Cleanroom HVAC design: loads, airflow, filtration, pressure, and controls

What this covers

• How the ISO class, the process, and the governing regulation each set a different part of the design.
• Why cleanroom airflow comes from three competing calculations and the largest one governs.
• Where fan heat, process heat, and makeup air sit in the load calculation.
• How filtration, pressurization, humidity, controls, and certification connect to one basis of design.

What changes the result

• A regulated minimum air change rate that exceeds what particle control or cooling alone would need.
• Fan heat from recirculation fans or fan filter units, which grows with airflow and lands directly in the room.
• Door gaps and wall construction, which set how much airflow offset each room needs to hold its pressure.
• A humidity target low enough to push the supply dew point below what a chilled water coil can reach.

Start from the process requirement, then the ISO class

An ISO class is a particle limit and nothing more. ISO 14644-1 defines each class by the maximum concentration of particles at or above stated sizes, so ISO 7 allows 352,000 particles of 0.5 micrometers and larger per cubic meter and ISO 8 allows ten times that. The standard does not say how much air to supply, what temperature to hold, or which way air should move between rooms.

Those decisions come from the process and from the regulation that governs it. A compounding pharmacy works under USP 797, and under USP 800 as well where hazardous drugs are handled. A sterile drug manufacturer answers to the FDA aseptic processing guidance. An electronics assembly room may answer to nothing but its own product yield. ISO 14644-4, the design and construction part of the series, is built around recording these requirements before design starts, with checklists of the performance parameters customer and designer have to agree.

State the class for both occupancy states. A room classified at rest, with equipment installed and nobody inside, and the same room in operation with a full shift working can differ by an order of magnitude in particle count, and the in-operation limit is the one the air system has to hold.

Airflow comes from three calculations, and the largest governs

Cleanroom supply airflow is the largest of three separate numbers. The first is the airflow that dilutes the particles people and processes release down to the class limit, with margin. The second is any minimum air change rate a regulation sets, such as the 30 air changes per hour USP 797 requires in an ISO 7 buffer room. The third is the airflow needed to carry away the room's sensible heat at a supply temperature the coil can deliver.

Which of the three wins changes from room to room. In the worked example on the air change rate page, particle dilution alone asks for about 15 air changes per hour in an ISO 7 room and under 2 in an ISO 8 room, so the regulatory minimum governs both. In a small test room full of electronics racks, cooling airflow can pass 150 air changes per hour and make the particle calculation irrelevant. Copying a figure from an air change table skips the two calculations that can decide the room.

Unidirectional zones for ISO 5 work are sized by velocity rather than by air changes. The FDA aseptic processing guidance cites 90 feet per minute (0.45 meters per second) with a range of plus or minus 20 percent at the work site, and the airflow follows from that velocity multiplied by the filtered area.

Loads the air system adds to the room it serves

A cleanroom load calculation carries every item an office calculation does, plus two that grow with the airflow itself. Recirculation fans and fan filter units turn nearly all of their electrical input into heat in the air stream, and at high air change rates that fan heat can reach a third of the room's sensible load. Makeup air, supplied continuously to replace exhaust and leakage, carries the outdoor moisture, which is why the latent side of the calculation belongs to the makeup air unit rather than the recirculation coil.

Enter process equipment at measured or data sheet operating power, not nameplate. A nameplate rating describes the largest current a circuit must carry, and adding nameplates overstates the load, which inflates airflow, fan heat, and cooling plant together. Where a tool rejects its own heat through process exhaust, that share leaves the room and comes off the sensible load. The heat load calculation page works through a 1,200 square foot ISO 7 room with every line entered.

Cleanroom walls that face other conditioned space keep envelope gain small, but the temperature difference runs the opposite way from comfort cooling. A room held at 68°F (20°C) inside a 75°F (24°C) building gains heat through its walls in every season.

Choosing between a ducted AHU and fan filter units

There are two basic ways to move recirculation air. A central air handling unit (AHU) pushes air through ductwork to terminal HEPA filter housings in the ceiling, with its coils in the same cabinet. A fan filter unit ceiling puts a small fan on every filter, draws return air from a plenum above the ceiling, and leaves cooling and dehumidification to a separate recirculation coil and a makeup air unit.

Neither is better as a category. A fan filter unit ceiling spreads fan duty across dozens of units, so one failure removes a small share of the airflow, and its return plenum sits below room pressure, so ceiling leakage runs out of the room. A ducted system keeps motors out of the ceiling, centralizes maintenance, and lets one large fan be selected at its best efficiency point. The fan filter unit versus air handling unit comparison sets out the trade on fan power, heat, redundancy, noise, and control.

