Fresh Air Systems vs. HEPA Recirculation vs. Vertical Flow Benches: Comparing Air Control in Pharmaceutical Clean Room Design

I'm the office administrator at a 120-person company, and I handle facilities and lab purchasing—about $380,000 a year across 14 vendors. I report to operations and to finance, which means when something fails a client audit I hear about it, and when it overshoots the budget I hear about that too.

When I took over purchasing in 2021, our air setup 'worked' in the sense that nobody had complained yet. Then a client audit happened, and I spent most of that year learning about particle counts and pressure relationships the hard way. Since then I've been through three air-control decisions in our clean room spaces, and every one of them looked simple going in. Each turned out to be a comparison I hadn't framed properly.

This is the version I wish someone had handed me.

The three decisions, and what I actually compare them on

If you're involved in pharmaceutical clean room design—new build or retrofit—most of the air argument collapses into three pairs:

Decision	Option A	Option B	What it really decides
Air path	Once-through fresh air system with exhaust	Recirculating with HEPA filtration	Where contamination ends up; energy cost
Room-level particle control	Standalone purifier HEPA filter units	Ducted high efficiency particulate air filter	Room average vs. critical zone; what you can document
Bench-level control	Vertical flow bench	Horizontal flow bench	Airflow obstruction and operator exposure

I compare on four things: capital cost, running cost (energy plus filter replacement), control at the critical zone rather than room average, and audit burden. That last one never shows up in a quote, and it's the one that has cost me the most time.

Decision 1: once-through fresh air vs. recirculating HEPA air

A once-through fresh air system pulls outdoor air in, conditions and filters it, pushes it through the space, and a fresh air exhaust fan pulls it back out of the building. None of it returns. A recirculating system sends most of the room air back through a high efficiency particulate air filter and returns it to the room, with a smaller outdoor-air fraction for makeup.

What most people assume this decision is about is cleanliness. It mostly isn't. Both designs can hold ISO Class 7 if they're built and balanced properly—ISO 14644-1:2015 is the classification standard, and ISO 14644-3 covers the test methods. What the choice actually decides is where airborne contamination ends up, and how much you pay to condition air.

Once-through means contamination leaves the building. If you're handling potent compounds, or anything you don't want accumulating, this isn't really a choice you get to make; it's the only defensible answer. I've been offered recirculating 'with good filtration' as a cost-saving alternative more than once, and the filters may well be good. The problem is what builds up in return plenums, ductwork and room surfaces over five years, and the fact that you can't easily verify any of it.

Recirculating means most of your air is already at temperature and humidity, so you're conditioning a makeup fraction instead of everything. In most climates that's the cheaper system to run, sometimes dramatically so.

The part that surprised me: the exhaust fan is a rounding error in the budget. The money goes to conditioning the makeup air. A $2,000 fan is nothing next to the cost of heating and cooling outdoor air eight hours a day, which is why the obvious efficiency move is usually to reduce the outdoor-air fraction rather than touch the fan at all.

My conclusion: if anything hazardous, potent or odorous is in the room, once-through—and document the reasoning in your contamination control strategy. EU GMP Annex 1 (revised, in force since 25 August 2023) requires a documented contamination control strategy for sterile products, and 21 CFR 211.42(c)(10) sets out design expectations for aseptic processing areas. If it's particulate control in a classified space, recirculating with properly tested HEPA filtration is usually the cheaper lifetime option.

Honestly, I'm not sure why filter lead times swing the way they do—six weeks one quarter, sixteen the next, same specification, same vendor. My best guess is raw media supply and how much of the production line gets committed to a handful of large projects. If someone knows the real mechanism, I'd like to hear it.

Decision 2: standalone purifier HEPA filters vs. ducted high efficiency particulate air filters

This is where I got burned, so I'll be specific about it.

A purifier HEPA filter unit is a box with a fan, a pre-filter and a HEPA filter, on casters. A ducted high efficiency particulate air filter is a component of the room's air handling system, sitting in a housing with a gasketed frame and a test port. The filter media inside them can come from the same factory, which is exactly why the comparison is misleading.

From the outside, a room with four purifiers running looks controlled. The reality is that you're reducing the airborne particle count in the middle of the room, not at the fill zone, not at the bench, not wherever your product is actually exposed.

What the portable units can't do: hold a pressure relationship to the corridor, sustain a defined air change rate, or accept a validated installation leak test. The question isn't whether a purifier can lower particle counts. It's whether the number it lowers is the number you're being judged on.

