How the two differ in kind
Every heating and cooling system is sized to a number, and there are two ways to get one. The rule of thumb multiplies floor area by a figure — so many square feet per ton, so many watts or BTU per square metre — and returns a capacity in seconds. A load calculation adds up where the heat actually goes: through each wall, roof and window at its own U-value and area, through infiltration at the building's measured or estimated leakage, and out of the internal gains from people, lighting and appliances, all at a defined design condition.
The difference is not simply accuracy. The rule of thumb carries a systematic bias, because the figures in circulation were calibrated on a housing stock with single glazing, little insulation and considerable air leakage. Apply them to a house built or upgraded to any modern standard and they return a capacity substantially larger than the building needs — and the error goes in the direction that is hardest to argue against, because the equipment plainly has plenty of power.
Oversizing is where the cost lands, and it is a comfort cost rather than an energy one. A cooling system removes moisture only while it runs: the coil has to stay cold long enough for water to condense on it and drain away, and that takes a sustained run rather than a burst. An oversized unit drives the room temperature down to the setpoint in a few minutes and shuts off with the moisture still in the air — so the house is cold and clammy, the occupant lowers the thermostat, and the runs get shorter still.
Heating oversizing is less dramatic and still real: short cycling, temperature swings, wear on components that pay for every start, and on a heat pump the loss of the low, steady output where it is most efficient. The honest middle ground is that a screening load calculation — real areas, real U-values, real design temperatures — gets you close enough for most ordinary buildings, and a full room-by-room calculation earns its fee when the building is unusual, when the ducts are being designed, or when the equipment is expensive enough that getting it wrong matters.
The factors that actually differ
Show
| Load calculation from the building | Area-based rule of thumb | |
|---|---|---|
| What it uses | Areas, U-values, orientation, infiltration, internal gains and the local design temperatures — the building as it actually is. | Floor area and a coefficient, with no input from the envelope at all. |
| Direction of the error | Either way, and bounded by how good the inputs are. | Almost always high on anything built or upgraded to a modern standard, because the coefficients were calibrated on leaky buildings. |
| Sensitivity to glazing and orientation | Captured directly. A west-facing glass wall and a blank north wall of the same area produce very different numbers. | Invisible. Two houses of the same footprint return the same answer whatever their windows do. |
| Room-by-room results | A full calculation gives per-room loads, which is what duct design and radiator sizing need. | One number for the whole building, which cannot size anything downstream. |
| Effort | Minutes for a screening pass with measured areas; hours for a full room-by-room calculation. | Seconds. |
| Design condition | An explicit design temperature — the condition exceeded only a small percentage of hours, not the record extreme. | Implicit and usually unstated, so the same coefficient is used in two different climates. |
| Latent load | Separated from sensible load, which is what decides whether the equipment will actually dehumidify. | Not addressed. The entire moisture question is outside the method. |
| What it is good for | Selecting equipment, designing distribution, and defending the choice afterwards. | A sanity check — knowing within seconds whether a proposal is in the right order of magnitude. |
| After an envelope upgrade | Reflects it immediately. New windows and insulation change the inputs and the answer. | Does not move at all, which is why replacement equipment is so often sized to the old house. |
| Cost of being wrong | Bounded by input quality — a wrong U-value moves the answer by a knowable amount. | Unbounded in the oversize direction, and oversizing is the expensive kind of wrong. |
Which one, and when
Choose load calculation from the building when…
- The building is new, recently upgraded, or unusual in any way — the cases where the rule of thumb is furthest out.
- The ducts or emitters are being designed, which needs room-by-room loads rather than a total.
- Humidity is a concern, so the sensible and latent split matters.
- The equipment is expensive enough that a fee for a proper calculation is small against the capital.
Choose area-based rule of thumb when…
- A quick sanity check on a quote — is this proposal roughly right, or wildly out?
- An early feasibility question where nothing is known about the envelope yet.
- Budget planning before design, where the order of magnitude is all that is needed.
- As the second opinion that prompts a proper calculation when the two disagree sharply.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- Why is oversized air conditioning a problem?
- Because dehumidification depends on how long the equipment runs, not on how powerful it is. Moisture is removed when air passes over a coil cold enough to condense water out of it, and that takes a sustained run: the coil has to get cold, stay cold, and shed the condensate. An oversized unit pulls the room temperature down to the setpoint in a few minutes and shuts off, having removed very little water, so the house ends up cold and damp — and the natural response, turning the thermostat lower, makes it worse by shortening the runs further. Add to that the wear from constant starting, the temperature swings from short bursts, and the higher purchase price, and the case for buying capacity you do not need falls apart entirely.
