DW/143 duct leakage limits: how to work out the pass figure for each class
The leakage limit is k × p0.65 litres per second per square metre of duct surface, where p is the test pressure in pascals and k is 0.027 for Class A, 0.009 for Class B, 0.003 for Class C and 0.001 for Class D. Multiply that rate by the section's surface area to get the pass figure, so a 46 m² Class B section tested at 1,000 Pa may leak up to 36.9 L/s.
Where the limits come from
DW/143 is the BESA guide to ductwork leakage testing and sets out how the test is done, while the limits themselves are set in DW/144, which is why specifications usually cite the two together. DW/144 groups ductwork into four leakage classes, with Class A for low pressure, Class B for medium pressure and Classes C and D for high pressure, and gives each class a limit that rises with the test pressure, since a duct held at a higher pressure pushes more air through the same gaps.
The project specification tells you which class each system belongs to and what pressure to test it at, so the first step is to read both values for the section on the rig from the specification rather than assume them from the system type. With those two values and the section's surface area, the pass figure is a short calculation.
The formula for each class
The limit is a rate of leakage per square metre of duct surface, worked out as k × p0.65, where p is the test pressure in pascals and k is the coefficient for the class. The table gives the formula for each class and the limit it produces at one example pressure.
| Class | Leakage limit, L/s per m² | Example test pressure | Allowed leakage |
|---|---|---|---|
| A (low pressure) | 0.027 × p0.65 | 400 Pa | 1.33 L/s per m² |
| B (medium pressure) | 0.009 × p0.65 | 1,000 Pa | 0.80 L/s per m² |
| C (high pressure) | 0.003 × p0.65 | 2,000 Pa | 0.42 L/s per m² |
| D (high pressure) | 0.001 × p0.65 | 2,000 Pa | 0.14 L/s per m² |
p is the test pressure in pascals.
The example pressures are the ones used to work out these values and are not a rule, so a section tested at another pressure has a different limit, and the formula has to be run with the pressure from your specification each time.

Worked example for a whole section
Take a section of medium pressure ductwork in Class B with a surface area of 46 m², tested at 1,000 Pa. Raising 1,000 to the power of 0.65 and multiplying the result by 0.009 gives a limit of 0.80 L/s per m², and multiplying that rate by the 46 m² gives an allowable leakage of 36.9 L/s for the whole section. The rig reading for the section is compared with that figure, so a measured leakage at or below 36.9 L/s passes and anything above it fails.
Carry the unrounded rate into the multiplication and round only the final figure, which is why a rate shown as 0.80 still gives 36.9 L/s for this section. The difference is small, but a section that reads close to its limit is exactly where a rounding step can turn a pass into a fail, and the witness will check the arithmetic.
When published coefficients do not match
Some published tables of DW/144 coefficients show figures roughly a tenth of the ones above, and a limit worked out from them is far stricter than one worked out from the figures in this post. The mistake that follows is taking a coefficient from one source and a limit or rig reading from another, which can fail a section that should pass or pass one that should fail.
Before comparing any coefficient, limit or reading, check the unit printed next to it and check the coefficient against the table in your project's copy of DW/144, keeping the whole calculation in litres per second, square metres and pascals. If a specification quotes coefficients that do not match the ones above, raise it with the consultant before testing so everyone works from the same figures.
Questions about DW/143 leakage limits
How is the duct surface area of a test section worked out?
From the drawings, by multiplying each straight run's perimeter by its length, which is twice the sum of width and height for rectangular duct and π times the diameter for circular duct, then adding the surface of each fitting in the section. Work in metres so the area comes out in square metres, and mark it on the drawing so the witness can check it.
What test pressure should be used?
The one your project specification sets for the system, which normally relates to the pressure the ductwork works at in service. If the specification gives a class but no test pressure, ask the consultant to confirm it in writing before testing starts, because the limit changes with the pressure.
Which ductwork has to be tested?
High pressure ductwork must be tested, while low and medium pressure ductwork is tested when the project specification asks for it. Check the specification for each system rather than relying on the pressure class alone.
Is "acceptable leakage percentage" the same thing?
No, a percentage of the system's design airflow is a different measure that some specifications use, and it cannot be turned into a DW/144 class without the section's airflow and surface area. If your specification quotes a percentage, confirm with the consultant which criterion governs before testing.
Is DW/143 the same as SMACNA leakage classes?
No, SMACNA leakage classes are expressed in cfm per 100 ft² of duct surface at 1 in. w.g., so a SMACNA class cannot be read across to a DW/144 class. If a specification mixes the two, ask the consultant which one governs before the method statement is approved.
Book a duct leakage test
Send the drawings, the specification and your target dates, and we'll reply with a test plan and a quote.
