Manual J vs Manual S vs Manual D: What Each One Decides

Three calculations stand behind every right-sized HVAC job: Manual J, Manual S, and Manual D. They get blurred together more than any other part of the design. A contractor who can rattle off a load calculation will still wave a hand at where the load stops and the equipment selection starts, or size a duct to a number nobody ever actually computed. The three are not interchangeable steps you pick from a menu. Each one answers a different question, and each one hands its answer to the next.

This guide lays out what each of the three decides, the order they have to run in, and why you cannot skip a step. There is a worked thread that carries one house from the load all the way to the duct sizes, a look at which calculations a permit actually requires, and the five mistakes that quietly throw the chain off. If you size air conditioners in your sleep but have never been walked through where J ends and S begins, this is that walkthrough.

Want to follow along with your own numbers? You can run a free Manual J estimate in the BuildSolver calculator with no signup for the first one. Everything here is for preliminary, sales-phase work; permit-grade calculation comes up near the end.

What's the difference between Manual J, Manual S, and Manual D?

Short answer: three different jobs. Manual J finds the load, how much heating and cooling the building needs. Manual S picks the equipment that meets that load without oversizing. Manual D sizes the ducts that deliver the air. The order is fixed, J then S then D, and each calculation feeds the one after it.

The cleanest way to keep them straight is to tie each calculation to the question it answers and the thing it produces. Manual J is about the house, Manual S about the equipment, and Manual D about the duct system that connects the two.

Each row of that table is a handoff. The load from J is the input to S. The equipment from S is the input to D. Treat them as one undivided "sizing" step and the seams are exactly where the errors hide, because a mistake in the load quietly travels all the way down to the duct.

What does Manual J decide?

Short answer: Manual J decides the load, how much heating and cooling the building needs at the design conditions, and nothing about the equipment. It returns a heating load and a cooling load in BTU/hr, with cooling split into sensible and latent, and it can run whole-house or room by room. Square footage is not an input; the envelope, the climate, and the internal gains are.

A Manual J load calculation measures the building, not the machine. Allison Bailes draws the line that the rest of the chain depends on:

"The first thing you need to know is that the term loads refers to how much heating and cooling the building needs and capacity refers to how much heating and cooling the equipment can supply."

• Allison Bailes, PhD, Energy Vanguard

The inputs are the physical reality of the house: conditioned area and ceiling height, the heating and cooling design temperatures, wall and ceiling and floor R-values, window area with U-factor and SHGC by orientation, the infiltration rate, the occupants, and the internal gains from appliances and lighting. Run those and you get two numbers that matter: a heating load and a cooling load. The cooling load splits further into sensible (the temperature part) and latent (the moisture part), and that split is what carries the dehumidification problem forward into equipment selection.

Manual J runs in two modes. A block load treats the house as one zone and returns the whole-house number. A room-by-room load returns the load for each room, which is what you need to size the branch ducts later. Both come from the same inputs.

The conversion to nominal tonnage is simple arithmetic once you have the cooling load: divide BTU/hr by 12,000 to get tons. The hard part is getting the inputs right, which is the entire point of the 2,000 sq ft sizing walkthrough. What Manual J never does is pick a unit. That is the next calculation.

What does Manual S decide?

Short answer: Manual S decides the equipment. It takes the Manual J loads and picks an actual unit from the manufacturer's expanded performance data so the capacity meets the sensible, latent, and total load at your design conditions, and stays inside the ACCA oversizing caps. For cooling, the cap is tight: 15 percent over the load for a single-stage compressor, 20 for two-stage, 30 for variable-speed.

Manual S is where a load becomes a model number. It cannot run until Manual J has produced the loads, because the loads are its input. You take the cooling and heating numbers to the manufacturer's expanded performance data and find a unit whose capacity at your design conditions covers the load. The sensible capacity has to cover the sensible load, and the latent capacity the latent load. The capacity you read is the one at design temperature, not the nameplate rating, because a unit puts out a different number on a 95 degree afternoon than it does at the AHRI rating point.

