Every HVAC system reaches a moment when the tech puts the panel back on, walks out, and gives you the number. Sometimes that number is $600 and the answer is obvious. Sometimes it's $3,800 and the calculation gets harder — especially on a unit that's been in your home for a decade or more. This is the framework we walk homeowners through when the question isn't binary.
It has four inputs: age, cost ratio, refrigerant status, and comfort symptoms. Each one contributes to the decision; none of them alone answers it.
Input 1: How old is the unit?
Front Range design temperature swings, high elevation, and dry air all contribute to somewhat harder service conditions than a mild coastal climate. Expected lifespans for residential HVAC equipment in Colorado:
- Gas furnaces: 15 – 20 years typical. High-quality condensing units at the top of the range with regular maintenance; older 80 % AFUE units drift toward the bottom.
- Air conditioners (split-system condensers): 12 – 15 years typical. Coastal AC lifespans stretch further, but the freeze-thaw cycles on Front Range pads take a toll.
- Heat pumps: 10 – 15 years typical. They work harder than an AC (both cooling and heating), so somewhat shorter than an AC-only unit.
- Boilers: 15 – 25 years typical. Cast-iron boilers routinely last 25 – 30; modern high-efficiency condensing boilers land closer to 15 – 20.
If your unit is past 75 % of its expected lifespan and needs a major repair, the age input alone tips the calculation toward replace.
Input 2: The 50 % rule and its limits
The industry-standard rule of thumb: if the repair cost is more than 50 % of a comparable new-system install, and the unit is more than half its expected life, replace. If it's less than 50 % and the unit is under half its expected life, repair.
In practice, we use a slightly more nuanced version:
- Compute the repair cost.
- Compute the net cost of a new equivalent system after incentives (the incentive stack matters here — see below).
- Compute the annual operating-cost delta between the current unit and a new one at your utility rates.
- Divide the net install cost by the annual savings to get a payback period.
- Compare payback to remaining lifespan if the unit is repaired.
Concrete example: 12-year-old 14 SEER AC needs a $2,400 compressor. New 16 SEER2 replacement, net after Xcel + 25C: $9,800. Annual electric savings: $180. Payback: 54 years. Even if the repaired unit lasts another 5 years, the repair pays back better than the replacement at this SEER2 level. But — if the customer is also considering heat-pump conversion (adding the state SB23-016 stack), the payback math shifts dramatically.
This is why we don't quote replacements without also quoting the repair alternative side-by-side. The math should be transparent, not persuasive.
A good HVAC contractor writes both the repair quote and the replacement quote on the same page, so you can read the trade-off directly. If yours doesn't, ask why.
Input 3: Refrigerant status (R-410A vs newer refrigerants)
R-410A — the refrigerant in almost every AC and heat pump installed from ~2010 through 2024 — began its EPA phase-down in January 2025. New equipment sold in 2026 uses R-454B or R-32 (lower-GWP replacements). What this means for repair-vs-replace:
- R-410A remains legal to service and repair. If your existing unit needs a refrigerant top-off or a valve replacement, that's fine — no need to replace over refrigerant status alone.
- R-410A availability is declining. Refrigerant costs are trending up as production ramps down. A minor recharge that would have cost $80 in 2020 is closer to $200 in 2026 and will keep climbing.
- Major R-410A repairs shift the math. A compressor failure requiring a full refrigerant recharge (roughly 7 – 12 lbs on a residential system) can add $1,500 – $3,000 in refrigerant alone. That pushes the 50 % rule toward replace.
- R-454B service is different. The newer refrigerants are mildly flammable (A2L classification). Techs need certification and slightly different service practices. This is why your 2026 install feels like a bigger step than a like-for-like replacement — the equipment is genuinely different.
Input 4: Comfort symptoms and system-condition red flags
Sometimes the repair-vs-replace decision is really a fixing-comfort-vs-fixing-equipment decision. Symptoms that indicate the whole system deserves rethinking, not just this one part replacement:
- Short cycling — indicates the system was oversized to begin with (see Manual J sizing). Replacing a compressor doesn't fix bad sizing.
- Rooms at consistently different temperatures — points to ductwork problems or missing zoning, both of which are addressed as part of a full-replacement redesign.
- Rising humidity indoors during cooling season — points to an oversized AC or a failing dehumidification cycle. A fresh Manual J at replacement fixes this.
- Rising utility bills without behavior change — points to declining efficiency or a partial refrigerant leak. Sometimes worth fixing, sometimes worth replacing with a modern SEER2/HSPF2 tier.
- Cracked heat exchanger — hard replace signal. See below.
Cracked heat exchangers specifically
A cracked heat exchanger can pass combustion byproducts, including carbon monoxide, into the airstream. A confirmed crack means the unit doesn't run again until it's replaced or the exchanger is swapped. In practice, on a furnace over 15 years old, the labor to replace a heat exchanger approaches the cost of a new high-efficiency furnace — so the decision usually resolves to replace, even when the exchanger itself is warranty-covered.
Newer high-efficiency furnaces carry lifetime heat-exchanger warranties in many cases. If your furnace is under 8 years old and the exchanger fails, the warranty covers the part; you pay labor. That's usually a repair.
The Xcel + state + federal stack changes the math
In 2026, a qualifying heat-pump or dual-fuel install stacks:
- Xcel heat-pump rebate (POS, deducted at install).
- Colorado SB23-016 state credit ($1,500 in 2026 program year).
- Federal 25C tax credit (30 % up to $2,000).
For a homeowner debating a major repair on an aging system, that stack often lands $3,500 – $6,500 of value on the replacement side of the ledger. That doesn't change the repair-vs-replace framework — but it changes the arithmetic in a way that materially favors electrification for a lot of decisions that would have been clear "repair" answers in 2020.
Three worked scenarios
Scenario A. 6-year-old Trane XR16 AC, capacitor and contactor failed simultaneously. Repair quote: $560. Replacement cost delta over just letting a good unit stay in place: massive. Answer: repair.
Scenario B. 11-year-old Rheem AC, compressor failed. Repair quote: $2,600 (compressor + refrigerant + labor). Replacement cost net after incentives: $10,400. The unit is 75 % through expected life, R-410A refrigerant, and the customer has been complaining about uneven cooling upstairs for two summers. Answer: replace with a matched-Manual-J heat pump, gain incentive stack, solve the upstairs problem via a better airflow design.
Scenario C. 18-year-old furnace, blower motor failed, tech noted heat exchanger looks fine but discoloration suggests moderate wear. Repair quote: $800 (motor + labor). At 18 years the unit is essentially at end-of-life. Refrigerant not a factor (gas heat). Answer: soft-lean toward replace unless the customer is planning to sell within 2 – 3 years, in which case repair and let the next owner make the call.
The default Colorado Bear stance
We don't sell replacements when repairs make sense. That's not a marketing sentence — it's an operational one. Roughly 60 % of the service calls we run resolve to a repair. About 25 % resolve to "this repair works, but here's what a replacement conversation would look like when you're ready." The remaining 15 % resolve to a replacement recommendation because the unit is genuinely past the point where continued repair is a good use of your money.
If you've been given a repair-or-replace question by another contractor and it feels weighted toward replace regardless of the numbers, bring both quotes to a Free 2nd Opinion. We'll walk through the framework above, measure the specific inputs on your equipment, and hand you a written second opinion — annotated on your existing quote if you have one.
The right answer to "repair or replace?" is the one that comes with math you can see. If the math isn't showing up on your quote, that's the first fix.