Air conditioner BTU calculator
The kilowatts printed on a unit are its output under a standard test condition, typically around 35 °C outdoors and 27 °C in the room. Capacity falls as the outside air gets hotter, because the condenser has less of a temperature difference to dump heat across, so the same unit delivers less on the day you most want it. If your summers regularly run past the test condition, read the manufacturer’s capacity table at your own design temperature rather than the headline figure, and expect a few per cent less than the box claims.
A rule-of-thumb estimate for a single room. A whole-house system or a heat pump should be sized by a proper heat-loss survey.
This sizes on room volume at about 25 watts per cubic metre, then adjusts for insulation, sun and occupancy. A 5 × 4 m room with 2.4 m ceilings is 48 m³, which comes to 1,200 W or 4,095 BTU per hour: a small 1.2 kW unit, or roughly a third of a ton.
How to size an air conditioner
The 25 watts per cubic metre this starts from is a temperate-climate rule for a normally built room, and what it leaves out is the part to understand before trusting it. It has no idea how much of your wall is glass, which is usually the single biggest term in a real cooling load. It does not know which way that glass faces, so a west-facing window baking through a summer afternoon and a north-facing one of the same size score identically. It does not ask whether the room sits under an uninsulated roof or between two heated floors. It has no air-change rate, no equipment load beyond a headcount, and no outdoor design temperature at all. The insulation and sun settings are coarse corrections layered over a volume, not a heat-loss calculation.
That matters most when you compare the answer to the American sizing table, because the two disagree by a lot. A 150 ft² room at an 8 ft ceiling comes out of this tool at about 2,900 BTU per hour. The AHAM table published by ENERGY STAR puts the same room at 5,000, which is 33 BTU per square foot against roughly 19 here. Neither is wrong so much as aimed differently: the American table is sized for a hotter, far more humid design day and for a window unit expected to strip moisture as well as heat.
The familiar "20 BTU per square foot" rule turns out to be an average of that table rather than a property of it. Run the numbers down the rows and the ratio falls steadily, from 33 BTU per square foot at 150 ft², through 24 at 250, 20 at 700, 18 at 1,000, and down to 13.6 at 2,500. A small room carries far more wall and window per square metre of floor than a large one, so the load per unit of floor genuinely does fall as rooms grow. The rule of thumb happens to be right at around 700 square feet and drifts in both directions from there.
A cooling load also has two halves, and only one of them appears in any of these figures. Sensible load is dropping the air temperature. Latent load is condensing moisture out of it, and in a humid climate that can be a quarter of the work. An oversized unit satisfies the thermostat on sensible load quickly and stops, before it has run long enough to pull much water out of the air, which is how a room ends up cold and clammy at the same time. Sizing to the actual load and letting the unit run longer at lower output is the correct answer, and it is also why an inverter unit outperforms a fixed-speed one of the same rating.
What people use it for
- Choosing a portable or split air conditioner for a bedroom
- Checking a quoted unit is not oversized before you commit
- Comparing BTU ratings against kilowatt ratings across brands
- Sizing cooling for a home office with equipment running in it
- Understanding why an American sizing chart gives a different answer
- Working out whether one unit can cover two connected rooms
- Judging a conservatory, which this rule handles badly on purpose
Questions
On this rule, about 19 BTU per square foot at a 2.4 m ceiling. The common "20 BTU per square foot" figure is an average taken across the American sizing table, and it only holds near 700 ft².
Because it answers a different question. The AHAM table published by ENERGY STAR assumes a hotter, more humid design day and a unit that also dehumidifies, so it puts 100 to 150 ft² at 5,000 BTU where this volume rule gives about 2,900.
In a temperate climate the volume figure here is closer to reality and will not leave you with an oversized unit. In a hot, humid climate, or where the room is heavily glazed, follow the higher table.
12,000 BTU per hour, about 3.5 kW. It is the rate at which a ton of ice absorbs heat as it melts over a day, which comes to 3.51 kW, rounded to a tidy 12,000.
No, and it is the most common mistake. An oversized unit short-cycles, never runs long enough to dehumidify, and leaves a room that is cold and damp at once.
Very much, and it is the reason this works on volume. A room with 3 m ceilings needs 1,500 W where the same floor area at 2.4 m needs 1,200, a full 25 per cent more.
Yes, and you have to add it yourself, because there is no kitchen setting here. The American guidance adds 4,000 BTU, roughly 1.2 kW, for cooking load.
Two. Only the third person onward adds anything, at 100 W each. The American table adds 600 BTU, about 176 W, per extra person, so it is more generous here too.
From good to poor is a factor of 1.6: multipliers of 0.8, 1.0 and 1.3. Sun exposure moves it by a further ten per cent either way.
Because it is almost entirely glass, and glass is where the load comes in. Nothing on this panel represents glazed area, so a conservatory can need two or three times what the volume rule suggests.
A single-hose portable does not, in practice. It exhausts room air outdoors and pulls unconditioned outside air in through every gap to replace it, so a meaningful share of the cooling is spent on air you just imported. A dual-hose model avoids that.
Nothing, directly. They describe how much electricity a unit uses to deliver its capacity, not how much capacity you need. Size first, then compare efficiency between units of the same rating.
Only where the opening between them is genuinely large and the far room has no separate load of its own. Air conditioning cools the air it is blown through, and a doorway is a poor duct.
No. Sizing to a rare extreme means running oversized for the other 360 days, with all the short-cycling that brings. Design conditions are deliberately set below the annual peak.
It does not, and it should be nudged upward for one. A room directly under a poorly insulated roof takes a large solar gain through the ceiling that no wall-insulation setting represents.
When the answer is expensive or permanent. A whole-house system or a heat pump you will live with for fifteen years deserves a room-by-room heat-loss calculation to a recognised method, not a multiplication.