Two work orders arrive on the same morning. Both are for air conditioning. Both carry a repair quote of 1,200 in the local currency.

Figures are illustrative.

Unit A leans clearly toward repair, assuming the quoted repair is technically sound and nothing else is wrong. Unit B makes a much stronger case for replacement, timed to the move-out, though the actual replacement cost and the condition of the equipment still matter.

Same quote. Very different answers. Which is the first thing worth understanding about this decision: the repair quote, on its own, tells you almost nothing. It is the one number every work order carries, and the least useful one for deciding.

When the rules of thumb disagree

There are two rules of thumb you'll hear repeated most often. Apply both to Unit B.

• Age multiplied by repair cost. Multiply the unit's age by the repair quote, and replace if the answer exceeds a fixed threshold, often quoted as 5,000 US dollars. For Unit B: 14 × 1,200 = 16,800. Well over. Replace.

• The half-of-replacement test. Replace if the repair costs more than half the price of a new unit. If Unit B would cost 6,000 to replace, the 1,200 repair is 20%. Well under. Repair.

Two commonly repeated rules of thumb, one unit, opposite verdicts.

• Neither should be treated as a universal decision standard. Each captures one real factor and ignores the rest. The first weighs age heavily and ignores what replacement actually costs. The second weighs the cost ratio and ignores age entirely, so it would happily keep repairing a 20-year-old unit forever, one cheap repair at a time.

• And the first rule has a problem specific to anyone operating across countries. A threshold fixed in US dollars doesn't adjust for inflation over the years it has been repeated, and it means nothing at all for a property priced in dirhams, pounds or Australian dollars. A rule that depends on a currency you don't use is not a rule you can use.

The honest position is that these are shortcuts, useful for a quick first read, and not a substitute for the decision itself.

Expected life is a median, not a deadline

The other number people reach for is expected lifespan. It needs handling with care, because of what it actually measures.

ASHRAE, the international society for heating, refrigerating and air-conditioning engineering, maintains a public database of equipment service life built for exactly these decisions. Its figures are medians, and a median means half of all units last longer.

ASHRAE's own material shows how much this matters. For centrifugal chillers, its data found survival falling to 50% at 31 years, but because the sample dropped below the statistically relevant number of 30 units at 25 years, ASHRAE says the median can only be stated with confidence as greater than 25 years.

So if you treat "expected life" as a replacement date, you will replace roughly half your equipment while it still has years left in it. Age tells you where a unit sits on the curve. It doesn't tell you whether this particular unit has reached the end of it.

Commonly cited ranges give a rough orientation, and they vary with maintenance, installation quality and climate:

Harsh conditions shorten these. Salt air on the coast, sustained extreme heat, heavy humidity and dust all wear equipment faster. A condenser in coastal Dubai and one in inland Calgary are not living the same life, even if they share a model number.

The four inputs that actually decide it

1. Age against expected life. Not as a deadline, but as context. A unit well inside its range with a first failure is a strong repair candidate. A unit well past its range is living on borrowed time, and repairs tend to buy less.

2. Repair history on that unit. An important input, and one that is often missing. One repair in fourteen years says something very different from three in two. Repeated or increasingly costly repairs can be evidence that the equipment is becoming less economical to keep.

3. Running cost. Older equipment can cost more to run, particularly where its efficiency has declined or newer equipment would be materially more efficient. Where replacement would meaningfully cut energy use, part of the replacement cost comes back over the equipment's life. This input matters most in climates where systems run hard for long seasons.

4. Refrigerant. Increasingly relevant, and increasingly misunderstood. It gets its own section below.

None of these is decisive alone. Together they make the call. And notice that only one of them, age, is visible on the unit itself. The other three live in records.

What the refrigerant phasedown actually changes

Refrigerant regulation is shifting the repair-or-replace math, but not in the way urgent sales messaging sometimes suggests.

The global framework is the Kigali Amendment to the Montreal Protocol, agreed in 2016, which commits signatory countries to phasing down HFC refrigerants. Each country implements it on its own timeline, so the details differ across the markets where you operate.

The US shows how it works in practice. Under the AIM Act, the EPA's rules restrict the manufacture, import and installation of certain new residential and light commercial air conditioning and heat pump equipment that uses refrigerants above set global-warming-potential limits, such as R-410A, with those restrictions phasing in from 2025. Harvard Law School's Environmental and Energy Law Program tracks these rules, including a federal appeals court upholding the EPA's HFC phasedown program in 2025.

The part that matters for existing equipment:

• Phased down is not banned. Existing R-410A systems can generally continue to be repaired and serviced. The restrictions target new equipment.

• What changes is cost. As production is capped, less of the older refrigerant is made each year, and service refrigerant tends to become more expensive over time.

So for a unit in good health, the phasedown alone is not a reason to replace it. For a unit that is old, failing repeatedly and leaking refrigerant, rising service costs add weight to a replacement decision that was already leaning that way. It tips close calls rather than making easy ones.

