Why process safety needs more than a risk register
Occupational safety deals with slips, trips and dropped objects. Process safety deals with loss of containment, fires, explosions and toxic releases, events that are rare but can be catastrophic. A simple risk register is not enough for these hazards. You need to know which barriers stand between each threat and the top event, whether those barriers are working, how much risk reduction each one really gives, and whether a change to the plant has quietly removed one of them.
In many facilities that knowledge is spread across HAZOP spreadsheets, LOPA workbooks, BowTie drawings in separate tools and a paper change control log. Studies are done once, filed, and not revalidated on time.
What good process safety management looks like
- Hazard studies such as HAZOP, PHA, BowTie and LOPA are kept as live documents with owners, revisions and review dates.
- Barriers are named, owned, verified at a defined frequency and given an honest status.
- Risk reduction claims are calculated openly: initiating frequency, protection layers, conditional modifiers and a target.
- Changes to plant, process or people go through a formal management of change review before they are made.
- Degraded barriers and overdue revalidations are visible on one dashboard.
Six study types in one workspace
The Process Safety module in SafetyPro holds six study types: BowTie Analysis, Process Hazard Analysis (PHA), HAZOP, Risk Assessment, Job Safety Analysis (JSA) and LOPA/SIL Analysis. Each study has a code, title, study leader, facility or area, methodology, scope and next review date. Studies move through Draft, In Progress, Under Review, Approved, Closed and Revalidation Due, and carry a revision number. A study can be duplicated as the starting point for a similar unit, and each worksheet has a print view for sign-off and distribution.
Behind the studies sit shared libraries: a Barrier Library and eight reference libraries for risk assessment items, hazards, HAZOP guide words, precautions and controls, scenarios, consequences, threats and independent protection layers (IPLs). Library items can be brought into studies, and libraries can be exported to PDF or Word.
BowTie analysis and barrier integrity
The BowTie builder draws the hazard, the top event, the threats on the left and the consequences on the right, with barriers placed on each pathway. Every barrier records:
- Type: preventive, mitigative, detective or recovery.
- Effectiveness: High, Medium, Low or Failed.
- Status: Active, Degraded, Failed, Not Verified or Planned.
- Owner, verification frequency, category and description.
- A Critical Control flag for the barriers you cannot afford to lose.
Barriers are colour coded on the diagram by status and effectiveness, so a failed or degraded barrier stands out at a glance. Each barrier can also carry escalation factors, the conditions that could defeat it such as a bypassed alarm or a missed test, with their own controls. Barriers can be picked from the Barrier Library, so the same pressure relief valve or gas detection system is described the same way across studies.
The process safety dashboard brings the main warning signs together: total studies by type and by status, open and overdue process safety actions, degraded barriers (those marked Degraded or Failed) against the total number of critical barriers, and studies due for revalidation, either flagged Revalidation Due or past their next review date. Lists show the most recent studies, overdue actions and the degraded or critical barriers with their owners.
LOPA and SIL determination, calculated openly
The LOPA/SIL worksheet follows the familiar layer of protection analysis method. For each scenario you enter the initiating event and its yearly frequency, the independent protection layers and their probability of failure on demand (PFD), and any conditional modifiers. SafetyPro then calculates:
Mitigated frequency = initiating event frequency x product of IPL PFDs x product of conditional modifiers
The mitigated frequency is compared with a target frequency set by consequence category. The default criteria are 1e-5 per year for fatality, 1e-4 for serious injury, 1e-3 for major environmental and 1e-2 for major financial consequences. Each scenario shows PASS or FAIL against its target, the risk reduction factor (RRF) and the matching SIL band, from No SIS Required through SIL 1 to SIL 4. A summary shows how many scenarios meet target and how many need a safety instrumented function.
Eight IPL categories come with default PFD values, for example 0.1 for basic process control and for alarm with operator response, and 0.01 for a safety instrumented function, a relief device or a physical barrier such as a dike. Five conditional modifiers are available, including probability of ignition, occupancy and fatality. Each IPL can be marked as independent, given an audit or test frequency from monthly to every five years, and linked to a barrier in the Barrier Library or on a BowTie diagram. That link is what turns a number in a spreadsheet into a named, owned barrier whose status you can follow.
HAZOP, PHA, risk assessment and JSA worksheets
The HAZOP worksheet is organised by node and deviation, with eleven default guide words: No or Not, More, Less, As Well As, Part Of, Reverse, Other Than, Early, Late, Before and After. PHA items are grouped into Process, Equipment, Human, Environmental and External categories. Risk assessments use a 5 by 5 likelihood and consequence matrix with levels from Very Low to Extreme.
The JSA worksheet breaks a job into steps with hazards, controls, PPE, tools and required permits, and records the permits needed for the whole job, such as hot work or confined space entry. These permits are recorded as text in the JSA. The permits themselves are raised and approved in the Permit to Work module.
AI assistance that you review before it counts
Every study has an AI panel with quick actions tailored to the study type. On a BowTie you can ask it to suggest threats, consequences or barriers, run a gap analysis that looks for unprotected pathways, single-barrier reliance, missing escalation factors, degraded barriers and missing critical controls, or run a completeness review that returns a completeness score. HAZOP quick actions suggest deviations and check guide word coverage. LOPA quick actions suggest IPLs, review SIL determinations and check LOPA quality against IEC 61511 and CCPS practice.
AI output never goes straight into the study. Suggested items appear as a list with checkboxes. You choose which ones to apply, and only those are added. Inside the barrier form, AI Suggest can also propose a description, type, effectiveness, category and verification frequency for a barrier you are adding.
Management of change and PSM audits
Management of change sits in the QHSE module, on its MOC tab. Each change is recorded as Temporary or Permanent, with a category (Equipment, Process, Personnel, Material, Electrical, Structure, Environmental or Other), a risk level, affected personnel, and required descriptions of the hazards and control measures. Changes follow a controlled workflow: Under Review, Reviewed, Approved, In Progress for implementation and Closed, with the option to reject. Creating and approving changes are separate permissions, Manage MOC and Approve MOC, and each record keeps its approval history.
For audits, the inspection template library includes a Process Safety Management (PSM) Audit template referencing OSHA 1910.119. It covers process safety information, operating procedures, management of change and incident investigation, and runs like any other inspection.
Safeguards and honest limits
- Barrier status and effectiveness are set by people. SafetyPro does not read barrier health from plant sensors or the control system.
- LOPA results depend on the frequencies, PFDs and credits you enter. The worksheet makes the calculation transparent, but a competent person must validate it.
- AI suggestions are drafts. Nothing is added to a study until a user selects and applies it.
- MOC records, JSA permit fields and studies are linked by people and process, not by automatic cross-references.
- Viewing, editing and deleting studies follow separate process safety permissions.
A practical rollout plan
- Load your critical barriers into the Barrier Library with owners and verification frequencies.
- Rebuild your top major accident hazards as BowTies and set each barrier status honestly.
- Move existing LOPA scenarios into the LOPA/SIL worksheet and link each IPL to its barrier.
- Set a next review date on every study so revalidation shows on the dashboard.
- Route plant and process changes through the MOC workflow and schedule a PSM audit.
- Review degraded barriers and overdue actions at every process safety meeting.