Repair or Replace Commercial HVAC: A Decision Guide

A rooftop unit fails in the second week of July. The service company says the compressor is gone, quotes a repair, and mentions that a new unit would of course be better. It is the middle of a heat wave, the tenant on that floor is calling every few hours, and somebody has to decide quickly whether to spend a moderate amount now or a large amount now.

That is the worst possible moment to make a capital decision, and it is when many commercial HVAC replacement decisions are actually made. The decision is not really hard — it is just being made under conditions that guarantee a poor answer, with incomplete information, no comparison, and a countdown running. This guide is about making it earlier and better: what actually drives the repair-or-replace answer for commercial equipment, why the popular rules of thumb are unreliable, and how the decision changes for buildings in DC, Virginia, and Maryland specifically. The same analysis belongs in any HVAC maintenance and capital planning program.

Why the common repair-or-replace rules of thumb do not work

You will find two heuristics repeated everywhere. The first is that you should replace when a repair costs more than half the price of new equipment. The second multiplies the equipment’s age by the repair cost and replaces above some threshold.

Both are rules of thumb, not standards, and both fail on commercial equipment for the same reason: they treat the unit as though it exists on its own. Commercial equipment does not. A rooftop unit sits on a roof that has its own remaining life and its own access constraints. It serves a tenant with a lease that has a term. It is one of several identical units that will all reach end of life within a few years of each other. It runs on a refrigerant whose availability is changing. It may be the difference between meeting an energy performance standard and not.

None of that is captured by a percentage. The heuristics also quietly assume the repair restores the equipment to its previous condition, which for an aging unit is rarely what happens — you have replaced one failed component in a system whose other components are the same age as the one that just failed.

The useful framing is not “is this repair worth it” but “what is the total cost of owning this unit for the next five years, under each option?” That question has an answer, and it usually differs from what the percentage rule suggests.

What actually drives the decision?

Seven factors, roughly in order of weight.

Age relative to expected service life. Commercial equipment types have broadly different lifespans, and it is worth being honest that any service-life figure is a planning assumption rather than a guarantee. ASHRAE’s Service Life and Maintenance Cost Database (opens in a new tab) is free to query by equipment type, region and building type, and it shows how wide the spread is: as of October 2026, the 30 centrifugal chillers it lists as replaced were between 7 and 42 years old when they went, with a median of 25. Where any individual unit lands depends on runtime hours, maintenance history, and how hard the equipment has been driven. A well-maintained unit late in its planning range may have real life left; a neglected one well short of it may not.

Repair history, not repair cost. A single large repair on an otherwise reliable unit is usually worth making. A run of service calls on the same unit is a different signal entirely, and it is the more important one. How often a unit fails says more about its future than the size of any one invoice. This is the factor most often unavailable at decision time, because nobody has the history in front of them — which is an argument for keeping it, not for ignoring it.

Refrigerant. This has become genuinely decisive, and we cover it separately below.

Energy cost. An old unit running at degraded efficiency costs money every hour it operates, and that operating delta compounds across the remaining years in a way a one-time repair does not. For a unit with long annual runtime hours, the energy difference alone can carry a replacement case that the repair-cost comparison would have lost. Our energy management ROI guide walks through how to actually size that number rather than assume it.

Downtime and business impact. What does an outage cost this particular tenant? A general office floor tolerates a warm afternoon. A tenant with a server room, a lab, temperature-sensitive inventory, or a healthcare use does not. The tolerance for another failure should scale the willingness to replace preemptively.

Capital timing and lease term. A planned replacement in the off-season generally costs less than the same replacement next July, unplanned. And who benefits matters: capital spent on a suite whose lease expires next year has a different return than the same spend serving a tenant with most of a long lease remaining.

Coupled work. The unit sits on a roof. If that roof is due for replacement in a few years, replacing the unit first and then craning it off again to redo the roof beneath it is a sequencing error that costs real money. The same coupling exists with electrical capacity, structural curbs, and controls. Our roof preventive maintenance checklist is the companion piece here, because these two capital items are more entangled than they look.

How does the refrigerant transition change the math?

Significantly, and it is a factor many owners have not yet priced in.

