If you're new to ACCA certified Manual J software like Amply, it's common to see a heating or cooling load that looks a lot smaller than what you'd get from a rule-of-thumb estimate (like "25 BTU per square foot" for heating or "1 ton per 500 sq ft" for cooling) or from practices like sizing based on the equipment currently installed in the home. This doesn't mean something is broken — it usually means Amply is doing exactly what Manual J is supposed to do.
Here's why the numbers often don't match, and how to know if your result is trustworthy.
Rule-of-thumb methods are for gut checking and nothing else
Square-footage rules of thumb are short cut to get a "gut-check" on loads. They DO NOT account for a projects unique design conditions, specific insulation levels, window performance, orientation, or air leakage, etc. so they almost always overshoot, usually by a lot.
Manual J does the opposite: it calculates loads from the actual measured geometry, unique location, orientation and construction of the specific home in front of you. When Amply's number comes in well under a rule-of-thumb estimate, that's the rule of thumb being conservative, not Amply being wrong.
From Manual J 8, v2.5, p.25:
"Do not use rules of thumb. The idea that the required equipment capacity equals the floor area multiplied by some magic number has resulted in many customer complaints and legal actions. Heating and cooling loads depend on individual circumstances. Floor area-to-tonnage ratios for the U.S. housing stock can range from less than 500 sq. ft./ton to more than 1,200 sq. ft./ton... an energy efficeny home built to 2014 specification may range from 1,300 sq ft per ton to 2,500 sq ft per ton."
Existing equipment size isn't a reliable benchmark either
It's very common for the system currently installed in a home to be oversized. This happens for a lot of ordinary reasons: the original installer used a rule of thumb, the home has since been renovated (new windows, added insulation, air sealing) since the last system was sized, or the previous contractor, since they didn't use a accurate sizing methodology sized generously "to be safe."
Also, remember that load calculations are about the home and are one part of figuring out the proper equipment. Load calculations are the problem that needs to be solved. The Manual S procedure helps us actually pick the right equipment, the right solution for that problem.
Sadly, it is quite common to come across old boilers and furnaces that are oversized for heating by more than double than the actual loads and similar mistakes on the cooling side.
Manual J already includes a safety margin
Manual J isn't a bare-minimum calculation — it has built-in conservatism already: roughly a 10% margin on cooling loads and 10–20% on heating loads, depending on the specific factors involved. That means you should NOT pad your inputs "just to be safe" on top of what Amply calculates. Doing so tends to compound safety factors on top of safety factors, which is exactly the kind of oversizing Manual J is designed to prevent.
From Manual J 8, v2.5, p.23:
"Be honest and aggressive. Manual J is an engineering tool that has an inherent and appropriate factor of safety. Any attempt to add other safety factors or to manipulate the procedure may result in unacceptable performance, especially at part load."
Before you trust a low number, remember garbage in = garbage out
While Amply includes smart starting assumptions based on the year built and the product location, no software can know that a customer recently replaced all their windows or blew in R50 into their attic. A genuinely low number is usually correct but you must verify the handful of inputs that swing loads the most:
Ducts: if the system is ducted and any part of the duct run passes through unconditioned space (an attic, crawlspace, or unconditioned basement), make sure you've modeled the duct system's location, insulation, and sealing level. This is one of the most common things new users forget to set, and it can add several thousand BTU to a load once it's modeled correctly. See Ducts - verify and edit.
Infiltration: Be really thoughtful here, especially in colder climates. Learn more about infiltration loads.
Windows: Low-E glass has much higher performance "clear" glass. This one inout can swing window heating and cooling by 50% or more. If you're not sure which one a home has, see How to determine if you have "low-e" or "clear" windows.
Insect screens: Full outdoor insect screens can reduce cooling load by up to 20%! Take credit for them if you've got them. Learn how to edit window assumptions.
Overhangs: Roof overhangs, wrap-around porches, and awnings (anything that helps externally shade a window) can significantly reduce solar gain through windows. Especially for south, east, or west-facing windows. Small changes in overhangs have an especially big impact on southern latitudes.
Insulation: Verify wall, floor and attic insulation to match what's actually in the home rather than relying on the year-built default, especially in older or renovated homes.
Pro Tip: Select "Preview Manual J" to review the report and make live edits inside if you find any issues
Review the second page of the Amply Manual J report, which highlights the loads by component in the building.
See zero loads for doors? Probably means you forgot to mark an exterior door.
See zero loads for ducts, but you're doing a ducted project? You probably forgot to model the ducts
See another number that's high than your used to? Maybe that's an assumption you should double-check
Review the 3rd page of the Manual J report to see all the assumptions made for insulation in the walls, type of windows, and look for inconsistencies that might need to be fixed.
If you've checked these and you still have questions, that's a great candidate to send us for a second look — see below.
Want a second opinion?
If you're not sure whether a result is right, export your Manual J report and send it to us at [email protected]. We're happy to review it and walk through the specific assumptions with you, especially on your first few projects.