An air conditioner that is too small never quite gets there; one that is too large short-cycles, leaves the room clammy and costs more to run than a correctly sized unit. Sizing is mostly arithmetic about the room: how big it is, how much glass it has, how high the ceiling is and how much heat the people and equipment inside add. This sizer walks through those four things and lands on a capacity in BTU/h, then rounds it to a unit size you can actually buy.
Room cooling sizer
Four questions about the room, then a capacity you can buy.
1 · Which room?
2 · Insulation and glazing
3 · Ceiling
4 · Sun
5 · Floor area
6 · People in the room
7 · Kit giving off heat
Result
Capacity needed, and the nearest unit size.
What “size” actually means here
Capacity, in this context, is the rate at which a unit can move heat out of a room — not how much electricity it draws, and not how loud it is. Getting that distinction straight before you start comparing boxes removes most of the confusion, because the big number on the packaging is printed in whichever unit the manufacturer thinks sells best.
BTU/h is the traditional figure: the capacity to move 1 BTU of heat per hour. kW is the same thing in metric, and 1 kW of cooling equals about 3,412 BTU/h. Ton is a refrigeration unit that survives in the trade: 1 ton equals 12,000 BTU/h, which is why a 1-ton machine and a 12k machine are the same size.
So 9,000 BTU/h, 2.6 kW and 0.75 ton all describe roughly one machine. If a listing gives you a kW figure and this calculator gives you BTU/h, divide by 3,412 rather than assuming you have found a different product.
A room air conditioner is rated by how much heat it can move out of a room per hour, in BTU/h. The room dictates the number: a bigger floor area, a higher ceiling, more glass and more equipment all push it up. The sizer starts from the floor area and multiplies through those four factors, then adds a fixed allowance per person and whatever you type in for equipment.
The coloured bar under the result is the part worth studying. It splits the total load into its five contributions, so you can see whether the size is being driven by the room itself or by what is inside it. A living room that jumps a size because of a bank of equipment is a different problem from one that jumps because it faces west.
Reading the number on the box
Specifications sheets are written to be compared, not to be understood, and the two figures that matter most are the two most often confused with each other. Capacity is heat moved per hour; power input is electricity consumed per hour. Confusing them makes a more efficient machine look like a weaker one.
What to take off the label before comparing models
- Cooling capacity in BTU/h or kW. This is the number to compare against your result.
- Power input in W. Smaller is better for the same capacity, and it is not the same as capacity.
- The efficiency rating — a ratio of the two above. It tells you running cost, not how cold the room will get.
- Whether it modulates. A fixed-speed machine is either on or off; a variable-speed one can run at perhaps a third of its capacity, which is where comfort comes from.
- The stated coverage area. Use it as a sanity check only. Manufacturers assume average rooms, and your west-facing top-floor room is not average.
Capacity turns up in three units and it is worth knowing all three, because listings mix them freely. BTU/h is the traditional figure and the one this sizer reports. Kilowatts are the same thing divided by 3,412, so 12,000 BTU/h is about 3.5 kW of cooling. A “ton” of cooling, which you will meet on larger systems, is 12,000 BTU/h — the historical amount of cooling you got from melting a ton of ice in a day.
| BTU/h | kW of cooling | Typically covers |
|---|---|---|
| 9,000 | 2.6 | Bedroom, small study, up to about 18 m² |
| 12,000 | 3.5 | Living room, larger bedroom, 18–30 m² |
| 18,000 | 5.3 | Open-plan living and dining, 30–45 m² |
| 24,000 | 7.0 | Whole floor or a large open space, 45–65 m² |
One trap: the headline number on a unit is its nominal cooling capacity, measured under standard test conditions. The useful output in your room depends on the actual conditions, and the efficiency ratings — EER for steady running and SEER across a season — tell you how much electricity it draws to deliver that output rather than how much cooling you get. A unit with a better SEER costs less to run for the same capacity, which is usually worth more than an extra size step.
Why the load factors are ranges, not constants
The multiplier applied to floor area stands for how hard the room is to cool. A well-insulated, shaded room with small windows sits near 80 W/m²; an average room with normal glazing is around 100; a poorly insulated room with a lot of glass is nearer 130; a top-floor room with the roof directly above it can reach 150. That is a twofold spread from the same floor area, which is why any rule of thumb quoted purely in square metres is unreliable.
