AC's been running nonstop this summer — when it's time to replace vs. just repair?

[JimmyMechanic]

Our unit is about 12 years old and just needed another repair (capacitor this time, not too bad) but it's the third service call in two years. Compressor and coils are still original.

I keep going back and forth — the repairs individually aren't crazy expensive, but they add up, and I know newer units are way more efficient (SEER2 ratings weren't even a thing when ours was installed). At what point does it stop making sense to keep patching it? Is there a rule of thumb on repair cost vs. age, or did you all just wait until it died completely?

Curious what pushed people here to finally pull the trigger on a new system vs. squeezing a few more years out of the old one.

[handyandy]

I usually tell people not to replace an AC just because it's 10–15 years old. I'd look at the whole picture.

A few things I consider:

• Frequency of repairs. Three service calls in two years starts getting my attention, even if each repair isn't terribly expensive.
• Type of repairs. A capacitor is a wear item and isn't a reason to replace the system. A failed compressor, leaking evaporator coil, or major refrigerant leak is a different story.
• Repair cost vs. replacement cost. A common rule of thumb is that if a repair is 30–50% of the cost of a new system, replacement is worth serious consideration—especially on a unit over 10 years old.
• Performance. If it's still cooling well, humidity is under control, and your electric bills aren't outrageous, there's no reason to rush into replacing it.

At 12 years old, you're kind of in the "start planning" phase. I wouldn't replace it because of a capacitor, but I would start getting quotes so you're ready if the next repair is something major. Replacing it on your schedule is usually a lot less stressful (and sometimes less expensive) than doing it in the middle of a 100° heat wave after a compressor fails.

If it were my system, I'd probably keep running it, set aside money for a replacement, and reassess if the next repair is a compressor, evaporator coil, or another expensive component.

[HankBriggs]

A rule of thumb a lot of HVAC techs use: multiply the age of the unit by the repair cost — if that number is over roughly $5,000, it's leaning toward replace. So a 12-year-old unit needing a $400 repair (5,000 ÷ 400 ≈ way under 12... math the other way: 12 x cost) — simpler version some use is the "50% rule": if a repair costs more than 50% of what a full replacement would run, replace instead.

A few other things worth weighing besides the repair cost itself:

• Efficiency gap — anything installed before SEER2 ratings (2023) is very likely well below current minimums. Depending on how old and what SEER your unit started at, a new system can mean a real, noticeable drop in your cooling bill, not just fewer repair calls.
• Refrigerant type — if your unit runs on R-22 (common up until around 2010), that refrigerant is being phased out and any repair involving a recharge gets more expensive every year since supply is shrinking. R-410A units don't have this urgency yet.
• Frequency, not just cost — three service calls in two years on a 12-year-old unit is a signal worth paying attention to on its own. Once one component starts failing, others at the same age/wear level tend to follow — it's not usually just bad luck.
• Get a load calculation before you replace — if you do decide to replace, insist on a proper Manual J load calc rather than a straight swap of the same tonnage. A lot of older systems were oversized to begin with, which itself causes short-cycling and extra wear.

At 12 years old with rising repair frequency, you're right in the zone where it's worth getting a real replacement quote to compare against — not necessarily to pull the trigger immediately, but so the repair-vs-replace math is based on real numbers instead of a rule of thumb.