A load calculation can be wrong in a very specific direction:
Too high.
That usually happens because the model makes the house look harsher, leakier, larger, or more exposed than it really is.
The danger is obvious.
An inflated Manual J can justify oversized equipment while still looking technical and professional.
First rule: do not trust the result because it came from software
Software performs the math you asked it to perform.
If you tell it:
- The house is extremely leaky
- The windows are terrible
- The design weather is extreme
- The attic is barely insulated
- The ducts lose enormous capacity
It will calculate a large load.
That does not mean the house actually has one.
Mistake 1: Using record temperatures
This is one of the fastest ways to inflate load.
Manual J should use recognized design conditions, not the hottest or coldest temperature ever recorded.
If normal winter design temperature is 15°F and you enter -5°F "for safety," heating load rises.
The same problem occurs in cooling.
Check climate inputs first.
Mistake 2: Overstating infiltration
Marking every existing home as extremely leaky can add a large heating and latent load.
Use:
- Blower-door data when available
- Reasonable tightness assumptions when not
Do not use infiltration as a hidden sizing factor.
Mistake 3: Modeling good windows as bad windows
If a house has modern low-E windows but the model uses old clear double-pane or single-pane performance, cooling and heating loads can both rise.
Check:
- U-factor
- SHGC
- Area
- Orientation
Window assumptions can move cooling load substantially.
Mistake 4: Exaggerating window area
Simple data-entry errors can be brutal.
Examples:
- Entering width/height in wrong units
- Counting a window twice
- Treating a glass door as a door and a window
- Copying repeated windows to the wrong room
If one room's cooling load is bizarrely high, inspect its glazing first.
Mistake 5: Treating interior walls as exterior
A wall between two conditioned rooms should not be modeled like a wall to outdoors.
This can happen when auto-generated room geometry is not cleaned up.
The more rooms affected, the larger the artificial load.
Mistake 6: Double-counting the thermal boundary
Common examples:
- Modeling both roof deck and ceiling as exterior boundaries in an encapsulated attic
- Counting a basement surface both as below-grade and outdoor wall
- Duplicating floor exposure
Sketch the thermal boundary if the geometry is confusing.
Mistake 7: Assuming no insulation when insulation is merely unknown
Unknown does not mean worst possible.
In an old home, use the best evidence available.
If wall insulation is uncertain, run plausible scenarios.
Do not automatically choose an extreme assembly unless evidence supports it.
Mistake 8: Excessive duct penalties
Ducts in a hot attic create real load.
But unrealistic leakage or loss assumptions can inflate the system requirement.
Inspect:
- Duct location
- Insulation
- Leakage condition
- Supply/return configuration
If the ducts are inside conditioned space, do not model them like attic ducts.
Mistake 9: Double-counting outdoor air
Infiltration and mechanical ventilation are different.
A model can become too high if the same outdoor airflow is effectively counted multiple times.
Review how the ventilation strategy interacts with infiltration assumptions.
Mistake 10: Wrong conditioned area or ceiling height
If the house is 2,000 ft² but the model accidentally includes an unconditioned garage, load rises.
If 8-foot ceilings are entered as 18 feet, volume and exposed area can explode.
Basic geometry errors often matter more than fine material details.
Mistake 11: Adding manual safety factors
Some workflows inflate multiple inputs:
- Lower insulation
- Higher leakage
- Worse windows
- More extreme temperatures
- Then round equipment up
Each adjustment compounds the others.
Manual J should model the building.
If reserve capacity is needed, address it transparently during equipment selection within the applicable rules.
How to troubleshoot a high result
Work from biggest drivers to smallest.
- Confirm floor area and ceiling height.
- Confirm design temperatures.
- Check windows and orientation.
- Check attic/roof.
- Check infiltration.
- Check duct location and losses.
- Check foundations.
- Review room-by-room outliers.
- Look for duplicated surfaces.
Do not tweak random values until the total "looks right."
Find the physical reason.
Compare with operating history carefully
If the existing 3-ton system maintains setpoint on design days and the new Manual J says the house needs 5 tons, that is a reason to investigate.
It does not prove the old system is delivering exactly 3 tons or that the new model is wrong.
But the mismatch is valuable QA information.
Bottom line
A Manual J that is too high usually contains assumptions that make the house look more demanding than reality.
The fix is not to choose a smaller unit despite the calculation.
The fix is to correct the calculation.
A defensible load should survive review of geometry, climate, windows, infiltration, envelope, and ducts without relying on hidden safety factors.