Ask ten people in an Indian lab how they know their fume hood is safe, and nine will give you the same answer: the face velocity is 100 feet per minute or 0.5 m/s. It is on the validation sticker. It was in the tender. The service engineer measures it once a year with an anemometer and signs the fume hood test report.
Here is the problem. Face velocity is a measure of how much air is moving into the hood. It is not a measure of whether the hood is keeping fumes away from the person standing in front of it. Those are different things, and the gap between them is where exposure happens.
If your specification stops at 100 fpm, you are buying an air velocity number, not protection.
Where the 100 fpm number came from
The number is a historical convention that hardened into a rule.
Older editions of the standards did prescribe a range. ANSI/AIHA Z9.5-1992 required each hood to maintain an average face velocity of 80 to 120 fpm. An older edition of NFPA 45 said that 80 to 120 fpm “generally provide containment”, but only if the hood’s location and the laboratory’s ventilation met other criteria. Note the conditional. Even then, the number was never the whole answer.
What is striking is how much the standards have moved since, and how little Indian procurement practice has moved with them.
OSHA’s laboratory standard, 29 CFR 1910.1450, does not set a number at all. Its non-mandatory Appendix A describes adequate hood performance as “typically 60-100 lfm” and adds that airflow “should not be excessively turbulent”. Those two phrases sit next to each other for a reason, and we will come back to that.
The current ANSI/ASSP Z9.5-2022 goes further. Its requirement reads, in full: “Face Velocity shall be adequate to provide containment.” That is a performance requirement, not a number. The same standard asks for cross-draft measurements and tracer gas containment testing, because those are the things that actually tell you whether the hood works.
SEFA 1-2026, the current laboratory fume hood standard, calls 60 to 100 fpm “the most widely accepted range of average face velocities”. It defines a high performance hood as one that delivers containment equal to or better than ASHRAE 110 ratings of AM 0.05 and AI/AU 0.10 at a face velocity of 60 fpm or less, with the sash fully open.
Read that again. The standard explicitly contemplates a high performance hood will contain better or equal at 60 fpm, than a conventional hood does at 100.
One fair counterpoint, because the picture is not uniform. California’s Cal/OSHA regulation still requires an average of 100 fpm with a minimum of 70 fpm at any point, and 150 fpm for certain listed carcinogens. Some US universities still hold their own hoods to 100 fpm at an 18-inch sash. Where a local regulation sets a number, that number is binding regardless of what the hood could achieve. The point is not that 100 fpm is wrong. It is that 100 fpm by itself proves nothing.
More air does not mean more containment
This is the part that surprises most procurement teams.
Mills and Sartor at Lawrence Berkeley National Laboratory put it plainly in their 2006 study: increasing the amount and rate of airflow “does not tend to improve containment”. Above a certain point, pushing more air through the face of a hood creates turbulence rather than control. The air tumbles over the sash edge, rolls around the operator’s body, and generates eddies and vortexes at exactly the place you least want them, in the breathing zone.
SEFA 1-2026 builds this into its design rules. It limits the face velocity at a 6-inch sash opening to no more than three times the fully open velocity, specifically because “excessive face velocity can cause significant turbulence within the hood”. The older SEFA 1-2010 said it even more directly: on a properly designed hood, a lower face velocity can actually enhance performance.
Prudent Practices, the National Research Council’s reference text, recommends 80 to 100 fpm and warns that velocities approaching or exceeding 150 fpm should not be used.
What does drive containment, then? Three things, roughly in order:
- Internal aerodynamics. The shape of the baffle, the airfoil, the sash edge and the exhaust plenum decide whether air moves through the hood in a single controlled path or breaks into multiple vortexes. A hood designed around a single pass airflow path contains at 60 fpm. A box-baffle hood often does not contain reliably even at 100.
- Cross-drafts. A corridor door, an air conditioning diffuser, a person walking past, a pedestal fan. Z9.5-2022 states that cross-draft velocities “shall not be of such magnitude and direction as to negatively affect containment”. Commonly cited engineering guidance puts the tolerable disturbance at roughly 30 to 50 percent of the hood’s face velocity. In many Indian labs the room air distribution has never been measured at all.
- The user. An open sash, clutter inside the hood, equipment blocking the rear baffle slots, and the operator’s own body creating a wake. None of this shows up on an anemometer reading.
The energy cost of the number
There is a second reason this matters, and it will interest your CFO more than your QA head.
Exhaust volume scales directly with face velocity. At a fixed sash opening, exhaust (CFM) equals face velocity multiplied by the open area. Run a hood at 60 fpm instead of 100 and you exhaust roughly 40 percent less air. Every one of those cubic feet has been cooled, dehumidified and moved by a fan, and in an Indian lab running 24 hours a day on 100 percent outside air, that is the single largest line item in the ventilation bill.
So the 100 fpm convention is not a free safety margin. You pay for it every month, and the standards do not require it.
One warning before anyone reaches for the VFD to reduce the face velocity. Lab Manager’s guidance is blunt: never simply turn a standard hood down to 60 fpm, because it will leak. Low face velocity only works in a hood that was designed and tested for it. Turning down a conventional hood converts an energy problem into an exposure problem.
What to specify instead
Replace the number in your specification with evidence. Four things are worth asking for, and all four are grounded in the standards above.
- A third-party ASHRAE 110 test report at your intended operating face velocity, not just a certificate. The report should show the testing laboratory, the date, the conditions and the measured leakage at each ejector position. SEFA 1-2026 asks for the test at both the design sash opening and the maximum sash opening, and it explicitly encourages independent third-party testing.
- As-installed commissioning, which is tracer gas testing of the hood in your room, not in the factory. This is where cross-drafts reveal themselves.
- A cross-draft analysis of the room, so that hood placement, doors and diffusers are assessed before the hood is installed rather than after someone reports a smell.
- A continuous airflow monitor with audible and visual alarm. SEFA 1-2026 is unambiguous on the alternative: “A ribbon taped to the bottom of the sash is not acceptable.”
Where this led us
LabGuard built India’s first dedicated ASHRAE 110 and EN 14175 test rig at our Pen facility in 2007, which means we have spent close to two decades watching hoods pass and fail under tracer gas. The pattern is consistent. Containment is won in the baffle and the airfoil, not in the fan.
That is the principle behind AeroGuard Dynamic Airflow in the Ray fume hood. A single pass vortex path and an aerodynamic internal baffle allow Ray to operate at 60 fpm with low pressure drop and laminar flow, rather than compensating for turbulence with volume.
Victor Neuman, co-author of ANSI Z9.5, tested Ray under SF6 tracer gas at 60 fpm with the sash fully open, the worst case condition. The result at the left, right and centre ejector positions was 0.000 ppm against an allowable limit of 0.05 ppm. Ray also runs a self-validation test using a real-time laser and smoke check, so containment is verified between annual certifications rather than assumed.
If you are drafting a fume hood specification this quarter, we are happy to review the containment clauses with your engineering team. No commitment, and the review is vendor-neutral.