SizingKitHVAC · the standard itself
Manual J calculator — what the standard asks for, and where it stops
Free and with no account: this page checks a load calculation against the ten inputs a certified ACCA Manual J is assembled from, then checks the equipment selection against the Manual S sizing figures — 90 to 115% of the load for cooling, 100 to 140% for heating. Two of the ten items cannot come from any free calculator, this one included, and the page names them rather than implying it has done them. If a permit office has asked you for a Manual J, what it wants is a report from ACCA-approved software with the input assumptions attached.
- 100% free
- No signup
- 10-item readiness check
- Manual S bounds applied
- Names what it cannot do
- Design percentiles in hours
Ten things a certified calculation is assembled from
Tick what you have. The point of the list is the two items at the bottom, which nothing on the open web can give you.
Where the calculation stands
In hand
3 / 10
Still to gather
5
Anyone with the plans can get these.
Needs a professional
2
Approved software and an expanded-data selection.
The two items at the bottom of the list are the ones that turn arithmetic into a Manual J. Whatever you assemble here, a jurisdiction asking for a load calculation is asking for a report from approved software — an equivalent worked by hand or by this site is not the same document and does not have to be accepted.
The check that comes after the load
A load calculation is only worth doing if the selection respects it. Enter the design load and the capacity from the manufacturer’s expanded performance data at those same conditions.
BTU/h unless you write another unit.
At the design condition, not the nominal size.
The selection against the sizing figures
Selection as a fraction of the load
120%
above the 115% upper figure
Window the table quotes
27,000 BTU/h – 34,500 BTU/h
2.25 tons to 2.88 tons for air conditioner, cooling
The table's upper figure for air conditioner, cooling is 115% of the load. At 120% the selection is past it. Total capacity against the total cooling load, at the design conditions. The upper bound is about humidity: a machine much above the load stops before it has dehumidified.
The sizing figures are not a pass mark. ACCA Manual S (Residential Equipment Selection), whose limits on how far selected capacity may exceed the Manual J load are the reason a load calculation is worth doing at all. Manual S has been revised and the exact bounds and their conditions differ between editions and between climates — the humid-climate rules on latent capacity are stricter than the numbers here suggest, and there are separate provisions for variable-capacity equipment, which can modulate below its rated output and is not well described by a single ratio. Check the current edition before quoting a limit as a requirement. The limits apply to capacity at the design conditions from the manufacturer's expanded data, not to nominal tonnage.
The occupancy default, since it is the one input people argue about
The calculation is run for 4 people, and it is a default rather than a count. A default for a house nobody has moved into yet. Where the real occupancy is known, use it.
A design temperature comes from the ASHRAE Handbook—Fundamentals climatic design tables, which publish, for each of several thousand weather stations, the 99.6% and 99% heating dry-bulb temperatures and the 0.4%, 1% and 2% cooling dry-bulb temperatures with their mean coincident wet bulbs. ACCA Manual J uses the 99% heating and 1% cooling values. They are percentiles of the hours in a year, not extremes: at the 99% heating design temperature the weather is colder for roughly 88 hours a year, and the heating system is expected to fall behind for some of them. Look yours up by station — a design temperature is local, and a city's airport figure can be several degrees off a site twenty miles away at a different elevation.
What each design percentile costs you in hours
A design condition is a percentile of the year’s 8,760 hours, so choosing one is choosing how many hours a year the equipment is allowed to fall behind.
| Column | Hours a year beyond it | Where it is used |
|---|---|---|
| 99.6% heating | 35 | The stricter winter pair, chosen where failing to hold temperature costs more than the equipment does. |
| 99% heating | 87.6 | What Manual J uses for winter. |
| 0.4% cooling | 35 | The stricter summer condition, used for critical spaces rather than dwellings. |
| 1% cooling | 87.6 | What Manual J uses for summer, paired with its mean coincident wet bulb. |
| 2% cooling | 175.2 | The loosest of the published summer columns. |
ACCA Manual J, 8th edition: the winter outdoor design condition is the ASHRAE 99% heating dry-bulb temperature. ASHRAE also publishes 99.6%, which is used where a failure to hold temperature matters more than first cost. ACCA Manual J, 8th edition: the summer outdoor design condition is the ASHRAE 1% cooling dry-bulb temperature with its mean coincident wet bulb. ASHRAE also publishes 0.4% and 2%.
