Incident & loss investigation
Every kind of loss. One standard of fact-finding.
People, assets, operations, information and reputation.
All investigationsPeople & health
Serious Incident & HiPOActual or potential major consequenceInjury & FatalitySequence, causes and preventionOccupational IllnessExposure, health effects and controlsPatient Safety ReviewIndependent, non-adversarial clinical reviewHarassment & Workplace ConductFair, confidential HR investigationTransport & projects
Road & FleetVehicle, driver and routeMarine & PortVessels, terminals and cargoConstruction & ContractorMulti-party site eventsProject Incident ReviewOne account from every partySecurity & Cargo LossTheft, robbery and hijackingAssets & operations
Fire & ExplosionOrigin, spread and conditionsIndustrial & ProcessReleases, upsets and start-upsMachinery & EquipmentFailure, guarding and maintenanceElectrical IncidentsShock, arc flash and isolationEnvironmental ReleaseSpills, effluent and emissionsProperty & Asset DamageDamage and business interruptionInformation, reputation & time
Information LossData, records and system failuresReputation LossWhat caused the damage to trustProcess LossYield, waste and reworkTime LossNon-productive time and delayProduct ContaminationFood, pharma and packagingRegulatory & Licence BreachFindings, sanctions and scrutinyBeyond accidents
Problems repeat when the real cause stays hidden.
Independent fact-finding and root cause analysis for customer, process, quality and operational breakdowns.
All business investigationsAnalytical methods
The event chooses the method — not the method the event.
We do not force every event through one RCA tool. Complexity and evidence determine the approach.
Explore methodsOur approach
The P3i Causation ModelFrom loss to control gapRoot Cause AnalysisFrom symptom to preventionMethod SelectionChoose the analysis to fit the caseGathering the facts
Evidence & Scene ManagementPreservation, documentation and custodyInvestigative InterviewingStructured, non-accusatory fact-findingSector knowledge
Local operating reality. Global discipline.
Teams and methods selected to fit the sector, event and decision required.
All industriesEnergy & process
Oil, Gas & EnergyChemicals, Petrochemicals & FertilizersCement, Steel & Heavy IndustryManufacturing & food
ManufacturingTextiles & ApparelPharmaceuticalsFood, Beverage & FMCGSugar Mills & Agro-ProcessingBuild investigators
Practical capability, not classroom theory.
Nationals of more than 100 countries trained; seminars led in over 20 countries.
Explore trainingInvestigator pathway
Introduction to Incident Investigation 1dIncident Investigation Techniques 2dIncident Investigation & Root Cause Analysis 5dLead Investigator 5dRCA Specialist & Facilitator 5dInvestigation for Leaders 1dSkills & reporting
Investigative Interviewing Skills 2dHuman Factors in Accident Investigation 2dEvidence Collection & Scene Management 2dHiPO Investigation 2dBarrier-Based Root Cause Analysis 2dInvestigation Report Formulation & Writing 2dPresenting RCA Findings to Management 1dCorrective Action Plans That Work 1dFor insurers & risk teams
Independent cause and circumstance support.
Technical fact-finding for complex losses without stepping into policy interpretation or adjusting decisions.
Insurance investigation servicesPOPULAR SEARCHES
Tip: press Ctrl + K or / anywhere to search.
No pages match “”.
Try a simpler word such as fire, road, training or insurance, or ask an investigator directly.
Home / Investigations / Industrial & Process
Process incidents rarely start with a failure. They start with a change.
A gasket swapped during a shutdown. A bypass fitted to keep production running. A detector left awaiting repair. We investigate releases, runaway reactions, utility failures and start-up events by testing every barrier that should have stopped them, and by finding the change that nobody reviewed.
Since the 1990s
30
years investigating
250+
investigations closed, incl. HiPOs
15
countries served
20+
countries of lead seminars
100+
nationalities trained
Switch the barriers off
A flange leaks gas during start-up. Tap each barrier to fail it, and see how the outcome changes. In real incidents, each of these failures is usually already in place before the day.
Barriers failed: 0 of 6
Contained
Every barrier held. Nobody notices anything happened.
An illustration of the layered-barrier model, often called the "Swiss cheese" model after the work of psychologist James Reason.
Had a release or an upset? A first call costs nothing, and tells you which logs and parts to secure today.
The most dangerous hour
Industry studies consistently find that a disproportionate share of process incidents happen during start-up, shutdown and recovery from upsets. These are the times procedures are used least often, equipment has just been opened, and more people and contractors are on the unit.
In Pakistan, power interruptions and changes in gas pressure mean some plants restart many more times a year than their designers ever planned for. Each restart is a non-routine operation, and each one carries that extra risk.
