What this is
What is a repeat failure review?
What is a repeat failure review?
A structured examination of an asset that has failed the same way more than once. Its subject is the sequence, not the single event. It gathers every prior occurrence, tests whether the interval between them is shortening, and asks whether the earlier repairs addressed the cause or only the symptom.
What counts as a repeat failure rather than two separate failures?
Two failures are the same failure when they share the asset and the failure mode code, not merely the component. A pump seizing twice for unrelated reasons is two failures; a pump seizing twice from contamination in the seal is one unresolved problem surfacing twice. Consistent mode coding is what makes the distinction possible.
How does this differ from a root cause analysis?
The review decides whether an investigation is warranted and at what depth; the analysis performs it. The review is short, evidence-led and raised on the second occurrence. It produces a decision — quick debrief, five why, full RCA or cross functional RCA — and hands off to an RCA record with its own ID.
Scope
When is a repeat failure review required?
This review sits between the individual failure records and the formal investigation machinery. Its job is to look across occurrences and decide what happens next — not to record one occurrence, and not to conduct the investigation it may call for.
Use this template when
- The same asset has failed with the same failure mode code for a second time, whatever the interval
- A bad actor list or MTBF review has surfaced an asset whose failures cluster around one mode
- The same component has been repaired three times and a technician has said so informally
- A temporary repair has been left in place and the asset has failed again around it
- Production has raised the same asset repeatedly as the cause of unplanned stoppages
Do not use it for
- Equipment Failure Report, which records one failure event as it happened, and is the source data this review reads across
- Failure Mode Record, which assigns the standard mode code; without it applied consistently, repeats cannot be detected at all
- Repair Record, which captures what was physically done and which parts changed on a single job
- Root Cause Analysis, which is the full investigation this review may trigger, not a substitute for it
- Bad Actor Report, which ranks assets by aggregate cost across all modes rather than examining one mode on one asset
Compliance mapping
Which ISO 14224 requirements does this satisfy?
ISO 14224 is a data standard rather than a management system standard: it defines what to collect about equipment, failures and maintenance so reliability figures mean the same thing across sites and over time.
| Clause | Requirement | Where it lands |
|---|---|---|
| ISO 14224 cl.6 | Equipment must be identified within a defined boundary and taxonomy, so a failure is attributed to the right item and comparisons are like for like. | Header |
| ISO 14224 cl.7 | Failure data must record the failure mode, the date and the maintenance activity carried out in response, as a linked set rather than free text. | Failure history |
| ISO 14224 cl.9 | Data is collected in order to be analysed, including for trends in time between failures rather than counts alone. | Pattern analysis |
| ISO 9001 cl.10.2 | Where a nonconformity recurs, the organisation must evaluate the need for action to eliminate the cause so it does not recur again or elsewhere. | Root cause |
| ISO 55001 cl.8.2 | Changes to assets or to the way they are maintained must be assessed for risk and controlled before implementation. | Root cause |
| ISO 14224 cl.7 | Maintenance data must include downtime and repair effort, so the consequence of a failure is quantifiable rather than narrative. | Cost of repetition |
What it does not cover
- The investigation itself, which this review only sizes and triggers; a cross functional RCA decision with no RCA ID against it is an unmet commitment, not a completed review.
- The corrective action, which lives in the CAPA record referenced from the cost of repetition section, with its own owner, due date and verification.
- Change control, which the design change required flag points at; the MOC record carries the technical review, and this template only records that it was raised.
- Maintenance strategy revision, which is where a confirmed pattern must land — a changed PM interval, an added monitoring route or a revised job plan.
- Failure data quality, which this review depends on entirely; if modes were coded inconsistently, the occurrences count and the pattern analysis are both fiction.
Global
Repeat Failure Review requirements by country
ISO 14224 is voluntary and carries no legal force. What gives repeat failure review teeth is a separate duty to keep work equipment in efficient working order, or to correct known deficiencies rather than repeatedly patch them.
Provision and Use of Work Equipment Regulations 1998, reg.5
Work equipment must be maintained in an efficient state, in efficient working order and in good repair, with maintenance logs kept up to date where they exist.
Efficient working order is not satisfied by a machine running between frequent identical breakdowns. A documented review with a recorded decision evidences that the duty holder recognised the pattern and acted rather than absorbed it.
