Infrastructure AdvisoryTwo measures, one asset
Infrastructure operators are routinely expected to improve availability while controlling maintenance expenditure. Taken separately both objectives look reasonable. The difficulty is that they are connected, and the connection is where the information is.
Availability holds while maintenance cost rises. Cost falls while availability holds. Either divergence is telling the organisation something, and neither is self-explanatory. The two measures are usually reported to different people, on different cycles, against different targets — so the gap between them, which is the most useful thing either of them produces, is frequently the one thing nobody is looking at.
Four patterns, and the one that gets investigated
There are four ways the two measures can move against each other, and they do not carry equal attention.
Cost up, availability falling is investigated, because it is visibly bad. Cost up, availability held is usually reported as expensive success. Cost down, availability held is reported as efficiency and frequently rewarded. Cost down, availability falling is investigated late, because the cost signal arrived first and looked like progress.
Three of the four are capable of being wrong in ways that will not surface for years. Only one of them looks like a problem at the time.
Availability that has been bought
Sustained availability under rising cost is often a purchased result rather than an asset characteristic.
The purchase is real and frequently correct: more inspection, more preventive intervention, more spares, shorter replacement intervals, faster response, standby capability. Each buys down the probability or duration of failure. The availability figure records the outcome and says nothing about the price, because it was never designed to.
The question is not whether the availability was bought. It is whether the organisation knows that it was, what it is paying per increment, and whether the increment is worth what it costs. An availability target met by escalating intervention is a different position from the same target met by an asset that does not fail — and the two are indistinguishable in the reported number.
Cost that has been deferred
The reverse pattern is the one most likely to be reported as an achievement.
Frequencies are extended, preventive activities removed, spares holdings reduced, renewal deferred. The maintenance budget improves immediately and availability does not move, because deterioration takes time to reach the point of failure. The saving is banked in one period and the consequence arrives in another — as increased failures, longer restoration, accelerated deterioration or more expensive corrective work.
A maintenance saving is therefore not necessarily a whole-life saving. The question is what consequence has been transferred, and where.
What a temporary control does to the signal
There is a third possibility that neither measure exposes on its own.
A control introduced to hold an unresolved condition can keep an availability figure steady while the condition beneath it continues to develop. The measure then reports the control rather than the asset. Cost rises quietly, availability holds, and the reporting shows an asset performing well under management that is doing more each year to keep it there.
How such a control becomes an enduring maintenance activity is examined elsewhere. For this purpose the point is only that its effect on the signal is to make a deteriorating position look stable.
What the measure actually counts
Before any divergence is interpreted, the measure itself needs interrogating — and this is where most analysis should start rather than finish.
What counts against availability, and what is excluded? Have the exclusions grown? Is the period comparable? Two assets with identical annual availability can have very different failure frequencies, restoration times and operational consequences, and an asset can achieve a high figure while failing at particularly damaging moments. System redundancy can make an individual asset’s unavailability almost irrelevant, or make simultaneous unavailability critical.
A divergence between a cost line and an availability figure whose definition has changed is not a finding. It is an artefact, and it is a common one.
Isolating the cause before resetting the budget
A divergence narrows the candidates. It does not identify a cause, and it should not be allowed to look as though it has.
The useful sequence is to establish that the measure is comparable across the period; to identify whether the divergence is distributed across the asset base or concentrated in a small population; to decompose the maintenance demand into planned, corrective and condition-driven; and only then to ask which explanation the evidence actually supports.
That is more work than adjusting a budget, and it is the reason it is often skipped. But a regime changed on the strength of an uninterrogated divergence is as likely to move the problem as to solve it — and the organisation will not find out for several years.
Criticality is what makes the answer actionable once it is found. The same expenditure carries very different value on an asset with redundancy, low failure consequence and a short replacement lead time than on one without. Divergence tells you where to look; consequence tells you what to do about it.
The decision point
The useful question is not what availability the organisation is achieving, or what maintenance is costing. Each answers on its own terms and neither is wrong.
It is what the relationship between them is describing — a regime that fits the asset, a result being purchased, a cost being deferred, a control holding something in place, or a measure that has quietly changed what it counts.
Availability and maintenance cost are two readings of one decision. Acting on either without understanding the other usually moves cost or risk somewhere the organisation is not currently looking.
Westheath reviews maintenance regimes where cost and availability have stopped explaining each other.