The backlog number that doesn't survive scrutiny
Ask five operators on the UKCS how they calculate maintenance backlog and you will get five different answers. Some count every overdue planned maintenance task regardless of criticality. Others exclude anything still inside a grace period. A few strip out contractor-executed work entirely because it sits in a separate system. When the number reaches the monthly integrity meeting, nobody quite trusts it, and when it reaches an audit, that mistrust becomes a finding.
This is the problem OEUK's Maintenance Backlog Measurement, Interpretation and Management Guideline (Issue 3) is trying to close. The guideline sets out an agreed suite of backlog KPI definitions so that a figure from one asset means the same thing as a figure from another. For HSE managers and integrity leads who have spent years explaining why their backlog isn't comparable to the platform next door, that is a meaningful shift — but only if the underlying data can actually support the definitions.
Why maintenance backlog keeps growing across the UKCS
Backlog rarely grows because teams ignore maintenance. It grows because deferred work has nowhere consistent to live. A planned preventative maintenance task slips past its due date, gets rescheduled twice, and eventually loses its link to the original justification for deferral. Multiply that across thousands of tasks on an ageing asset with a mixed contractor workforce, and the backlog stops being a list you can interrogate. It becomes a number you report and hope nobody asks about too closely.
The deferral itself is often defensible — a valve that needs a shutdown window, a joint that needs scaffold access, a job competing with higher-priority work. The problem is that most maintenance management systems were not built to preserve the reasoning behind that deferral over months or years. By the time an auditor asks why a safety-critical element has sat in backlog for longer than its revalidation interval, the original risk assessment is buried in an email thread or not recorded at all.
Where common backlog definitions expose weak underlying data
A shared KPI definition only helps if the data feeding it is complete and current. This is where the OEUK guideline will be uncomfortable for some operators. Standardising the calculation doesn't fix a maintenance system where:
- Overdue tasks are tracked in a spreadsheet that sits alongside, rather than inside, the maintenance management system.
- Criticality tagging is inconsistent, so safety-critical elements and routine housekeeping tasks are counted the same way.
- Contractor-executed maintenance is recorded separately from operator-executed maintenance, with no single register showing true asset-wide backlog.
- Deferral justifications exist as free text with no link back to a risk assessment or management of change record.
Applying a common formula to fragmented inputs produces a number that is comparable in name only. The real value of the guideline is that it forces operators to ask whether their maintenance data can support a defensible calculation at all — and for many assets, particularly those with a long history of contractor turnover, the honest answer is not yet.
What auditors and regulators actually check
HSE and OPRED inspectors reviewing backlog do not simply want the headline percentage. They want to see the trail behind it: which safety-critical elements are overdue, for how long, what compensating measures are in place, and who signed off the deferral. This sits alongside the wider expectations set out in OEUK's process safety leadership principles, which push operators towards demonstrable ownership of asset integrity rather than reactive reporting. If your backlog figure can't be broken down by criticality, by duration, and by justification on request, that is where the finding lands, regardless of how the top-line KPI compares to the OEUK benchmark.
The same scrutiny applies to how backlog interacts with control of work. A permit issued against equipment that already has an open, aged maintenance deferral is a common audit trail — and one that contractor permit management systems are frequently unable to flag automatically, because the permit system and the maintenance backlog live in separate places.
A checklist for testing your own backlog reporting
Before adopting the OEUK definitions as your standard, it's worth testing whether your current system could answer these questions without a manual data pull:
- Can you produce, within minutes, every safety-critical maintenance task overdue by more than one revalidation cycle, with the reason for deferral attached?
- Is contractor-executed maintenance visible in the same register as operator-executed maintenance, or does it require reconciling two systems?
- When a task moves from "planned" to "backlog", does that change carry a timestamp, an owner, and a justification, or does the history disappear?
- Can you filter backlog by criticality tier rather than reporting a single blended percentage?
- If an auditor asked for the deferral rationale behind a specific overdue task from eighteen months ago, could you find it today?
If any of these are difficult to answer, the gap isn't in your understanding of backlog — it's in the system recording it. That distinction matters, because no amount of KPI standardisation fixes a maintenance register that can't produce its own audit trail.
Turning a standard definition into standard practice
OEUK's Issue 3 guideline gives the sector a shared language for backlog. What it can't do is guarantee that the number behind the language is trustworthy. That depends on whether planned preventative maintenance, deferral justification, and criticality tagging live in one system that produces the same answer every time it's asked — which is the same discipline that underpins operational assurance more broadly. If your current backlog figure would struggle to survive the checklist above, get in touch with Elisian to talk through how a connected Planned Preventative Maintenance system closes that gap before the next audit does it for you.
Source: Maintenance Backlog Measurement, Interpretation and Management Guideline - Issue 3