Installed filter leakage testing—the scan of filter face, frame and gasket with an aerosol—is defined in ISO 14644-3, and it's the test an auditor asks for. Why does that matter? Because a certificate for the filter in its box tells you nothing about the seal in your ceiling. For the filter itself, IEST-RP-CC001 defines HEPA as at least 99.97% removal at 0.3 µm, and EN 1822 grades H13 and H14 by efficiency at the most penetrating particle size. Ask for the individual filter test certificate, not the datasheet.

What I did: we had a particle excursion in a lab. I bought two portable units with H13 filters—about $900 each, if I remember right, though I might be misremembering the exact figure—and the room-average counts came down nicely. The excursion stayed open anyway, because settle plates near the bench hadn't changed. Root cause turned out to be a door gasket and a pressure imbalance, which two purifiers fixed precisely not at all. Worse, they made the monitoring look better, and that's a worse outcome than failing loudly.

Conclusion: purifiers for offices, unclassified labs, general air quality, and as a supplement. Ducted HEPA with documented leak testing for anything classified or product-contact. And if you have an excursion, measure pressure differentials and run airflow visualization before you buy anything at all.

Decision 3: vertical flow bench vs. horizontal flow bench

First, the thing that gets lost: both of these are clean benches, and clean benches protect the product, not the person. If there's any hazard in what you're working with, neither is the right answer, and what you want is a biosafety cabinet tested to NSF/ANSI 49 or EN 12469. People treat vertical flow as a synonym for safer. It isn't.

A horizontal flow bench pushes HEPA-filtered air from the back of the hood across the work surface toward the operator. Product protection is good, because nothing stands between the filter and the work. Anything you aerosolize at the bench travels toward whoever is sitting there.

A vertical flow bench delivers air downward from the top through the filter and sweeps it down over the work surface. The geometry keeps the operator's face out of a direct airstream, which is a real practical difference, and it's part of why vertical benches get specified so often. But the same geometry has a weakness that doesn't show up on the spec sheet: vertical flow is easy to obstruct.

That's the counter-intuitive conclusion I landed on. Put a tall rack of samples under a vertical flow bench and you've built a baffle. The airstream splits around it and creates a wake downstream, and the critical zone behind that rack isn't the clean zone you think you paid for. A horizontal bench on the same task keeps a clear path across the work surface. Vertical is not automatically better—it depends entirely on what you're putting on the bench.

I found this the expensive way. I specified a four-foot vertical flow bench for a staining workflow without running a smoke test first. Airflow visualization took about twenty minutes with a fog generator. The flow broke around the tallest rack every time, in every position we tried. I still kick myself for that one. If I'd asked for a visualization test before signing the order, I'd have spent the same money on a horizontal unit and saved a year of arguing about whether the samples were getting contaminated or weren't.

That said—if the operator is working with anything hazardous, the answer isn't horizontal, it's a cabinet. Don't let an airflow discussion talk you into the wrong class of equipment.

So which one do you pick

Three scenarios, three different answers.

New build with product-contact or potent materials

Once-through fresh air with dedicated exhaust, ducted HEPA, and a biosafety cabinet at the bench. Insist on the room balance report and the installed filter leak test at handover, not three weeks later. The commissioning documentation is as much the deliverable as the room is.

Existing classified room with failing counts or failed settle plates

Don't buy purifiers. Measure first—pressure differentials between rooms, airflow visualization, door gaskets, filter leak test. In our case, two of the three excursions we've had since 2021 came down to pressure relationships rather than filtration. The purifier reflex is expensive and it hides the problem.

Bench work in an unclassified lab with bulky, non-hazardous items

Horizontal flow bench, almost every time. The airflow path stays clear, tall items don't matter, and you save the vertical premium. Run a smoke test after installation anyway, so you know which positions to avoid.

Why I stopped treating this as a pure purchasing decision

Clean rooms are visible. Clients walk through them on tours. Auditors photograph them. A visitor sees filter housings with no test stickers, a portable unit in the corner with its cord taped down, and forms an opinion about your whole operation in about ten seconds. That opinion isn't really about air.

I'm not saying the cheap option is wrong. Some of our spaces are unclassified, and in those rooms the portable purifier was the correct purchase. What I mean is that on the equipment visitors, auditors and clients actually see, the gap between the documented version and the undocumented version is smaller than the cost of explaining the undocumented one later—which is to say, the difference between it works and it works and here is the certificate tends to shrink every year you own the room.

The best part of finally getting our air documentation in order: I stopped getting audit questions forwarded to me at 4pm on a Friday. There's something satisfying about handing over a binder with filter certificates, balance reports and visualization footage, and watching the questioning stop.

We run maybe 40 filters a year across two buildings—35, I'd have to check the log—and I now know the certificate number for most of them. That's roughly the level of detail this stuff rewards.