To calculate air changes per hour (ACH), divide the supply airflow by the room volume: ACH = supply airflow (m³/h) ÷ room volume (m³), or ACH = CFM × 60 ÷ room volume (ft³) in imperial units. A 144 m³ cleanroom supplied with 4,320 m³/h of filtered air therefore has 30 air changes per hour. In a pharmaceutical cleanroom, that number is only a starting point: the design rate must then pass a heat-load check, a recovery-time check and a pressure balance.
Result in brief: in the worked example below, a 144 m³ Grade C preparation room starts at a rule-of-thumb 20 ACH (2,880 m³/h). The heat load is covered, but the estimated 100:1 recovery time of 19.7 minutes misses a 15-minute design target. Raising the rate to 30 ACH (4,320 m³/h, about 2,543 CFM) brings recovery to 13.2 minutes. With +15 Pa to the corridor, the return airflow is set at about 4,100 m³/h.
The air changes per hour formula
Air changes per hour is the number of times the volume of air in a room is replaced by supply air in one hour. It is a ratio, not a measure of cleanliness: two rooms with the same ACH can perform very differently depending on where the air enters and leaves.
Metric: ACH = Q (m³/h) ÷ V (m³)
Imperial: ACH = Q (CFM) × 60 ÷ V (ft³)
Q = supply airflow delivered through the terminal filters
V = room volume = length × width × clear heightRearranged, the same formula gives the airflow a design needs: Q = ACH × V. Use the clear height below the ceiling grid, not the slab height, and the supply airflow actually delivered by the HEPA filters, not the fan nameplate. For conversions, 1 m³ = 35.315 ft³ and 1 m³/h = 0.5886 CFM.
What the standards actually fix
Neither EU GMP Annex 1 nor ISO 14644 sets a mandatory air change rate for non-unidirectional cleanrooms. The regulations fix outcomes (particle limits, clean-up time, pressure differentials, airflow velocity in Grade A) and leave the air change rate to the designer, who must justify it. The only widely quoted numerical ACH value comes from the US FDA.
| Parameter | Value | Source |
|---|---|---|
| Air change rate, ISO 8 (Class 100,000) supporting rooms | At least 20 ACH "typically acceptable" | US FDA aseptic processing guidance (2004) |
| Clean-up period from "in operation" to "at rest" particle limits | Less than 20 minutes (guidance value) | EU GMP Annex 1 (2022), 4.29 |
| Unidirectional airflow speed at the working position (Grade A) | 0.36–0.54 m/s (guidance value) | EU GMP Annex 1 (2022), 4.30 |
| Pressure difference between adjacent rooms of different grades | At least 10 Pa (guidance value) | EU GMP Annex 1 (2022), section 4 |
| Recovery performance test | Time to reduce a particle concentration 100:1, measured | ISO 14644-3:2019 |
For the full regulatory text, see EU GMP Annex 1 (2022). Annex 1 also states that for non-unidirectional grades, recovery testing should replace velocity testing during qualification, which is why recovery is the check that usually decides the final air change rate. The broader design context is covered in our guide to cleanroom HVAC design.
Worked example: sizing a Grade C preparation room
The example below follows the order a designer would use: volume, trial airflow, heat-load check, recovery check, then pressure balance. The inputs are illustrative but realistic for a Grade C solution preparation room next to a Grade D corridor.
| Input | Value | Note |
|---|---|---|
| Room dimensions | 8.0 m × 6.0 m × 3.0 m clear height | Grade C, non-unidirectional airflow |
| Trial air change rate | 20 ACH | Starting point from the FDA ISO 8 value |
| Sensible heat gains | 4.5 kW | 4 operators, process equipment, lighting |
| Supply-to-room temperature difference | 8 K | Room 20 °C, supply 12 °C |
| Mixing (ventilation) effectiveness ε | 0.7 | Ceiling HEPA supply, low-level returns |
| Design target for 100:1 recovery | 15 minutes | Margin below the Annex 1 clean-up guidance of less than 20 minutes |
| Pressure to adjacent Grade D corridor | +15 Pa | Above the Annex 1 minimum of 10 Pa |
| Door leakage area | 0.02 m² | Single door with about 3 mm perimeter gaps |
-
Step 1: Calculate the room volume
V = 8.0 × 6.0 × 3.0 = 144 m³, or about 5,085 ft³. Deduct large fixed equipment only if it occupies a significant share of the volume; most designers keep the gross volume for a conservative result.
