Pharma Biotech Cleanroom HVAC Duct Guide — GMP Annex 1, ISO 14644, Stainless Welded Duct Specification

Why pharma duct is different — and why it matters

Pharmaceutical and biotech cleanroom HVAC ductwork is not building services. It is part of the manufacturing process. Under EU GMP Annex 1 (2022 revision) and FDA cGMP 21 CFR Part 211, the air that contacts a sterile product is treated with the same rigour as the stainless reactor that produces it. This single regulatory premise drives every material, finish, joint, gasket, damper and weld decision on the duct package — and it is the reason a perfectly competent commercial HVAC contractor can spend an entire project trying and failing to deliver a cleanroom duct system that an Annex 1 inspector will sign off.

The contrast is sharp. Office-building HVAC duct in galvanised steel with TDF flanges, slip-and-drive joints, mastic sealant and PEPS class B leakage performance is engineered for the comfort of occupants. The duct walls can shed zinc, the seams can leak, the mastic can off-gas — none of these are patient-safety risks for an office worker. Move that same system into a Grade A aseptic fill suite and every one of those characteristics becomes a non-conformance. The fix is not better commercial duct. It is a different category of duct: fully welded 316L stainless, Ra ≤ 0.5 µm electropolished internal surface, leakage class D, qualified welds by qualified welders to a documented WPS, surfaces certified by profilometer, joints sealed with silicone-free EPDM, and a turnover package thick enough to satisfy an FDA Form 483 audit defence.

The framework that drives this is layered. ICH Q9 establishes the principle that quality risks must be managed proportionally — a Grade A risk demands a Grade A response. EU GMP Annex 1 (2022 revision) sits as the most prescriptive global standard on contamination control for sterile manufacturing, with the contamination control strategy (CCS) now formalised as the central document tying facility, equipment and process. FDA cGMP under 21 CFR 211 overlays a US-specific layer with its own inspection methodology, including the well-known scrutiny of HVAC commissioning under 211.46 Ventilation, air filtration, air heating and cooling. MHRA applies UK-specific application of GMP and PIC/S applies in most non-EU regulated markets. Behind all of these sits ISO 14644-1, which classifies the cleanroom by airborne particle concentration and gives the duct designer the only quantitative target the system has to meet.

This guide takes a duct fabricator's perspective on those frameworks, then translates them into the specific materials, processes, machines and documentation that will pass an inspection. It is the reference SBKJ engineers walk through with pharma engineering project managers when we are asked to supply the cleanroom-grade fabrication machinery, the stainless coil and the technical knowledge to build the duct system in-house — or when we are asked to supply finished welded 316L duct packages turnkey from the SBKJ factory.

ISO 14644-1 cleanroom classifications and HVAC implications

The cleanroom class is the single most consequential parameter for the duct designer. It sets the air change rate, drives the filter selection, defines the recovery time after a breach and propagates straight into duct cross-section, fan kW and the kilometres of stainless that have to be welded. Every other decision in the duct package is downstream of the class.

The current ISO 14644-1:2015 classification is based on the maximum allowable concentration of airborne particles ≥ 0.5 µm per cubic metre. The bands map onto the older US Federal Standard 209E "Class N" nomenclature and onto the EU GMP Grades A, B, C and D. The mapping is not exact in the cleanest classes but the working approximation that pharma engineers use day-to-day is:

• ISO Class 5 / Class 100 / Grade A. Maximum 3 520 particles/m³ ≥ 0.5 µm. The aseptic core — open product exposure, sterile filling, lyophilisation loading. Unidirectional laminar flow at 0.36 to 0.54 m/s face velocity. Air change rate practically meaningless because the volume turnover is dominated by the laminar curtain — typical equivalent ACH 250 to 600.
• ISO Class 7 / Class 10 000 / Grade B. Maximum 352 000 particles/m³ ≥ 0.5 µm in operation, 3 520 at rest. Background environment for Grade A activities. Turbulent or mixed flow. Air change rate 30 to 60 ACH.
• ISO Class 8 / Class 100 000 / Grade C. Maximum 3 520 000 particles/m³ ≥ 0.5 µm in operation, 352 000 at rest. Less critical activities — solution preparation, component prep, terminally sterilised processing. ACH 5 to 25.
• ISO Class 9 / Grade D. Maximum 35 200 000 particles/m³ ≥ 0.5 µm. Warehousing of components, primary packaging materials, weighing of unopened raw materials. ACH 5 to 20.