Filtration from the outdoor intake to the terminal filter

The terminal HEPA filter sets the particle quality of the supply air, and the filters upstream decide how long it lasts. HEPA filters in United States practice are rated at 99.97 percent efficiency for 0.3 micrometer particles, the figure the CDC cites for protective environment rooms. Prefilters and intermediate filters on the makeup air and recirculation paths catch the coarse load that would otherwise plug HEPA media early.

Every filter enters the static pressure budget twice, once clean and once loaded. Fans are selected at the loaded condition and controlled back to design airflow while the filters are new, which the AHU external static pressure calculation shows step by step.

An installed HEPA filter needs a leak test once it is in place, because the media is only one path a particle can take. The FDA aseptic processing guidance asks for leak testing at installation to find breaches around gaskets, through frames, and across the media, and gives twice a year as an example interval for an aseptic processing room.

Pressure differentials and the airflow offset behind them

Pressurization keeps air moving from cleaner spaces toward less clean ones through every crack while the doors are closed. The published figures are close without being identical: the FDA aseptic guidance asks for at least 10 to 15 pascals between adjacent rooms of different classification, USP 797 requires at least 0.020 inches of water column (about 5 pascals) between compounding rooms, and ISO 14644-4 gives a range of 5 to 20 pascals.

Pressure is the result; airflow offset is what the design controls. Each room's offset equals the leakage flowing out through its door gaps and penetrations at the chosen pressure difference, less any transfer air flowing in from a higher-pressure neighbor. Sizing that offset from leakage area rather than as a flat share of supply is the subject of the pressure cascade design page.

Containment reverses the direction. A hazardous drug buffer room, an airborne infection isolation room, and the room around a hot cell are held negative to their surroundings, so exhaust exceeds supply. The negative pressure room design page covers the offset arithmetic, and the USP 800 ventilation page covers the regulated values for pharmacies.

Temperature and humidity as two separate control loops

Room temperature is a sensible load problem and room humidity is a moisture balance, and asking one coil to satisfy both setpoints is where cleanroom humidity control goes wrong. The workable arrangement is a makeup air unit that dries outdoor air to a dew point low enough to absorb the room's latent load, and a recirculation coil that runs dry and removes only sensible heat.

The dew point target follows from the room condition and the occupant moisture. At 68°F (20°C) and 40 percent relative humidity, the example on the humidity control page needs makeup air at a dew point near 37°F (3°C), which is colder than a 42°F (5.6°C) chilled water supply, so no coil on that water can produce it. USP 797 sets 20°C or cooler and 60 percent relative humidity or below for compounding rooms, and the USP 797 HVAC requirements page sets those values beside the air change and pressure rules.

Controls, monitoring, and planned failure

A cleanroom control system does more than hold setpoints. It monitors pressure differentials continuously, alarms when they drift, and follows written responses when a supply fan stops, an exhaust fan trips, a fire alarm shuts air handlers down, or power returns and fans restart in the wrong order. A containment room whose supply fan restarts before its exhaust can push air outward for as long as the gap lasts, which is the event a startup sequence exists to prevent.

Those responses belong in the design, not in an operations document written after handover. The failure sequences page writes them out step by step for four events, and the redundancy page covers how much standby fan and cooling capacity a room needs so that a single failure does not end production.

Commissioning and certification close the design

The design is finished only when its numbers are measured in the built room. ISO 14644-3 describes the test methods: airflow volume or velocity, pressure difference, airflow direction and visualization, installed filter leakage, recovery time, and particle counts for classification. Commissioning under ASHRAE Guideline 0 runs alongside certification and checks each control response against the written sequence of operations, including the failure responses.

Record acceptance criteria before testing starts. ISO 14644-3 treats the recovery test as optional, with acceptance agreed between customer and supplier, so a recovery time nobody wrote down has nothing to pass against.

A basis of design written before equipment selection

The basis of design is the document every later calculation refers back to. For each room it states the ISO class at rest and in operation, the governing regulation, the design temperature and humidity with tolerance, the pressure relationship to each neighbor, the occupancy and process loads with their source, the airflow and which calculation set it, the filtration stages, the redundancy level, the failure responses, and the tests that will prove them.

Writing the source beside each number keeps the design honest when something changes. If a tool vendor revises process heat, the basis of design shows at once whether airflow was set by cooling, in which case airflow moves, or by an air change minimum, in which case only the coil does. The table below lists each item, what decides it, and where its calculation lives.

Questions people ask about this