Repair vs Replace — A 7-Question Framework for HVAC Decisions

Repair-versus-replace is a capital-allocation decision, not a comfort decision. It's the choice between spending a smaller amount now to keep a depreciating asset running, or a larger amount to reset the depreciation clock and (usually) upgrade efficiency, refrigerant, capacity, and warranty coverage in one motion.

The framework we walk every Cypress Cooling customer through is seven questions:

1. What's the specific failed component, and what does that part cost?
2. How old is the system, and how does that compare to its expected service life?
3. What refrigerant is the system on, and what's the trajectory of that refrigerant's price and availability?
4. Is the equipment correctly sized for the home today?
5. Is the existing duct distribution healthy enough to support either path?
6. What's the realistic energy-bill delta of a new system at this home's actual load?
7. What rebate, tax-credit, and financing windows are currently open?

Answer those seven and the decision usually answers itself.

Repair wins when the failed component is well-defined, the cost lands under roughly 30% of a comparable new-system install, the system runs on R-410A (or newer R-454B), and the rest of the cabinet — compressor, condenser coil, evaporator coil, blower motor, control board — has no other open failures. A homeowner with a 7-year-old 16 SEER system that needs a capacitor, a contactor, or a single sensor is almost always a repair.

Replace wins when two or more of these stack up: the system runs on R-22; the compressor or evaporator coil itself has failed; the system is past its expected service life (typically 12–15 years in South Louisiana's salt-air environment); a Manual J load calculation shows the existing equipment is materially over- or under-sized for the current home; or the repair quote crosses 40% of a comparable new install.

The middle ground — "it depends" — usually involves a 10-to-12-year-old R-410A system with one significant failure (a TXV, a blower motor assembly, a condenser fan motor). At that age, we look next at duct health, refrigerant charge trends from prior service tickets, and whether the homeowner expects to stay in the house another five years. If yes to all three, repair. If any one is no, we lay out both numbers and let the homeowner choose.

The "default to replace" trap. Most of the private-equity-backed HVAC chains operating in Greater New Orleans run a replacement-quota model: technicians are scored, paid, and promoted on dollar value of jobs sold per dispatch. Their decision tree starts at replacement and works backward only when the homeowner pushes hard. That model is structurally hostile to a 7-year-old system that needs a $400 part.

The "over-spend on a dying system" trap. The opposite mistake is equally real — homeowners pour $2,800 into a 15-year-old R-22 system because each individual repair "wasn't that bad," and the compressor still fails six months later. Money wasted. The trip-counter on small repairs is the warning sign nobody tracks.

The "bad diagnosis" trap. A homeowner gets quoted "the coil is shot, you need a new system," when the actual problem was a low-side refrigerant leak at the line-set flare nut that a 20-minute reseal would have fixed. Without measured subcooling and superheat on the work order, you can't tell whether the diagnosis is real.

The "right answer, wrong equipment" trap. Replacement is correct, but the replacement quote is for a 4-ton system in a home that actually loads at 2.5 tons. A correctly sized smaller system would have cost less, run longer cycles, dehumidified better in Gulf air, and lasted longer. Replacement without a fresh Manual J is just transferring a sizing error to the next decade.

Five measurements separate a real repair-versus-replace recommendation from a guess:

1. Component cost in writing. The failed part's OEM (or equivalent) price, labor hours, and parts warranty — itemized. If a quote doesn't list the part, treat it as a sales document, not a service document.
2. Refrigerant charge and leak status. Measured subcooling at the liquid line, measured superheat at the suction line, and an electronic-detector or UV-dye leak check if charge is low. A system 30% short on charge with no documented leak isn't a repair candidate yet — it's an incomplete diagnosis.
3. Static pressure across the air handler. Total external static pressure above 0.8 inWC means the duct system is choking the equipment. Replacing the cabinet without fixing duct restriction guarantees the new system has a short, unhappy life.
4. Manual J load calculation. A current Manual J using the home's actual envelope, window glazing, infiltration, and orientation. Houses change — windows get replaced, attics get re-insulated, additions go on, occupancy shifts. The original installer's sizing is a guess about the 2010 house, not yours.
5. Twelve-month billing data. Pull Entergy or Cleco bills for the past year. If summer kWh has crept up 15–25% year-over-year on a 10+ year system, you're paying the replacement in efficiency-loss installments either way.

If a quote doesn't show those five inputs, ask for them in writing before you sign anything — including from us.

The framework, condensed: diagnose first, do the math, then choose. Don't let calendar age alone decide. Don't let a quota-driven technician decide. And don't decide during a 95-degree dispatch when the only thing on your mind is getting cool again.

If you've already been quoted a replacement and the price feels like a stretch — or the diagnosis feels thin — a second-opinion HVAC diagnosis with documented measurements will tell you whether the original quote stands up. We'd rather lose the replacement sale than sell you the wrong answer.

For background, our how air conditioning actually works explainer walks through what each component does and which failures are catastrophic versus routine. Pair that with this framework and the seven questions become an hour of homework instead of a $7,000 coin flip.

The criteria, side by side. A single row landing in the "Replace" column doesn't decide it. Two rows is a coin flip. Three or more is your answer.