For the duct designer the consequences cascade. A 200 m² Grade A fill line at 500 ACH and 3 m ceiling height moves 300 000 m³/h. Sized at 8 m/s in the main supply trunk this is a 1 200 mm diameter round duct, in 316L stainless, fully welded with orbital GTAW, on a leakage budget that allows roughly 3 L/s of total system leakage at 1 000 Pa. By contrast a 200 m² Grade D warehouse at 10 ACH on the same ceiling height moves 6 000 m³/h — a 500 mm spiral lockformed stainless duct with no welding requirement. The room is the same size. The duct package differs by an order of magnitude in cost and complexity.

Pressure cascade design — the cleanroom's defence in depth

ISO classification tells you what is allowed inside a room at steady state. The pressure cascade is what keeps the dirtier air outside. The principle is simple: air must always flow from a cleaner zone to a less-clean zone, even when a door opens, even when a single HEPA fails, even when the operator is moving between rooms. The duct system has to deliver this on every operating scenario in the URS — and Annex 1 inspectors increasingly run real-time challenge tests on the cascade during inspection.

Industry practice in regulated pharma is a 15 Pa step gradient between adjacent grades. A typical aseptic fill cascade looks like:

• External corridor: 0 Pa reference.
• First gowning room (Grade D): +15 Pa.
• Second gowning room (Grade C): +30 Pa.
• Grade B background corridor: +45 Pa.
• Grade B fill suite: +60 Pa.
• Grade A laminar flow zone over filling line: localised unidirectional flow, suite background +60 Pa.

Three airlock topologies are in common use. Cascade airlocks step down progressively from clean to less-clean, used for personnel transit into aseptic suites. Sink airlocks are at lower pressure than both adjacent zones — used to contain potent or biohazardous material so that the airlock acts as a contamination trap rather than letting the contamination flow with the cascade. Bubble airlocks are at higher pressure than both adjacent zones, used for transferring sterile components into a Grade A zone. The airlock topology drives the duct schedule for that interlock — the supply and exhaust to a sink airlock must be sized for negative pressure operation under all door-position scenarios, which usually means the exhaust side is on contained ductwork with isolation dampers.

The duct designer has three knobs to make the cascade stable: supply fan modulation through VFD, room exhaust modulation through stainless control dampers, and the leakage class of the ductwork itself. A leaking duct upstream of a HEPA terminal can pressurise the ceiling plenum and undermine the cascade — which is why class D leakage performance is the default on the supply branch serving Grade B and Grade A.

Air change rates and how they drive duct sizing

Air change rate is the cleanroom designer's blunt instrument. It is the parameter that buyers and engineers debate first, because it sets both the operating cost (fan kW, chiller load, filter replacement) and the capital cost (duct cross-section, AHU footprint, HEPA face area).

The ACH bands published by ISPE Baseline Guide and accepted by FDA and EMA inspectors as defensible:

• ISO 5 / Grade A. 250–600 ACH equivalent. Practical design driven by laminar flow face velocity 0.36–0.54 m/s and the ceiling area covered.
• ISO 6. 100–180 ACH. Used in some semi-continuous processing zones.
• ISO 7 / Grade B. 30–60 ACH. Background to Grade A activities.
• ISO 8 / Grade C. 5–25 ACH. Solution prep and non-sterile processing.
• ISO 9 / Grade D. 5–20 ACH. Warehousing.

Sized on these rates a single Grade A fill suite of 8 × 6 × 3 m carries roughly 50 000 to 60 000 m³/h on its laminar flow ceiling alone. A medium-large vaccine fill-finish facility with a Grade A core, Grade B background, Grade C solution prep and Grade D warehousing typically runs 350 000 to 600 000 m³/h of total conditioned air through the HVAC plant. Of that volume, perhaps 60 to 80 % is in stainless ductwork — the supply runs to Grade A and B, the contained exhaust from open-product zones, the recirculation through HEPA back-walls. The remaining 20 to 40 % can be in galvanised or coated commercial duct on the building-services side.

The duct sizing follows from the volume. SBKJ engineers work to a velocity cap of 8 m/s on main supply trunks to Grade A and B, 10 m/s on Grade C and D supply, 12 m/s on common return, 15 m/s on contained exhaust. These limits keep the regenerated noise level below NC 50 in the cleanroom and keep duct erosion negligible over a 25-year service life.