Where to get one. A Manual J that a permit office will accept comes from software on ACCA’s approved-software list, run by a mechanical contractor, an energy rater or a design firm — and it is ordinarily a few hundred dollars against equipment costing tens of times that. Ask for the input report as well as the load summary: the assumptions are what a reviewer reads. If what you need instead is a defensible number for a conversation rather than a document, do the room-by-room arithmetic, get the winter fabric and ventilation loss separately, and keep the square-foot figure as nothing more than a check that neither of them is wildly out.
How to find out whether you need a Manual J or already have what one needs
The list is the tool. The arithmetic is on the sibling pages.
Tick off what you actually hold
Go down the ten items and check only the ones you have in a form somebody else could read: a station design temperature rather than a state, a measured area rather than a total, an NFRC label figure rather than a description of the windows. Anything you would have to estimate on the spot is not held.
Read the two counts separately
The middle count is work anybody with the plans can finish this afternoon. The right-hand count is the part that needs approved software and a selection from expanded performance data, and it does not shrink by being careful — it shrinks by hiring somebody.
Run the selection through the sizing figures before you agree to it
Enter the design load and the capacity the proposed equipment actually makes at your design conditions, not its nominal size. The page returns the ratio and says which figure it sits against, with the caveat that those figures move between editions of the manual and are stricter in humid climates than a single ratio suggests.
Technical specifications
| Readiness items | 10, of which 8 are gatherable by anyone with plans and a measure-up and 2 are not: an approved-software calculation and an expanded-performance-data selection. |
|---|---|
| Manual S sizing figures | Air conditioner cooling 90–115% of the load, heat pump cooling 90–125%, furnace or boiler heating 100–140%. The cooling bounds are tighter because the upper one is about humidity rather than about cycling. |
| Design percentiles reported | 99.6% and 99% heating, 0.4%, 1% and 2% cooling, with the annual hours beyond each: 35.0 hours for the 99.6% and 0.4% columns, 87.6 hours for the 99% and 1% columns, 175.2 for the 2%. |
| Occupancy default | Bedrooms plus one, applied unless the owner states a real number — a default for a house nobody has moved into yet rather than a description of a household. |
| Load units accepted | BTU/h, MBH, kW and tons of refrigeration, in either field, in any mixture. MBH is a thousand BTU/h — the M is the Roman numeral — so an 80 MBH furnace is 80,000 BTU/h. |
| Edition quoted | ACCA Manual J 8th edition for the design conditions and occupant figures, ACCA Manual S for the selection bounds. Which edition is in force where you are building is set by the code your jurisdiction has adopted, and this page does not assume it. |
| What this page never does | Compute a load. That is a separate operation on separate pages, and combining the two here would produce exactly the thing this page exists to warn about: a number that looks like a Manual J and is not one. |
| Verdicts it will not render | Compliant, passing, or safe. It reports which sizing figure the ratio sits against and prints the caveat that comes with those figures, because the person who signs a design off is not a web page. |
Frequently asked questions
Will a permit office accept a printout from this page?
No. A jurisdiction asking for a load calculation is asking for a report generated by software on ACCA's approved list, with the input assumptions attached so a plans reviewer can check them. That is a procedural requirement about provenance, not a claim about arithmetic — a hand calculation that reached the same BTU/h by the same method is still not the document, and the reviewer has no way to audit it.
What is the difference between Manual J, Manual S, Manual D and Manual T?
They are four steps of one procedure and they run in that order. J produces the room-by-room heating and cooling loads at your design conditions; S selects equipment against those loads from the manufacturer's expanded performance data; D designs the duct system that delivers each room's share of the air; T places and sizes the registers and grilles so the air arrives where people are. Skipping S is the common failure: the load gets calculated and the equipment gets picked on nominal tonnage anyway.
