A tight-looking flange is not evidence that its gasket, bolts and temporary supply train are suitable. This revision separates purging, pressure testing, leak location and return to service so that each change of state has a defined boundary and release condition.
Control stored pressure, gas discharge and the evidence needed to proceed
This is reference content written and revised with AI. It has not been verified or approved by a professional. Review by a suitably qualified professional and site authorisation are required before use.
Stop work
Stop gas admission from a safe position if pressure behaves unexpectedly, the boundary is entered, a joint moves, a leak is suspected or atmosphere monitoring fails or alarms. Keep people out, use the planned controlled vent route and re-establish safe conditions before investigation. Do not tighten a leaking joint under pressure.
1. Legal and technical scope
| Basis | Application limits |
|---|---|
| Safe Work Australia — confined spaces model code | Official model-code portal and jurisdiction qualification reviewed; the linked full model-code PDF was not reviewed. |
| Safe Work Australia — exposure guidance, July 2025 | Section 5.8 simple-asphyxiant passage reviewed through official PDF search extraction; no claim of full-document audit. |
| HSE GS4 — Safety in pressure testing | Fourth edition, 2012, all nine pages reviewed: pressure-test planning, compatible components, protection, controlled release and safe inspection. Legal references apply in Great Britain; elsewhere this is foreign technical guidance. |
| US CSB — Hazards of nitrogen asphyxiation | Official investigation overview and selected sections of the 2003 bulletin reviewed: oxygen displacement, outdoor exposure, monitoring and rescue hazards. CSB guidance is not a test code. |
| Dixon — King Safety Cables | Manufacturer instructions reviewed for this product family. Its pressure limits and fitting method must not be transferred to a different restraint. Actual site equipment manuals were not supplied. |
Australia: determine the state or territory and applicable WHS/OHS law. Safe Work Australia model guidance is not an identical enforceable rule in every jurisdiction.
The official confined-spaces model-code page, updated 14 November 2024, requires checking local legal effect. The July 2025 exposure-guidance section on simple asphyxiants identifies 19.5% oxygen as a minimum at normal atmospheric pressure, with specialist assessment at different pressures. This is not a recommended alarm setting at that boundary or permission to enter a nitrogen-purged pipe. The complete exposure guide, local confined-space and construction provisions, and AS/NZS and ASME test standards were not audited. Specify breathing-zone controls separately from the process purge endpoint.
2. Scope and required inputs
Included scope
Planned inert-gas purging and pneumatic leak testing of an identified piping circuit by competent personnel under an approved procedure. Includes temporary connections, isolation, exclusion, atmosphere control, staged testing and restoration.
Separate review
Entry into an inerted system, live leak repair, strength-test design, oxygen service, cryogenic liquid transfer and emergency rescue require separate engineered arrangements. This sheet does not authorise replacing a hydrostatic test with a pneumatic test.
| Required input | What to verify |
|---|---|
| Purpose | Distinguish inerting endpoint, leak tightness and strength proof; identify applicable code and why gas is justified. |
| Boundary | Marked P&ID, all branches and dead legs, blinds, valve states, removed instruments and limits of temporary equipment. |
| Supply | Gas identity and SDS, inlet pressure, regulator, independent overpressure protection, rated fittings, restraints and safe discharge. |
| Evidence | Calibrated instruments, sampling locations and delays, atmosphere limits, barriers, remote controls, communication and emergency arrangements. |
The design code, actual pressure and volume, materials, gas supply, relief capacity and dispersion study are not supplied. The responsible specialist must set the medium, temperature range, pressure stages, hold times, acceptance criteria, inspection conditions and exclusion boundary. No fixed 15 m radius, 10-minute hold, Schedule 80 pipe or 1.5-times component rating is a universal prescription.
