Definition
Deferral Compounding Index (DCI) measures how many times the same maintenance task has been deferred, weighted by the consequence of failure for the affected asset and accelerated by elapsed time beyond the original schedule.
The Measurement Gap
Your schedule optimizer is doing exactly what you told it to do. Every week, it ranks all schedulable work by priority, matches work against available crew and parts, and builds the best possible schedule given the constraints. And every week, the same fourteen work orders get pushed to next week.
Not because they're unimportant. Because something is always more important. The emergency from Tuesday displaces the PM from Wednesday. The production window that opened for Thursday gets claimed by the job that was already deferred twice. The optimizer sees each week independently: it solves the one-week resource allocation problem and moves on. What it doesn't track is that it just pushed that gearbox inspection for the fourth time since January, and the gearbox serves a production line that represents $180,000 per shift in output.
You might think your existing metrics catch this. They don't.
Backlog Age
shows you how old your work orders are, but a work order that was created six months ago and scheduled three times isn't the same as a work order that was created six months ago and never planned. One is deferred risk; the other is planning latency. They show the same backlog age.
Schedule Compliance
tells you whether you executed the schedule, but the deferred work wasn't on this week's schedule, so it doesn't count against compliance. You can have 95% schedule compliance and a growing mountain of repeatedly deferred critical work, and both numbers look fine.
Maintenance managers know this problem intuitively. They talk about "the list": the work that keeps getting bumped, that everyone knows about, that nobody can find resources for because the urgent always displaces the important. But "the list" lives in someone's head or on a whiteboard. It's not quantified, it's not weighted by consequence, and it's not tracked systematically.
The optimizer makes it worse. A human scheduler has memory: they remember that the gearbox inspection has been pushed three times and start advocating for it. An optimizer has no memory. Every week is a fresh optimization. The prior deferrals have zero influence on the current schedule. The algorithm doesn't compound risk because it doesn't see risk: it sees priority scores and constraint matrices.
There is no metric that tells you how many times the same maintenance task has been deferred by optimization or prioritization decisions, which means compounding deferral risk accumulates invisibly until something fails.
Formula
Deferral Compounding Index (DCI) measures how many times the same maintenance task has been deferred, weighted by the consequence of failure for the affected asset and accelerated by elapsed time beyond the original schedule. It produces a per-work-order risk score that grows every time the work gets bumped, and an aggregate portfolio score that tracks total deferred-risk exposure.
In plain language:
Every time a work order's scheduled date gets moved to a later date, that's a deferral. Count the deferrals. Multiply by how critical the asset is. Multiply again by how long the work has been sitting relative to when it was originally supposed to happen. The result is a number that gets bigger every time the work gets pushed and every day it sits past its original date.
The formula:
DCI_i = D_i × C_i × A_i
Where:
- D_i = number of times work order i has been deferred (rescheduled to a later date)
- C_i = consequence of failure weight for the asset (1–5 scale from asset criticality register)
- A_i = age acceleration factor = 1 + (T_now − T_original_schedule) / T_original_lead
Portfolio-level:
- DCI_portfolio = Σ DCI_i (for all open WOs with D_i ≥ 1)
- DCI_max = max(DCI_i) (worst single item)
- DCI_count = count of WOs where D_i ≥ threshold
The age acceleration factor is what makes DCI more than a simple deferral counter. A work order deferred 3 times over 6 months is worse than one deferred 3 times over 3 weeks: the asset has been running without the maintenance for longer, and degradation compounds with time.
Worked example:
Three work orders are open, all deferred at least once:
WO-28471: Replace coupling on Pump P-1104
Criticality: 4 (critical production pump, no installed spare). Created January 15. First scheduled: February 5 (original lead = 21 days). Deferred Feb 5 → Feb 19 (emergency displacement). Deferred Feb 19 → March 12 (production window cancelled). Deferred March 12 → April 2 (parts reassigned). Current date: March 28.
D = 3, C = 4, A = 1 + (51 days / 21 days) = 3.43
DCI = 3 × 4 × 3.43 = 41.1
WO-28533: Inspect gearbox on Conveyor CV-0312
Criticality: 2 (redundant conveyor). Created February 1. First scheduled: February 12 (lead = 11 days). Deferred twice. Current date: March 28, overdue.
D = 2, C = 2, A = 1 + (44 / 11) = 5.0
DCI = 2 × 2 × 5.0 = 20.0
WO-28612: Calibrate pressure relief valve PRV-0089
Criticality: 5 (safety-critical, regulatory). Created March 1. First scheduled: March 15 (lead = 14 days). Deferred once. Current date: March 28.
D = 1, C = 5, A = 1 + (13 / 14) = 1.93
DCI = 1 × 5 × 1.93 = 9.6
Portfolio: DCI_portfolio = 41.1 + 20.0 + 9.6 = 70.7
DCI_max: WO-28471 at 41.1
Notice what the numbers reveal: WO-28471 has the highest DCI despite being criticality 4, not 5, because it has been deferred three times and is sitting 51 days past its original schedule. WO-28612 is criticality 5 but has only been deferred once and is still close to its original timeline. DCI correctly weights accumulated deferral history, not just point-in-time criticality.
Edge cases:
A rescheduled-EARLIER date is not a deferral. Don't increment D for schedule advances. Emergency or breakdown work orders are excluded since they were never scheduled for deferral. If an asset has no criticality rating, assign C = 1 but flag it for assessment. If T_original_lead = 0, treat every elapsed day as one unit of acceleration.
The most important edge case is
deferral laundering: closing a work order and opening a new one for the same scope to reset the deferral count. DCI must carry forward to replacement work orders covering the same scope on the same asset. If a WO is cancelled and a new WO is created for the same asset and same general scope within 30 days, the new WO inherits the deferral count. Without this rule, organizations can game DCI by administrative action.
Proposed formula, pending validation pilots.