If you want to attack the number rather than buy around it, the envelope is where the leverage is — the insulation board calculator prices the material for a wall or loft upgrade, and the room volume calculator gives you the cubic figure that heating and cooling specs are quoted against.
Ceiling height quietly doubles the problem
A load per square metre assumes a standard ceiling. Raise it to 3 m and you are conditioning a fifth more air for the same carpet area; at 3.5 m and above, roughly a third more. This is the factor people forget when they size from floor area alone, and it is why a converted loft with the same footprint as the bedroom below it needs a noticeably larger unit.
Which input moves the answer most
It is worth knowing where the leverage is, because that tells you what to spend money on. Starting from a 24 m² living room with average insulation, a 2.4 m ceiling, average sun, three people and 400 W of equipment, the sizer gives about 11,400 BTU/h. Change one thing at a time:
| Change | New result | Movement |
|---|---|---|
| Insulation: average to poor | about 13,700 BTU/h | +20%, a full size step |
| Ceiling: 2.4 m to 3.0 m | about 13,700 BTU/h | +20%, a full size step |
| Sun: average to west-facing glass | about 12,500 BTU/h | +10% |
| Occupancy: 3 to 6 people | about 12,400 BTU/h | +9% |
| Equipment: 400 W to 1,200 W | about 14,300 BTU/h | +25%, a full size step |
The envelope dominates, and equipment is the one internal factor that can rival it. That ordering matters when you are deciding what to change: insulating a poorly insulated room, or closing a west-facing room to afternoon sun with a blind or external shading, can drop you a size for less than the price difference between units. Adding occupancy rarely moves the answer by much unless the room is genuinely crowded.
Humidity is the half of the load nobody budgets for
A room feels cool because of two things at once: the air temperature falls, and water vapour is condensed out of the air on the cold coil. The second job is called the latent load, and it is why a unit can be technically large enough and still leave a room feeling clammy. Sizing purely on temperature ignores it, which is the most common reason a correctly-calculated number still feels wrong in practice.
The practical consequence is about run time, not capacity. Removing water takes time on the coil, so a unit that reaches the target temperature quickly and then switches off never gets around to drying the air. That is the mechanism behind the advice to avoid oversizing, and it is also why a variable-speed machine usually feels better than a nominally larger fixed-speed one: it runs for longer at a lower output, doing both jobs.
Oversizing is not the safe choice
The instinct to buy the next size up is understandable — nobody wants a machine that cannot keep up on the hottest afternoon of the year. But air conditioning does not work like a heater, where extra capacity is simply spare. A cooling coil that is too large for the room reaches the set temperature in a few minutes and stops, and every one of the problems people blame on an undersized unit is actually caused by that short-cycling.
Size it like this
- Size to the calculated load, then round to the nearest available unit rather than the next one up.
- Prefer a variable-speed (inverter) machine, which can run at part load instead of cycling on and off.
- Confirm the headline figure is cooling output. Electrical input is a different, much smaller number.
- Treat a sunny west-facing room as its own case rather than averaging it in with the rest of the floor.
Not like this
- Adding 20–50% on top of the result for hot days. The load factors already carry a margin.
- Sizing an open-plan space on the largest single zone and ignoring the rest of the volume.
- Expecting a bigger unit to cool the room faster. It cools at the same rate; it just stops sooner.
- Comparing machines on electrical watts and assuming the bigger number is the stronger one.
The instinct is to buy the next size up “just in case”. It is the wrong direction. A unit that is too large reaches the set temperature quickly and switches off before it has run long enough to remove moisture, so the room ends up cool but clammy, and the compressor cycles far more often than it was designed to. Short-cycling costs more electricity and wears the unit out sooner. If the estimate lands close to a boundary, look at which factor pushed it there instead of defaulting upward.
