Fume hoods and local exhaust ventilation: what protects the operator and what does not
Face velocity, containment testing, ducted against filtered recirculating hoods, and the failures a hood cannot save you from. What to write into a tender so the hood you receive is the hood you specified.
A fume hood that has never been tested is a cupboard with a fan on it. That is not a rhetorical flourish: containment is a property of the whole installation, not of the box, and it is entirely possible for a well-made hood from a good manufacturer to fail to contain because the room has no make-up air, or because a ceiling diffuser blows across its face. Every laboratory we survey has hoods. Rather fewer have hoods with a test report, and fewer still have a test report from after installation rather than from the factory.
Face velocity is a proxy, not the goal
The goal is containment: contaminant generated inside the hood stays inside it and does not reach the operator's breathing zone. Face velocity became the specification because it is cheap to measure and containment is not. It remains a useful indicator, provided you remember what it does not tell you.
Too little face velocity lets contaminant escape at the sash opening. Too much is also a defect, because a fast inward stream separates around the operator's body and forms a vortex at the sash that pulls contaminant back out. Above a certain velocity, adding more makes containment worse while making the energy bill and the noise worse as well. Design practice for a conventional hood sits in a fairly narrow band, and your supplier should state the design face velocity and the design sash height together, because one is meaningless without the other.
When you measure, measure on a grid across the sash opening at the design height with a calibrated thermal anemometer, and record every reading rather than just the average. A hood with an acceptable average and a wide spread between the highest and lowest readings is a hood with a bad baffle setting or an obstruction, and the average is hiding it. The spread is the more informative number.
Containment testing: the test that actually means something
Type testing to EN 14175 part 3 is done in the manufacturer's laboratory on that model of hood, using a tracer gas released inside the hood and sampled in front of a mannequin. It proves the design. Ask for the report and check that it covers the model and sash configuration you are buying, not a related one.
On-site testing to EN 14175 part 4 is done on your hood, in your room, after installation. It proves the installation. This is the report that matters and it is the one most commonly missing.
The ASHRAE 110 tracer gas method reports results in three conditions that are worth understanding even if you specify the European standard: as manufactured, as installed and as used. 'As used' means with the equipment, bottles and clutter that are actually in the hood, and it is routinely the worst of the three. A hood full of apparatus is a different hood.
Smoke visualisation is qualitative, takes two minutes and costs almost nothing. Run a smoke pencil across the sash plane and along the work surface monthly. It will not give you a number, but it will find a reverse flow at the sash long before the annual test does.
Airflow monitor on every hood. A visual and audible low-flow alarm at the sash is the only thing standing between a belt failure on the roof and an operator who has no way of knowing. Treat it as part of the hood, not an accessory.
Ducted or filtered recirculating
Ductless hoods are attractive in Dhaka for real reasons: rented floors where you have no rights to a riser or a roof, buildings with no spare shaft, and refurbishments where a duct route simply does not exist. They are a legitimate answer to a narrow question, and a poor answer to a broad one.
Consideration
Ducted hood
Filtered recirculating hood
Chemistry it suits
Anything within the liner's chemical resistance, including changing and unknown work
A known, limited, low-volume list matched to a specific filter medium
Where the contaminant goes
Out of the building, treated by a scrubber if the discharge requires it
Onto a filter, then back into the room once the filter is exhausted
Fan energy plus conditioning the make-up air, which is the bigger figure in this climate
Filter replacement, and filters have a shelf life as well as a service life
The failure mode
Visible: fan stops, alarm sounds, work stops
Invisible: filter saturates and the hood discharges into the room
Who has to own it
Building maintenance, on a planned schedule
A named person tracking usage, breakthrough tests and filter dates
Filtered hoods are defensible for a teaching bench doing one known preparation, or a weighing enclosure for powders. They are the wrong choice for a research laboratory whose chemistry changes with each new student, and they are unsuitable for perchloric acid, for high-boiling acid digestions, for radionuclides, and for low molecular weight gases that carbon adsorbs poorly. If you buy one, buy the breakthrough test with it and put the filter change into a maintenance schedule with an owner, or it will be running on an exhausted filter within two years and nobody will know.