The part contractors skip is the ceiling. ACCA does not let you round up "to be safe." Ed Janowiak, ACCA's manager of HVAC design education, ties the cap to the compressor technology:

ACCA Manual S § oversizing limits

"The correct airflow for an air conditioner will be one that meets the sensible, meets the latent, and doesn't exceed the total by 15, 20, or 30 percent, depending on the compressor technology." - Ed Janowiak, ACCA, ACCA HVAC Blog

Those three numbers map to the compressor: 15 percent for a single-stage, 20 for two-stage, 30 for variable-speed. The DOE Building America program frames the cooling selection the same way from the capacity side: the listed capacity at design conditions should land between 95 and 115 percent of the design load, or the next nominal size (PNNL Building America).

Heating is looser. A gas furnace gets sized to the heat loss, and Manual S allows the output to run well over the load when a bigger blower is needed to move the air. Either way, the cap exists to protect comfort.

An oversized cooling unit satisfies the thermostat fast, shuts off, and never runs long enough to pull moisture out of the air, which is the dehumidification failure covered in the heat pump sizing guide. Skipping Manual S is how a correct load calculation still ends in a clammy house. Doing it right is the whole job of the Manual S equipment selection step.

What does Manual D decide?

Short answer: Manual D decides the ducts. It sizes every trunk and branch to deliver each room's required airflow at a friction rate the selected equipment can actually push. That friction rate comes from the available static pressure and the total effective length of the runs, not from a default like 0.10. Manual D needs the equipment first, because the blower data sets the static pressure it has to work with.

Manual D is duct design, and it starts at the equipment, which is why it comes last. The first thing you read is the selected unit's external static pressure at the design airflow, straight off the manufacturer's blower data. From there the math is three steps.

First, the available static pressure: take the external static pressure and subtract the pressure lost to the coil, filter, registers, grilles, and balancing dampers. What is left is the pressure the duct system actually gets to use. Second, the total effective length: the longest supply run plus the longest return run plus the equivalent length of every fitting along the way, pulled from the Manual D fitting tables. Third, the friction rate, which is just the available static pressure spread over that length:

Friction rate = (available static pressure x 100) / total effective length

ACCA wants that friction rate to land inside what the standard calls the wedge. Luis Escobar puts the range on it:

"You want the friction rate to fall within 'the wedge' (a friction rate between 0.06 and 0.18 IWC)."

• Luis Escobar, ACCA, ACCA HVAC Blog

Once you have the friction rate, the equal-friction method sizes every duct to that single number, and you sanity-check that the velocity does not run too high. The mistake the standard is built to prevent is plugging in a default. Bryan Orr says it plainly:

"A common friction rate rule of thumb is 0.10, but it's something we should really calculate."

• Bryan Orr, HVAC School

Size the ducts to a guessed 0.10 and they will not match the blower they are paired with. The right-sized box from Manual S then cannot move its rated air, and the system underperforms even though every number upstream was correct.

Running all three by hand is where the half hour on every quote goes. Describe the house in plain words and BuildSolver runs the Manual J first, carries the loads into the Manual S selection, and returns a first-pass duct sizing, with the standard cited on each step.

Why do Manual J, S, and D have to run in that order?

Short answer: because each calculation needs the output of the one before it. You cannot pick equipment with Manual S until Manual J gives you the load, and you cannot size ducts with Manual D until Manual S gives you the equipment's airflow and static pressure. The friction rate at the heart of Manual D traces straight back to the blower in the unit Manual S selected. Reorder the steps and you are guessing.

The order is not a convention you can bend for convenience. It is a chain of dependencies. Bailes lays out the sequence:

"The load calculation is just the first part of the design process ... Manual J. But to determine the proper size of the system, there's Manual S. And that's not the end either ... Manual D."

• Allison Bailes, PhD, GreenBuildingAdvisor

Follow the data and the lock-in is obvious. The friction rate in Manual D is derived from the available static pressure, which in turn starts from the equipment's rated external static pressure at the design airflow. That airflow and that rating only exist after Manual S has chosen the unit, and Manual S could only run because Manual J handed it a load.

Pull out the load and the equipment selection has nothing to work from. Pull out the equipment and the duct design has no static pressure to spend. Each missing piece collapses everything downstream of it, which is why "I sized the ducts first and worked backward" produces a system that does not hang together.