The clearest case is R-22. Under EPA’s HCFC phaseout schedule (opens in a new tab), production and import of R-22 in the United States ended as of January 1, 2020, so any equipment still running on it is being serviced from recovered, reclaimed and stockpiled supply. EPA’s FAQ for owners of commercial air-conditioning equipment (opens in a new tab) confirms you may keep using R-22 equipment, but warns that R-22 supply will decline and that new and reclaimed R-22 is likely to rise in price as existing supplies are used up. A major refrigerant-circuit repair on an R-22 unit today is a bet that you can keep sourcing a refrigerant nobody is making — and it means paying to recharge a system that will need the same refrigerant again. For R-22 equipment, a significant compressor or coil repair is often the moment to replace instead.

The broader HFC phasedown is the newer part. Under the AIM Act, EPA is cutting HFC supply in steps: allowances are capped at 60% of the historic baseline for 2024–2028, 30% for 2029–2033 and 15% from 2036, according to EPA’s HFC phasedown FAQ (opens in a new tab). Separately, EPA’s Technology Transitions restrictions (opens in a new tab) bar refrigerants with a global warming potential (GWP) of 700 or more in new residential and light commercial air-conditioning systems and new comfort-cooling chillers from January 1, 2025, with other dates for other system types. As of EPA’s August 2026 table, residential and light commercial systems built entirely from components made or imported before 2025 are exempt. R-410A has a GWP of 2,088 on EPA’s GWP reference table (opens in a new tab), so new equipment in these categories uses lower-GWP refrigerants such as R-454B and R-32, which EPA describes as mildly flammable (the A2L safety class). Two practical consequences follow, and it is worth separating them clearly.

For existing R-410A equipment, this is not an emergency. EPA’s FAQ confirms that existing systems remain legal to operate and to repair, including replacing a major component such as a compressor or condensing unit, and R-410A components can still be made and imported for servicing existing systems. The change is directional rather than immediate: EPA itself notes that the availability of HFCs and components may lessen over time as the phasedown proceeds, so a large refrigerant-circuit repair on an aging R-410A unit deserves more scrutiny than it once did.

For replacement decisions, it means new equipment likely uses a different refrigerant than what it replaces, which carries real implications: technician training, service tooling, and — where A2L refrigerants are involved — charge-size, ventilation, and leak-detection requirements that come with a mildly flammable refrigerant. This is a reason to plan a replacement rather than improvise one, and a reason to ask a prospective contractor directly what they are certified and equipped to service. Anyone who opens the refrigerant circuit must hold EPA Section 608 technician certification (opens in a new tab), and refrigerant activity should be logged per unit. For many chillers and other large systems that log is now a legal duty: EPA’s Section 608 rules (for appliances holding 50 pounds or more of an ozone-depleting refrigerant such as R-22) and, since January 1, 2026, the AIM Act leak-repair rule (opens in a new tab) (for HFC appliances holding 15 pounds or more, outside the residential and light commercial subsector) require a leak-rate calculation each time refrigerant is added. The log is also one of the better predictors of when a unit is becoming a replacement candidate, because a system that keeps needing charge is a system that keeps leaking.

What does the DMV specifically add to this decision?

Three things.

Energy performance standards. The District’s Building Energy Performance Standards (opens in a new tab) and Montgomery County’s BEPS (opens in a new tab) put covered buildings under an obligation to reach a performance standard rather than merely to report a number; Montgomery County’s program covers most buildings of 25,000 square feet and larger. Maryland’s statewide BEPS (opens in a new tab) covers buildings of 35,000 square feet or more elsewhere in the state, with performance standards arriving in the 2030s. That changes the calculus: an efficiency upgrade that was optional becomes part of a compliance pathway, and HVAC is often one of the largest levers available. If your building is covered, the replacement case may be materially stronger than the equipment condition alone suggests. We cover the timelines in BEPS deadlines for DC and Maryland.

Climate loading. This region runs both a real cooling season and a real heating season, with high summer humidity. Equipment here does not get the easy life it would in a milder climate — latent load on cooling equipment is genuine, condensate systems work hard, and units cycle across a wide annual range. Service-life planning ranges should be applied with that in mind rather than optimistically.

Access and logistics. Urban buildings in DC and inner Northern Virginia frequently need street closure permits, crane or helicopter placement, after-hours work windows, and coordination with neighboring properties for a rooftop unit replacement. Those costs are not in the equipment quote and can be a substantial share of the project. They are also strongly seasonal in availability, which is another argument for deciding in the fall rather than in July.

How should this decision actually be made?

Not during the outage. The point of everything above is to move the decision upstream, and there is a repeatable way to do that.

Keep the equipment list current. Every unit, with make, model, serial, install date, refrigerant, tonnage, location, and the tenant or area it serves. This sounds obvious and is often missing. You cannot plan replacements for equipment you have not inventoried.