OSHA Process Safety Management, 29 CFR 1910.119(j)
Mechanical integrity programmes must inspect and test covered equipment, and deficiencies outside acceptable limits must be corrected before further use or in a safe and timely manner.
Repeated identical failures on covered equipment prove a deficiency was known. Where the record shows previous repairs addressed the symptom only, an inspector has a trail showing it was recognised and not corrected.
Petroleum Safety Authority Activities Regulations s.47, with NORSOK Z-008
Maintenance programmes must be based on the consequence of failure and updated when experience data shows the assumed failure behaviour is wrong.
This is the regime ISO 14224 was written to serve. A repeat failure is experience data contradicting the interval assumed in the programme, and the expectation is that the programme changes.
How to complete it
How to complete a repeat failure review, step by step
Most of this template is transcription from work order history. Four judgement calls decide whether the record holds up a year later, and in all four the easy answer is the wrong one.
Shortening carries zero. With two or three occurrences, ordinary variation looks like a trend either way. State the actual gaps in the failure history rows; do not smooth a 90, 40, 45 day sequence into shortening because it feels worse, or into stable because the last gap grew.
The honest answer is usually no, symptom only — and the reviewer is often the person who did those repairs. A candid no scores zero and escalates, which is correct: the score measures the state of the asset, not the competence of the fitter.
Design, installation, operation, maintenance, environment or specification. This unscored field decides who owns the fix, and maintenance is the default chosen when nobody wants an argument. If a correlation question returned yes for a production change or cleaning regime, the category is operation or environment and the owner sits outside maintenance.
Set the permanent fix against accumulated downtime, repair cost and production loss across every occurrence, not one failure. Marginal is for where the fix is affordable but the loss figures are estimates nobody will defend; clearly requires numbers finance would accept.
What auditors find
Most common repeat failure review findings
These are the patterns that show up when repeat failure reviews are audited a year on, when someone asks why the asset is still failing.
| Finding | Clause | What fixes it |
|---|---|---|
| Occurrences recorded as two when work order history shows six, because only the current financial year was counted | ISO 14224 cl.9 | Set period covered explicitly and pull history across year boundaries. If CMMS history is unavailable beyond a date, say so there rather than letting the count imply a newly troublesome asset. |
| Every correlation question answered no, with no evidence any of them were tested against production or cleaning records | ISO 14224 cl.8 | Each correlation field scores two for no, so a wall of nos produces a comfortable score and no action. Require failure dates to be checked against the production change log and cleaning schedule first. |
| Investigation required set to no on an asset whose failures have caused repeated unplanned production stoppages | ISO 55001 cl.8.2 | The help text is explicit that the decision is set by potential outcome, not by what happened. Where downtime accumulates across occurrences, the potential outcome is already demonstrated and no is not defensible. |
| Investigation level set to full RCA with RCA ID blank and no RCA record ever raised | ISO 9001 cl.10.2 | RCA ID is conditional and not mandatory, so the record can be signed with the commitment unfulfilled. Raise the RCA record first, then close the review carrying its reference. |
| Design change required set to yes with no MOC ID, so the modification is later done as a maintenance job with no engineering review | ISO 55001 cl.8.2 | Treat the MOC reference as mandatory in practice whenever the flag is yes. A modification made under a work order leaves no record of who assessed its consequences. |
| Production loss value left blank because the figure sits with finance, leaving payback answered on repair cost alone | ISO 14224 cl.7 | Repair cost is almost always the smallest of the three numbers and alone makes any permanent fix look unjustified. Get the loss figure, as a defensible estimate if necessary, before answering payback. |
Case in point
Case in point: the seal that failed five times before anyone counted
A filling line rotary valve on a dairy site failed with contamination in the seal in March, then again in June, September, November and mid December. Five work orders, five seal kits, five closures marked complete, five different fitters. Each job took under three hours and none individually triggered anything. The valve never reached the bad actor list because aggregate repair cost was modest and downtime was booked against the line, not the asset.
When a review was finally raised on the sixth failure, the pattern analysis answered in ten minutes. Same component: yes. Same mode: yes. Interval: shortening. Correlates with cleaning regime: yes — the wash down had moved to a higher concentration caustic in February to shorten the cycle. The cause category was environment, not maintenance, and the fix was a seal material change costing less than two of the six repairs.
