Use the commonly cited benchmarks of 4–6 crew-weeks for total backlog and 2–4 crew-weeks for ready backlog as references when planning site-specific capacity, asset criticality, and outage strategy [S5].
Measure backlog as estimated labor hours divided by available productive crew hours per week, while separating executable work from jobs held for parts, permits, vendor support, or equipment access.
Target Range and Measurement Basis
The commonly cited process-industry target is 4–6 crew-weeks of total backlog, supported by a ready backlog of approximately 2–4 crew-weeks [S2][S3][S7]. These are planning benchmarks, not regulatory limits. Asset criticality, outage timing, staffing model, workload seasonality, and the mix between preventive and corrective work determine whether the range is appropriate for a specific site [S2][S3].
Measure total backlog as:
Backlog weeks = total estimated labor hours of open work ÷ available productive crew hours per week [S2][S6][S8].
Open work-order counts are not a substitute because jobs can vary greatly in labor content [S5].
Capacity Must Match Productive Availability
Gross labor capacity overstates the hours available for backlog work. For example, 4 technicians working 40 hours each provide 160 gross hours per week, and productive availability must be used as the capacity denominator when calculating backlog [S3].
Use historical productive hours, not the organization chart, when establishing the denominator [S2]. The relevant schedule restrictions include preventive-maintenance commitments, training, meetings, average reactive load, and execution constraints. Capacity should also account for technician trade, skill, shift pattern, and planned absence, because aggregate hours cannot be spent on every job equally.
Do not mix unlike measures. Wrench time, paid time, straight-time hours, and available crew hours are different denominators. Record the selected definition with each weekly calculation so changes in availability do not masquerade as backlog growth [S3].
Ready, Waiting, and Total Backlog

Ready backlog contains identified and estimated work that is planned, has the required parts and instructions, and can execute as soon as a slot is available [S2][S3]. Waiting backlog is held by spare parts, vendor support, permits, or equipment access [S3]. Total backlog includes both categories and overdue preventive-maintenance tasks that were not completed by their due date [S3].
A workable division is 2–4 ready crew-weeks inside a 4–6 crew-week total [S3][S7]. The ready portion provides the scheduler with valid choices, while the waiting portion exposes constraints that should be attacked separately. Backlog aging matters more than backlog size, and old safety work can matter more than a larger fresh queue [S2][S5].
Exclude future preventive maintenance that is not yet due, vague improvement ideas, unestimated requests, completed work, and reactive repairs that do not enter the queue [S3][S5]. In a fleet, include overdue PM, approved open work orders, unresolved inspection defects, parts-hold work, and follow-up work from incomplete repairs [S5].
Symptom: Target Met but Work Still Slips
A nominal 5-crew-week backlog can conceal a scheduling failure if most jobs are blocked or lack complete estimates. The root cause is using paid or nominal capacity instead of productive availability.
Correct the denominator by subtracting planned preventive work, meetings, training, average reactive demand, and other commitments from gross labor hours. Estimate backlog work and exclude unestimated requests so the backlog remains meaningful [S3]. Use priority scoring, asset criticality, and scheduling to triage backlog work rather than treating every job equally [S4].
Replace or escalate a metric that cannot be reconciled to CMMS records rather than publishing a target-looking number. The minimum acceptance check is that the work-order register, estimate field, capacity roster, and calculated result use the same reporting week and approved scope.
Symptom: Backlog Grows Despite High Technician Activity

Overdue preventive tasks contribute to backlog, and a growing backlog of deferred repairs tends to drive more unplanned failures later [S3].
Protect planned work by separating categories: overdue PM, corrective repairs, inspection findings, parts holds, vendor waits, and follow-up work [S3][S5]. Track backlog direction, PM compliance, and the ready-to-schedule share together. A stable total with high PM compliance is different from a rising total with falling PM compliance because the latter indicates corrective work crowding out planned work [S3].
Do not treat the backlog itself as the failure symptom. If a backlog exceeds 6 crew-weeks because a planned outage was removed, reschedule the affected work rather than declaring an immediate capacity shortage. Escalate when asset safety, regulatory obligations, or critical-service restoration cannot be assigned an executable date [S4].
Symptom: Backlog Falls but Reliability Does Not Improve
A falling backlog can reflect weak work identification rather than better maintenance performance. Zero backlog is not a healthy target, because a crew with no queued work may lack identified work or have an inadequate inspection process [S2][S3]. A near-zero queue can also indicate overstaffing or that inspections are not detecting enough defects [S3].
Validate completion rather than closure status. Confirm that work orders closed with repair evidence, that inspection findings were translated into approved work, and that no follow-up labor was omitted. For fleet work, keep unresolved driver inspection defects visible until repair is complete [S5]. The acceptance criterion is not an empty queue, but a stable 4–6 crew-week total with 2–4 ready crew-weeks and documented asset-condition coverage [S2][S3].
Where backlog data supports condition monitoring, predictive maintenance alerts should create or update traceable work rather than bypass the backlog calculation. The practical predictive-maintenance ROI check remains completed corrective work and verified risk reduction, not a reduction in reported maintenance labor.
Burn-Down Plan and Weekly Governance

Start the weekly process with a work-order scrub: identify open approved work, confirm estimates, assign a status, and separate ready work from constrained work [S2]. Then protect capacity for backlog reduction rather than consuming every available hour on incoming requests. Facility guidance reserves 20–30% of weekly capacity for backlog reduction and supports a monthly reduction objective of 10–15% where the backlog is materially oversized [S4].
Measure the result weekly using a fixed formula and fixed scope. Track total backlog in crew-weeks, ready backlog in crew-weeks, the ready-to-total percentage, the 8–12 week trend, PM completion percentage, and work aging. Use age buckets such as 0–7, 8–30, 31–60, 61–90, and 90+ days to identify stale work rather than allowing a large fresh queue to conceal old safety items [S4][S5].
Review the target quarterly [S4]. Reapprove it only after testing actual capacity, outage plans, asset criticality, and seasonal workload. Maintain the commonly cited 4–6 week total and 2–4 week ready ranges when the trend is flat, and act when the total exceeds 6 weeks for multiple reporting periods, the ready share falls below 2 weeks, or aging safety work lacks an owner [S2][S3].