3. Eight-stage JSA
| Job stage | Hazards and effects | Failure factors | Technical and procedural controls | Site verification | P | S | R | A |
|---|---|---|---|---|---|---|---|---|
| 1. Define the test | 1. Burst2. Wrong gas3. Unplanned entry4. Wrong acceptance | 1. Purpose confused2. Volume unknown3. Code unconfirmed4. Crew unbriefed | 1. Confirm test purpose2. Justify gas use3. Set procedure limits4. Brief the team | 1. Test pack revision2. Pressure and volume3. Competent authority4. Acceptance criteria | ||||
| 2. Protect the area | 1. Fragments strike2. Oxygen deficiency3. Fall from height4. Tools fall | 1. Zone too small2. Vent near people3. Access unprotected4. Rescue unplanned | 1. Assess exclusion zone2. Route discharge safely3. Provide protected access4. Prepare rescue response | 1. Zone and controls2. Vent destination3. Access protection4. Atmosphere plan | ||||
| 3. Assemble supply | 1. Hose whips2. Gasket fails3. Overpressure4. Cylinder falls | 1. Parts mismatched2. Seal unverified3. Relief missing4. Source unsecured | 1. Check ratings and seals2. Fit suitable restraints3. Verify regulator and relief4. Secure gas source | 1. Ratings at temperature2. Connections and locks3. Relief path and capacity4. Supply identity | ||||
| 4. Isolate circuit | 1. Backflow2. Unexpected pressure3. Blind ejects4. Pocket trapped | 1. Branch missed2. Valve passes3. Blind underrated4. Gauge isolated | 1. Walk down the circuit2. Fit planned isolation3. Secure valve states4. Identify every pocket | 1. Marked boundaries2. Isolation evidence3. Blind specification4. Vent and gauge paths | ||||
| 5. Purge and sample | 1. Asphyxiation2. Flammable discharge3. Overpressure4. False endpoint | 1. Exhaust obstructed2. Wrong sample point3. Dead leg remains4. Sensor unsuitable | 1. Control purge flow2. Keep outlet safe3. Sample defined points4. Verify analyzer method | 1. Discharge condition2. Sample locations3. Composition record4. Analyzer response | ||||
| 6. Raise and hold | 1. Boundary bursts2. Fragment impact3. Cooling damage4. Limit exceeded | 1. Pressure raised fast2. Temperature ignored3. Control fails4. Zone entered | 1. Raise by approved stages2. Monitor remotely3. Maintain exclusion4. Stop on deviation | 1. Actual pressure2. Temperature and time3. Relief available4. Stage release | ||||
| 7. Locate leakage | 1. Gas jet strikes2. Oxygen falls3. Joint fails4. Unsafe repair | 1. Close approach2. Method unsuitable3. Leak persists4. Bolts tightened | 1. Use protected method2. Keep atmosphere watch3. Stop for suspected leak4. Depressurise for repair | 1. Inspection conditions2. Detector suitability3. Leak record4. Safe repair state | ||||
| 8. Vent and restore | 1. Trapped gas ejects2. Vent line recoils3. Asphyxiation4. Wrong reconnection | 1. Gauge reads falsely2. Outlet restricted3. Gas route unsafe4. Blind left fitted | 1. Isolate gas source2. Vent each trapped part3. Prove safe zero pressure4. Restore and hand over | 1. Verified vent paths2. Atmosphere safe3. Restored configuration4. Receiver and limits |
PDF A covers basic task controls; B focuses on technical failures; C covers supervision, records and handover. The article has eight stages with four items per main cell; each PDF has six key stages with three items per main cell. P/S/R/A, measurements, confirmations and signatures remain blank for actual site assessment.
4. Fictional scenario and failure mechanisms
The original educational image is retained. It is not evidence of an actual incident or site suitability.
The scenario and record comparisons are fictional teaching examples, not records of real incidents, measurements, inspections, signatures or approvals.
| Item | Fictional scenario |
|---|---|
| Initial situation | Teaching example: a temporary manifold is assembled and its bolt marks appear complete. |
| Changed condition | The gasket identification and the pressure rating at the test temperature cannot be reconciled with the test pack. |
| Failure path | Gas admission is held. Tight bolts and a past successful test do not resolve the missing material evidence. |
| Possible outcome | The team checks the installed parts, replaces incompatible items after safe isolation, and rechecks relief, boundaries and discharge routing. |
| Prevention barrier | Only the authorised procedure can release the next stage. Purge composition, pressure-test outcome and safe restoration are recorded separately. |
Failure mechanisms
Pressure acting across an effective area produces a separating force. Gasket performance also depends on flange geometry, bolt preload, material, assembly and temperature; a torque mark is not a material certificate. A regulator controls pressure but is not automatically the independent relief protection needed if it fails.
Compressed gas can release substantial expansion energy. The expression P·V/(γ−1) describes ideal-gas internal energy under assumptions; it is not by itself the available blast work relative to ambient pressure or a formula for the safe distance. A competent assessment uses absolute pressures, volume, expansion assumptions and credible fragment paths. Low-pressure inspection still requires protection.
Purging displaces the original gas and can create a hazardous exhaust mixture. Specify a representative sampling strategy, including dead legs, analyzer suitability and response time. A combustible-gas sensor may not read correctly in oxygen-deficient gas. The required oxygen level inside an inerted pipe is different from the acceptable atmosphere around people; neither proves the other.
