What a design load is, and what it is not
A design load is the rate at which heat crosses the building envelope at a defined outdoor condition, with the indoor condition held at setpoint. It is not an energy consumption figure and it is not an average. Manual J heating loads are computed at the 99% winter design dry bulb, the temperature that only 1% of hours in an average winter fall below; cooling loads use the 1% summer design dry bulb and its coincident wet bulb. Design for those and the equipment holds setpoint in all but a handful of hours a year.
The heating load has two terms and no gains. Conduction is UA × ΔT, where UA is the sum over every assembly of its area divided by its whole-assembly R-value. Infiltration is 1.08 × CFM × ΔT. Manual J deliberately ignores solar and internal gains in the heating calculation, because the design condition is a winter night and you cannot depend on people or sunshine being there.
The cooling load has five terms and splits into two answers. Conduction and infiltration reappear at the summer temperature difference, then solar gain through the glass, occupant sensible gain, and appliance gain are added. All of those are sensible: they raise air temperature. Separately, moisture carried in with infiltration air and given off by the occupants forms the latent load, which the coil has to condense out. Total cooling is the sum, and the ratio of sensible to total, the SHR, tells you what kind of machine you need.
The distinction matters because equipment is rated on both. A three-ton system does not deliver three tons of sensible cooling; it delivers perhaps 2.3 tons sensible and 0.7 tons latent at rated conditions, and that split shifts with entering wet bulb and with airflow. If your load needs an SHR of 0.65 and the equipment delivers 0.78, the house will hit setpoint and stay humid.
Where 1.08 and 0.68 come from
Both constants are the properties of standard air folded into one number, and knowing their derivation tells you when to change them.
The sensible constant starts from mass flow. One cubic foot per minute of standard air at 0.075 lb per cubic foot moves 60 × 0.075 = 4.5 pounds of air per hour. The specific heat of air is 0.24 BTU per pound per degree Fahrenheit, so 4.5 × 0.24 = 1.08 BTU/h for every cfm and every degree of temperature difference. That single number carries every airflow calculation in this trade, including the supply airflow in the room CFM calculator.
The latent constant starts from the same 4.5 pounds per hour. The latent heat of vaporisation of water near room temperature is about 1,061 BTU per pound, and humidity ratios in the field are quoted in grains, of which there are 7,000 to the pound. So 4.5 × 1,061 ÷ 7,000 = 0.68 BTU/h for every cfm and every grain per pound of humidity difference.
Both assume sea-level air density. At 5,000 ft the density is roughly 83% of standard, so both constants fall by the same proportion, and using 1.08 at altitude overstates the load. The occupant figures, 230 BTU/h sensible and 200 BTU/h latent, are the residential defaults Manual J publishes for a person at light domestic activity, and Manual J's default occupancy is the number of bedrooms plus one.
Worked example: a 2,000 sq ft single-storey house
A 2,000 sq ft slab-free house on a vented crawlspace, 8 ft ceilings, R-38 ceiling, R-19 floor, R-13 walls with 1,600 sq ft of net wall area, 300 sq ft of double-glazed low-e glass at U-0.30 and SHGC 0.30 spread on all four sides, 0.35 natural air changes per hour. Design conditions 10 °F winter and 90 °F summer, setpoints 70 and 75 °F, 30 grains of humidity difference, four occupants and 1,200 BTU/h of appliance gain.
- Assemble UA. Ceiling 2,000 ÷ 38 = 52.63. Floor 2,000 ÷ 19 = 105.26. Walls 1,600 ÷ 13 = 123.08. Glass 300 × 0.30 = 90.00. Total UA = 370.97 BTU/h·°F.
- Infiltration airflow. Volume is 2,000 × 8 = 16,000 cu ft. 0.35 × 16,000 ÷ 60 = 93.33 cfm, so the air-side sensible term is 1.08 × 93.33 = 100.8 BTU/h·°F.
- Heating load. ΔT = 70 − 10 = 60 °F. Conduction 370.97 × 60 = 22,258 BTU/h. Infiltration 100.8 × 60 = 6,048 BTU/h. Total 28,306 BTU/h, or 14.2 BTU/h per square foot.
- Cooling conduction and infiltration. ΔT = 90 − 75 = 15 °F. 370.97 × 15 = 5,565 and 100.8 × 15 = 1,512 BTU/h.
- Solar. 300 sq ft × 0.30 SHGC × 45 BTU/h·sq ft = 4,050 BTU/h. This single term is 31% of the sensible load.
- Internal sensible. 230 × 4 = 920 from people, plus 1,200 from appliances.
- Sensible total. 5,565 + 1,512 + 4,050 + 920 + 1,200 = 13,247 BTU/h.
- Latent. 0.68 × 93.33 × 30 = 1,904 from infiltration, plus 200 × 4 = 800 from people. 2,704 BTU/h.