- Is undersizing better then?
- No, and the argument is not that small is good — it is that RIGHT is good and the rule of thumb cannot find it. An undersized system runs continuously in design weather and never quite reaches setpoint, which on the hottest few days a year is uncomfortable and on a heating system in a cold snap can be a genuine problem. What makes undersizing less catastrophic than the equivalent oversizing is that it fails only at the extremes, whereas an oversized unit fails at its comfort job every single day. The design target is a system that runs nearly continuously at the design condition and modulates or cycles gently the rest of the time — which is exactly what a load calculation is trying to identify.
- What design temperature should I use?
- The published design condition for your location, not the record extreme — and the distinction is the whole point. Design temperatures are defined as the condition exceeded only a small percentage of hours in a year, so sizing to them means the system just meets the load in the weather it will actually see, and runs comfortably long the rest of the time. Sizing to the coldest or hottest temperature ever recorded guarantees a system oversized for the thousands of hours that are not that day, which produces the short-cycling problems described above in exchange for a handful of hours of marginal benefit. If those extreme hours genuinely matter — a process, a vulnerable occupant — the answer is supplementary capacity that engages only then, not a permanently oversized main system.
- How far out is the rule of thumb, typically?
- Far enough to change the equipment selection, and the gap widens the better the building is. The coefficients in general circulation come from a housing stock with little insulation, single glazing and substantial air leakage, so they encode a heat loss and gain per square metre that a modern envelope simply does not have. A well-insulated, airtight house with good glazing can need a fraction of what the rule returns, which is why the pattern of a new low-energy house being fitted with equipment sized for its predecessor is so common. The direction is consistent enough to be useful in itself: if a rule-of-thumb figure and a load calculation disagree, the rule is almost always the higher of the two, and the building is almost always the reason.
- Does a screening calculation replace a full Manual J?
- It replaces the rule of thumb, which is the improvement that matters most, and it does not replace a full room-by-room calculation where one is needed. A screening pass with real areas, real U-values and a real design temperature gets the whole-building load close enough to select equipment sensibly for an ordinary building — and it is enormously better than an area coefficient because it responds to the envelope. What it does not give you is per-room loads, and those are what duct sizing, radiator sizing and zoning all depend on: without them the distribution is being guessed even though the equipment is not. Use the screening calculation to decide whether the proposal is sane and what size bracket you are in; commission the full calculation when the distribution is being designed or the building is unusual.
- What comes after the load calculation?
- Equipment selection and then distribution design, and both are separate exercises that the load number feeds rather than answers. Selection is not simply picking the nearest nominal capacity: a unit's actual output varies with outdoor temperature, and its split between sensible and latent capacity varies with indoor conditions, so the selection is made against performance data at the design condition rather than against the badge rating. Distribution then has to deliver each room's share, which means duct sizes, static pressure and register selection, or pipe sizes and emitter outputs. A correctly sized unit on a distribution system that cannot deliver its output is a common and frustrating outcome, and it is why the room-by-room numbers matter beyond the headline total.
- Does a heat pump change the sizing question?
- It sharpens it, because a heat pump's capacity is not a fixed number — it falls as the outdoor temperature drops, which is precisely when the heating load is highest. That makes the selection a matter of matching two curves rather than one number to another, and it introduces a balance point: the outdoor temperature below which the unit alone can no longer meet the load and supplementary heat engages. Sizing from a rule of thumb cannot participate in that conversation at all. Oversizing has a specific penalty here too: a variable-capacity heat pump is most efficient turned down and running steadily, so a unit too large for the building spends the season cycling in the part of its range where it performs worst.
- The existing system was this size and it worked — why change?
- Because it is evidence about the old building, not the new one, and it may not even be good evidence about the old one. If the house has had windows, insulation or air sealing done since, the load has genuinely fallen and replacing like for like installs an oversized unit deliberately. If nothing has changed, the existing size still only tells you that the old unit was big enough, not that it was right — an oversized system that short-cycled for twenty years looks identical, from the outside, to one that was sized correctly. What the old system does provide is a useful cross-check: if a load calculation comes out wildly different from the incumbent and nothing about the building has changed, that disagreement is worth investigating before anybody orders equipment.