Scope. This guide describes ACCA Manual J methodology in general terms, for residential work in the US. It is not ACCA approved or certified, it does not reproduce the manual, and it is not a stamped design. Manual J is a registered mark of ACCA; this site is independent of it.
The short answer
Manual J is the Air Conditioning Contractors of America’s standard procedure for calculating a home’s peak heating and cooling loads — the heat it loses on a cold design day and the heat it gains on a hot one, in Btu per hour. The current edition is the 8th Edition, published as the ANSI/ACCA 2 Manual J standard, and it is what people mean when they say “MJ8.”
Two things it is not. It is not an equipment size: picking the furnace, air conditioner or heat pump that fits those loads is a separate step, Manual S. And it is not square feet per ton. Floor area is one input among dozens, and rarely the one that moves the answer most.
What it actually calculates
A load calculation walks around the house surface by surface and adds up every path heat takes in or out. The physics is old and unglamorous; the value is in doing all of it, with the real numbers for this house.
- Conduction through the envelope. Walls, ceilings, floors, doors and glass, each with its own area and U-factor (the inverse of the assembly’s R-value). For heating the core is simple: Btuh = U × area × temperature difference. Cooling adds the sun-warmed surface effects.
- Solar gain through glass. On the cooling side, windows are usually the biggest single line. Manual J treats each window by orientation, glass type and shading, which is why a west-facing slider costs far more than the same glass facing north.
- Infiltration. Outdoor air leaking in through the envelope. Sensible load is about 1.1 × CFM × temperature difference; on the cooling side there is also a latent (moisture) load of about 0.68 × CFM × the difference in grains of moisture.
- Ventilation. Any mechanical fresh air the house brings in on purpose, treated the same way as infiltration.
- Internal gains (cooling only). People and appliances. Manual J uses a standard occupancy, the number of bedrooms plus one, at 230 Btuh sensible and 200 Btuh latent per person, not a headcount of who lives there today.
- Duct loads. Heat lost or gained by ductwork running through attics, crawlspaces or garages, including leakage. Ducts in a hot attic can add a large share to the cooling load; ducts inside conditioned space add almost nothing.
Cooling is split into sensible load (temperature) and latent load (moisture). That split matters later, because equipment has to cover both, and a humid climate can need more latent capacity than the total-tonnage number suggests.
Design conditions: why not the coldest day ever?
Loads are calculated at design conditions, not record extremes. Outdoors, Manual J uses the 99% heating dry-bulb (only about 1% of winter hours are colder) and the 1% cooling dry-bulb with its coincident wet-bulb (only about 1% of summer hours are hotter). Indoors, the standard conditions are 70 °F for heating and 75 °F at 50% relative humidity for cooling. The design temperatures by city table lists the outdoor values for 28 US cities.
Sizing for the worst hour on record would leave the system oversized for every other hour of the year. On the handful of hours beyond design, a correctly sized system runs continuously and the house drifts a degree or two, which is the intended trade.
Design temperatures come from published ASHRAE climate data for the nearest representative weather station. Getting the station right matters more than people expect: a mild coastal station standing in for a hot inland ZIP understates the cooling load before a single wall is measured.
A worked example: part of a heat loss
Here is a slice of the heating side for an illustrative house at a 5 °F design temperature, 70 °F indoors, so a 65 °F temperature difference. The U-factors are typical round numbers, not values from the manual’s tables.
| Component | Area or flow | U-factor | Heat loss |
|---|---|---|---|
| Walls, net of windows | 1,400 sq ft | 0.08 | 7,280 Btuh |
| Windows | 250 sq ft | 0.32 | 5,200 Btuh |
| Ceiling under the attic | 1,500 sq ft | 0.03 | 2,925 Btuh |
| Infiltration | 90 CFM | — | 6,435 Btuh |
Two things jump out. The windows are less than a fifth of the wall area and lose almost three quarters as much heat. And air leakage, which no floor-area rule can see, is close to the walls on its own. A full calculation adds floors or foundation, doors, ducts and ventilation, then does the cooling side with solar and internal gains — but the pattern holds: glass, leakage and ducts are where loads hide.
