- The quick screen: repair quote × system age > $5,000 → lean toward replacing. It’s triage, not an answer.
- The real test: 15-year cumulative cost of keeping (energy + rising repairs) vs. replacing (net install + lower energy). Break-even under ~6 years is a clear replace; past ~12, repair and bank the difference.
- Replacing a ~SEER 10 unit with 16–18 SEER2 cuts that system’s electricity use roughly 40–50% for the same cooling.
- Typical service lives: central AC 12–17 yrs, heat pumps 10–15, gas furnaces 15–25 — the final years carry the worst repairs and efficiency.
The $5,000 rule — useful, but incomplete
Contractors use a quick screen: multiply the repair quote by the system's age. If the product tops roughly $5,000, lean replace; below it, lean repair. An $800 compressor repair on a 12-year-old AC scores $9,600 — replace territory. A $300 capacitor on a 7-year-old unit scores $2,100 — repair it and move on.
The rule works as triage because age predicts two things at once: more failures ahead, and worse efficiency behind. What it can't see is your energy picture — climate, utility rates, and how inefficient the old unit actually is. Those routinely swing the answer by thousands of dollars in either direction.
The three costs of keeping an old system
The repair quote is the only number a failing unit shows you — but it's the smallest of three. Total all of them and "keep it" often turns out to be the expensive choice wearing a cheap disguise:
Energy, at a degraded efficiency
A 2005-era AC was rated ~SEER 10 and has drifted lower since. A new 16–18 SEER2 unit does the same cooling on 40–50% less electricity — $500–$1,200/yr in hot climates. Same story for an 80% vs. 96% AFUE furnace.
The largest cost, and the least visibleRepairs that climb every year
Capacitors and contactors early; fan motors and boards next; then the compressor or heat exchanger — the repairs that cost a third of a new system. It's a rising curve, not a flat fee.
Single-year budgets understate the trendRisk you can't schedule
Old units die in the July heat wave or the January freeze — peak prices, booked contractors, no time for three bids. Planned shoulder-season replacement runs 10–20% cheaper. Add R-22 scarcity (pre-2010 ACs) and heat-exchanger CO risk (15–20-yr furnaces).
Unplanned failure carries a premiumThe break-even method
The honest comparison is cumulative cost over time. Keep = this year's energy at the old unit's efficiency, escalating with utility prices and degradation, plus the rising repair curve. Replace = the net install cost up front (after rebates), then the new unit's lower energy use, nearly repair-free under its parts warranty. Plot both for 15 years; the year the keep line crosses above the replace line is your break-even.
A rough decoder for where your break-even lands:
Strong yes. You'll bank savings for most of the new system's 15–20-year life — waiting is the expensive option.
Solid if the unit is past mid-life, unreliable, or runs R-22 — avoided repairs and avoided emergencies tip the scale.
Marginal on money alone. Comfort, humidity, quiet, and CO₂ may still justify it — decide with full information.
A legitimate result — common when replacing an already-efficient unit in a mild climate. Repair, invest the difference, and reassess in a few years.
When repair is the right call
Replacement isn't always the answer. Any one of these tilts the math back toward fixing what you have:
Under ~10 years oldA cheap, isolated failure on a young unit — repair it and move on.
Already efficientSEER 14+ / 90%+ AFUE leaves thin energy savings to harvest.
Selling soonBuyers rarely pay full freight for a new system — let the payback belong to someone who'll stay.
Mild climate, small billsA 45% cut of a small number is a small number. The calculator will say so honestly.
Sometimes the calculator's answer is a payback that never arrives — and that's worth knowing before you spend five figures, not after.
Frequently asked questions
Is the $5,000 rule (repair cost × age) reliable?
It's a useful first screen but not a decision. It captures failure risk and age, yet ignores energy costs — the largest line item over time. A unit that passes the $5,000 test can still be worth replacing in a hot, high-rate climate, and one that fails it can be worth repairing in a mild one. Run the break-even math.
How long do AC units, heat pumps, and furnaces last?
Typical service lives: central AC 12–17 years, heat pumps 10–15 (they run year-round), gas furnaces 15–25. Coastal salt air, poor maintenance, and undersized ducts shorten these; the last few years of life usually bring the worst repair bills and efficiency.
Should I replace my AC before it dies?
Often yes, once it's past ~12–15 years old. A planned replacement lets you compare quotes in shoulder season, capture rebates, and choose the right size — an emergency replacement in a heat wave usually costs more and gets less scrutiny. The keep-vs-replace chart shows whether waiting is costing you money in the meantime.
Does an old system really use that much more energy?
Yes, on two counts: the rating gap (SEER 10 vs. 17 SEER2 is roughly a 45% cut in cooling energy) and degradation (older coils and drifting refrigerant charge typically cost another ~0.5–1% of efficiency per year). Together, replacing a 20-year-old unit commonly cuts that system's energy use by 40–55%.