A pharmaceutical plant asks three different things of its air: low cross-draft velocity at fume hoods and biosafety cabinets in the lab, particle control and full-room flushing with no stagnant zones in production, and temperatures held within setpoint limits throughout the storage volume in the warehouse. What the three share is a critical point that supply airflow alone does not settle, which is how that air is distributed. This article walks through each zone against the international guidance and explains where a fabric duct answers, and under what conditions.
Airflow under regulatory scrutiny
In a 2025 warning letter, the U.S. Food and Drug Administration (FDA) described smoke studies on an ISO 5 (Grade A) aseptic filling line in which the airflow was observed to swirl and stagnate. The baseline requirement sits in 21 CFR 211.46(b): equipment for adequate control over air pressure, micro-organisms, dust, humidity and temperature shall be provided when appropriate for the manufacture, processing, packing or holding of a drug product.
Zone 1: the laboratory
Research, development and innovation (R&D&I), quality control, microbiology and stability testing happen here. The World Health Organization (WHO) states that laboratories and production areas should have separate air supply, and that microbiological laboratories need separate air-handling units (TRS 986, Annex 2).
The CDC/NIH biosafety manual (BMBL, 6th ed.) describes the air curtain of a biological safety cabinet as quite fragile, at a nominal velocity of about 1 mph (0.45 m/s). In tracer gas tests, Dunn and colleagues (2014) found leakage much more prominent on the side of a fume hood closest to the room supply diffuser. For its own buildings, the National Institutes of Health (NIH) caps cross-draft velocity at 50 fpm (about 0.25 m/s), measured 2 ft (0.6 m) from the hood or cabinet face and 5 ft (1.5 m) above the floor. That is an owner requirement, not a code limit.
Zone 2: production
Weighing, granulation, compression and coating release dust. The WHO says it should be removed at source wherever possible, and that the HVAC system should not normally serve as the primary mechanism of dust control (TRS 1010, Annex 8).
For sterile products, EU GMP Annex 1 (2022) defines four grades: A for high-risk operations such as aseptic filling, B as its background, and C and D for less critical stages. Clause 4.14 calls for at least 0.04 in. w.g. (10 Pa) between adjacent rooms of different grades, as a guidance value.
More air changes do not guarantee a clean room. Writing about non-unidirectional airflow rooms, Whyte and colleagues (2018) note that 20 air changes per hour does not mean the room is free of particles in 3 minutes. The WHO asks for supply and extract grilles located to flush the room effectively and prevent zones of stagnant air.
### Dust reaches the ductwork
A person in cleanroom garments disperses on average 1,020,000 particles of 0.5 µm and larger per minute (Whyte and Hejab, 2007). In DUCTECOL’s field experience, that dust, together with process dust and infiltration, settles on top of exposed ducts. Inside supply ducts, Pasanen and colleagues (1997) cite estimated accumulation rates of 0.7 to 1.8 g/m² per year, measured in office buildings rather than pharmaceutical plants. In a properly maintained system that dust matters little; once surfaces get wet, it offers an opportunity for microbial growth (Pasanen and colleagues, 1992).
### Conventional ductwork falls short
The NIH does not permit internal duct liner at any of its facilities, because lined ductwork is prone to bacterial growth that is impossible to sanitize once contaminated. For homes, the U.S. Environmental Protection Agency (EPA) warns that once fiberglass liner is contaminated with mold, cleaning is not sufficient to prevent regrowth. With pre-insulated panel duct, DUCTECOL’s field crews find elbows and transitions with uncoated internal edges and sealing tape that lifts. Galvanized or stainless sheet metal can be cleaned, but in that same experience, in pharmaceutical plants it almost never is. When they are cleaned from inside the room, a small study at a university hospital recorded a particle spike that took about an hour to return to baseline (De Waegemaeker and colleagues, 2025). None of these ducts can be taken down and washed outside the room.
The WHO principle is that no HVAC material should become a source of contamination. Annex 1 goes further for sterile manufacturing: ducts should not be present in cleanrooms (cl. 6.6), unless an alternative approach is justified (cl. 2.2).
Zone 3: the warehouse
As a minimum, the WHO expects a system able to continuously maintain air temperatures within setpoint limits throughout the validated storage volume (TRS 961, Annex 9); for controlled room temperature it recommends 59 to 77 degrees Fahrenheit (15 to 25 degrees Celsius) (TRS 1025, Annex 7, recommended value). High-bay pallet racking is particularly susceptible to stratification, with warm air collecting over the top levels where product also sits. Qualification should define zones not to be used for storage, for instance near cooling coils or cold air streams. Temperature mapping on a grid and at several heights exposes the problem, and the WHO lists altering the air distribution or retrofitting new distribution equipment as possible remedies.