What a fume hood cannot protect you from
It does not protect the sample. A fume hood protects the operator by pulling room air inwards across unfiltered work. A biological safety cabinet protects operator and sample; a laminar flow bench protects only the sample and blows air at the operator. These three look similar from the corridor and are not interchangeable. Using a laminar flow bench for chemical work puts the contaminant straight into the operator's face.
It does not contain an explosion or a serious fire. A sash is a splash and spatter guard. Pressure work needs a blast shield and a hood rated for the duty.
It is not a storage cupboard. Bottles standing in a hood block the rear baffle slots, ruin the airflow pattern and are the single most common cause of a hood that passes an empty test and fails a real one. Solvents belong in a vented under-bench chemical store.
It does not work with the sash raised. The design sash height is part of the specification. Above it, face velocity falls and containment goes with it. Sash stops and a clearly marked working height are cheap.
It does not work without make-up air. This is the failure that bites hardest during load-shedding here, because the extract fans are usually on the generator and the air handling unit supplying the room is usually not. The hoods keep pulling, the room goes strongly negative, corridor doors become hard to open, and the hoods themselves starve and lose containment. Put extract and supply on the same essential supply, or interlock them so that losing supply air ramps the extract down in a controlled way.
It does not handle perchloric acid without a wash-down hood and dedicated ductwork, and it does not make radioactive work safe without the additional controls that work requires.
Commissioning, and the annual test after that
Balance the room first. Establish the supply air and the total extract, and confirm the laboratory sits slightly negative to the corridor. Slightly. If the door needs a shove, the balance is wrong and the hoods are competing with the door seals for air.
Check the discharge before you accept the installation. The stack should discharge upwards at a velocity high enough to carry the plume clear of the roof, and it must be away from fresh air intakes, opening windows and neighbouring roofs. In Dhaka the roof is shared with condensers, water tanks and often a caretaker's quarters, so this needs to be drawn and agreed rather than assumed.
Set and record the baffles. Then mark the setting, because somebody will move it.
Measure face velocity on a grid at the design sash height and record every reading, then run a smoke test and a tracer gas containment test to EN 14175 part 4 with the hood in its as-used state.
Prove the airflow monitor by shutting the fan down and confirming the alarm is audible and visible at the hood, and that a low-flow condition is reported to whoever is responsible for the building.
Label each hood with its design sash height, the date of the last test and the date of the next. A sticker on the sash frame does more for daily safety than a folder in an office.
Retest annually and after any change to the fan, ductwork, room ventilation or adjacent hoods. Adding one hood to a shared manifold changes the balance of every hood on it, which is why the annual test should cover the system rather than each hood in isolation.
Writing the tender so you get the hood you specified
In a competitive e-GP evaluation, anything not named in the bill of quantities is a cost the lowest bidder has already removed. These lines are worth the space they take:
An EN 14175 part 3 type test report for the offered model, from a third party, submitted with the bid rather than promised after award.
An on-site containment test to EN 14175 part 4 named as a handover deliverable, with the certificate as a condition of the final payment.
Liner and work surface materials matched to the chemistry: polypropylene where hydrofluoric acid is used, epoxy resin for general chemistry, stainless steel where the work is radioactive, and a wash-down liner and dedicated duct where perchloric acid is involved.
An airflow monitor with visual and audible alarm on every hood, plus the volt-free contact needed to report to the building management system.
Fan, ductwork, discharge stack, make-up air and any exhaust treatment in the same package as the hood. Splitting them across packages is how a laboratory ends up with hoods, a fan and nobody responsible for the fact that they do not work together.
Twelve months of filter or media where a filtered hood is offered, plus the breakthrough test method.
Operator training and a written test schedule handed over at commissioning, naming who tests, how often and against which standard.