What an air conditioner BTU rating actually means
A BTU rating is a rate of heat removal, not a quantity of cold. One British thermal unit raises one pound of water by one degree Fahrenheit, and an air conditioner rated 12,000 BTU/h removes heat from a room at 12,000 BTU every hour. That figure is also one ton of refrigeration, a unit inherited from the ice trade: melting one short ton of ice over 24 hours absorbs about 288,000 BTU, which is 12,000 BTU/h. Every capacity you will see quoted, from a 5,000 BTU/h window unit to a 60,000 BTU/h five-ton condenser, is that same rate.
Sizing means matching that rate to the rate at which heat enters your room at design conditions. Heat arrives four ways: conducted through walls, ceilings and glass; radiated through windows as sunlight; carried in by outdoor air that leaks through the envelope; and released inside by people, lights and appliances. The area-based method on this page is a compressed proxy for all four. It assigns a base rate per square foot for your climate, then adjusts it for the parts of the picture that vary most between otherwise identical rooms: how tall the space is, how well it is insulated, how much sun it takes, and how much heat is generated inside it.
The output you want is a rate that is close to the room's peak load, not comfortably above it. Oversizing is the more common and the more damaging error, because a unit that satisfies the thermostat in six minutes never runs long enough to dehumidify. You end up with a cold, clammy room and a compressor that wears out from short-cycling.
How the formula is built, term by term
Start with the base term, A × q. The rate q is capacity per square foot of floor, and it lands between 18 and 25 BTU/h per sq ft for typical residential rooms. That range is a proxy for outdoor design temperature: a room in International Energy Conservation Code climate zone 6 sees a design dry bulb near 85 °F, while the same room in zone 2 sees 95 °F or higher, and the conduction and infiltration terms scale roughly with that temperature difference.
The ceiling factor H ÷ 8 exists because the published rates are quoted per square foot but the load actually depends on volume for infiltration and on wall area for conduction. Both grow with height. A 400 sq ft room with a 10 ft ceiling gets 400 × 20 × 1.25 = 10,000 BTU/h rather than 8,000.
The insulation factor fins and the sun factor fsun are multiplicative because they scale the conduction and solar terms respectively, both of which are already proportional to area. ENERGY STAR's published guidance is to reduce capacity by 10% for a heavily shaded room and increase it by 10% for a very sunny one, and those are the values in the drop-down.
The last two terms are additive rather than multiplicative, because they do not depend on room size at all. A person at rest gives off roughly 600 BTU/h of combined sensible and latent heat, and the published method charges you for occupants beyond the first two only, since the base rate already contains a nominal couple. The kitchen allowance of 4,000 BTU/h covers the cooktop, oven and refrigerator that a kitchen has and a bedroom does not.
Once you have Q, divide by 12,000 for tons. If you need the airflow the coil should move at that capacity, take the result to the CFM per ton calculator, which converts tonnage into the 350 to 450 CFM per ton band a coil needs to hit its rated performance.
Worked example: a 500 sq ft sunny kitchen with four people
An open kitchen and dining space in a mixed climate measures 500 sq ft with a standard 8 ft ceiling. It is average construction, it faces west with unshaded glass, four people are usually in it, and it contains the cooking appliances.
- Base envelope load. 500 sq ft × 20 BTU/h per sq ft = 10,000 BTU/h.
- Ceiling factor. 8 ÷ 8 = 1.00, so nothing changes.
- Insulation factor. Average construction, so fins = 1.00.
- Sun factor. Very sunny, so fsun = 1.10. The adjusted envelope load is 10,000 × 1.10 = 11,000 BTU/h.
- Occupant adder. Four people, two of them already in the base rate: 600 × (4 − 2) = 1,200 BTU/h. Running total 12,200 BTU/h.
- Kitchen adder. +4,000 BTU/h. Total required capacity 16,200 BTU/h.
- Convert to tons. 16,200 ÷ 12,000 = 1.35 tons.
- Pick a catalogue size. The next standard capacity at or above 16,200 is 18,000 BTU/h, a nominal 1.5 ton unit.
Note how much of the answer is the two additive terms: 5,200 of 16,200 BTU/h, or 32%, comes from people and cooking. Size the same room as a bedroom and you would be looking at 11,000 BTU/h and a 12,000 BTU/h unit. That is why a per-square-foot rule of thumb applied without adders gets kitchens badly wrong.
How to read the result and choose a unit
Round up to the next catalogue size, and no further. Manufacturers list room air conditioners in discrete steps, and the gap between 12,000 and 14,000 BTU/h is 17%. Taking the step above that, to 18,000, puts you 50% over the load, which is deep into short-cycling territory. If your result sits just above a catalogue size, say 12,400 BTU/h, re-examine the inputs before you jump a size: a shaded rather than average exposure moves that room to 11,300 and back under the 12,000 unit.
