Repair or replace?.

The four questions

Committees usually arrive at this decision holding two quotations that answer different questions, and an instinct that the system is “old”. Age is a poor guide: a well-installed system with available parts can run far longer than its nominal life, and a young system on failing cable can be beyond economic repair.

Two or more answers in the right-hand column is a replacement conversation, whatever the age of the equipment.

When repair is clearly right

A single failed device, in a system whose parts are current, on cabling that is sound, with no history of the same failure elsewhere. This is most faults, most of the time, and the correct response is to repair it and record it.

Repair is also right when the building is mid-way through a funding cycle and the system is functional but tired. A planned replacement, properly scoped and funded, beats an emergency replacement forced by a failure — and holding a system together deliberately for eighteen months while that is arranged is a legitimate decision, provided it is a decision rather than a drift.

When replacement is clearly right

• Parts no longer exist. A system that can only be repaired by cannibalising itself has a finite number of repairs left, and nobody knows the number.
• The same class of fault keeps returning from different parts of the installation. That is the infrastructure failing, not the devices.
• The cabling limits the requirement. If the building needs coverage or retention the existing installation cannot carry, replacing devices onto it spends money without meeting the requirement.
• The building has changed. A car park that filled up, an entrance that was added, a resident population that doubled.

The grey middle: partial and staged

Most real buildings sit between the two columns, and the useful answer is usually neither a repair nor a full replacement but a staged programme — replacing the layer that has actually failed while retaining the ones that have not.

The sequence matters. Cabling and containment first, then head end, then devices. Replacing devices onto cabling that is itself replaced a year later means paying for the same access twice, and it is the most common way a staged programme costs more than doing it at once.

What discontinued support really means

“End of support” from a manufacturer is not a switch. It usually means no new firmware, then no spares, then no technical assistance, spread over several years. The practical trigger for a building is the spares date, because that is the one that converts a two-day repair into a system replacement carried out under pressure.

It is worth asking a maintaining contractor, in writing and annually, which parts of the installation are approaching that point. That single question turns an unpredictable capital risk into a line a committee can plan for — see specifying and funding the work.

How each option is funded

Under Malaysian strata practice the distinction usually falls out as: repairs and scheduled servicing from the maintenance account, end-of-life replacement and new coverage from the sinking fund. The awkward cases — a repair so extensive it amounts to replacement, or an upgrade that is technically a repair because the original part no longer exists — are decisions for the committee with its managing agent.

What a contractor can usefully do is price the elements separately so the allocation can be made properly rather than argued about afterwards.

Putting the decision to owners

A replacement is easier to approve than to defend afterwards, and the difference is whether the paper showed the alternatives. Owners are generally willing to fund a replacement whose case is: here is what failed, here is what repair would cost and how long it would last, here is why replacement is the better value, and here is what the building holds at the end of it.

The commonest failure at a general meeting is not a weak case but an incomplete one: a single price, with no repair option costed beside it and no statement of what happens if nothing is done. Owners asked to approve a number with no alternatives will usually defer, and deferral is itself a decision — one that leaves a failing system in service for another year.

What the decision should look like in writing

Whichever way it goes, the reasoning should survive the committee that made it. Membership turns over, and a successor committee that cannot see why a system was repaired rather than replaced will re-open the question from the beginning, usually at the point the equipment fails again.

• The condition found, by layer, and the date it was surveyed.
• Both options costed — repair and replacement — even where one was obviously right.
• The expected remaining life under the option chosen.
• What was excluded from the decision and deliberately left for later.
• The fund it was drawn from, and the approval that authorised it.

That is a page, not a report, and it is the difference between a building with a maintenance history and a building with a drawer of invoices.

Occhio Tec works on the Extra Low Voltage layer of Kuala Lumpur buildings, and describes its niche as high-end condominiums in the Mont Kiara, Bangsar and KLCC areas, much of that work being service and maintenance. The company was established in 2010.

Next step

All four questions are survey questions. A site survey answers them with a record rather than an opinion, and that record is what makes three quotations comparable. Occhio Tec lists consulting, installation and maintenance among its fourteen ELV services, and a survey is where each of them starts.