SEER2 Savings Calculator

Divide by the factor published for that equipment class, and multiply to go the other way. The factors were developed by AHRI and appear, in the SEER = SEER2 × factor direction, in the California Energy Commission's Rating Conversion Table: 1.049 for a split air conditioner under 45,000 Btu/h, 1.051 for a split air conditioner at or above that, 1.049 for a split heat pump, 1.045 for a packaged air conditioner or heat pump, 1.026 for a space-constrained air conditioner, 1.008 for a space-constrained heat pump, and exactly 1.000 for a small-duct high-velocity system. So a 15.0 SEER2 split system is 15.73 SEER, the same rating on a packaged unit is 15.67, and on small-duct high-velocity it is 15.00 unchanged. RESNET publishes the same conversions in MINHERS Addendum 71f in the opposite direction and more coarsely rounded, and the two reconcile on every class to rounding.

Not as such, and it would be unfair to call it invented. 0.95 is exactly what RESNET MINHERS Addendum 71f publishes for a ducted split system, and it is the reciprocal of AHRI's 1.049 to within half a percent — for that one class it is right. The error is applying it to everything. RESNET's own table, on the same page, gives 1.00 for a ductless system, 0.97 for a ducted space-constrained air conditioner and 0.99 for a ducted space-constrained heat pump. Use 0.95 on small-duct high-velocity equipment and you have moved the rating by five percent for no reason. The weaker rule in circulation is "divide by 1.05", equivalent to multiplying by 0.9524, which corresponds to no published class at all — it is a rounding of a rounding.

The homeowner's own bills, if you can get them. The default here is 1,000 hours, which is the representative average annual cooling use in the DOE test procedure at 10 CFR 429.16 — the same national assumption that produced the SEER2 figure on the nameplate, which makes it the internally consistent starting point rather than a guess. It is emphatically not any particular house's number: a system in Phoenix runs far longer than that and one in Seattle far less. There is no city or climate-zone dropdown on this page on purpose. Equivalent-full-load-hour tables do exist, but the published figures sit inside individual state Technical Reference Manuals built on methodologies that disagree with one another, no two independent sources were found that agree on a single national table, and ACEEE has published a critique of the concept. Rather than ship a table that looks authoritative and is not, the hours are an editable field and the guidance is to derive them from actual summer consumption.

Whatever appears on the customer's bill, including delivery charges, not the supply rate alone. The default of 18.44¢ per kWh is the US residential average for May 2026 from the EIA's Electric Power Monthly, and it is dated on the page because it changes every month. More importantly it is a national average across an enormous range: residential rates run from roughly 12¢ per kWh in North Dakota to about 52¢ in Hawaii, so a savings figure computed on the average can be off by a factor of two either direction. No state table is baked in here, because a table of rates that moves monthly quietly rots and a stale number on a customer-facing sheet is worse than an obviously editable one.

The difference between the two options you quoted, not the price of the job. If the baseline system installs for $9,200 and the higher-efficiency one for $11,000, the incremental cost is $1,800 — because the homeowner is replacing the equipment either way, and the only money the efficiency has to earn back is the money the efficiency added. Divide $9,200 of unavoidable replacement cost by an annual saving and you get a payback period of decades, which is arithmetically fine and answers a question nobody asked. Note also what a simple payback leaves out: it is undiscounted, ignores financing cost, and applies no rebate or tax credit. Utility rebates in particular can move a break-even by years and are the first thing worth checking before the number is presented at all.

Rarely exactly, and the errors mostly run in the customer's favour. This assumes both systems deliver the same cooling over the same season at the same thermostat setting, which is the same assumption behind every published figure on this subject, and it makes no allowance for four things that matter. An existing system has usually lost efficiency to age, dirty coils and drifting refrigerant charge, so it is performing below its own nameplate rating and the real gap is wider. Leaky or undersized ductwork wastes some of the improvement on both systems. If the old equipment was oversized, the new one will short-cycle less and dehumidify better in ways a seasonal rating does not capture. And a new system commissioned properly — charge weighed in, airflow measured — performs closer to its rating than one that is not. What this page will not do is inflate the number to account for any of that, because a figure that has been quietly padded is one you cannot defend in six months.

No, and this tool is built to say so. Payback is the extra cost divided by the annual saving, so it lengthens when electricity is cheap, when the cooling season is short, when the equipment is small, or when the efficiency premium is large. Past roughly fifteen years the break-even has run past the service life of the machine and there is no economic case left — the page flags that explicitly rather than printing a confident-looking figure. There is also the case where the proposed rating is not higher than the existing one at all, which happens more than people expect: a failed compressor, a refrigerant transition, capacity that was wrong from the day it was installed. Then the annual saving is zero or negative, and instead of an infinite or negative payback period this returns nothing and says why. Comfort, humidity control, noise, staging and warranty are all real reasons to install better equipment, and they hold up better than an economic argument that does not.