People and equipment are real heat
An adult at rest adds roughly 100 W, which is why occupancy is its own input rather than a rounding error — a living room used by two people and the same room hosting eight are different loads by about 600 W, or over 2000 BTU/h. Equipment is often larger still: a desktop workstation and monitor is 150–300 W, a hob and oven in use far more. If the room is a shared space, the room occupancy calculator gives a defensible headcount before you set this input, and if you are still deciding where the unit will physically go, the furniture fit calculator checks the wall run you have available.
Which form of unit you are actually buying
The capacity number is only half the decision. The same 12,000 BTU/h behaves quite differently depending on how the machine is built, because some forms throw away part of their capacity before it ever reaches the room. This matters most for portable units, where the published figure and the useful figure are genuinely not the same thing.
One box, one opening
- Cheapest per BTU/h and the simplest mechanism, so there is less to go wrong.
- Needs a suitable opening or a wall sleeve; not every window can take one.
- The compressor sits in the room, so it is the noisiest option by some distance.
- Good for a single room where an opening already exists.
Nothing to fix, plenty to give up
- No installation and no structural work, which is the whole appeal.
- Single-hose units pull room air out through the exhaust, drawing warm replacement air back in through every gap — real capacity is well below the number on the box.
- Dual-hose designs avoid that, and are worth insisting on.
- Best treated as a stopgap or for rooms where nothing can be mounted.
The default for a reason
- Quiet, because the compressor lives outside, and efficient because the refrigerant run is short.
- Needs an outdoor unit, a wall to mount it on, and a professional to install and commission it.
- Modulates down to part load, which is what makes it comfortable rather than merely cold.
- Best for bedrooms, home offices and any room used daily.
| Form | Typical range | What it needs | Where the number misleads |
|---|---|---|---|
| Window / through-wall | 5k–18k BTU/h | A suitable opening or wall sleeve | Noise is rarely reflected in the capacity figure |
| Portable, single-hose | 8k–14k BTU/h | A window for the exhaust hose | Useful output is well below the published figure |
| Portable, dual-hose | 9k–14k BTU/h | A window kit with two ducts | Closer to published output, but still below |
| Mini-split | 9k–24k BTU/h | Outdoor unit and professional install | Rated at full output; part-load behaviour is what matters |
Where the unit goes changes how the size performs
Sizing assumes the unit can actually work the air in the room. Two planning decisions affect that, and neither is about capacity. The first is short-circuiting: if the indoor unit blows cold air straight back into its own return within a metre or two, the room thermostat is satisfied while the far corner stays warm. Placing the unit so it throws air across the longest dimension of the room avoids this.
The second is stratification. Warm air collects at the ceiling, so a unit mounted very high on a tall wall is cooling air that has already left the occupied zone, and a room with a 3.5 m ceiling needs more than a proportional increase in capacity to feel right at head height. Ceiling cassettes and floor-mounted units exist precisely because wall units struggle in those rooms. Both of these are worth raising with an installer before you buy, because they can change which form factor you need rather than which capacity.
How you use the room matters as much as how big it is
Two rooms with identical dimensions and identical glazing can still need different machines if one is used for two hours in the evening and the other all day. The difference is not the steady load — it is how quickly the room’s fabric and contents have to be pulled back down each time, and whether the unit ever gets a long enough run to deal with humidity.
| Pattern | What it does to the load | What it means for sizing |
|---|---|---|
| Evening only, two hours | Fabric and furniture start warm each time | Slightly more capacity helps; it shortens the pull-down |
| All day, steady occupancy | Steady load with continuous latent work | Correct size matters more than extra size |
| Weekend and occasional | Long idle periods, then a big pull-down | Do not oversize; short runs leave the room clammy |
| Door to a hallway open often | Effectively a larger volume, refreshed with warm air | Add the adjoining volume rather than the unit size |
The last row is the one that gets missed. A room whose door stands open to an unconditioned hallway is not the room you measured; it is somewhere between that and the whole floor. There is no factor for it here because it is a habit, not a property of the building — but if you know the door is open all day, measure the combined footprint into the room area calculator and use that figure instead.
A unit does not instantly deliver its rated capacity the moment it switches on; the first part of a run goes into cooling the fabric, the furniture and the air that leaked in while it was off. That is why a room used in short bursts feels worse than the same room used continuously, and why buying a size up to fix an evening-only room sometimes works — but it works by shortening the pull-down, not by adding real capacity. If the room is used all day, that trick costs you comfort.
