On this page ▾
Why the rules of thumb miss
Rules of thumb fail because they collapse a four-variable problem down to one number. "Age times repair cost equals $5,000" gives the exact same answer for a 10-year-old system needing a $500 repair and a 5-year-old system needing a $1,000 repair — but those are wildly different situations. A 5-year-old system with a $1,000 repair is almost always worth fixing (it's still under some warranties, has 10+ years of expected life ahead, uses current refrigerant). A 10-year-old system with a $500 repair might also be worth fixing — or might be signaling the next failure is around the corner.
The honest framework needs four inputs. Each variable is either "good" or "bad" (with a gray zone we'll flag). If two or more are bad, replacement is usually the right call. If just one is bad, repair wins almost every time.
The four variables
Variable 1 — Age
Age is the strongest single predictor of remaining useful life. A modern residential AC has a design life around 15 years in the River Valley — hot-humid climates run equipment harder than dry climates, which is why our lifespans skew a bit shorter than the national average.
- Under 8 years old: good. Repair almost always wins.
- 8-12 years: gray zone. Depends on the other three variables.
- 12-15 years: bad. Leaning toward replace unless the repair is trivial.
- 15+ years: bad. Replace unless it's a $50 capacitor and you're truly buying just one more summer.
Variable 2 — Repair cost as % of new install
Not the raw dollar amount — the ratio. A $2,000 repair on a $15,000 new install is a different decision from a $2,000 repair on a $9,000 new install. Rough guidance:
- Under 25% of new install: good. Repair almost always makes sense.
- 25-40%: gray zone. Depends on age and remaining life.
- 40-60%: bad. Repairing starts to feel expensive relative to a fresh 15-year warranty.
- 60%+: bad. Almost always replace, unless the equipment is under 5 years old.
Variable 3 — Refrigerant type
This one changed dramatically in the last few years. See our R-454B vs. R-410A refrigerant change guide for the full story, but the short version:
- R-22 (pre-2010 equipment): bad. Refrigerant has been phased out; new R-22 is unavailable and reclaimed R-22 runs $150+/pound. Any refrigerant-side repair is throwing money at a system that's on borrowed time. Replace.
- R-410A (2010 to 2024 equipment): gray zone. New R-410A is still available but is scheduled to phase down through 2035, and prices have already climbed. A capacitor repair? Fine. A compressor or coil replacement (both of which involve refrigerant handling)? That's the point where repair economics get thin.
- R-454B (2025+ equipment): good. Current-generation refrigerant. Repair, obviously.
Variable 4 — Comfort and energy trend
Even a system that mechanically "still works" can be signaling the end. Three trends to watch:
- Rising utility bills: if summer electric bills have crept up 15-25% over the last three years with no rate change and no lifestyle change, the equipment is losing efficiency.
- Persistent comfort complaints: hot upstairs, clammy in humidity, cold spots downstairs, rooms that never quite match the thermostat. Sometimes a duct problem, but on aging equipment often a capacity problem too.
- Short-cycling or long-cycling: either running very briefly and shutting off, or running continuously without hitting setpoint on hot days. Both point to a system that's lost the capacity it once had.
Any one of those three going bad is enough to score this variable as "bad."
The framework: count the bad variables
Score each variable independently. Add up how many are in the "bad" column:
| Bad variables | Usual recommendation |
|---|---|
| 0 | Repair, obviously |
| 1 | Repair — a single weak variable is not a strong replacement signal |
| 2 | Replace unless one of them is borderline. This is the tipping point. |
| 3 | Replace |
| 4 | Replace, no ambiguity |
The framework isn't rigid — a heavily borderline "bad" (age 8 years, or repair cost 26% of new) counts differently than a solidly bad one (age 16 years, or repair cost 65% of new). But used honestly, it gets to the right answer far more often than any single-variable rule of thumb.
Four real scenarios from our service area
Scenario 1 — Young system, expensive repair
6-year-old R-410A split system. Needs a capacitor. $200 all-in.
- Age: 6 years → good
- Repair cost: 2% of new → good
- Refrigerant: R-410A → gray (but the repair isn't refrigerant-side)
- Comfort trend: no complaints → good
0 bad variables. Obviously repair. Anyone recommending replacement on this system for a $200 capacitor is selling, not diagnosing.
Scenario 2 — Middle-aged system, major failure
10-year-old R-410A system. Compressor has failed. $3,200 for a warranty compressor plus labor and refrigerant recovery/recharge. New system install would run around $11,000.
- Age: 10 years → gray, leaning bad
- Repair cost: ~29% of new → gray
- Refrigerant: R-410A + compressor repair = full refrigerant recovery/recharge on aging refrigerant → bad
- Comfort trend: no other complaints → good
Two variables leaning bad, one solidly bad — call this 2 of 4. Replace. Spending $3,200 on a 10-year-old system is buying maybe 4-5 more years of life on a refrigerant platform that's phasing out. The same $3,200 applied to a new install (via financing or as effective down payment) gets 15 years of a modern refrigerant with a full warranty.