Nitrogen can create oxygen deficiency outdoors as well as in enclosed areas. Temperature, release momentum, ventilation and geometry determine dispersion; monitoring only low points is insufficient. Pressure change during a hold can reflect temperature as well as leakage. A zero gauge reading can conceal blocked sensing lines or isolated pockets, so every relevant volume needs a proven safe vent path before opening.
Decision sequence
| Decision stage | What to verify |
|---|---|
| Known facts | Identify the test purpose, circuit revision, gas inventory and people exposed. |
| Unresolved matters | Separate observed bolt condition from unknown gasket rating, trapped volumes and exhaust destination. |
| Reason to hold | Hold admission because a pressure boundary and safe atmosphere are not yet demonstrated. |
| Required evidence | Obtain component evidence, isolation checks, relief verification, sampling plan and exclusion assessment. |
| Restart decision | Release each stage separately; repeat affected checks after changes and record actual results before restoration. |
5. Conditional decisions and common errors
| Condition | Action | Evidence | Proceed when |
|---|---|---|---|
| Unknown gasket | Hold supply; identify or replace | Material, rating and assembly record | Parts match the test |
| Purge endpoint | Sample designated locations | Composition and analyzer response | Endpoint demonstrated |
| Unexpected pressure | Stop and investigate safely | Pressure, temperature and valve state | Cause resolved |
| Opening the line | Isolate and prove depressurisation | All pockets vented; safe atmosphere | Zero pressure proven |
Similar situations, different decisions
The scenario and record comparisons are fictional teaching examples, not records of real incidents, measurements, inspections, signatures or approvals.
| Comparison | Situation A | Situation B |
|---|---|---|
| Pressure stable | A stable display is treated as proof of no leaks. | Temperature, resolution, duration and acceptance criteria are considered. |
| Oxygen acceptable | One outdoor reading is used to approve the inerted pipe for entry. | Breathing-zone checks and process-gas endpoints serve different decisions. |
| Gauge at zero | Bolts are loosened to see whether gas remains. | Isolation and effective venting of every relevant pocket are verified before opening. |
6. Roles and responsibilities
| Role | Responsibility |
|---|---|
| Test authority | Select the applicable procedure, justify test medium and set limits and acceptance criteria. |
| Test supervisor | Verify boundaries, briefing, access control and stage releases; stop on deviations. |
| Assembler | Check compatible parts, seals, connections, supports and restraints without pressure present. |
| Test operator | Control admission and discharge from a protected location and record pressure and temperature. |
| Atmosphere / handover lead | Check instruments and representative atmosphere locations, coordinate rescue readiness and record restoration. |
Supervision, records, handover
| Job stage | Technical and procedural controls | Site verification |
|---|---|---|
| Accept procedure | 1. Confirm test scope2. Review test medium3. Set release points | 1. Approved procedure2. Circuit volume3. Acceptance basis |
| Accept boundary | 1. Walk down boundary2. Review isolation3. Identify pockets | 1. Circuit revision2. Blind register3. Vent locations |
| Accept test train | 1. Verify component data2. Confirm relief duty3. Review restraints | 1. Ratings and seals2. Relief evidence3. Supply setup |
| Control purge | 1. Review discharge route2. Check sampling plan3. Confirm endpoint | 1. Gas composition2. Analyzer checks3. Atmosphere records |
| Control testing | 1. Release each stage2. Keep people clear3. Stop deviations | 1. Pressure and time2. Temperature record3. Inspection method |
| Accept handover | 1. Prove safe pressure2. Check restoration3. Record restrictions | 1. Pocket verification2. Blind and gauge status3. Receiving person |
7. Hold points and record comparison
| Hold point | Evidence before release | Confirmation / approval |
|---|---|---|
| Start | Planned inert-gas purging and pneumatic leak testing of an identified piping circuit by competent personnel under an approved procedure. Includes temporary connections, isolation, exclusion, atmosphere control, staged testing and restoration. | |
| HP-1 Isolate circuit | Marked boundaries · Isolation evidence · Blind specification · Vent and gauge paths | |
| HP-2 Purge and sample | Discharge condition · Sample locations · Composition record · Analyzer response | |
| HP-3 Raise and hold | Actual pressure · Temperature and time · Relief available · Stage release | |
| HP-4 Vent and restore | Verified vent paths · Atmosphere safe · Restored configuration · Receiver and limits |
Complete after actual checks. Recheck after changes, defects or shift handover.
Record comparison examples
The scenario and record comparisons are fictional teaching examples, not records of real incidents, measurements, inspections, signatures or approvals.