A rough running-cost check
Capacity tells you what to buy; the electricity is what you pay for afterwards. The arithmetic is rough but useful, and it mostly serves to show that the size decision and the bill are only loosely connected — how long the machine runs matters more than how big it is.
Multiply the electrical draw by the hours you expect to run it and by your tariff, and you have a seasonal figure worth having. The reason it belongs on a sizing page is that oversizing makes this number worse rather than better: a machine that cycles on and off at full output spends more of its life at maximum draw than one that settles to a third of its capacity and stays there. Choosing the smaller of two candidate sizes is usually also choosing the lower bill.
Capacity is not the same as consumption, and the two get conflated constantly. A correctly sized 12,000 BTU/h unit in a room that needs 11,400 does not run flat out all day; it modulates, and an inverter unit spends most of its time well below full output. The arithmetic for a sanity check is simple enough to do by hand:
- Take the cooling load in kW from the result — about 3.3 kW for the default living room.
- Divide by the unit’s EER, typically between 3 and 4 for modern inverter units, to get electrical input in kW. At EER 3.5, that is about 0.95 kW.
- Multiply by the hours you expect to run it. Four hours an evening is 3.8 kWh.
- Multiply by your electricity unit price for a daily figure, then by the days in the season.
Treat it as an order-of-magnitude check, not a bill. What it does show is that the difference between EER 3 and EER 4 on a unit that runs 400 hours a season is a meaningful amount of electricity, usually more than the price gap between two capacity steps. Efficiency is the decision worth spending time on once the size is settled.
What the installation actually involves
Sizing is a paper exercise; installing is not. None of the following changes the number on this page, but all of it changes whether the unit you buy on the strength of that number performs the way the calculation assumed. This is a list of things to ask about, not a guide to doing any of them yourself.
Things worth confirming before you order
- Where the outdoor unit will sit, and whether it has clearance to move air on all sides.
- How the condensate will drain. A gravity drain is silent; a pump adds a part that can fail.
- That the wall you plan to mount on is what it looks like — stud, block, or a cavity with services in it.
- Whether the circuit it needs exists, or whether one has to be added by a qualified electrician.
- That the refrigerant line run is within the manufacturer’s stated length for the capacity you chose.
- Who will commission it. A unit that is installed but not commissioned rarely delivers its rated output.
Everything on this page is arithmetic about a room, and it is deliberately kept on that side of the line. Anything that involves fixing equipment to a wall, running a new electrical circuit, joining refrigerant pipework or handling refrigerant is work for a licensed installer, and in most places it is also a legal requirement. Bring the number you calculated to that conversation; it makes the quote comparable between contractors.
If your calculation and the installer’s assessment disagree by more than one unit size, ask which input they treated differently. It is almost always insulation, glazing or the ceiling — the three things a surveyor can see and a web form cannot.
When to stop estimating
This sizer is for choosing between the sizes on a shelf, not for specifying plant. Get a proper room-by-room load calculation when any of these apply: the space is open-plan or spans more than one room; you are sizing for a whole floor or a house; the room has glazing on more than one aspect or a large area of it; there is cooking equipment or a server rack in the space; or the result came out above 24,000 BTU/h. In those cases the factors this tool groups together — orientation, glazing area, infiltration, internal gains — need to be measured separately, which is what a site survey does.
Where this estimate stops being reliable
- This is a planning estimate, not a Manual J style load calculation. It does not treat glazing area, orientation, infiltration or internal gains separately.
- Open-plan spaces that flow into a kitchen or hallway need the whole connected volume counted, not one room.
- It does not model humidity load, which in a damp climate can be the dominant factor.
- Multi-room or whole-home systems change the economics completely; they are sized and commissioned on site.
- Electrical supply, condenser siting, drainage and refrigerant handling are installation matters for a qualified installer.
A second worked example: a top-floor bedroom
The preset rooms cover the common cases. The one that catches people out is the room that is small but punishing: modest floor area, poor insulation underneath the roof, and glass facing the afternoon sun. Here is that case, worked through the same steps the calculator uses.