-
Step 2: Derive a trial supply airflow
Q = ACH × V = 20 × 144 = 2,880 m³/h (about 1,695 CFM). This is the airflow the room would need if 20 air changes per hour were enough. Steps 3 and 4 test whether it is.
-
Step 3: Check the airflow against the heat load
The supply air must also remove the sensible heat in the room. The airflow needed for cooling is:
Q_heat (m³/h) = P (W) × 3600 ÷ (ρ × cp × ΔT) = 4,500 × 3600 ÷ (1.2 × 1,005 × 8) ≈ 1,679 m³/hWith ρ = 1.2 kg/m³ and cp = 1,005 J/(kg·K), the cooling airflow of about 1,679 m³/h is below the 2,880 m³/h trial airflow, so cleanliness, not heat, governs this room. In rooms with ovens, autoclaves or dense equipment the result is often the opposite, and the heat load then sets the air change rate.
-
Step 4: Check the recovery time
For a well-mixed room, particle concentration decays exponentially once the source stops. The time to reach a 100:1 reduction, the ratio used in the ISO 14644-3 recovery test, is:
t(100:1) = ln(100) ÷ (ε × ACH) hours = 276.3 ÷ (ε × ACH) minutesAt 20 ACH and ε = 0.7, t = 276.3 ÷ 14 = 19.7 minutes. That fails the 15-minute design target. Solving for the target gives a minimum of 276.3 ÷ (0.7 × 15) = 26.3 ACH. Adding about 10% for filter loading and balancing tolerance, the design rate becomes 30 ACH, for which t = 276.3 ÷ 21 = 13.2 minutes.
-
Step 5: Set the pressure offset and the return airflow
The room is held at +15 Pa to the corridor by supplying more air than is returned. The difference leaks out through the door gaps, estimated with the orifice equation:
Q_leak = Cd × A × √(2 × ΔP ÷ ρ) = 0.61 × 0.02 × √(2 × 15 ÷ 1.2) = 0.061 m³/s ≈ 220 m³/hThe return airflow is therefore 4,320 − 220 = 4,100 m³/h. Add every other leakage path, such as pass-through hatches and service penetrations, and size the airlocks so that the cascade holds when doors open, as described in our article on cleanroom airlock design.
How recovery time falls as the air change rate rises
Recovery time is inversely proportional to the air change rate, so each additional air change buys less improvement than the one before. Above roughly 30–40 ACH, the gain in recovery is small compared with the extra fan energy and cooling it costs.
| Air changes per hour | Recovery, ε = 1.0 (min) | Recovery, ε = 0.7 (min) |
|---|---|---|
| 10 | 27.6 | 39.5 |
| 15 | 18.4 | 26.3 |
| 20 | 13.8 | 19.7 |
| 25 | 11.1 | 15.8 |
| 30 | 9.2 | 13.2 |
| 40 | 6.9 | 9.9 |
| 50 | 5.5 | 7.9 |
| 60 | 4.6 | 6.6 |
Results of the worked example
The final design for the Grade C room is 30 air changes per hour, set by the recovery check rather than by the heat load or the rule of thumb.
| Quantity | Trial (20 ACH) | Final design (30 ACH) |
|---|---|---|
| Supply airflow | 2,880 m³/h (≈ 1,695 CFM) | 4,320 m³/h (≈ 2,543 CFM) |
| Airflow needed for 4.5 kW at ΔT 8 K | ≈ 1,679 m³/h (covered) | ≈ 1,679 m³/h (covered) |
| Estimated 100:1 recovery (ε = 0.7) | 19.7 min (fails 15 min target) | 13.2 min (meets target) |
| Door leakage at +15 Pa | ≈ 220 m³/h | ≈ 220 m³/h |
| Return airflow | ≈ 2,660 m³/h | ≈ 4,100 m³/h |
The room airflows then roll up to the air handling unit: the AHU must deliver the sum of the supply airflows of every room it serves, plus margin for filter loading, and the fresh air share must cover all exhaust and leakage to outside.