The boxes are static record prompts, not interactive or saved approvals. A tick does not establish suitability or permission to restart. Record the evidence, person, time and unresolved issues; retain the change history.
| Record type | Insufficient record | Record checklist |
|---|---|---|
| Boundary | Valves checked | □ Circuit revision□ Blind and valve IDs□ Trapped sections□ Isolation verifier |
| Temporary train | Manifold ready | □ Component ratings□ Gasket identity□ Regulator and relief□ Hose restraints |
| Purge and test | Test passed | □ Sampling locations□ Actual composition□ Pressure and temperature□ Acceptance decision |
| Restoration | Zero pressure | □ Supply isolated□ Vent paths proven□ Blinds and instruments restored□ Receiver and restrictions |
8. Site checklist and emergency response
| Check item | Result | Person / action |
|---|---|---|
| □ Define the test: Test pack revision | ||
| □ Protect the area: Zone and controls | ||
| □ Assemble supply: Ratings at temperature | ||
| □ Isolate circuit: Marked boundaries | ||
| □ Purge and sample: Discharge condition | ||
| □ Raise and hold: Actual pressure | ||
| □ Locate leakage: Inspection conditions | ||
| □ Vent and restore: Verified vent paths | ||
| □ Release / handover authority |
Stop work
Stop gas admission from a safe position if pressure behaves unexpectedly, the boundary is entered, a joint moves, a leak is suspected or atmosphere monitoring fails or alarms. Keep people out, use the planned controlled vent route and re-establish safe conditions before investigation. Do not tighten a leaking joint under pressure.
Emergency response
Raise the alarm and withdraw to a verified safe atmosphere. Isolate supply remotely only when safe. Prevent unprotected rescue or entry into the discharge plume; trained responders need the specified breathing supply and rescue plan. A filter respirator does not supply missing oxygen.
9. Sources and review limits
2026-10-08
| Source | Scope and limitations |
|---|---|
| Safe Work Australia — confined spaces model code | Official model-code portal and jurisdiction qualification reviewed; the linked full model-code PDF was not reviewed. |
| Safe Work Australia — exposure guidance, July 2025 | Section 5.8 simple-asphyxiant passage reviewed through official PDF search extraction; no claim of full-document audit. |
| HSE GS4 — Safety in pressure testing | Fourth edition, 2012, all nine pages reviewed: pressure-test planning, compatible components, protection, controlled release and safe inspection. Legal references apply in Great Britain; elsewhere this is foreign technical guidance. |
| US CSB — Hazards of nitrogen asphyxiation | Official investigation overview and selected sections of the 2003 bulletin reviewed: oxygen displacement, outdoor exposure, monitoring and rescue hazards. CSB guidance is not a test code. |
| Dixon — King Safety Cables | Manufacturer instructions reviewed for this product family. Its pressure limits and fitting method must not be transferred to a different restraint. Actual site equipment manuals were not supplied. |
This is reference content written and revised with AI. It has not been verified or approved by a professional. Review by a suitably qualified professional and site authorisation are required before use.
10. PDF guidance and FAQ
PDF A covers basic task controls; B focuses on technical failures; C covers supervision, records and handover. The article has eight stages with four items per main cell; each PDF has six key stages with three items per main cell. P/S/R/A, measurements, confirmations and signatures remain blank for actual site assessment.
Content group ID
jsa-leak-testing
Does a leak test prove strength?
Its purpose and acceptance criteria depend on the governing procedure. Purging and leak location do not automatically replace the required strength test.
Is a whipcheck enough protection?
No. The selected device must suit the hose and application; it does not stop gas flow or replace exclusion and engineered shielding.
Can an oxygen monitor find a nitrogen leak?
It assesses oxygen deficiency where sampled. It does not identify every leak or prove a pressure boundary is tight. Use a method suitable for the test gas and task.
May a leaking flange be tightened?
Stop, isolate and depressurise safely first. Investigate the cause, repair under the applicable procedure and repeat the required checks.
Has a professional verified or approved this document?
No. Public sources and editorial, layout and consistency checks were reviewed using AI. No site inspection or professional technical approval has been performed. Obtain appropriate professional review before use.
11. Revision history and authorship
This is reference content written and revised with AI. It has not been verified or approved by a professional. Review by a suitably qualified professional and site authorisation are required before use.
| Date | Version | Change | Review status |
|---|---|---|---|
| 2026-07-06 | Original version unmarked | Original preserved | No professional approval |
| 2026-10-08 | 1.1 | Decisions, records, JSA and PDFs revised | No professional approval |
The original universal pressure factors, fixed holds, exclusion radius, scaffold tag/harness rules and automatic compliance claims are withdrawn. The incident narrative is replaced with an explicitly fictional decision example. Zero gauge pressure is distinguished from absolute zero; loosening bolts is not a venting method. Gas-energy and hypoxia explanations are corrected without inventing measurements or expert approval.