You run the load.
The house has a 5-ton AC.
Manual J says the design cooling load is 31,000 BTU/hr.
Now what?
The wrong reactions are:
- "Manual J must be wrong."
- "The old unit must be wrong."
Either could be true.
A large mismatch is a reason to investigate before selecting equipment.
Step 1: Verify the Manual J inputs
Start with the model.
Check the highest-impact inputs:
Geometry
- Conditioned floor area
- Ceiling height
- Exterior walls
- Vaulted spaces
- Basements and bonus rooms
Design conditions
Make sure the project is using the correct local summer and winter design data.
Windows
Check area, orientation, U-factor, and SHGC.
Insulation
Confirm major roof, wall, floor, and foundation assumptions.
Infiltration
Make sure the house was not modeled unrealistically tight.
Ducts
If ducts are in a hot attic or other unconditioned space, ensure the system load accounts for them appropriately.
If the model survives that review, take the result seriously.
Step 2: Ask whether the house changed
The old equipment may have been selected before:
- Window replacement
- Attic insulation
- Air sealing
- Crawlspace encapsulation
- Addition removal or reconfiguration
- Spray-foam retrofit
- Duct repairs
The old tonnage may describe an older version of the house.
Step 3: Look at operating history
How did the old system behave?
Useful clues:
- Very short cycles during hot weather
- High indoor humidity
- Frequent starts and stops
- Large temperature swings
- The system rarely reaching long runtime
- Cold supply blasts followed by long off periods
Those symptoms can be consistent with oversizing.
But be careful.
A bad thermostat location, poor airflow, or refrigerant problems can mimic some of them.
Step 4: Do not force the Manual J up to the old tonnage
This is a common failure mode.
The contractor expects 5 tons, sees 31,000 BTU/hr, and starts changing inputs:
- Make the house leakier
- Lower the insulation
- Use harsher design temperatures
- Increase window loads
- Add arbitrary duct penalties
Eventually the calculation "proves" 5 tons.
That is backwards.
Inputs should describe the building, not justify an equipment choice.
Step 5: Evaluate the actual replacement equipment
A 31,000 BTU/hr cooling load does not automatically mean "install a 2.5-ton unit."
Manual S still matters.
Check actual equipment performance at design conditions.
For cooling, verify:
- Total capacity
- Sensible capacity
- Latent performance
- Sizing limits
For heat pumps, also verify:
- Low-ambient heating capacity
- Heating deficit
- Auxiliary heat
Nominal tonnage is not the final answer.
Step 6: Check the ducts
If the old system was oversized, the duct system may also be poorly matched.
Possible problems include:
- Excessive airflow
- High static pressure
- Noisy registers
- Undersized returns
- Poor branch balancing
A smaller system can sometimes fit the existing ductwork better.
But if you change airflow, confirm room distribution still works.
What if the homeowner says the old system worked perfectly?
That does not prove it was correctly sized.
Oversized systems can maintain thermostat setpoint easily.
Ask more specific questions:
- Did humidity stay comfortable?
- Were cycles long or short?
- Were rooms even?
- Was the system noisy?
- Did it run continuously on the hottest days?
- Did the homeowner make major envelope upgrades?
"Kept the house cold" is only one performance metric.
What if the old system ran constantly?
That deserves attention.
If the existing unit runs nearly continuously during real design weather and maintains setpoint, a load calculation showing dramatically less required capacity should be reviewed carefully.
Possible explanations:
- Existing equipment is degraded
- Duct losses are large
- Actual airflow is low
- Refrigerant performance is poor
- Manual J inputs are too optimistic
- The outdoor conditions were beyond design
Runtime is evidence. Use it to QA the model.
Oversized furnace vs. oversized AC
Furnace nameplate capacity is especially weak evidence of building heating load.
Residential furnaces are commonly installed with large margins.
Do not assume a 100,000 BTU/hr furnace means the home needs 100,000 BTU/hr of heat.
Heat-pump conversions should start with a new Manual J rather than translating furnace size into heat-pump tonnage.
Bottom line
When Manual J says the existing system is oversized, do not ignore the result and do not blindly trust it.
Verify the model.
Compare it with building history and runtime.
Then perform Manual S on the actual replacement equipment.
The goal is not to make the calculation agree with the old nameplate.
The goal is to determine what the current house actually needs.