Check the capacity-per-square-foot output as a sanity test. For a bedroom or living room it should land between 18 and 28 BTU/h per sq ft. Anything above about 35 means the additive terms are dominating, which is legitimate for a small kitchen or a crowded office but is a red flag for a bedroom. Anything below 15 usually means the room area or ceiling height has been entered wrong.
Latent capacity matters as much as total capacity in a humid climate, and this method does not report it separately. A unit sized to this figure will remove moisture adequately if it runs in long cycles. If the room is already humid and the unit satisfies the thermostat quickly, the fix is a smaller unit or a variable-capacity one, not a bigger one. Check the dew point in the space with the dew point calculator: if the indoor dew point stays above roughly 60 °F while the room holds 75 °F, the equipment is not running long enough to dry the air.
Finally, treat this number as a screening estimate. It is a proxy method, and its error bar on any individual room is wider than a proper load calculation's. If you are buying ducted equipment, replacing a system, or pulling a permit, run the Manual J load calculator instead and use the areas and R-values of the actual assemblies.
Base capacity by floor area and climate band
| Floor area (sq ft) | Cool, 18 BTU/h/sq ft | Mixed, 20 | Warm, 22 | Hot, 25 |
|---|---|---|---|---|
| 150 | 2,700 | 3,000 | 3,300 | 3,750 |
| 250 | 4,500 | 5,000 | 5,500 | 6,250 |
| 400 | 7,200 | 8,000 | 8,800 | 10,000 |
| 600 | 10,800 | 12,000 | 13,200 | 15,000 |
| 800 | 14,400 | 16,000 | 17,600 | 20,000 |
| 1,000 | 18,000 | 20,000 | 22,000 | 25,000 |
Every cell is floor area multiplied by the climate rate at the head of its column. Multiply by ceiling height over 8, then by the insulation and sun factors, before adding the occupant and kitchen allowances.
Mistakes that produce the wrong size
- Sizing the whole floor when you are cooling one room. A single head or window unit conditions the space it can throw air into. If the doorway to the next room stays open, include that room's area; if it does not, do not.
- Ignoring the adders and then rounding up anyway. The published method already builds in headroom through the climate rate. Adding a safety margin on top of a rounded-up catalogue size is how a 1.5 ton room ends up with a 2.5 ton unit.
- Treating BTU/h and tons as interchangeable with SEER or EER. Capacity is how much heat the unit moves; SEER2 and EER2 are how much electricity it uses doing so. A 12,000 BTU/h unit at SEER2 15 and one at SEER2 22 cool the same room equally well.
- Using a nominal tonnage as if it were the real capacity. A nominal three-ton condenser matched to a particular coil and blower may deliver 34,600 BTU/h, not 36,000. Use the AHRI-rated capacity of the matched system when it matters.
- Forgetting that a portable unit exhausts through a hose. Single-hose portables pull conditioned air out of the room and pull replacement air in through every crack, which cuts delivered capacity materially. Size generously for single-hose units or choose a dual-hose design.
- Assuming heating and cooling loads match. A heat pump sized to this cooling figure may be well short of the heating load in a cold climate. Check the heating side with the furnace size calculator or a full load calculation.
Where this method comes from and where it stops
The area rate, the ±10% shade and sun adjustments, the 600 BTU/h per additional occupant and the 4,000 BTU/h kitchen allowance are the sizing guidance ENERGY STAR publishes for room air conditioners, developed with AHAM, whose ANSI/AHAM RAC-1 standard defines how those units are rated and tested. That guidance is written for a single room served by a single unit.
It is explicitly not a substitute for ACCA Manual J, which is what building codes and most utility programmes require for ducted equipment. Manual J works from measured assembly areas, U-factors, window solar heat gain coefficients by orientation, a measured or estimated infiltration rate, and separate sensible and latent totals. Where the two disagree on a real house, Manual J is the one to trust.
Where this sits among the other sizing tools
Use the area method when the question is which window unit or which single mini-split head to buy, and the room is a normal residential space with normal glazing. It takes a minute and it is accurate enough to keep you from buying two sizes too big.
Move to a full load calculation when any of these is true: the building is being permitted; the equipment is ducted; the glazing area exceeds roughly 20% of the floor area; the room has a cathedral ceiling or a large exposed floor; or the answer will decide between two catalogue sizes and you cannot afford to guess. A load calculation also gives you the separate sensible and latent split, which is what you need to choose between a standard-efficiency unit and one with a lower sensible heat ratio for a humid climate.
Once the capacity is settled, the rest of the design follows from it. Airflow comes from the sensible load and the supply air temperature difference, which the room CFM calculator handles. Duct sizes come from that airflow and the friction rate available from the blower, which is the job of the duct size calculator and the available static pressure calculator. Envelope improvements feed back into the load through their U-values, which the R-value to U-value converter will give you. Nothing downstream can rescue a badly sized machine, so it is worth getting this number right first.