What comes out of a Manual J
| Output | What it is for |
|---|---|
| Design heating load (Btuh) | Sizing the furnace, boiler or heat pump plus backup heat. |
| Sensible cooling load (Btuh) | The temperature part of the cooling job. |
| Latent cooling load (Btuh) | The moisture part, checked separately in humid climates. |
| Total cooling load (Btuh) | Sensible plus latent; the number tonnage is chosen against. |
| Room loads and CFM (room-by-room only) | How much air each room needs, which drives duct sizing. |
A whole-house block load gives the first four. A room-by-room calculation adds the last one, which is what duct design needs. The block load vs room-by-room guide covers when each is enough.
When code asks for one
The International Residential Code ties equipment sizing to these procedures: section M1401.3 calls for heating and cooling equipment to be sized with ACCA Manual S, from loads calculated with ACCA Manual J or another approved method, and the energy code (IECC) carries a matching equipment-sizing requirement. Duct design under Manual D sits in the IRC’s duct chapter.
What that means on your job depends on what your jurisdiction adopted and enforces. Some building departments want Manual J, S and D reports at plan review and will not accept plans without them; others ask only on new construction; some do not ask on a like-for-like replacement at all. Check with the authority having jurisdiction, and check whether it wants the calculation prepared or stamped by a particular professional.
Reading a Manual J report: the inputs to check
A Manual J is only as good as its inputs, and the common way to get a big number out of an honest procedure is to pad the inputs one at a time. If you are reviewing a report — your own, a contractor’s, or one attached to a permit — these are the lines worth a look:
- Design temperatures. Do the outdoor numbers match the published 99% / 1% values for the location? A few degrees colder in winter or hotter in summer quietly adds capacity.
- Indoor conditions. 70 °F heating and 75 °F cooling are the standard. A report designed to 68 °F for cooling is sizing for a preference, not the method.
- Orientation. Is the front of the house facing the right way? Swapping east and west glass moves the cooling load.
- Occupants. Bedrooms plus one. A report that loads in ten people for a three-bedroom house is padding.
- Window area and type. Measured, or estimated generously? Double-pane low-e entered as single-pane clear is a big swing.
- Infiltration. Is it from a blower-door test, or the loosest category in the menu? Leakage is one of the largest inputs, and one of the easiest to overstate.
- Ducts. Are ducts in the attic modeled as leaky and uninsulated when they are sealed and wrapped, or the other way round?
- Extra “safety factors.” Manual J is designed to be used without added margin. Equipment selection rules already allow some oversizing. Margin stacked on margin is how a 3-ton house gets a 5-ton unit.
None of those items needs an engineering degree to check. They need the report and ten minutes, which is a fair thing to ask of anyone selling you equipment.
Running one
There are three practical routes, in rising order of detail:
- A free block load in your browser. The free Manual J calculator on the homepage takes eight inputs and a ZIP code, pulls published design temperatures, and returns whole-house heating and cooling loads with a tonnage. It is a sanity check, and it says so.
- A Manual J on your phone, at the house. The HVAC Load Calculator app runs Manual J 8th Edition methods offline, from a quick 8-input estimate up to a detailed 16-input calc with overrides, shows Manual S equipment sizes that fit the result, and produces a PDF report.
- A designer’s full room-by-room Manual J, when the job needs duct design or the building department wants a stamped submission.
Not sure what to measure first? The Manual J inputs checklist is the site visit, in order.
FAQ
What is a Manual J load calculation?
ACCA’s procedure for calculating a home’s peak heating loss and heat gain, in Btuh, at design conditions. The heating load and the sensible and latent cooling loads are then used to pick equipment with Manual S.
Is a Manual J the same as an equipment size?
No. Manual J gives loads in Btuh. Manual S picks equipment whose real capacity fits those loads, and Manual D designs the ducts for it.
Does code require a Manual J?
The IRC (M1401.3) points equipment sizing at Manual S and Manual J or another approved method. Whether your building department actually asks for one, especially on a replacement, varies, so check locally.
What design temperatures does Manual J use?
The 99% heating and 1% cooling outdoor values for the location, and 70 °F heating and 75 °F / 50% RH cooling indoors. Not record extremes, on purpose.
Can I do a Manual J myself?
A whole-house estimate, yes, and it is a good check on a quote. A full room-by-room calc takes careful measurements and software, and a permit may need it from a qualified professional.
Is this calculator ACCA approved?
No. They use Manual J 8th Edition methods as estimating aids. They are not ACCA approved or certified, and not a stamped calculation.