Check the rules in each country where you operate before planning around them. The mechanism is similar, but the dates are not.

The fifth input: timing

Everything above is how an engineer would frame it. Property management adds a fifth input that changes the decision more than any of them: whether the unit is occupied.

Look at Unit B again. The resident moves out in five weeks.

If you repair it now, you're likely to be spending on a unit that may need replacing anyway, and you risk a larger failure before the move-out. If you replace it during the vacancy, the replacement can be planned into a turn that's already happening: in an empty unit, on your schedule, without displacing anyone.

Compare that with the alternative: the same unit failing in peak summer, occupied, with the resident without cooling while you source equipment and a contractor at the busiest time of year. The job becomes an emergency, with less control over price, availability and timing.

That's why the replace decision for an ageing unit is often better made at the turn than at the breakdown. We covered how to plan the work of a vacancy in the turn starts at notice, not at move-out. And knowing when units will come vacant, months ahead, is what makes it possible to plan replacements into those windows, which we covered in your leasing season was decided twelve months ago.

Replacing early at a vacancy costs you the remaining life of the old unit. Replacing late at a failure costs you an emergency. For a unit already near the end of its expected service range, replacing during a planned vacancy can be more practical than waiting for an occupied-unit failure.

How system type changes the answer

The same four inputs apply everywhere, but the weighting shifts with the kind of system.

Individual split systems and packaged units. One per unit or small group. Replacement is contained and can be timed to each unit's own vacancy. This is where the timing input is most powerful.

Heating boilers. Common where heating dominates, including many buildings in the UK and colder parts of Canada and the US. Often longer-lived, and often serving many units at once, so a failure affects the whole building and can't be timed to a single vacancy.

Central plant and chillers. Common in larger buildings and in hot climates, including many buildings in the UAE. Long-lived and expensive, serving the whole building. Replacement is a capital project that needs years of planning, not a work-order decision.

The bigger and more shared the system, the less the single-unit decision applies, and the more it becomes a question of portfolio capital planning, which sits in its own guide.

The information problem underneath

Go back to the four inputs. Age you can read off the unit. The other three, repair history, running cost and refrigerant type, live in records. And timing depends on knowing which units are about to come vacant.

In many operations those facts are scattered. The repair history sits in past work orders, possibly under different vendors. The refrigerant type is on a nameplate nobody photographed. The move-out date sits with leasing. So the technician looking at Unit B sees a broken unit and a 1,200 quote, and very little of what would actually decide it.

That's how the wrong call gets made with the best of intentions. Nobody decided to repair a failing 14-year-old unit a month before it went vacant. They just couldn't see that it was 14, failing and about to go vacant, at the moment they approved the quote.

Unit A and Unit B had the same repair quote. What separated them was everything the quote left out: age, history, refrigerant and the move-out date five weeks away.

That's the real difficulty with this decision. The information that decides it rarely sits in one place, and rarely sits in front of the person approving the work.

RIOO's property setup holds building systems and assets across units and communities, and maintenance planning and scheduling keeps the service history behind them. Because move-ins and move-outs run in the same environment, an approaching vacancy is visible alongside the equipment that might be replaced during it, and dashboards and reports show the pattern across a portfolio.

Next time a repair quote lands for an older unit, ask four questions before approving it: how old is it, how often has it failed, what refrigerant does it run on, and is anyone moving out soon? If those take more than a minute to answer, the decision is being made on the quote alone.

This article sits alongside our annual preventive maintenance calendar, the seasonal readiness guides for winter and spring and summer, and covers the replacement decision those guides leave out.

Frequently asked questions

Q1. How do you decide whether to repair or replace an HVAC unit?
Weigh the unit's age against its expected life, its repair history, its running cost and the refrigerant it uses, then consider timing. For rental property, whether the unit is occupied or about to become vacant often matters as much as any of the technical factors.

Q2. What is the rule of thumb for replacing HVAC?
Two are widely repeated: multiply the unit's age by the repair cost and replace above a fixed threshold, or replace if the repair costs more than half the price of a new unit. They can give opposite answers for the same unit, so treat them as quick first reads rather than decisions.

Q3. How long does HVAC equipment last?
Commonly cited ranges are roughly 10 to 15 years for split systems and heat pumps, 15 to 20 for gas furnaces and 20 to 25 for boilers, with ASHRAE stating the median for centrifugal chillers as over 25 years. These are medians, so half of all units last longer, and climate and maintenance change them considerably.

Q4. Do I have to replace my R-410A air conditioner?
Not because of the phasedown alone. Under the Kigali framework and national rules such as the US AIM Act, restrictions target new equipment, and existing R-410A systems can generally continue to be repaired and serviced. What rises over time is the cost of service refrigerant, which can tip close decisions toward replacement.

Q5. When is the best time to replace HVAC in a rental unit?
Often during a vacancy, when the resident has moved out. The replacement can be planned into the turn, in an empty unit on your schedule, rather than becoming an emergency in an occupied unit during peak season.