The template
The template, field by field
The form exactly as it installs. Every field, option, score and conditional rule is editable, and the links to other templates come with it.
5 sections
- Reference
- MNT-017
- Archetype
- Review
- Record ID
- RFR-2026-000
- Scoring
- Not scored
- Direction
- n/a
- Singleton
- No
- Basis
- ISO 14224
- Links
- Links Asset, Failure Modes; feeds RCA
- Tags
- Maintenance, Reliability
- Sections
- 5
- Fields
- 47
- Follow up fields
- 6
- Repeating sections
- 1
- Links out
- 6
Header
14 fieldsReview ID*
Auto sequence. Format RFR-2026-000.
The record's own ID. Other templates point at this value.
Status*
Drives who this goes to next.
- Planned2 pts
- In progress2 pts
- Complete3 pts
- Deferred0 pts
- Open0 pts
- Closed3 pts
- Overdue0 pts
Date and Time*
Raised By*
Site*
Site ID*
Format SITE-000.
Links to FDN-001 Site ID
Area
The area within the site.
Exact Location
Drop a pin for anything hard to find.
Asset*
Asset ID*
Format AST-0000.
Links to FDN-002 Asset ID
The Last Repair Treated A Symptom
A second identical failure means the first repair addressed what broke, not why it broke. This review exists to stop the third one.
Failure Mode*
Occurrences*
Period Covered*
Failure history
Repeats6 fieldsDate*
Work Order ID*
Links to MNT-002 Work Order ID
Downtime Hours
Repair Done*
Component Replaced
Time Since Previous Failure
Pattern analysis
6 fieldsInterval Between Failures*
Shortening intervals mean the underlying condition is getting worse.
- Lengthening3 pts
- Stable1 pt
- Shortening0 pts
Same Component Every Time*
- No1 pt
- Yes0 pts
Same Failure Mode Every Time*
- No1 pt
- Yes0 pts
Correlates With Production Change*
- No2 pts
- Yes0 pts
Correlates With Cleaning Regime*
Wash down chemistry and pressure destroy bearings and seals in food plants.
- No2 pts
- Yes0 pts
Correlates With A Particular Shift Or Operator*
- No2 pts
- Yes0 pts
Root cause
8 fieldsPrevious Repairs Addressed The Cause*
- Yes3 pts
- Partly1 pt
- No, symptom only0 pts
Underlying Cause Now Identified*
- Yes3 pts
- Suspected1 pt
- No0 pts
Cause Category
Design, installation, operation, maintenance, environment or specification.
Investigation Required*
Set by potential outcome, not by what actually happened.
- No3 pts
- Yes0 pts
Investigation Level
RCA ID
Format RCA-2026-00000.
Links to FDN-013 RCA ID
Design Change Required*
MOC ID
Links to FDN-020 MOC ID
Cost of repetition
13 fieldsTotal Downtime Hours*
Total Repair Cost
Production Loss Value
Cost Of Permanent Fix
Payback Justified*
Put the cost of doing nothing next to the cost of fixing it properly.
- Clearly3 pts
- Marginal1 pt
- No0 pts
Action Required*
Raise the action record, then enter its reference here.
- No2 pts
- Yes0 pts
Priority
- High0 pts
- Medium1 pt
- Low3 pts
CAPA ID
Format CAPA-2026-00000.
Links to FDN-014 CAPA ID
Action Owner
Reliability*
Signature*
Maintenance Manager*
Second Signature*
MNT-017 · record IDs look like RFR-2026-000 · Links Asset, Failure Modes; feeds RCA
Open in KnowellaRun it with agents
From a document you fill in to a programme that runs itself
The review takes an hour. What fails is everything around it — noticing the second occurrence, pulling coherent history from the CMMS, and making sure the references it commits to are real.
Holds the asset register and work order history this review reads from, and raises the review when a second failure carrying the same mode code lands against the same asset — the step that most often never happens.
Carries the production change log and shift records the correlation questions must be tested against. Without it, correlates with production change is answered from memory, and the answer is always no.
Owns the CAPA record referenced from the cost of repetition section, so the corrective action carries a due date, an owner and verification rather than a line in a closed review.
Assembles the failure history rows from existing work orders, flags where a stated investigation level has no matching RCA reference, and holds every write for approval.