HEPA and ULPA terminal filter integration

The HEPA filter is what actually delivers the cleanroom class. The duct system's job is to deliver the right air volume at the right velocity to the upstream face of the filter, with enough filter face area, enough seal pressure and enough physical access for in-place leak testing. The duct classification chain ends at the terminal — which is why HEPA terminal housings are specified, sealed and tested as duct components, not as filter accessories.

Filter classification is governed by EN 1822-1, which defines classes by the Most Penetrating Particle Size (MPPS) penetration:

• H13. ≥ 99.95 % MPPS retention. Typical for Grade C / D supply terminals where the filter is the primary contamination-control element.
• H14. ≥ 99.995 % MPPS retention. The default for Grade B background supply and most Grade A laminar flow ceilings.
• U15. ≥ 99.9995 % MPPS retention. Used in restricted access barrier systems (RABS) and isolators on Grade A.
• U16. ≥ 99.99995 % MPPS retention. Specialist applications including high-potency manufacturing and certain cell-therapy isolators.
• U17. ≥ 99.999995 % MPPS retention. Niche applications in semiconductor cleanrooms and specialised aseptic isolators.

The duct connection to the terminal housing is one of the highest-risk integration points in the system. Three rules apply on every Grade A / B terminal SBKJ fabricates: the duct must enter the housing through a fully welded, leak-tested transition; the upstream sample / injection port for in-place DOP / PAO challenge testing must be accessible without breaking the cleanroom envelope; and the downstream side of the filter must allow scan testing per ISO 14644-3 or IEST-RP-CC034. A poorly designed terminal connection that requires breaking ceiling tiles to access the test ports turns every annual re-qualification into a multi-day shutdown.

Material selection — 304L vs 316L stainless

The stainless grade decision is one of the few in cleanroom duct where the cost-benefit calculation is genuinely contestable. 316L has a chromium content of 16–18 %, nickel 10–14 %, molybdenum 2–3 %, with carbon ≤ 0.03 %. 304L has chromium 18–20 %, nickel 8–12 %, no molybdenum. The molybdenum in 316L confers significantly better resistance to chloride pitting and to the trace organic acids generated by some pharmaceutical processes — and it is the molybdenum that drives the price premium of roughly 25–35 % over 304L on equivalent gauges.

The decision tree SBKJ engineers apply:

• Active pharmaceutical corridor, Grade A and B. 316L always. The cost of an in-service corrosion failure that contaminates a sterile batch is orders of magnitude greater than the material premium.
• Sterile water for injection (WFI) loop ductwork. 316L always. Chloride exposure from cleaning chemistries is the failure mode.
• Biotech process exhaust (cell culture, fermentation off-gas). 316L. Acidic condensate is the failure mode.
• Solid dose Grade C cleanroom (OSD). 304L is acceptable on supply and return where there is no liquid contact and no chloride exposure. Some clients standardise on 316L across the board for inventory simplicity — both choices are defensible.
• Grade D warehousing and ancillary cleanroom. 304L acceptable, sometimes galvanised steel with epoxy coating where the URS allows.
• Outdoor weatherheads, building exhaust. 316L for marine sites, 304L acceptable inland.

The L designation (low carbon, ≤ 0.03 %) matters because cleanroom duct is welded, and welding higher-carbon grades risks chromium carbide precipitation at the heat-affected zone — a phenomenon called sensitisation that destroys the corrosion resistance exactly where it is needed. Specifying 304L and 316L is a non-negotiable on welded cleanroom duct.

On material certification, SBKJ requires 3.1 mill certificates per EN 10204 on every heat — these are inspection certificates issued by the steelmaker's own quality department, not by an independent third party (which would be 3.2). 3.1 is the pharma industry standard. Heat numbers are stamped on the material and traced through fabrication to the as-built weld map. Positive Material Identification (PMI) by handheld XRF is performed on every heat at the fabricator's incoming inspection — the analytical fingerprint is recorded against the heat number. Counterfeit or mis-graded stainless is a known supply-chain risk and PMI is the non-negotiable defence.

Welded round duct vs sealed rectangular vs spiral lockformed

The construction method of the duct is downstream of three decisions: the cleanroom grade it serves, the cleanability requirement under Annex 1 paragraph 4.10, and the regulatory regime. The three options on the table are fully welded round duct, sealed rectangular duct and spiral lockformed duct.