Keep the service history attached to the unit. Not in a folder of invoices — attached to the asset, so that “how many times has this unit failed in two years” is a question with an immediate answer. This is exactly what a maintenance system is for, and it is the single input that most improves these decisions. In Levaru’s platform each unit carries its own history, and the platform is included with our management rather than sold separately.

Grade the fleet annually. Once a year, walk the list and sort every unit into: healthy, watch, plan to replace, replace now. The value is not precision — it is that the July failure now lands on a unit you already classified, with a decision already reasoned through.

Put the “plan to replace” units into the capital plan with dates and numbers. That is where this connects to capital planning proper: HVAC is often among the largest line items in a commercial building’s capital forecast, and a forecast built from an actual graded equipment list is dramatically more credible than one built from age alone. Our capital planning and reserve study guide covers how that forecast is assembled.

Bundle where it makes sense. Illustrative example: nine rooftop units of the same vintage will not fail on a convenient schedule, but they can be replaced on one. Phased fleet replacement — three a year for three years — can buy better pricing, three planned mobilizations instead of nine emergency ones, and a predictable budget line, and it takes the decision out of the emergency category permanently.

Does maintenance actually change the answer?

Yes, in both directions, and it is worth being precise about how.

Good preventive maintenance extends usable life, keeps efficiency from degrading, and catches failures while they are still small. Filter changes, coil cleaning, condensate treatment, belt and bearing service, and electrical connection checks are unglamorous and they are often the difference between a unit that outlasts its planning range and an identical one that fails well short of it. Our HVAC preventive maintenance checklist sets out the actual scope, and the broader case for running it as a program is in our guide to building a preventive maintenance program.

The second direction matters just as much: maintenance is what generates the information the replacement decision needs. A documented PM program produces the runtime observations, the refrigerant log, the failure history, and the photographs that let you decide in October rather than in July. A building with no maintenance record has no basis for a replacement decision other than age and whatever the contractor standing on the roof says — which is precisely how buildings end up replacing units that had life left and repairing units that did not.

Frequently asked questions

When should you replace rather than repair commercial HVAC equipment?

Replace when several factors point the same way: the unit is late in its expected service life, it has failed repeatedly rather than once, it runs on a refrigerant that is difficult or expensive to source, and its efficiency is measurably costing you. A single large repair on a well-maintained mid-life unit with no failure history is usually still worth making.

Is the 50% rule reliable for commercial HVAC?

No. It is a rule of thumb, not a standard, and it ignores everything that makes a commercial decision different — roof condition and access, lease term, fleet vintage, refrigerant availability, energy performance obligations, and the cost of downtime to a specific tenant. Use it as a rough prompt to look closer, never as the decision.

How long does commercial HVAC equipment last?

There is no single answer, and any figure is a planning assumption rather than a guarantee. ASHRAE’s public Service Life and Maintenance Cost Database shows how wide the spread is: as of October 2026, the centrifugal chillers it lists as replaced were between 7 and 42 years old when they went. Runtime hours, maintenance history, and how hard equipment has been driven move any individual unit substantially, so plan from the unit’s own condition and repair history, not from age alone.

Should I replace equipment that still uses R-22?

Usually, once it needs significant work. R-22 has not been produced or imported in the United States since 2020, so service depends on recovered, reclaimed and stockpiled supply, which EPA expects to decline and likely rise in price. A major refrigerant-circuit repair on R-22 equipment means paying to recharge a system with a refrigerant nobody is making — that is generally the moment to replace.

Do I need to replace working R-410A equipment because of the refrigerant phasedown?

No. EPA confirms that existing R-410A equipment remains legal to operate and to repair, including major components. What changes is that a large refrigerant-circuit repair on an aging R-410A unit deserves more scrutiny than it once did, and that replacement equipment will likely use a different refrigerant — which is a reason to plan the transition rather than be surprised by it.

What is the best time of year to replace commercial HVAC equipment?

Shoulder seasons — fall and early spring. Contractor availability is usually better, crane and permit logistics are easier to schedule, tenant disruption is lower because the equipment is not carrying peak load, and you have time to collect competing bids instead of accepting the first quote during a heat wave.

How do I plan HVAC replacements across a whole portfolio?

Build a graded equipment list, review it annually, and phase replacements deliberately rather than reacting to failures. Grouping same-vintage units into planned phases can produce better pricing, fewer mobilizations, and a capital line you can actually forecast — and it moves the decision out of the emergency category, which is where it tends to cost the most.