Grade A zones: velocity, not air changes
Unidirectional airflow zones are sized by air velocity across the filter face, not by air changes per hour. Annex 1 gives a guidance range of 0.36–0.54 m/s at the working position. For a 2.4 m × 1.2 m filling zone at 0.45 m/s, the airflow is 2.88 m² × 0.45 m/s × 3,600 = about 4,666 m³/h.
Expressed as ACH over the 7.2 m³ volume under a 2.5 m high canopy, that is about 648 air changes per hour, which shows why the metric is meaningless for unidirectional flow. What matters is uniform velocity, the protection of the critical zone in smoke studies and the barrier concept chosen, compared in our article on RABS vs isolators.
Limits: when this calculation is not enough
The ACH formula and the perfect-mixing recovery estimate are design tools, not proof of performance. They are not enough on their own when:
- The airflow pattern is uneven. Short-circuiting between ceiling supply and nearby returns, dead zones behind equipment or high-level returns reduce the real mixing effectiveness well below the assumed value.
- Particle generation is high or localised. Powder handling, open processes or many operators call for a source-based calculation, the approach encouraged by ISO 14644-16:2019 on energy efficiency, rather than a default rate.
- Containment reverses the cascade. Negative-pressure suites for potent compounds or live organisms need the leakage balance worked in the opposite direction.
- Qualification is the final word. Recovery, airflow volume and pressure differential are measured during cleanroom validation; the calculation must leave enough margin for the measured values to pass.
Frequently asked questions
What is the formula for air changes per hour?
Air changes per hour equals the supply airflow divided by the room volume. In metric units, ACH = airflow in m³/h ÷ volume in m³. In imperial units, ACH = CFM × 60 ÷ volume in cubic feet. For example, 4,320 m³/h supplied to a 144 m³ room gives 30 air changes per hour.
How many air changes per hour does a cleanroom need?
No regulation sets a mandatory figure for non-unidirectional cleanrooms. The US FDA's 2004 aseptic processing guidance considers at least 20 air changes per hour typically acceptable for ISO 8 supporting rooms. Higher grades usually need more, and the final rate should be justified by heat load, recovery time and in-operation particle results.
What does 12 air changes per hour mean?
Twelve air changes per hour means that the supply system delivers a volume of air equal to the room volume twelve times each hour, or once every five minutes. It does not mean the air is fully replaced every five minutes: with perfect mixing, reducing a contaminant concentration 100 times would take about 23 minutes.
How do you calculate cleanroom recovery time from the air change rate?
For a well-mixed room, the time to reduce particle concentration 100:1 is ln(100) divided by the product of the air change rate and the mixing effectiveness. That equals 276.3 ÷ (ε × ACH) minutes. At 30 ACH and an effectiveness of 0.7, the estimate is 13.2 minutes. ISO 14644-3 measures the real value.
Why are air changes per hour not used for Grade A zones?
Grade A zones use unidirectional airflow, where clean air sweeps the critical area in parallel streams rather than diluting contamination. Their performance depends on air velocity and uniformity, so EU GMP Annex 1 gives a guidance range of 0.36–0.54 m/s at the working position instead of an air change rate.
How PharmaTwin helps
PharmaTwin is a pharmaceutical facility simulator that runs these calculations room by room as the layout is drawn. Each room's grade, volume and heat gains feed air change rates, supply and return airflows and the pressure cascade, so the effect of changing one room on the air handling unit and its neighbours is visible before the design is frozen. The results still need confirmation by detailed HVAC design and qualification.