Fully welded round duct. 316L stainless, formed from sheet or strip on a roller machine, longitudinal seam orbital GTAW or laser welded, transverse joints orbital GTAW only. No mechanical seams, no sealants in the air path. Internal surface electropolished to Ra ≤ 0.5 µm. Pickle and passivate after welding to remove heat tint. This is the default for sterile-API supply, vaccine fill-finish supply, biologics aseptic processing supply and all Grade A / B contained exhaust. SBKJ supplies this duct in diameters 100 mm to 1 600 mm in 316L 1.0 mm to 3.0 mm wall.

Sealed rectangular duct. 304L or 316L stainless, formed on rectangular line tooling, transverse joints flanged with continuous seam weld or fully gasketed with EPDM and silicone-free sealant, longitudinal seams either Pittsburgh lockformed (lower duty) or seam welded (higher duty). Used where space constraints force rectangular sections — typical in retrofit projects, dense plant rooms and integration with existing AHUs. Acceptable for Grade C and below on most projects; acceptable for Grade B by exception with leakage testing per spool. SBKJ rectangular cleanroom duct uses our SBLD-V Auto Duct Line series re-tooled for stainless. See our auto duct line catalogue.

Spiral lockformed stainless duct. 304L or 316L coil fed through a spiral tubeformer, mandrel-formed seam locked and continuously sealed with silicone-free butyl. The seam is the most-discussed feature: SBKJ's stainless tubeformer mandrel-formed seam produces a flush internal surface with no internal leakage path, which makes it cleanable in place to the limits Annex 1 expects on non-sterile supply and on cleanroom return. Spiral duct is the right choice for cleanroom return, building exhaust, dust collection on solid dose, and any supply duty that does not specifically require fully welded construction. SBKJ produces this duct on the SBTF stainless tubeformer — see the spiral tubeformer catalogue.

The selection logic, applied per duct run on the isometrics:

• Supply to Grade A or B sterile zones → fully welded 316L round duct.
• Supply to Grade C / D non-sterile cleanroom → spiral lockformed 304L or 316L, or sealed rectangular.
• Cleanroom return (any grade) → spiral lockformed stainless acceptable.
• Building exhaust (any grade) → spiral lockformed stainless acceptable.
• Contained exhaust on OEB 4/5 substances → fully welded 316L, leakage class D, isolation dampers.
• Dust collection on OSD facilities → spiral lockformed stainless acceptable, with explosion relief if required.

SBKJ stainless spiral tubeformer — mandrel-formed seam vs lockformed seam

The seam construction on stainless spiral duct is one of the most consequential and most under-discussed engineering choices in cleanroom duct. SBKJ's SBTF stainless tubeformer uses a mandrel-formed seam, which produces a flush internal surface where the formed strip overlaps and is mechanically locked under the mandrel rolls. The internal surface presents no protruding seam edge to the air stream, which means no shadow, no debris harbour and no measurable cleanability penalty against welded duct on Grade C and D supply duties.

The contrast is with lower-cost lockformed seams that protrude into the duct interior by 1.5 to 3 mm, creating both a leakage path and a cleaning shadow. On commercial HVAC duct the protrusion is irrelevant. On a regulated pharma supply duct it is a non-conformance under Annex 1 paragraph 4.10 cleanability expectations.

SBKJ SBTF stainless tubeformer headline specifications:

• Material: 304, 304L, 316, 316L stainless coil. Wall thickness 0.5 mm to 1.5 mm. Up to 2.0 mm with extended tooling pack.
• Diameter range: 80 mm to 1 500 mm with quick-change mandrel sets.
• Forming speed: 18 to 35 m/min depending on diameter.
• Seam: mandrel-formed, flush internal surface, continuous silicone-free butyl injection on the seam line.
• Coil width: 137 mm slit coil typical, customisable to match diameter mix.
• Drive system: AC servo with closed-loop length feedback. Length tolerance ± 1 mm per metre.
• Cut-off: orbital plasma or shear with internal deburring ring.
• Control system: Siemens or Mitsubishi PLC with industrial HMI, recipe-based diameter changeover under 8 minutes.

For pharma cleanroom duct production lines SBKJ pairs the SBTF tubeformer with a stainless-grade automated welding station for transverse joint welding (orbital GTAW), a pickle-and-passivate station with closed-loop chemistry, and a profilometer-equipped surface inspection cell. The combined line produces 316L cleanroom-grade duct end-to-end from coil to certified spool.

