How Building Envelope Infiltration Sabotages HVAC Load Calculations
On paper, the building was airtight. Mechanical engineering teams know how that story usually ends: the chiller is sized exactly to the calculation, the owner moves in, and by July the offices feel clammy and the system never quite catches up.
The culprit is often air nobody counted. Uncontrolled air leaking through the building envelope, called infiltration, adds heat and moisture the load calculation never saw. When the assumed leakage is wrong, the equipment ends up oversized, undersized, or just badly matched to the real load.
This article explains how infiltration throws off HVAC load calculations, where the error creeps in, and how engineers get closer to the truth.

What Infiltration Is and Why Load Calculations Depend on It
Infiltration is outdoor air that leaks into a building through cracks, gaps, and openings in the building envelope, which is the shell of walls, roof, windows, doors, and floors that separates inside from outside. Its opposite, exfiltration, is indoor air leaking out.
It is not the same as ventilation. Ventilation is outdoor air you bring in on purpose, through ducts and equipment, so you can control and condition it. Infiltration is accidental. It comes and goes with wind, temperature, and pressure, and it bypasses your filters and heat recovery.
A HVAC load calculation is the process of estimating how much heating and cooling a building needs at design conditions. Think of it like sizing a bucket to catch the rain. If the roof leaks and you did not count the leaks, the bucket will be the wrong size. Infiltration is one of those leaks, and it affects both heat and humidity.
7 Ways Envelope Infiltration Throws Off HVAC Load Calculations
Infiltration does not fail in one dramatic way. It chips away at accuracy in several quiet ones. Here is where the damage happens.
1. The Leakage Number Is Usually a Guess
Most early-stage load calculations use an assumed air change rate or a rule-of-thumb leakage value, because the building does not exist yet. That assumption feeds straight into the cooling and heating load. If the real building leaks twice as much, the calculation is off by a large margin before the first duct is drawn.
2. The Latent Load Gets Underestimated
Sensible load is about temperature. Latent load is about moisture: the energy needed to pull water vapor out of the air. Humid outdoor air leaking in adds latent load that standard equipment may not handle well.
This is why buildings with leaky envelopes in humid climates can feel cold but sticky. The air temperature is fine, but the humidity is not.
3. Leakage Changes With Weather, Not Just Design Day
Load calculations use fixed design conditions, such as a hot summer afternoon or a cold winter night. Real infiltration moves with wind speed, wind direction, and the temperature difference between inside and out. A building may leak lightly most days and heavily during a storm, and the design number captures only one snapshot.
4. Stack Effect Varies by Building Height
In tall buildings, warm indoor air rises and escapes at the top, pulling cold air in at the bottom. This is called stack effect. It means infiltration is not uniform: lower floors in winter may see strong inflow, while upper floors see outflow. A single average value hides these differences, and so does a single zone assumption.
5. Leakage Is Not Spread Evenly
Calculations often assume air leaks evenly across all exterior walls. In real buildings, leakage concentrates at specific spots: shaft penetrations, roof-wall joints, loading dock doors, and window frames. Those hot spots can overload one zone while others sit fine, leaving comfort complaints in a few rooms and no explanation in the model.
6. Pressurization Can Hide or Worsen the Problem
Mechanical systems pressurize or depressurize parts of a building. Exhaust fans can pull air in through leaks, while excess supply air can push conditioned air out. If the calculation assumes a neutral building and the real one runs negative, infiltration increases. The model and the building end up describing two different places.
7. Safety Factors Cover the Error, but at a Price
When engineers cannot trust the leakage assumption, they often add a safety margin. That keeps occupants from freezing or sweating, but it can lead to oversized equipment. Oversized systems cost more upfront and may short-cycle, which wastes energy, wears components, and can leave humidity unresolved because the unit shuts off before it dries the air.
Now that you know where the errors come from, the next question is how big they can get.
The Infiltration Load Math, Explained in Plain Terms
The math behind infiltration is simple, which is exactly why a wrong input hurts so much. Three equations do most of the work, using the standard-air constants common in sea-level calculations.
Airflow: CFM = (ACH × building volume in cubic feet) ÷ 60. This converts air changes per hour (ACH) into cubic feet per minute.
Sensible load: Btu/h = 1.08 × CFM × temperature difference in °F.
Latent load: Btu/h = 0.68 × CFM × humidity ratio difference in grains of moisture per pound of dry air.