Three disasters, three small decisions
Overseas — chemical plant
A cracked reactor was removed and a temporary bypass pipe fitted to keep the plant running, without a proper engineering design. It failed, and the vapour cloud explosion killed 28 people. It is the reason "management of change" exists.
Overseas — offshore platform
A pump's safety valve had been removed for maintenance. The night shift did not know, and started the pump. The release and fires that followed killed 167 people. The paperwork existed; the handover did not.
South Asia — pesticide plant
Several of the plant's safety systems were out of service or unable to cope when a toxic gas release began. Thousands died. It remains the worst industrial disaster in history, in our own region.
A bypass, a handover, a system left switched off. Each one looked reasonable at the time, and each one is still found in plants today.
What we find in Pakistani plants
You commission one investigation. You get a review of the whole system.
If one barrier failed here, the same barrier is relied on in other units. After the cause is established, our process specialists check those barriers plant-wide and give you a short list of targeted actions, ranked by risk, instead of a generic register.
The event
A gas leak at a flange during start-up
No ignition. First recorded as "gasket failure — replaced".
Change control
No change review existed for "like-for-like" part swaps.
Stores
Gaskets of three different ratings sat in one unlabelled bin.
Instruments
Eleven detectors across the plant were bypassed or awaiting repair.
Start-up
The start-up procedure had no leak-test step at all.
Permits
Hot work and start-up permits were issued by different offices.
Staffing
Night shift covered two units with one operator during restarts.
A typical example, based on what we see across many process incident investigations. Not a single client’s case.
The P3i approach
Every P3i investigation is led by a senior investigator and supported by internationally qualified process specialists. You will not receive a list of people to blame, or a long list of generic actions. You will receive the facts, the proven causes, and a short plan that closes the gaps, in this case and across your system.
Independent
No products, repairs or contracts to sell you afterwards.
Senior-led
Thirty years of investigation, from the first call to the final briefing.
Beyond the event
Process specialists find the other cracks before they fail.
Ready to present
Written for boards, buyers, insurers and regulators.
How P3i does it for you
Secure trends and logs
Control-system data, alarms and the failed part.
Rebuild the upset
What the process did, minute by minute.
Test the barriers
Trips, alarms, relief and procedures: which held, which did not.
Find the unreviewed change
Settings, materials, staffing or equipment changed without review.
Fix process safety across units
Our specialists check other units for the same gaps.
What you receive
Call us when
Free 30-minute case assessment.
Tell us what happened. In confidence, we will tell you what the investigation should cover, how long it will take and what to preserve today.
How we work with you
After a process incident, the first call is about securing trends, logs and the failed part. From there, this is the usual pattern.
First hour
A call with a lead investigator. You leave it knowing what to preserve today and who to call next.
Day one
An investigator on site, anywhere in Pakistan. Scope and fixed fee agreed in writing.
Week one
A private briefing on early findings, so urgent risks are dealt with before the report is finished.
Weeks three to four
The final report, the wider system review and a remedial work plan, presented to your team.
After
We come back to check the actions were completed and actually worked.
Typical timings for a single-site event. Complex, multi-party matters take longer, and we tell you so on the first call.
What you receive
The summary is what leadership reads first. Behind it sits the complete file: every witness heard, every document and record examined, and every finding tied to numbered evidence, so a board, an insurer, a buyer or a regulator can test it.
1 · Established sequence
2 · Causes
3 · Wider system findings
4 · Remedial work plan
Illustrative summary page. Report length and structure depend on the scope of each case.
Inside the full report
Every action has an owner, a date and a test of whether it worked. The evidence stays indexed, so any reader can check how each conclusion was reached.
From our files
See all 16 cases → Client names, people, locations and dates removed.
Questions after a process incident
Any unplanned release of hazardous material or energy from a process: gas, vapour, chemical, steam or pressure, whether or not anyone is hurt. These events are rarer than personal injuries but carry far higher potential consequence.
Start-up and shutdown are non-routine. Procedures are used less often, equipment has just been opened, and more people are on the unit. Frequent power and gas interruptions mean some plants restart far more often than their design assumed.
It is the review of any change to equipment, materials, procedures or staffing before it is made. Many major process disasters trace back to a change that looked too small to review.
Yes. Where you already use an established process safety management system or incident method, we work within it and report against its elements.
No. Our process specialists check the same barriers across other units and give you targeted actions, ranked by risk, for the gaps that would cause the next release.
Yes. P3i has worked in 15 countries, including on process and energy facilities.
Had a release or an upset?