One worked thread: a single house from load to ducts

Short answer: take one house through all three calculations. Manual J returns the load, Manual S converts it into a unit inside the caps, and Manual D sizes the ducts to the selected blower. The numbers below are illustrative arithmetic to show how the handoffs work, not a sizing for any real house.

Here is the whole chain on one set of numbers. Treat it as a demonstration of the handoffs, since the real values depend entirely on the inputs for your specific house.

1. Manual J returns the load. For this house, the cooling load comes back at 30,000 BTU/hr total, split into 24,000 sensible and 6,000 latent, and the heating load at 36,000 BTU/hr.
2. Convert to nominal tons. 30,000 / 12,000 = 2.5 tons. That is the starting point for equipment selection, not the answer.
3. Manual S picks the unit. Find a unit whose capacity at design conditions lands between 95 and 115 percent of the 30,000 cooling load, so 28,500 to 34,500 BTU/hr. Say you select one rated about 31,000 total and 24,500 sensible at design. It clears the load on both the total and the sensible side, sits inside the single-stage cap, and its blower moves roughly 1,000 CFM.
4. Manual D sizes the ducts. Read the unit's external static pressure at 1,000 CFM, say 0.50 IWC. Subtract about 0.27 for the coil, filter, registers, grilles, and dampers, leaving an available static pressure of 0.23. With a total effective length near 200 feet, the friction rate is (0.23 x 100) / 200 = 0.115 IWC per 100 feet, comfortably inside the wedge. Size every duct to 0.115, and each room's branch carries its share of the 1,000 CFM from the room-by-room loads.

Change the load in step 1 and watch every number after it move. A higher cooling load pushes the tonnage up, which changes the unit, which changes the airflow and the static pressure, which changes the duct sizes. That cascade is the reason the order is fixed, and the reason a padded Manual J does not just oversize the box; it mis-sizes the entire system.

What comes after Manual D?

Short answer: Manual T. Once the ducts are sized, Manual T selects and places the registers and grilles so the air actually throws across the room instead of dumping at the boot. It is the most skipped step in the chain, and it is why a correctly sized system can still feel drafty or starved in spots.

The chain does not quite end at the duct. Manual T handles air distribution: which register goes where, and how its throw, spread, and terminal velocity mix the supply air into the room. Bailes calls register selection the most neglected part of HVAC design, and it shows up as the comfort complaint that survives a perfect load calculation, a right-sized unit, and clean duct sizing. For residential sales-phase work you rarely formalize Manual T. But it is worth knowing the chain runs J to S to D to T, so a comfort problem in a correctly sized system points you to the registers rather than back to the tonnage.

Which of these do you need for a permit vs a quote?

Short answer: for a permit, the building code requires the load from Manual J and the equipment selection from Manual S, and 2024 code editions are pushing Manual D into plan review too. For a quote, a phone call, or a comfort conversation, a fast preliminary estimate is the right tool, and a permit submission is not.

The code is explicit about the first two. The International Residential Code ties equipment sizing directly to Manual J and Manual S:

IRC § M1401.3

"Heating and cooling equipment and appliances shall be sized in accordance with ACCA Manual S or other approved sizing methodologies based on building loads calculated in accordance with ACCA Manual J or other approved heating and cooling calculation methodologies." - IRC Section M1401.3

The 2024 code cycle keeps that requirement and tightens the envelope and duct-leakage testing around it, with reviewers increasingly checking the design temperatures, window schedules, and duct details on the plans. Exact blower-door and leakage thresholds vary by the edition your jurisdiction has adopted, so confirm those against your local code rather than a national number. New construction almost always lands in the permit bucket, which is the path covered in Manual J for new construction.

A preliminary, sales-phase estimate is a different tool for a different moment. It is legitimate and useful for quoting and for comfort conversations, and it is not a permit submission. For permit work you need a documented calculation from approved software, which is the route covered in Manual J for permit. The fast estimate is for the driveway and the kitchen-table quote; the permit package is for the plan reviewer.

What are the most common mistakes with J, S, and D?

Short answer: the same five every time. Running Manual J and skipping Manual S, sizing ducts off a default friction rate, sizing the system off the old ductwork instead of a load, padding the Manual J inputs, and mixing up which calculation governs what. Each one quietly pushes the install off target.

