Cleanrooms — grades, particle counts and the air that holds them

A cleanroom is not a clean room. It is a room whose airborne particle concentration is specified, achieved by continuous air supply, and measured — and the whole discipline follows from the fact that the contamination it controls is mostly generated inside it, by the people working there.

What the class actually counts

ISO 14644-1 sets a maximum airborne particle concentration, in particles per cubic metre, at one or more stated particle sizes. That is all a class number is.

At the 0.5 µm size the limits step by a factor of ten per class, which makes the three classes that pharmaceutical work actually uses easy to hold in your head.

Class	Particles ≥ 0.5 µm per m³
ISO 5	3 520
ISO 7	352 000
ISO 8	3 520 000

Three and a half thousand particles in a cubic metre sounds like a lot until you realise ordinary room air runs in the tens of millions.

At rest and in operation

A classification is meaningless without the state it was taken in, and ISO 14644-1 recognises two.

At rest — the installation complete, air handling running, equipment installed and operating, and nobody in the room.

In operation — the room running as it does in production, with the process working and the normal number of operators present doing normal work.

The difference between the two numbers is the contribution of the people and the process. It is usually large.

Grades A to D

EU GMP Annex 1 does not use ISO class numbers as its primary language. It defines grades A, B, C and D, and each grade carries a particle limit for both states plus viable limits and a set of expectations about how the work is done.

The mapping is close to the ISO classes but not interchangeable, because Annex 1 also specifies the 5 µm size and applies the limits differently between states.

• Grade A — the critical zone. Open product, open containers, and any surface that will contact them. Unidirectional airflow, and the same stringent limit whether at rest or in operation, because there is no acceptable excursion over exposed product.
• Grade B — the background room immediately around Grade A for aseptic processing. Gowned, controlled, and limited in headcount.
• Grade C and Grade D — supporting areas, where the work is less critical and the in-operation allowance is correspondingly wider.

The design rule that follows is one grade step at a time. You do not open a door from Grade D into Grade A; you pass through airlocks that step the grade, and each step has its own gowning.

The pressure cascade

No room is sealed. Doors have gaps, cables and pipes have penetrations, hatches open. Air moves through all of them, and the only thing deciding which way is the pressure difference.

So classified suites are held in a cascade — each space positive with respect to the dirtier space next to it, typically by a differential in the region of ten to fifteen pascals between adjacent grades. Leakage then always runs outward, from clean towards less clean.

How the air is actually delivered

Supply air passes a HEPA filter at the point of entry to the room. Filtering upstream in the plant room would leave the whole duct run downstream as a source, so the final filter sits in the ceiling terminal.

Most of the air is recirculated. Conditioning outside air from ambient to cleanroom temperature and humidity is expensive, so only enough fresh air is brought in to serve the people and hold the cascade; the rest returns through low level grilles and goes round again.

The air change rate is the designer’s main lever. Dilution is what keeps the count down between filter and return, so supporting grades are designed to a rate — a figure the designer chooses to achieve recovery and classification, not a number any standard hands you.

Grade A is different. There, the air is not diluting, it is sweeping: unidirectional flow passing over the critical zone and carrying contamination away in one direction. That is specified as a velocity at the working position, within a range, because too little fails to sweep and too much generates turbulence that entrains room air into the zone it was meant to protect.

Filter integrity is not filter efficiency

A HEPA filter is rated for the efficiency of its medium. Installed, what matters is whether there is any path past it — a pinhole, a pinched gasket, a poor frame seal.

So the installed filter is leak tested by scanning the downstream face with a challenge aerosol, to the methods in ISO 14644-3. A filter can be fully efficient and still fail this, and the failure is the one that matters, because an unfiltered bypass puts whatever the duct contains directly into the room.

What gets monitored, and why continuously

ISO 14644-2 sets the requirements for demonstrating continued compliance, and Annex 1 expects a contamination control strategy rather than a schedule of tests. In practice a classified suite carries:

• Non-viable particle counts — continuously in Grade A and the Grade B around it, periodically elsewhere, with alert and action levels set from the room’s own history rather than from the class limit.
• Viable monitoring — settle plates, contact plates, active air and operator gloves and gowns, trended over time.
• Pressure differentials, temperature and humidity — continuously, alarmed.
• Recovery time — how long the room takes to return to its classified state after a disturbance, which is the test that tells you the air change rate is still doing its job.

What to take away

• An ISO class is a particle concentration at a stated size, and nothing more — not organisms, not air changes, not pressure.
• Always state the condition. At rest and in operation are different rooms commercially and technically, and confusing them at tender is a classic project failure.
• Grade A is for exposed product, with Grade B around it for aseptic work. Step one grade at a time through airlocks.
• The pressure cascade is the control of last resort, which is why a reversed differential is a contamination event rather than a defect to schedule.
• The final filter belongs at the room, and most of the air is recirculated because conditioning outside air is the expensive part.
• Grade A airflow sweeps rather than dilutes, and its velocity is a range because too much is as wrong as too little.
• Integrity testing is not efficiency testing. A perfect filter with a pinched seal fails the test that matters.
• Viable and non-viable are different measurements. When the particles are fine and the organisms are not, look at the people.