- Total and SHR. 13,247 + 2,704 = 15,951 BTU/h, which is 1.33 tons. SHR = 13,247 ÷ 15,951 = 0.831.
The house needs a nominal 1.5 ton cooling system, not the three tons a 500-square-feet-per-ton rule of thumb would have produced.
What to do with the four numbers
Take the heating load to equipment selection first, because it is the number a furnace or boiler is sized against. The furnace size calculator converts it to an input rating at your AFUE and checks the result against the ACCA Manual S ceiling of 140% of load.
Take total cooling to tonnage, but select on the sensible and latent split rather than the total. Manual S sizes cooling equipment against the total load with a much tighter window than heating, and it requires that the equipment's sensible capacity at your design conditions covers the sensible load. Expanded performance tables, not the nominal tonnage, are where that check is made.
Read the SHR as a specification for the machine. A load with SHR above about 0.80 is dry-climate work where almost any equipment will do. Between 0.70 and 0.80 is the normal residential range that standard split systems are designed around. Below 0.70 the latent load is unusually large relative to sensible, which happens in humid climates in very tight, well-insulated houses where the envelope no longer contributes much sensible gain. That is the case where a conventional system satisfies the thermostat and leaves the house at 60% relative humidity, and it usually calls for a variable-capacity system or standalone dehumidification. Confirm what is actually happening in the space by measuring conditions and running them through the dew point calculator.
The per-square-foot heating figure is a plausibility check on your inputs, not a design output. If it lands far outside what the vintage and climate would suggest, go back and look for an area entered in the wrong unit or an R-value that is a batt label rather than a whole-assembly value.
Heating load sensitivity for the example house
| Infiltration | dT = 40 F | dT = 60 F | dT = 80 F |
|---|---|---|---|
| 0.20 ACH | 17,143 | 25,714 | 34,286 |
| 0.35 ACH | 18,871 | 28,306 | 37,742 |
| 0.50 ACH | 20,599 | 30,898 | 41,198 |
| 0.75 ACH | 23,479 | 35,218 | 46,958 |
Each cell is (370.97 + 1.08 x ACH x 16,000 / 60) x dT. Moving from 0.20 to 0.75 ACH adds 37% to the load at every temperature difference, because both terms scale linearly with dT and the infiltration term scales linearly with ACH.
What this block load leaves out
- Room-by-room results. A block load sizes equipment. Duct design needs the load of each room, which is what a full Manual J produces and what Manual D consumes. Split the block total in proportion to room loads with the room CFM calculator only as an interim measure.
- Orientation-specific solar. Real Manual J applies a different glass load factor to each orientation, adjusted for latitude, external shading, internal shading and the hour of peak. One peak solar factor cannot reproduce that, and the error grows with glazing area.
- Slab-on-grade floors. A slab loses heat through its perimeter, not its area, and is calculated from exposed perimeter length times an F-factor. Set the floor R-value to zero here and add that load separately.
- Duct losses. Ducts in unconditioned space add load that no envelope calculation sees. Manual J has a duct load procedure; the simpler route is the percentage allowance in the furnace sizing calculator.
- Ventilation air. Mechanical ventilation required by ASHRAE 62.2 is an additional airflow beyond natural infiltration, and it carries both sensible and latent load. Add it to CFM using the air changes per hour calculator if the house has a ventilation system running at design conditions.
- Thermal mass and diversity. Manual J's cooling load factors already reflect the delay between a solar gain and its appearance as a cooling load. A steady-state calculation like this one does not, so it tends to be conservative on the solar term.
This is a screening calculation, not a submittable Manual J
ACCA Manual J, currently in its 8th edition, is a defined procedure with published tables, and jurisdictions that require a load calculation for a permit generally require output from ACCA-approved software running that procedure room by room. The calculation here follows the same physics and the same constants, but it collapses the building into one zone and the glass into one solar factor. Use it to check a contractor's proposal, to sanity-test an existing system, or to see how sensitive the answer is to the envelope. Do not submit it for permit.
How the load drives everything downstream
Once the load is known the whole design follows in order. Equipment selection comes next under Manual S, using the load and the local design conditions against the manufacturer's expanded performance data. Airflow follows from the sensible load and the supply temperature difference. Duct sizes follow from that airflow and from the static pressure the blower has left over, which is the sequence covered by the available static pressure calculator and the duct size calculator. Register selection and placement close the loop under Manual T.
The most productive thing this calculation usually reveals is not the equipment size but where the load lives. In the worked example, glass is 90 BTU/h·°F of a 371 UA, a quarter of the conduction from 15% of the surface area, and its solar term is nearly a third of the sensible cooling load. Improving the windows moves the number far more than adding attic insulation would. Run assemblies through the R-value to U-value converter and re-run the load to see which envelope change is worth its cost before you size anything.