- Start with the floor areaThe room measures 4.0 m by 3.5 m: 14 m². Small enough that the first instinct is a 9,000 BTU/h machine, and that instinct is what the rest of the calculation overturns.
- Pick the load factor for the fabricIt is directly under the roof with minimal insulation, so 150 W/m² rather than the 100 W/m² you would use for an average room: 14 × 150 = 2,100 W.
- Correct for ceiling heightA sloping ceiling averaging about 2.7 m: factor 1.1, giving 2,310 W.
- Correct for solar gainA west-facing window that takes the afternoon sun: factor 1.2, giving 2,772 W.
- Add the internal gainsTwo occupants at about 100 W each, plus a laptop and a television at roughly 200 W between them: 2,772 + 200 + 200 = 3,172 W.
- Convert and choose a unit3,172 × 3.412 = 10,823 BTU/h. A 9k unit is now clearly short, and the answer is a 12,000 BTU/h machine — the next size up from the instinct, for a room that looked small.
The floor area alone would have suggested 9k. The roof and the west window together added about 32% before a single person was counted, and they are both properties of the building rather than of the room’s size — which is why two rooms with identical floor areas can need different machines.
How the arithmetic works
Base load is floor area multiplied by the load factor. The ceiling factor and the sun factor are applied on top, in that order, because they scale the envelope rather than adding to it. Occupants contribute 100 W each and equipment contributes whatever you set. The total in watts is converted to BTU/h at 3.412 BTU/h per watt, and the unit size is that figure rounded up to the next 3000 BTU/h step, which is how the common 9k, 12k, 18k and 24k sizes are spaced. Everything runs in your browser; nothing is uploaded.
Questions people ask
What size for a 20 m² room?
Somewhere between 9000 and 12000 BTU/h depending on ceiling height, insulation and which way the room faces. Put your own conditions into the sizer — the spread between a shaded ground floor and a sunny top floor at that size is a full unit size.
Does an inverter matter?
Yes. An inverter unit modulates its output rather than switching fully on and off, which is exactly what you want when the load in a room changes through the day. It also softens the penalty of being slightly oversized.
Should the kitchen be included?
If it opens onto the space, yes. Cooking adds a large intermittent load, and a unit sized for the seating area alone will struggle every time the hob is on.
Can one unit cool several rooms?
Not well. Air does not move freely through doorways, so a single wall unit cools the room it is in. Multiple rooms need either one unit each or a multi-split system, and that is a design decision for an installer.
Is 12,000 BTU/h the same as 1 ton?
Yes. A ton of refrigeration is defined as 12,000 BTU/h, a figure that goes back to the cooling you got from a ton of melting ice over 24 hours. You will mostly see tons quoted on larger commercial or whole-home systems, and BTU/h on room units.
Does a higher ceiling just mean a bigger unit?
It means a bigger number here, but the better answer is often a different unit type. Above about 3 m the warm layer at the ceiling matters more than the extra volume, so a floor-mounted or cassette unit that works the occupied zone directly can outperform a larger wall unit mounted high.
How much difference does one extra person make?
About 100 W, or roughly 340 BTU/h. Small on its own, which is why a room used by two people and the same room hosting six differs by around 1,000 BTU/h — enough to matter only if you were already near a boundary between sizes.
Should I size for the hottest day of the year?
No. Sizing for a once-a-decade afternoon means the unit is oversized for the other ninety-nine days, and that is when the short-cycling problems show up. The load factors already carry a margin for ordinary hot weather. If you genuinely need the room comfortable through an extreme week, the better answer is a variable-speed machine that can run high when asked and low the rest of the time.
Will a bigger unit cool the room faster?
Marginally, and not in the way people expect. Cooling rate is governed by how much air the unit moves across its coil, and a larger unit moves more air but also switches off sooner. The felt difference between arriving at 24 °C in twelve minutes rather than sixteen is small; the felt difference between a room that is dry and one that is clammy is not.
Does closing the curtains actually change the number?
Yes, and it is the cheapest change available. Direct sun through glazing is one of the largest single terms in a domestic cooling load, so a room with the blinds down through the afternoon is a genuinely different room from the same one with them up. If you intend to do that, the shaded factor is honest rather than optimistic.