How HVAC Load Estimation Actually Works: The Engineering Behind System Sizing

Every honest system size starts in the same place: a load calculation. Before anyone talks tonnage or brands, the question is how much heat a building gains in summer and loses in winter at the worst reasonable conditions. That quantity is the load, measured in BTU per hour, and it is the foundation the whole design sits on. This is the engineering behind that number, written for homeowners who want to understand what a real contractor is actually doing when they size your system.

Why we calculate a load at all

Cooling and heating loads exist to answer a single practical question: what capacity of equipment holds the indoor design conditions when the outdoor weather is at its design value. Get the load right and the rest of the design follows. Get it wrong and no amount of good equipment fixes it.

The load is not the same as the equipment we install. We calculate the load, then choose the nearest standard equipment size that covers it. The gap between those two numbers is where oversizing creeps in, so a good installer keeps it small and deliberate rather than rounding up “to be safe.”

Heat gain is not the same as cooling load

This is the idea that separates someone who runs the math from someone who reads a chart. Heat enters a building two ways. Convective gains, like warm air and infiltration, become a load almost instantly. Radiant gains, like sun striking a floor or a wall heating up, are absorbed by the mass of the building first, then released hours later. So the heat that flows in at 2 PM does not all show up as a cooling load at 2 PM. Part of it is stored and re-radiated into the evening.

The consequence is that the peak cooling load lags and smooths the peak heat gain. The refined methods account for that storage and lag instead of pretending heat gain equals cooling load. On the heating side it is simpler. Heat loss is treated as instantaneous, so the heat loss at the design temperature can be used directly to size the furnace or heat pump.

Design conditions: the temperatures we size to

Indoor. Standard comfort targets are about 75°F and 50% relative humidity for cooling, and 70°F for heating. Those are the conditions the equipment has to hold.

Outdoor. Here is the part most people get wrong. We do not size for the hottest or coldest temperature on record. We size to a frequency-based design value. The cooling design temperature is the value the location exceeds only about 1% of the hours in a year, roughly 88 hours. The heating design value is the 99% value, the temperature it drops below only about 88 hours a year. These come from weather-station records published by ASHRAE.

Why work this way: it is neither economical nor practical to build a system around a peak that might last two or three hours once every few years. Covering that single peak oversizes the equipment for every normal hour. The accepted practice is to size to the design value and accept a small, brief comfort drift on the rare hours beyond it.

Across our service area the cooling design values run about 81°F on the East Bay coast and San Francisco, 92°F in the South Bay, and 99°F in the hot inland Tri-Valley and Diablo Valley. Heating design values land around 30°F to 36°F. That spread is the reason a coastal house and an inland house of the same size do not get the same system. We go deeper on this in what temperature your HVAC should be sized for.

Where the load comes from

Both loads break down into the same handful of sources. Knowing them is how we can tell you exactly what is driving your number.

• Conduction through the envelope. Heat moving through walls, ceiling, floor, windows, and doors, driven by the temperature difference. The rate for each surface is its area divided by its R-value, or its area times its U-value for glass. This is why insulation and window quality move the number so much.
• Solar gain through glass. Sun coming through windows, set by the glass area, its solar heat gain coefficient (SHGC), shading, and orientation. West and southwest glass dominates the late-afternoon peak when it is hottest outside. North glass adds little.
• Sol-air on opaque surfaces. The sun also heats the roof and walls well above outdoor air temperature, so conduction through them runs on an effective temperature difference higher than outdoor minus indoor. A dark roof over an attic is the big one. A proper calculation includes it; a plain temperature-difference shortcut misses it.
• Air leakage and ventilation. Infiltration through cracks plus any required mechanical ventilation on tight modern homes. Outdoor air carries both a sensible load from its temperature and a latent load from its moisture.
• Internal gains. People, lights, appliances, and electronics. Each occupant adds roughly 250 BTU per hour of sensible heat plus a latent component. On the cooling side these are heat we are trying to remove.

Sensible versus latent

Cooling does two jobs: drop the air temperature, which is the sensible load, and remove moisture, which is the latent load. The total cooling load is the sum, and the equipment coil has to handle both. This matters for comfort. An oversized system satisfies the thermostat on temperature so fast that it never runs long enough to wring out moisture, which is why an oversized AC can leave a house cold and clammy. In the dry inland Bay Area the latent load is small, so most of our cooling load is sensible, but the principle still drives why we do not oversize. Heating is treated as sensible only.

The methods, from crude to correct

Square-foot-per-ton rule of thumb. Floor area divided by a fixed ratio. Fast, and wrong often enough to matter. It ignores orientation, glass area, insulation, air leakage, the number of occupants, and the difference between a one-story and a two-story envelope. Useful for a ballpark phone conversation, never for an install contract.