A Sample Calculation
These numbers are illustrative, not from a real project. Imagine a 120,000 cubic foot building with an assumed infiltration rate of 0.5 ACH on a hot, humid design day.
Airflow: (0.5 × 120,000) ÷ 60 = 1,000 CFM
Sensible load at a 20°F difference: 1.08 × 1,000 × 20 = 21,600 Btu/h
Latent load at a 40 grains/lb difference: 0.68 × 1,000 × 40 = 27,200 Btu/h
Total: about 48,800 Btu/h, or roughly 4 tons of cooling
Now suppose the real building leaks twice as much, 1.0 ACH. The same math gives about 97,600 Btu/h, or roughly 8 tons. The load doubles, because it scales directly with leakage.
Notice something else: in this example, the latent load is larger than the sensible load. That is why ignoring moisture can matter more than ignoring temperature.
Where Leakage Hides in a Building Envelope
Air finds the weak points, and most of them are small. Here are the usual suspects and why they matter to a load calculation.
Location | Why it leaks | Effect on the load |
Window and door frames | Gaps around frames and worn weatherstripping | Adds both sensible and latent load near the perimeter |
Roof-to-wall joints | Hard-to-seal transition where two assemblies meet | Drives stack effect and top-floor loss |
Penetrations | Pipes, conduits, and ducts that cut through the air barrier | Creates concentrated leakage in specific zones |
Loading docks and large doors | Doors open often and seal poorly | Spikes infiltration in short bursts |
Elevator and stair shafts | Tall vertical paths between floors | Amplifies stack effect in tall buildings |
Curtain wall and panel joints | Joint sealants age and move | Increases leakage over time as the building ages |
The lesson is simple. Leakage is a collection of small paths, and the paths you cannot see are the ones that skew the number.
How Mechanical Engineers Get Accurate Infiltration Numbers
You cannot eliminate uncertainty, but good mechanical engineering practice can shrink it. Here are the tools that matter most.
Specify the air barrier early. A continuous air barrier, detailed at transitions and penetrations, reduces leakage before it happens. The design team should treat it as a system, not an afterthought.
Require fan pressurization testing. A blower door or large-building fan test measures how leaky the envelope really is at a set pressure. Results are often reported as air changes at 50 pascals (ACH50) or flow per square foot of envelope.
Convert test results carefully. Test pressure is higher than normal conditions, so engineers use conversion methods to estimate natural infiltration. A common starting rule divides ACH50 by roughly 20, but the right factor depends on climate, height, and exposure.
Use diagnostic tools. Smoke tests and thermal imaging help find where the leaks actually are, not just how large the total is.
Run sensitivity checks. Model the load at low, expected, and high leakage values. If the equipment selection swings wildly, the assumption deserves more attention.
Coordinate with commissioning. Verify that mechanical systems pressurize the building the way the calculation assumed.
For existing buildings, testing before a retrofit is even more valuable. The measured leakage replaces guesswork, and the equipment can be sized to the real load.
Count the Air You Cannot See Before You Size the Equipment
Infiltration is easy to overlook because it is invisible, but it can change a load calculation by thousands of Btu per hour. Treat the leakage assumption as a design decision, test it where you can, and check it against the equipment you select.
If you want a team that treats these details with care, reach out to Century Engineers Inc. We provide specialized engineering consulting, including mechanical engineering support for load calculations, system design, and building performance, backed by experts who have spent years helping clients get the numbers right. Call 323-847-2668 or visit https://www.centuryengineers.com/services to learn more about how we can help.
Frequently Asked Questions
What is infiltration in an HVAC load calculation?
It is the load created by outdoor air leaking into a building through the envelope. It adds sensible heat and moisture that the system must remove, and it is usually estimated using an air change rate.
How does infiltration affect HVAC sizing?
More leakage means a higher heating and cooling load. If the assumption is too low, equipment may be undersized. If it is too high, equipment may be oversized.
What is the difference between infiltration and ventilation?
Ventilation is intentional outdoor air supplied through the HVAC system. Infiltration is uncontrolled leakage through the envelope.
How is building envelope leakage measured?
Engineers use fan pressurization tests, such as a blower door test, which pressurize or depressurize the building and measure the airflow needed to hold a set pressure.
Can a tighter envelope reduce equipment size?
Yes, in many cases. A tighter envelope lowers the infiltration load, which can allow smaller equipment, provided ventilation is handled properly and the calculation reflects the measured performance.




Comments