence, numbered, existing or recommended, and takes no numeric credit for them. Credit, with independence and PFD, waits for LOPA in the HAZOP.
Safeguards · Overpressure of V-201
The recommendation leaves with the facilitator.
A review of common HAZOP problems found recommendations often not closed out properly, and external study leaders usually unable to take responsibility for the close-out.
Source: Munn, IChemE Hazards XXI
In UProcess Each recommendation names a colleague or an external person. Sent to Action Tracking, it becomes one action in the study’s plan, and the study counts those with no owner or not yet sent.
Action Tracking · AT-0139
Study V-201 overpressure in HAZOP, with LOPA
- Owner
- Sarah Mitchell
- Due
- 12 Oct 2026
- From
- Process Hazard Identification · R4
HAZOP-0007 node 2 analysed with linked LOPA · checker David Chen
Session 18 Sep 2026Scope Nodes 1–3Facilitator Emma Clarke
| Team member | 16 Sep | 18 Sep | Total |
|---|---|---|---|
Emma Clarke | 2 | ||
Sarah Mitchell | 2 | ||
Ana Ribeiro | 1 | ||
Mark HughesExternal participant | 1 |
The record an auditor asks for.
A screening study is evidence only if it shows who looked, and at what.
- The team: colleagues and external participants by name, facilitator marked.
- Every session with its date and scope, and attendance by session.
- Each guide word answered on each node, with the team’s comment.
- The documents the team worked from, kept as files with the study.
What’s in a study
- Structure
- Study → nodes → guide words → findings. A present finding opens scenarios, each with causes and consequences.
- Risk
- Initial and residual risk on each consequence, rated on the study’s published matrix version.
- Safeguards
- Recorded per consequence, numbered, existing or recommended. No numeric credit.
- Recommendations
- Numbered per study, owned by a colleague or an external name, linked to the consequences they answer, sent to Action Tracking.
- Team
- Facilitator and participants, sessions with date and scope, attendance per session, and study documents with their files.
Questions
Is HAZID a lighter HAZOP?
No. HAZID screens a node against prompts; HAZOP works deviation by deviation. A HAZOP study that continues a HAZID shows where it came from.
Can we use our own guide words?
Yes. The list belongs to your company, and a prompt added in a session stays for the next study.
Does HAZID take credit for safeguards?
No. It records them. Credit, with independence and PFD, happens in LOPA.
What happens if we send recommendations twice?
Nothing new is created. Each recommendation keeps its one action in the study’s plan.