CLTD/CLF, the Cooling Load Temperature Difference and Cooling Load Factor method. The most practical fully-manual method. It applies pre-computed temperature differences to walls and roofs that already include the sol-air and thermal-storage effects, so it captures the reality that heat gain is not cooling load without an hour-by-hour simulation.

Manual J for residential, Manual N for commercial. The simplified room-by-room load procedure that sizing should be based on. It works from first principles, conduction plus solar plus infiltration plus internal gains against the design temperature difference, and produces a defensible cooling and heating load for each room and the whole house. Manual J is required on every California HVAC permit. For the homeowner-facing version of this, with realistic tonnage ranges by climate zone, see our HVAC sizing guide.

The math in plain form

For anyone who wants to sanity-check a number, the core terms are straightforward:

• Conduction: for each surface, BTU per hour equals area divided by R-value times the temperature difference, or area times U-value times the temperature difference for glass.
• Sensible infiltration and ventilation: BTU per hour equals 1.08 times CFM times the temperature difference.
• Latent infiltration and ventilation: BTU per hour equals 4840 times CFM times the humidity-ratio difference.
• Solar through glass: glass area times peak solar gain times SHGC times shading times an orientation factor.
• Internal: roughly 250 BTU per hour of sensible heat per person, plus appliance and lighting gains.
• Tons: total cooling BTU per hour divided by 12,000, rounded up to the nearest standard size, with no padding on top.

How to get your system sized

You do not have to run any of this by hand. Our free HVAC load calculator runs exactly this math: envelope conduction by your home’s vintage and insulation, solar through glass by area and orientation, sol-air on the roof and walls, infiltration and mechanical ventilation, internal gains, sensible plus latent, all against the 1% cooling and 99% heating design temperatures for your climate zone. It reports the calculated Manual J load as the recommendation and shows the local square-foot rule of thumb only as a comparison, so you can see how much that shortcut would oversize you. For a single room, an ADU, or a garage conversion, the mini-split calculator sizes per zone.

When you want binding numbers, call (925) 999-4095 and we will run a full Manual J on-site as part of your installation estimate. The one-line version of everything above: the design temperature already excludes the rare extreme days, and the load calculation already accounts for your house specifically, so adding a safety margin on top of it is just oversizing a system you then run badly the other 360 days of the year. We size to the load.

Key takeaways

• A load calculation answers one question: what equipment capacity holds your indoor target when the outdoor weather is at its design value. The load is calculated first; equipment is chosen to cover it.
• Heat gain is not the same as cooling load. Radiant heat is stored in the building's mass and released hours later, so the peak cooling load lags and smooths the peak heat gain. Real methods account for that; rules of thumb do not.
• Equipment is sized to the design temperature (the 1% cooling / 99% heating value), not the record extreme. Sizing to the record day is how systems get oversized.
• The load comes from envelope conduction, solar through glass, sol-air on the roof and walls, air leakage and ventilation, and internal gains, split into sensible (temperature) and latent (moisture).
• Square-foot-per-ton is a phone-call ballpark. CLTD/CLF and ACCA Manual J are the real methods. Manual J is required on every California HVAC permit.
• Our [free load calculator](/free-hvac-load-calculator/) runs this exact math, and reports the calculated load as the recommendation, with the rule of thumb shown only as a comparison.

Related questions

Written by Andrew Kuznetsov. Andrew is the founder and owner of Bay Area HVAC Service (ADRIUM Service Solutions). He holds a California Contractor License (CSLB #1136642), EPA 608 certification, and completed factory training at the Daikin/Goodman plant in Houston in 2025. He writes from direct field experience, not marketing copy.

Further reading

• Mitsubishi Hyper-Heat (H2i): How It Keeps Putting Out Heat When It Gets Cold in the Bay Area , Most people think heat pumps quit when it gets cold. Mitsubishi's Hyper-Heat (H2i) line is built to keep working as temperatures drop, which matters more here than folks expect, especially on cold mornings and up in the East Bay hills.
• Mitsubishi Mini-Split Blinking a Fault Light? What It Means and What to Do , A blinking light on your Mitsubishi indoor head is the system telling you something is wrong. Here is what the blink pattern means, the checks you can safely do yourself, and when to bring in a factory-trained tech.
• Why Your Mitsubishi Mini-Split Is Leaking Water Inside (and How to Fix It) , A Mitsubishi mini-split dripping water inside almost always comes down to condensate that can't drain. Here's what causes it, the safe checks you can do yourself, and when to call a tech.