Air change rates and airflow in cleanrooms

Air change rate is the most important number in cleanroom HVAC design. It sets the size of the fans, air handlers, ductwork and filters, and it drives most of the energy bill. It is also one of the least standardized numbers. Neither ISO 14644 nor EU GMP Annex 1 prescribes air change rates for non-unidirectional cleanrooms, so designers rely on industry guidance, experience and calculation.

This guide explains where typical air change rates come from, how to calculate airflow more rigorously, and what to consider before choosing a number.

Two kinds of airflow design

Cleanrooms use two basic airflow approaches:

• Unidirectional airflow areas are designed by velocity. Air moves in parallel streams across the whole area, and the airflow rate is the velocity times the area. ISO 5 and cleaner zones usually use this approach.
• Non-unidirectional airflow rooms are designed by air change rate. Clean air enters through ceiling filters, mixes with room air and dilutes particles, which leave through returns.

The two are compared in unidirectional vs non-unidirectional airflow. This guide focuses mainly on non-unidirectional rooms, where the air change rate is a choice.

Typical ranges and where they come from

Commonly cited ranges trace back to US industry practice, including IEST recommended practice IEST-RP-CC012, and the era of Federal Standard 209E. They are useful as a sanity check but are wide because real contamination loads vary widely.

These figures are for illustration only. Regulated industries have their own minimums: for example, USP 797 requires at least 30 ACH of HEPA-filtered air in ISO 7 buffer rooms and at least 20 ACH in ISO 8 ante-rooms. Hospital and pharmacy rooms in Canada may also be subject to CSA Z317.2 and provincial pharmacy regulator requirements.

Many existing cleanrooms operate well below the upper ends of these ranges and still meet their class, because actual particle generation is lower than the design assumed.

Calculating airflow from first principles

A more rigorous approach estimates how many particles are released in the room and how much clean air is needed to dilute them.

Steady-state dilution

For a well-mixed room supplied with HEPA-filtered air that contains essentially no particles, the steady-state concentration is approximately:

C = G / Q

where C is the room concentration (particles per m3), G is the particle generation rate (particles per second) and Q is the supply airflow (m3/s). Real rooms are not perfectly mixed, so a ventilation effectiveness factor is often applied.

Particle generation

People are usually the dominant source. Emission rates depend heavily on gowning type, gown condition and activity level. Published research reports emission rates that differ by an order of magnitude or more between garment systems and activities. Equipment, materials and processes add more. A design should use emission data appropriate to the facility's gowning and activity, and test assumptions where possible.

A worked example

Consider an ISO 7 room of 50 m2 with a 3 m ceiling (150 m3 volume) and four operators in cleanroom garments.

1. Assume each operator generates 10,000 particles per second at 0.5 um and larger, an illustrative assumption for good cleanroom garments and light activity. Total G = 40,000 particles per second.
2. The ISO 7 limit at 0.5 um is 352,000 particles per m3. Design to a safety margin, for example half the limit: target C = 176,000 per m3.
3. Required airflow Q = G / C = 40,000 / 176,000 = 0.23 m3/s.
4. Air change rate = 0.23 x 3,600 / 150 = about 5.5 ACH.

That result is far below the typical range for ISO 7. It shows how strongly the answer depends on emission assumptions and margin. With higher emission rates, poor mixing or heavier activity, the requirement rises quickly. Other factors also push airflow up, as described below.

Factors besides dilution

Recovery time

The time a room takes to recover after a contamination event often governs. For a well-mixed room, the theoretical time for a 100:1 reduction in concentration is:

t (hours) = ln(100) / N = 4.6 / N

where N is the air change rate per hour. At 20 ACH, a perfect room recovers 100:1 in about 14 minutes; at 60 ACH, about 5 minutes. Real rooms recover more slowly than this ideal. Recovery testing is described in cleanroom recovery and contamination control testing.

Heat removal

Supply air must also remove sensible heat from process equipment, lighting and fans. In rooms with high equipment loads, cooling airflow may exceed the airflow needed for cleanliness.

Pressurization and exhaust

Rooms with large process exhaust or high leakage need enough supply to replace exhaust and maintain pressure.

Diffuser and return layout

The same airflow can perform very differently depending on where it enters and leaves. Filters distributed evenly across the ceiling, with low-level returns spread around the perimeter, produce better particle removal than a few filters clustered in one area with a single return. Smoke studies and particle mapping reveal dead zones.

Unidirectional zones

In unidirectional areas, velocity at the working position is the design parameter. EU GMP Annex 1 gives a guidance range of 0.36 to 0.54 m/s for Grade A. Many semiconductor and electronics cleanrooms operate lower velocities in ISO 5 bays, sometimes around 0.3 m/s, when testing shows the class is met. Velocity uniformity across the filter face matters as much as the average.

Because airflow equals velocity times area, a large unidirectional ceiling is extremely expensive to operate. Limiting unidirectional flow to the critical zone, using hoods, minienvironments or isolators, is usually the most important energy decision.

Avoid automatic overdesign

Owners and designers sometimes choose the top of the typical range "to be safe". Because fan power rises sharply with airflow, this can multiply operating costs over the life of the facility. A better approach is to:

1. Calculate airflow from expected particle generation, heat load and recovery requirements.
2. Compare the result to industry ranges and regulatory minimums.
3. Design the system with variable speed fans so airflow can be adjusted after testing.
4. Verify cleanliness at the design airflow, then consider reducing it if results show a large margin.
5. Use reduced airflow during unoccupied periods where the risk assessment allows.

ISO 14644-16 supports this approach as part of energy-efficient cleanroom design, as covered in cleanroom energy efficiency.

Key takeaways

• Neither ISO 14644 nor EU GMP Annex 1 prescribes air change rates for non-unidirectional cleanrooms.
• Typical ranges by ISO class are wide and best used as a sanity check, alongside regulatory minimums such as USP 797.
• Airflow can be calculated from particle generation, target concentration, recovery time and heat removal.
• Diffuser and return layout affect performance as much as the air change rate itself.
• Design with adjustable airflow and verify by testing rather than defaulting to the top of the range.