Surface finish — 2B, #4 brushed, electropolish

Surface finish on cleanroom duct is specified, measured and certified — not assumed. The four finishes encountered on pharma duct, ranked from least to most stringent:

• 2B mill finish. The default cold-rolled, annealed and lightly skin-passed finish on stainless coil. Internal surface roughness Ra typically 0.3 to 1.0 µm. Acceptable for Grade C, D and most cleanroom return / exhaust duct.
• 2D mill finish. Cold rolled, annealed, no skin pass — slightly rougher than 2B (Ra 0.5 to 2.0 µm). Less common in cleanroom duct, mostly seen on heavy gauge feedstock.
• #4 brushed finish. Mechanical polish with abrasive belt, Ra 0.4 to 0.8 µm depending on grit. Used on duct exteriors where appearance matters and on some Grade C supply.
• Electropolish. Electrochemical removal of the surface micro-asperities, leaves a passive chromium-rich oxide layer. Achievable Ra ≤ 0.5 µm routinely, ≤ 0.25 µm on best practice. The default for Grade A and B sterile supply, for high-purity gas distribution and for any duct that has to be cleanable in place.

Finish is measured by contact profilometer per ISO 4287 or ASME B46.1, on a sample frequency of one read per spool minimum, three reads per spool on Grade A duct. The Ra value is recorded against the spool serial number, the heat number and the operator. The lot certificate goes into the turnover package. SBKJ uses Mitutoyo SJ-410 profilometers on the cleanroom duct line and issues finish certificates by spool with a calibration trace.

The relationship between Ra and microbiological cleanability is well established. Below Ra ≈ 0.8 µm, biofilm formation rate falls steeply because the surface micro-asperities that nucleate biofilm are smaller than the bacterial cell size. Below Ra ≈ 0.4 µm the rate is essentially zero on a routine cleaning cycle. This is the engineering basis for the Ra ≤ 0.5 µm specification on Grade A / B duct — it is not arbitrary, it tracks the size of the contamination problem.

For deeper specification background see SBKJ's separate guide on galvanised versus stainless steel duct.

Pickle and passivate — restoring the stainless after welding

Welding 316L creates two visible problems and one invisible one. The visible problems are heat tint (the rainbow oxide layer at the heat-affected zone) and weld scale (the dark slag on the weld bead). Both are concentrated chromium-depleted zones that present a corrosion initiation site if left in service. The invisible problem is that the chromium oxide passive layer that gives stainless its corrosion resistance has been disrupted across the heat-affected zone — even where no visible heat tint is present.

Pickle and passivate is the chemical process that fixes all three. It is required on every welded stainless cleanroom duct fabrication and it is governed by:

• ASTM A380. Standard practice for cleaning, descaling and passivation of stainless steel parts, equipment and systems.
• ASTM A967. Standard specification for chemical passivation treatments for stainless steel parts.
• ISO 16048. Passivation of corrosion-resistant stainless-steel parts.

The two-stage process: pickle with a nitric-hydrofluoric acid blend (typical 10–15 % HNO₃ + 1–3 % HF) at 20–40 °C for 15–60 minutes to remove heat tint, weld scale and surface contamination; then passivate with nitric acid (typical 20 % HNO₃) at 20–50 °C for 20–30 minutes to restore the chromium oxide layer. Triple rinse with deionised water between stages and after the passivate step. Drain and dry under filtered air.

Verification is by FerroxylTM test or copper sulphate test on a witness coupon per batch — both detect free iron contamination on the surface. A passivated surface shows no blue stain in the FerroxylTM test and no copper deposition in the copper sulphate test. Pass / fail is recorded against the batch and goes into the turnover package.

Pickle and passivate has to be done after all welding is complete. It is destructive to gaskets, sealants and any non-stainless components, so pre-installation passivation in the fabricator's shop is the only practical sequence for cleanroom duct. SBKJ's stainless duct shop has a dedicated pickle-and-passivate line with closed-loop chemistry, automated rinse, neutralised effluent treatment and certified batch records on every duct package.