This template lives in KnowMaintain — asset maintenance. Work orders, planned maintenance, calibration, reliability and shutdowns.
Meet KnowMaintain→Glossary
Repeat Failure Review definitions and key terms
- Failure mode
- The manner in which a failure occurs, expressed as a standard code rather than free text — contamination in seal, misalignment, insufficient temperature.
- Repeat failure
- A second or subsequent failure of the same item with the same failure mode. The defining characteristic is the shared mode, not the shared component.
- Time between failures
- The elapsed interval from one failure to the next on the same item. Its trend across occurrences carries more diagnostic weight than the count.
- Bad actor
- An asset consuming disproportionate maintenance resource or downtime across all failure modes. A repeat failure is narrower: one mode, one asset.
- Symptomatic repair
- A repair restoring function by replacing what failed without altering the condition that caused it to fail. Legitimate as an immediate response, inadequate as a settled answer.
FAQ
Frequently asked questions about repeat failure review
Should the review be raised on the second failure or after a threshold?+
On the second. A threshold of three or five means the cost of repetition is paid in full before anyone looks. The review is cheap, and its output can legitimately be that no investigation is required.
The asset failed twice but on different components. Is that a repeat failure?+
It can be, and the template catches it: same component every time scores one for no rather than zero, because a mode migrating between components usually points to a systemic cause such as misalignment.
Who should complete it, reliability or the technician who did the repairs?+
Reliability leads, with input only the technicians have — hence the reliability signature and the maintenance manager countersignature. A review written by the person who did the earlier repairs will rarely record no, symptom only.
What if the underlying cause is suspected but not proven?+
Record suspected. It scores one rather than three, which keeps the record escalating, and it is honest. A yes on a hunch closes the record and stops the investigation that would have confirmed it.
Does the review need financial figures if the fix is obviously worth doing?+
Yes, because obviousness does not survive a capital review. The cost of repetition section converts engineering conviction into a comparison a finance approver can act on, and estimates with stated assumptions beat blanks.
What should change in the maintenance strategy afterwards?+
Something visible. A confirmed pattern that produces a CAPA but leaves the PM interval, job plan and monitoring route untouched has not closed the loop. Those artefacts are where the learning has to land.
Keep going
Related templates and programmes
Industries this is written for
Programmes this belongs to
Used together in Reliability and Predictive Maintenance
Root Cause Analysis
Finds out why something happened rather than who was involved
Equipment Failure Report
Records that an asset has failed and stopped or degraded production
Failure Mode Record
Records how an asset failed, using standard failure mode codes rather than free text
Repair Record
Records what was actually done to fix the asset, including parts replaced and settings changed
Temporary Repair Record
Records a repair that is not permanent, with an expiry date and a plan for the proper fix
Production Impact Record
Records what the failure cost in lost output, scrap and late orders
More in Breakdowns
Equipment Failure Report
Records that an asset has failed and stopped or degraded production
Failure Mode Record
Records how an asset failed, using standard failure mode codes rather than free text
Repair Record
Records what was actually done to fix the asset, including parts replaced and settings changed
Temporary Repair Record
Records a repair that is not permanent, with an expiry date and a plan for the proper fix
Production Impact Record
Records what the failure cost in lost output, scrap and late orders
Breakdown Debrief
A short review straight after a significant breakdown, capturing what happened and what would have helped
Written and reviewed by
Siddarth Singh
Founder & Chief Executive Officer, Knowella
Certified Safety Professional and industrial and systems engineer with more than a decade inside food supply chain, freight and manufacturing operations. This page was written against the current text of the standards it cites, not against secondary summaries of them.
- Certified Safety Professional (CSP), Board of Certified Safety Professionals
- MBA, University of Chicago Booth School of Business
- MS and BS, The Ohio State University, Industrial and Systems Engineering
- Six Sigma Black Belt
Sources and last review. Reviewed 16 August 2026 against:
- ISO 14224:2016 — Collection and exchange of reliability and maintenance data for equipment, cl.6 to cl.9
- ISO 9001:2015 — Quality management systems, cl.10.2 nonconformity and corrective action
- Provision and Use of Work Equipment Regulations 1998 (SI 1998/2306), reg.5
- 29 CFR 1910.119(j) — OSHA Process Safety Management, mechanical integrity
This page is general guidance, not legal advice. Confirm requirements with your jurisdiction’s regulator.