Should You Repair or Replace Your HVAC System?

No single number settles the repair-or-replace question — you weigh these together, because any one of them can be outvoted by the others. A 12-year-old system with a $250 repair and no comfort complaints is an easy repair; the same age with a failed compressor and a hot upstairs is a different conversation.

It's a rule of thumb, not a verdict: multiply the system's age in years by the repair estimate in dollars. If the product is over $5,000, that points toward replacement; comfortably under, and a repair usually makes sense. A 5-year-old system with a $600 repair (5 × 600 = $3,000) leans repair. A 15-year-old system with a $2,000 repair (15 × 2,000 = $30,000) leans hard toward replacement. The test captures something real — money spent on an old system has a short runway — but it's only a starting point. A scary symptom is sometimes a $300 capacitor, and a good diagnostic can change the answer. Treat it as the opening question, not the closing argument.

Age doesn't just add years — it changes what a repair is worth. In the first third of a system's life, a repair restores something with most of its runway ahead of it, so almost any reasonable fix pays off. In the middle third, it depends on the size of the repair. In the final third — a central AC past roughly 15 years, a furnace deep into its teens or twenties — you're weighing a repair against a system that could need the next one at any time, often on parts that are getting harder to source. This is why the same $1,500 repair is obviously worth it at year five and obviously questionable at year eighteen. It isn't that old systems aren't fixable; it's that each fix buys less, and the odds of the next failure keep climbing.

Refrigerant type is the one factor that can settle the question by itself. If your system runs on R-22, any repair touching the refrigerant circuit is expensive dead-end money — R-22 is no longer produced, so the refrigerant itself is costly and getting scarcer, and you'd be pouring it into an old system. That combination usually points to replacement whenever the refrigerant circuit is involved. An R-410A system is a much weaker signal: those can still be serviced, and the 2026 low-GWP transition mainly affects what new equipment ships with, not whether you can keep running what you have. New systems now use lower-GWP refrigerants like R-32 or R-454B by default, so if you replace, that choice is made for you. The full picture — including what the transition does and doesn't require — is in the refrigerant transition guide.

Efficiency is the factor homeowners underweight most, and in Sacramento's long cooling season it matters. Cooling efficiency is now rated in SEER2 (the current test standard that replaced SEER), and heating efficiency for a furnace in AFUE. A system installed 15 to 20 years ago was rated far lower than today's minimums even when it was new — and it has lost efficiency since. On a home that runs the AC hard from June through September, the gap between an old, degraded system and a modern high-efficiency one shows up on the utility bill month after month. That doesn't mean efficiency alone justifies replacing a working system; it means that when a big repair already has you on the fence, the ongoing energy savings belong in the calculation, not as an afterthought. A heat pump can also change the equation by handling both heating and cooling on electricity, which is worth modeling if you're replacing anyway.

Rather than eyeball it, you can put your own answers into the repair-vs-replace calculator. It uses the same factors above — age, repair estimate, refrigerant, breakdown history, and comfort — applies the $5,000 test, and returns a plain-language lean toward repair, replace, or a genuine toss-up, with the reasoning for each point spelled out. It's transparent math you can see, not a black-box quote. If replacement is in play, the HVAC replacement cost calculator gives a planning range for what a new system might run, and the AC replacement cost guide and heat pump cost guide break down what actually drives the number.

Replacement isn't the default answer, and a good contractor shouldn't push it when the math doesn't support it. A repair usually wins when the system is newer (roughly 0 to 7 years), the fix is small and non-refrigerant, it's the first real breakdown, the ducts are sound, and comfort has otherwise been fine. In those cases a targeted AC repair is the smarter money, and replacing early throws away good remaining life. 916air's approach is to diagnose first, show you the options in writing, and recommend replacement only when the numbers actually favor it.

The picture flips when several factors stack up: a system in its late teens or beyond, a large repair on top of that age, R-22 refrigerant, a pattern of repeated breakdowns, rising summer bills, and ongoing comfort complaints. Any one of those alone might not be decisive, but together they mean you'd be pouring money into a system near the end of its life. At that point a written replacement scope — right-sized equipment per a proper load calculation, duct and electrical needs, permit requirements, and any rebate basics that may apply — is worth comparing against one more repair. When you compare quotes, weigh the whole scope, not just the sticker: sizing method, duct sealing, the refrigerant the new system uses, and warranty all change the real value.

Incentives can shift the replacement math, but only if they actually apply to your project — so treat them as a maybe, not a discount you can bank. Qualifying heat pump projects may be eligible for SMUD residential rebates, but eligibility and contractor participation requirements have to be met, and program amounts and availability change; verify the current program before counting on any figure. On the federal side, the Energy Efficient Home Improvement Credit (25C) applied to qualifying property placed in service through the end of 2025 and should not be assumed available in 2026 unless new guidance extends it. The honest way to use incentives in a decision is to confirm current eligibility first, then factor in only what's real.