Cleanroom-grade gaskets and sealants

The gasket and sealant decision is small in cost terms and large in regulatory consequence. Three categories of failure SBKJ has seen on inspection:

• Silicone migration. Silicone sealants migrate as molecular vapour and recondense on every surface in the cleanroom — including on the product. Silicone contamination on a finished oral solid dose tablet is a documented cause of Annex 1 deviations. Silicone-free sealants are the default on every pharma cleanroom duct project SBKJ runs.
• Gasket fibre shedding. Fibre-loaded gaskets (asbestos historical, glass fibre or aramid current) can shed micron-scale fibres into the air stream over time. EPDM monolithic gaskets, PTFE envelope gaskets and FKM are the cleanroom-grade choices. Always specify monolithic compounds, never composite.
• Plasticiser leaching. Some PVC-based sealants and gaskets contain phthalate plasticisers that leach over the service life. For drug-product manufacturing always specify phthalate-free formulations with a manufacturer's compliance statement.

Default gasket and sealant bill of materials for SBKJ cleanroom duct packages:

• Flange gaskets, Grade A / B sterile supply. EPDM monolithic, 70 Shore A, USP Class VI tested.
• Flange gaskets, Grade C / D supply. EPDM or NBR monolithic, 60–70 Shore A.
• Sealant on flanged joints. Silicone-free polymer sealant, typically polyurethane or modified silyl polymer. Shore A 30–40 cured.
• Spiral seam continuous sealant. Silicone-free butyl, food-grade, FDA 21 CFR 175.300 compliant where applicable.
• OSD high-potency exhaust. FKM (Viton) gaskets, USP Class VI, with manufacturer's chemical compatibility statement for the specific compound list.
• Biotech bio-process exhaust. EPDM with explicit bio-burden compatibility statement; some clients specify steam-cleanable gaskets where the duct is cleaned in place with pure steam.

For a deeper treatment of cleanroom-grade sealant chemistry and selection, see SBKJ's separate guide on HVAC duct sealants and gaskets.

Damper specification for pressure-cascade control

Dampers are the moving parts of the cleanroom duct system, and on a pressure-cascade-controlled facility they are also the active elements that defend the cascade in real time. Three damper categories appear in pharma cleanroom duct:

• Volume control dampers (VCDs). Modulating dampers that trim air flow per branch under building automation system (BAS) control. On Grade A and B branches SBKJ specifies stainless damper blades and shafts, low-leakage shaft seals (PTFE or graphite-impregnated), pneumatic actuator with positive position feedback (typically 4–20 mA loop), and EN 1751 leakage class 3 minimum.
• Pressure-relief dampers. Single-direction barometric dampers that protect against over-pressurisation under transient events. Spring-loaded or counterweighted. Stainless for Grade A and B, galvanised acceptable lower grades.
• Isolation dampers. Bubble-tight (zero leakage) dampers for contained exhaust on OEB 4 / 5 substances and for cross-contamination isolation between rooms during decontamination cycles. EN 1751 class 4 with positive-pressure shaft seal. Pneumatic actuator with limit switches confirming open and closed positions.

Leakage class is governed by EN 1751:2014 — Air terminal devices, dampers and valves. The classes:

• Class 1. Highest leakage allowance, ≤ 175 mL/(s·m²) at 1 000 Pa. Commercial only.
• Class 2. ≤ 56 mL/(s·m²) at 1 000 Pa. Commercial.
• Class 3. ≤ 18 mL/(s·m²) at 1 000 Pa. Pharma cleanroom standard for Grade A / B supply.
• Class 4. ≤ 5 mL/(s·m²) at 1 000 Pa. Contained exhaust, isolation dampers.

The actuation is as important as the damper hardware. Pneumatic positioning provides reliable feedback under fire-suppression scenarios where electric actuators may fail. SBKJ standardises on pneumatic actuators with 4–20 mA position transmitters and dual limit switches on every Grade A / B / contained exhaust damper, with a certificate of conformance per actuator into the turnover package.

High containment — OEB 4/5 active substances

High-potency manufacturing — cytotoxic anti-cancer agents, hormones, certain peptides, conjugated antibody-drug conjugates — drives an order-of-magnitude tighter exhaust duct specification than aseptic supply duct. The classification framework is the Occupational Exposure Band (OEB), published by Safebridge, ISPE and adopted as common language across the regulated pharma manufacturing community:

• OEB 1. Occupational Exposure Limit (OEL) ≥ 1 000 µg/m³. Standard pharmaceutical containment.
• OEB 2. OEL 100 to 1 000 µg/m³.
• OEB 3. OEL 10 to 100 µg/m³.
• OEB 4. OEL 1 to 10 µg/m³. High potency. Most cytotoxic agents, sex hormones.
• OEB 5. OEL ≤ 1 µg/m³. Highest containment band. Some ADCs, certain conjugated peptides, high-potency anti-cancer compounds.