How much does a Manual J cost and who performs one?
Ordinarily a few hundred dollars, from a mechanical contractor, an energy rater or a design firm that holds the software. Set against a system that costs tens of times more and lasts fifteen to twenty years, it is the cheapest decision in the project — and the one place it reliably pays for itself is in the equipment it lets you not buy, because a correctly sized system is frequently smaller than the one that was there before.
The contractor says a replacement does not need one. Is that right?
It depends entirely on what your jurisdiction has adopted, and the reasoning behind the claim is usually weaker than the claim. Replacing like for like assumes the original was sized correctly, and sizing by whatever was there before is how an oversizing error propagates through a house's whole service life. If the building has since had windows, insulation or air sealing done, the old load is definitively wrong — every one of those changes it downward.
Why is there an upper limit on how large the equipment may be?
Because oversized equipment fails in a way people misread as the machine being faulty. A cooling system much larger than the load satisfies the thermostat before the coil has run long enough to condense water out of the air, so the house sits at the setpoint and feels clammy — the answer to which is usually to lower the setpoint, which makes it worse. On the heating side the penalty is cycling rather than humidity, which is why the furnace bound is far more generous than the cooling one.
Is a Manual J the same thing as an energy model or a HERS rating?
No, and the two are computed for opposite conditions. A load calculation is a peak: what the building needs at one extreme hour, which is what equipment has to be sized for. An energy model is an integral: what it consumes across all 8,760 hours, which is what a utility bill reflects and what a HERS index scores. A house can have a modest annual consumption and a large peak, and sizing from the annual figure would leave it cold on the coldest night of the year.
Where does the load get turned into a machine?
In the selection, using capacity at your design conditions rather than the number in the model code. Nominal tonnage is measured at a rating point that is not your design condition, so a nominal three-ton unit does not make 36,000 BTU/h at 100 °F outdoor — it makes rather less, and the expanded performance table in the manufacturer's literature is where the real figure lives.
Why a permit office asks for this, and what the document contains
The requirement exists because equipment sizing was, for decades, done by rule and by habit, and the result was a housing stock in which oversizing is the norm rather than the exception. Requiring a load calculation moves the decision from a driveway conversation to a document with named assumptions in it: which weather station, which wall construction, which window U-factor, how many air changes, where the ducts run. A reviewer does not check the totals — they cannot, without redoing the work — they check whether the assumptions are plausible for the building described on the plans. That is why the input report matters more than the output summary, and why a bare BTU/h figure, however correct, is not what is being asked for.
What a certified calculation contains that arithmetic on the open web does not is mostly tables. The manual carries its own construction numbers for walls, roofs, floors and doors, its own glass-load multipliers by orientation, latitude and shading, its own infiltration model, its own duct-loss method, and the climatic data for several thousand stations. Reproducing any one of those on a web page would mean transcribing a copyrighted table, and this site does not do that — where a figure is not in its own standards library, it says the table is elsewhere and stops. The consequence is straightforward: the room-by-room schedule on this site does the structure of the procedure honestly and leaves the glass-load column empty, and the winter fabric and ventilation calculation does the heating half completely, because heating has no solar term to be missing.
The step after the load is the one that most often gets dropped, and it is where the money is. Manual S selects equipment from expanded performance data at the design conditions, which is a different number from the size on the model code — nominal tonnage is a rating-point measurement and the tonnage conversion that turns BTU/h into tons is arithmetic, not a selection. The unit's real capacity at 100 °F outdoor and a 78 °F indoor wet bulb is in the manufacturer's table and is lower, sometimes by a full step of nominal size. Running that comparison, in the units the tables are printed in, is what the second half of this page is for; converting between those units is what the BTU to tons calculator handles.
What the checklist knows about your project
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type — loads, lengths, nameplate ratings, the rates your utility charges you — is uploaded, logged or kept, which is also why the calculators carry on working in a mechanical room with no signal.
The ticks, the bedroom count and the two load figures are held in the page’s own state for as long as the tab is open and are not written to storage, so nothing here identifies a property or a job, and closing the tab is the whole of deleting it.