Duct leakage testing finds and seals gaps in the air distribution network by pressurising the duct to 250 to 500 Pa and measuring the flow needed to hold that pressure; a well-sealed supply duct should lose under 5 percent of design flow, while SMACNA class 6 permits about 6 cfm per 100 ft² at 1 inch w.g. Leaks above 10 percent waste fan energy and usually point to unsealed joints that a 30-minute test can locate and fix.
Why Duct Leaks Waste Real Money
The fan delivers a fixed volume, say 10,000 m³/h; if 8 percent escapes through seams, that air is heated or cooled and then lost in a ceiling void. At 250 to 500 Pa static, a single 10 mm gap leaks 15 to 25 cfm, and a typical poorly sealed plant room can lose 15 to 25 percent of supply. Closing those gaps cuts fan power 5 to 12 percent and improves delivered cooling, so the test pays back in one to two cooling seasons.
Testing Methods Compared
Two methods dominate: the fan-pressurisation orifice method for whole-duct certification, and the tracer or ultrasonic scan for locating individual leaks. The table sets out when each fits.
| Method | Pressure used | Measures | Best for |
|---|---|---|---|
| Fan pressurisation | 250 to 500 Pa | Total leakage cfm | Commissioning cert |
| Ultrasonic scan | None, listens | Leak location | Existing duct hunt |
| Soap bubble | 100 to 250 Pa | Joint by joint | Small fixes |
| Thermal image | Operating | Cold spot leak | Supply leaks |
Step-by-Step Leakage Test
Isolate the section and seal outlets with blanking plates. Connect a calibrated fan and orifice to one end and raise pressure to the test point, typically 250 Pa for low-pressure and 500 Pa for medium-pressure duct. Record the make-up flow in cfm; divide by duct surface area to get the class. For a 100 m² supply duct, class 6 at 6 cfm per 100 ft² equals about 65 cfm of allowed loss, so readings above that fail and need sealing.
Finding the Leaks
With the duct at 100 to 250 Pa, walk the run with a soap solution and brush it on every transverse and longitudinal joint; bubbles show the path. An ultrasonic detector picks up the 30 to 50 kHz hiss through insulation without opening it. Mark each leak losing over 1 cfm; on a 50 m run you commonly find 5 to 15 such joints, and sealing them typically recovers 40 to 60 percent of the total loss.
When to Test vs When to Just Reseal
- Test and certify when commissioning a new AHU plant or the spec requires SMACNA class 3.
- Test existing ducts when energy use is high but filters and coils are clean.
- Reseal all transverse joints on sight during a major overhaul without a full test.
- Skip the full test on short, exposed, positive-pressure runs under 10 m that you can visually inspect.
Sealing Right the First Time
Apply UL-181 mastic to longitudinal seams and joint flanges, covering the full width, and let it cure before pressurising. Use foil tape only for gaps under 3 mm. Never rely on cloth duct tape; it embrittles in 1 to 2 years and is responsible for up to 20 percent of measured leakage on old plants. After sealing, repeat the pressurisation test; a pass shows leakage below 5 percent and confirms the fan is no longer paying for lost air.
Keruite manufactures this equipment at an 80,000 m² facility in Dezhou, China, with an in-house AMCA-210-style test laboratory; every unit is balanced to G2.5 and performance-tested before shipment, and duct static-pressure performance is verified during factory acceptance.
SMACNA Leakage Classes Explained
SMACNA classes set the allowed leakage per 100 ft² of duct surface at 1 inch w.g.: class 1 allows under 1 cfm, class 3 about 3 cfm, and class 6 about 6 cfm, with higher classes for low-pressure residential-style work. Specifying class 3 on a supply plant typically caps loss at 3 to 5 percent of design flow, while leaving it to class 6 commonly lets 8 to 12 percent escape. The tighter class costs a little more mastic and labour at build but returns 5 to 12 percent lower fan power for the life of the system, which on a 10,000 m³/h plant is the cheaper choice.
Leakage vs Duct Pressure Class
Match the test pressure to the duct construction: low-pressure duct at 250 Pa static is tested near 250 Pa, while medium-pressure at 500 Pa is tested near 500 Pa, because leaks scale with pressure and a low test hides real losses. A joint that passes at 250 Pa can leak 40 percent more at the 500 Pa the system actually runs, so always test at or above operating pressure. Record the class and the pressure together on the commissioning sheet so a later auditor can see the result means what it says.
Case Study: A 10,000 m³/h Supply Plant
A new 10,000 m³/h air-handling plant failed its first test at 65 cfm of leakage against a class 3 target of about 45 cfm. The team walked the run with a soap solution, found 11 transverse joints weeping at 1 to 4 cfm each, and resealed them with UL-181 mastic. A retest read 28 cfm, a 57 percent cut, and the fan static pressure dropped 6 percent for the same delivered flow, saving roughly 4 percent of fan power every hour the plant runs and proving the test paid for itself in the first year.
Retesting and Accepting the Sealed Duct
Sealing is not finished until you prove it, so repeat the pressurisation test after the mastic has cured for the maker's stated time, usually 24 hours, and compare the new make-up flow against the class target. A pass is leakage below 5 percent of design flow, or under the SMACNA class written in the spec, whichever is tighter; a fail means another walk with the soap solution to find the joint that still weeps. Sign off the commissioning sheet with both the before and after cfm and the test pressure, because that pair is the evidence the fan is no longer paying to condition a ceiling void and the auditor will ask for it years later.
FAQ
What is an acceptable duct leakage rate?
SMACNA classifies ducts by leakage: class 6 allows about 6 cfm per 100 ft² at 1 inch w.g., class 3 about 3 cfm, and class 1 under 1 cfm. A well-sealed supply duct should lose under 5 percent of design flow; above 10 percent the fan energy is wasted and the test has failed.
How do you find a duct air leak?
Pressurise the section to 250 to 500 Pa and walk it with a soap solution or an ultrasonic leak detector that hears the 30 to 50 kHz hiss. Thermal imaging shows the cold spot at a supply leak. Mark every joint losing more than 1 cfm and seal it; a single 10 mm gap can leak 15 to 25 cfm.
Should supply or return ducts be tested?
Test supply ducts because they carry conditioned air at positive pressure and waste energy directly. Return ducts at negative pressure pull in unconditioned, unfiltered air, so test them for indoor-air-quality leakage. On a balanced system test both if the spec calls for SMACNA class 3 or tighter.
What sealant is best for duct leaks?
Use UL-181 listed mastic on longitudinal seams and joints; it stays flexible and beats tape long term. Foil tape rated UL-181A is fine for small gaps under 3 mm. Avoid cloth duct tape, which dries and fails within 1 to 2 years and can account for 20 percent of total leakage.