1. Running the load and skipping the selection. A contractor does a real Manual J, then jumps straight to a model number without the Manual S oversizing check. The load was right and the box is still two sizes too big.
2. Sizing ducts to a default friction rate. Plugging 0.10 into Manual D instead of deriving the friction rate from the available static pressure and total effective length. The ducts end up matched to a blower that does not exist.
3. Sizing the system off the existing ductwork. Telling a customer the house "has four tons of ductwork, so it gets four tons" reasons backward from the tin to the load. Run a fresh calculation instead, the same discipline a replacement sizing demands.
4. Padding the Manual J inputs. Stretching the window area or adding occupants inflates the load a little at a time. Bailes puts a number on where it ends up:

"Contractors doing these load calculations often feel compelled to stretch a little bit here and a little bit there ... you may be looking at putting in a 4 ton air conditioner where 2.5 tons could work."

• Allison Bailes, PhD, ACCA HVAC Blog

5. Mixing up which calculation governs what. Reaching for Manual J to settle a question about airflow or equipment, when that answer lives in Manual S, or arguing duct sizes from the load instead of the friction rate. Keep each question with its own manual and the confusion clears.

You can pull every Manual J input from a walkthrough even without plans, the same way you would on a retrofit with no blueprints, which removes the usual excuse for skipping the load and guessing.

How does BuildSolver run J, S, and D in one chat?

Short answer: you describe the house in plain words and BuildSolver runs each calculation as its own ACCA step in a single chat, the Manual J load, the Manual S selection inside the oversizing caps, and a first-pass Manual D duct sizing, then returns the numbers with the standard cited on each one. It is a chain of separate, checkable steps, not a single black-box answer.

The workflow keeps the three calculations distinct, in the order the standards require:

1. Describe the job in words, the way you would tell a client on the phone. No blueprint upload, no 167-field form, no LiDAR scan.
2. Get the Manual J load. It runs the heating and cooling load and returns both, with the sensible and latent split, following the ACCA Manual J procedure in deterministic code.
3. Carry the loads into Manual S. It selects equipment against the loads, applies the oversizing caps by compressor type, and reads capacity at your design temperature. The Manual S selection and the room-by-room and full Manual D steps are on the paid tiers.
4. Size a first-pass duct system to the selected airflow, so you leave with a friction rate and trunk and branch sizes, rather than only a tonnage.
5. Send a branded PDF with your logo, the results, the formulas, and the assumptions listed, formatted as the quote you hand the client.

Every step shows its work and cites the standard it followed, so a reviewer or a customer can trace each number back to its source instead of taking one bundled answer on faith. Each result carries the same disclaimer the standard deserves: a fast estimate for the sales phase, not a stamped design. You can try the flow in the BuildSolver chat or start from the Manual J calculator.

Sources

• ACCA HVAC Blog (Ed Janowiak), Correct Airflow for an Air Conditioner: hvac-blog.acca.org
• ACCA HVAC Blog (Luis Escobar), Definitive Design: ACCA's Manual D Revised: hvac-blog.acca.org
• ACCA HVAC Blog (Allison Bailes, PhD), Making Mistakes With Manual J: hvac-blog.acca.org
• Energy Vanguard (Allison Bailes, PhD), How to Read Manual J Load Calculation Reports: energyvanguard.com
• GreenBuildingAdvisor (Allison Bailes, PhD), The Most Neglected Part of HVAC Design: greenbuildingadvisor.com
• HVAC School (Bryan Orr), When Ducts Get Too Big: hvacrschool.com
• PNNL Building America, Air Conditioning Equipment Sizing: basc.pnnl.gov
• IRC Section M1401.3, Equipment and Appliance Sizing: up.codes
• ASHRAE Terminology, Ton of Refrigeration: terminology.ashrae.org

Three calculations, three jobs, one fixed order: Manual J finds the load, Manual S picks the box inside the caps, Manual D sizes the tin to the box. Keep them separate in your head and the design holds together; blur them and the errors hide in the seams.

Try BuildSolver free, with no signup for the first calculation, at buildsolver.com. Describe the job, get the load, the equipment selection, and a first-pass duct sizing with the standard cited on each step, and hand the client a branded quote in about the time it took to read this.

For estimation purposes only. Not a substitute for a licensed engineer, and not ACCA-approved for permit submission.