Scenario 3 — Older system, coil failure
14-year-old R-410A system. Evaporator coil has developed a leak. Replacement coil $1,800 + labor. No other complaints, still cools when it has refrigerant.
- Age: 14 years → bad
- Repair cost: ~20% of new → good in isolation
- Refrigerant: R-410A + coil repair (refrigerant recovery involved) → bad
- Comfort trend: no complaints → good, but at 14 years the compressor is on borrowed time
2 of 4 bad. Replace. Spending $2,500 on an evap coil in a 14-year-old system is a bet that the compressor holds for another 3-5 years. If it doesn't, the coil investment goes with it. Meanwhile the new-system economics are getting better every year the equipment ages.
Scenario 4 — Old R-22 system, small repair
17-year-old R-22 system. Contactor is pitted, blower capacitor is weak. About $350 in parts and labor to fix both. Otherwise the system still runs and hasn't lost efficiency noticeably.
- Age: 17 years → bad
- Repair cost: 3-4% of new → good
- Refrigerant: R-22 → bad
- Comfort trend: still comfortable → good
2 of 4 bad — but the honest call here is nuanced. The repair is trivial and the failed parts aren't refrigerant-side. Fix the $350 today, but start budgeting for replacement now, not later. The moment R-22 refrigerant is needed on this system, the math flips hard toward replacement. See our AC replacement cost guide for what to plan for.
The two myths that cost homeowners the most money
The July emergency-replacement trap
The other version of this same trap: "we can install tomorrow if you sign tonight, otherwise we're booked out three weeks." That's almost always false urgency. The reality is that every reputable installer is booked out in a July heat wave — the ones offering same-day installs are moving inventory, not managing a schedule.
When to get a third opinion
Get another set of eyes on a quote any time any of the following is true:
- The recommendation is to replace based on a single symptom ("the capacitor is bad, so you need a new system") without walking through the four variables
- The recommended repair is over $3,000 and the recommendation is to repair (not replace) — that's usually the tipping point where replacement math starts to win
- You don't understand the reasoning. If a contractor can't explain why replacement is the right call in terms of age + cost + refrigerant + comfort trend, that's a sign the reasoning is either bad or hidden
- You feel pressured to sign the day of the diagnostic
- The quote is dramatically different — either much higher or much lower — than your gut expectation
We'll happily give a second opinion on someone else's quote. It's part of the job.
Keep reading
If the framework points at replacement, our AC replacement cost in Arkansas guide walks through real 2026 pricing bands for the River Valley. If your equipment is R-410A and you want to understand the refrigerant transition better before deciding, see R-454B vs. R-410A. And if you're replacing anyway, the biggest one-time decision is AC-plus-furnace vs. heat pump — walk through the trade-offs in our AC vs. heat pump comparison. When you're ready to talk numbers on your specific system, request a service visit or ask about financing.
Frequently asked
What's a fair labor rate for a compressor replacement in the River Valley? +
Labor on a warranty compressor typically runs $800-$1,500 depending on system tonnage, access difficulty, and how much refrigerant recovery is required. Add the refrigerant recharge itself ($200-$600 depending on system size and refrigerant type). Add a compressor start kit if not included. All-in, a warranty compressor swap on a mid-size residential unit typically lands in the $1,500-$2,500 range. Above $3,000, ask for a written line-item breakdown.
If I replace the outdoor unit, do I have to replace the indoor coil too? +
Almost always, yes. The outdoor condenser and the indoor evaporator coil are a matched refrigerant system — the coil is engineered for the specific refrigerant pressure, flow rate, and temperature of the paired condenser. Replacing just the outdoor unit and expecting an older coil to work correctly is a common shortcut that voids the manufacturer warranty and usually delivers 20-30% less capacity than the rated spec. It also matters more than ever on the refrigerant transition — an R-454B condenser cannot legally be paired with an R-410A coil.
Is there a fifth variable I should be thinking about? +
Financing terms, honestly. If you can put a repair on a credit card at 0% for 12 months but a replacement requires committing to a 5- or 10-year loan, that shifts the math for some households. It doesn't change the technical case for repair or replace, but it can change the timing of when you make the switch. See our financing page for what we offer.
Can you rebuild an old system's efficiency with just a tune-up and a coil cleaning? +
Partly. A dirty coil, a low charge, or a failing capacitor can all rob 10-20% of a system's rated capacity, and fixing those can restore most of that. What you can't restore is fundamental aging — worn compressor bearings, degraded refrigerant flow, oxidized copper. A tune-up on a 15-year-old system can make it run as well as a 15-year-old system should run, but that's still noticeably less efficient than a new one. Worth doing if you're not ready to replace; not a substitute for eventual replacement.
What about repair-plus-service-agreement vs. replace? +
Bundling a repair with a multi-year service agreement is a real option for equipment in the gray zone. If the framework says "maybe replace," adding a maintenance plan (see our plans) can shift the balance toward repair, because tune-up catches at year 11 and year 12 catch the next failure early. It doesn't change the framework — it just adds one more input on the "repair a little longer" side. We'll flag when we think it applies.