Fabric duct, zone by zone
### Lab: jets that decay early
A fabric duct discharges through many small outlets. For the jet zone where velocity decays quickly, the ASHRAE Handbook—Fundamentals (Ch. 20, Space Air Diffusion) gives:
V_x = K_c3 · V_o · √A_o / X
where V_x is the centerline velocity at distance X, K_c3 the outlet’s centerline velocity constant, V_o the discharge velocity and A_o the outlet area. With the same constant, discharge velocity and distance, a smaller outlet delivers a slower jet at the hood face.
### Hygiene: cleaning leaves the room
The fabric duct is the only one that is taken down, washed and verified outside the room. DUCTECOL follows a traceable sequence: removal, washing, disinfection, chemical sterilization when required, inspection and a report for each unit. The fabric carries a laboratory-tested antimicrobial and antifungal treatment. Exposed, it still collects dust on top, but it goes out to be washed.
### Production: filtered air across the whole room
For non-sterile products, the WHO allows HEPA filters in the air-handling unit (TRS 1019, Annex 2); in that layout, the fabric duct spreads filtered air along the full length of the room, which helps flushing. It is suitable for all types and classes of rooms, with validation by DUCTECOL engineers in each case; in sterile plants, that validation supports the alternative approach under cl. 2.2.
### Warehouse: one duct per aisle
A duct over each aisle distributes airflow along its full length, without localized cold drafts on product. Mapping before and after is the proof.
CFD-based design
DUCTECOL models every project with computational fluid dynamics (CFD) to anticipate airflow patterns before fabrication, backed by high-tech manufacturing and international shipping worldwide. The 2022 revision of ISO 14644-4 states in its informative introduction that it focuses on ventilation effectiveness through control of airflow patterns and clean-up recovery rates. Compliance is still demonstrated by measurement.
Frequently asked questions
Can fabric duct be installed inside a cleanroom in a sterile manufacturing plant?
Yes, as a justified alternative approach validated case by case. EU GMP Annex 1 states that ducts should not be present in cleanrooms and that, if unavoidable, they should be installed without recesses, unsealed openings or surfaces that are difficult to clean (cl. 6.6). It accepts alternative approaches supported by rationale, risk assessment and mitigation that meet the intent of the Annex (cl. 2.2). DUCTECOL considers fabric duct suitable for all types and classes of rooms, always with validation by its engineers in each case. There is no blanket approval, and in sterile plants that validation supports the manufacturer’s rationale, whose contamination control strategy should assess the effectiveness of all controls, design controls included (Annex 1, cl. 2.3).
What cross-draft velocity is acceptable at a fume hood?
There is no universal figure. The NIH Design Requirements Manual sets a maximum of 50 fpm (15.2 m/min, about 0.25 m/s) at its own facilities, measured 2 ft (0.6 m) from the hood or cabinet face and 5 ft (1.5 m) above the finished floor. It is a useful reference for a specification, but each project sets its own criterion and verifies it with field measurement.
How do you prove that new air distribution fixed warehouse stratification?
With temperature mapping before and after. WHO Supplement 8 (2015) suggests data loggers on a grid every 16 to 33 ft (5 to 10 m) and, where the ceiling is higher than about 12 ft (3.6 m), vertical arrays; for a 20 ft (6 m) store, for example, at roughly 1, 6, 12 and 18 ft (0.3, 1.8, 3.6 and 5.4 m). In ambient warehouses a study typically runs at least seven consecutive days, and where seasons matter at least two studies may be needed, one in the warmest season and one in the coldest.
Does taking the duct down for washing trigger requalification?
Not necessarily; under the guidance, the plant’s change management decides. For non-sterile facilities, the WHO says any HVAC change should go through change control, that the extent of requalification should depend on the scope and impact of the change, and that maintenance should normally be scheduled at appropriate times, such as outside production hours (TRS 1010, Annex 8, cl. 12.10 and 13.5). For sterile facilities, EU GMP Annex 1 lists interruption of air movement and special maintenance, such as final filter changes, as examples of changes to consider, and sets periodic requalification at a maximum interval of 6 months for Grades A and B and 12 months for C and D (cl. 4.32). Scheduling washing as preventive maintenance, aligned with that periodic requalification where possible, keeps interventions to a minimum.
In a pharmaceutical plant, the design question is not only how many air changes the equipment delivers but how the air reaches each zone: low, measured velocity at hoods and cabinets in the lab; full-room flushing in production, with materials that do not become a contamination source and can actually be cleaned; and temperatures within setpoint limits throughout the validated volume, confirmed by mapping, in the warehouse. A fabric duct engineered for each zone, modeled in CFD and validated case by case can meet those conditions.
Planning or expanding a plant? Request zone-by-zone fabric duct sizing with CFD modeling at fabricduct.co.