On OEB 4 and OEB 5 substances the duct system is the secondary containment behind the isolator or RABS primary barrier. Failure of the duct integrity is a regulatory and an occupational health event simultaneously. The duct specification SBKJ delivers on contained exhaust at OEB 4 / 5:

• Material: 316L stainless, fully welded, no spiral seam, no slip joints.
• Surface finish: Ra ≤ 0.8 µm internal, electropolish on Grade B-adjacent zones.
• Welding: orbital GTAW, 100 % visual and 10 % radiographic inspection on transverse joints, all welds to qualified WPS.
• Leakage class: D under EN 1507, pressure-decay testable for periodic re-qualification.
• Filtration: bag-in / bag-out HEPA terminal housings, in-place leak testing capability per ISO 14644-3.
• Isolation: bubble-tight dampers (EN 1751 class 4) at every cross-suite tie-in.
• Pressure transducers: continuous monitoring with alarm to BAS on duct integrity loss.
• Documentation: full traceability per spool with photographic record of every weld and surface finish reading.

The cost implication is significant. A contained exhaust duct package for a high-potency oncology API facility runs 2 to 3 times the per-metre cost of a conventional Grade B aseptic supply duct of the same diameter. The justification is straightforward: the consequence of failure is patient harm, occupational exposure or both.

Validation — leakage testing per IEST-RP-CC006

Once the duct is fabricated, installed and pickled-and-passivated, it has to be tested. Leakage testing is mandatory on every pharma cleanroom duct system and the protocol is the documented evidence that the leakage class promised in the specification was actually achieved on the as-built system.

The two governing standards in current pharma practice:

• IEST-RP-CC006. Recommended Practice from the Institute of Environmental Sciences and Technology, the standard reference for testing cleanrooms in the United States. Defines test pressures, instrumentation accuracy, calculation method and reporting format.
• EN 12237 / EN 1507. European Standards for ductwork strength and leakage. EN 12237 covers circular metal duct; EN 1507 covers rectangular sheet metal ducts.

The test method, in summary: the duct section under test is sealed at both ends, pressurised by a calibrated fan to the test pressure (typical 1 000 Pa for supply, 1 500 Pa for contained exhaust, 500 Pa for general extract), and the leakage flow is measured by an orifice plate or a calibrated rotameter on the inlet line. Leakage is reported in L/s/m² of duct surface area. Pass / fail is against the applicable leakage class.

Class D under EN 1507, the default for pharma supply, allows ≤ 0.009 L/s/m² at 1 000 Pa. To put that in context, on a 100 m² duct surface area at 1 000 Pa, the maximum allowable leakage is 0.9 L/s — slightly less than one litre per second, distributed over an area roughly the size of a small tennis court. Achieving class D requires welded seams, gasket-and-sealant flanged joints applied correctly, and damper-and-filter housing leak tightness designed in from the start.

Test instrumentation must be calibrated and the calibration traceable. SBKJ uses Setra and Dwyer calibrated manometers with NIST-traceable certificates. The test report includes: duct identification, surface area, test pressure, measured leakage, calculated L/s/m², leakage class, ambient conditions, instrument calibration reference, witness signatures and the date. Every report goes into the IQ documentation.

HEPA terminal integrity testing is a separate process under ISO 14644-3 and IEST-RP-CC034. The terminal is challenged upstream with DOP (di-octyl phthalate, increasingly replaced by PAO due to phthalate concerns) at a defined upstream concentration, and the downstream side is scanned with a photometer. Penetration must be ≤ 0.01 % at any point on the filter face. Terminals that fail the scan are repaired (gel-seal injection on the gasket interface) or replaced. This is an annual re-test on most pharma facilities.

Documentation — material certs, WPS, WPQ, PMI, weld maps

The duct package's audit defence is its documentation. A cleanroom-grade pharma duct turnover package SBKJ delivers contains, at minimum:

• 3.1 Mill Test Certificates per EN 10204. One per heat of stainless used. Includes chemical composition, mechanical properties, heat number and steelmaker stamp.
• Positive Material Identification (PMI) Reports. XRF spectroscopy results per heat number, performed at goods-in and again on representative finished spools.
• Weld Procedure Specifications (WPS). One per joint configuration / process / base material / thickness combination. Qualified per ASME Section IX or AWS D18.1.
• Procedure Qualification Records (PQR). The supporting test record for each WPS, including mechanical and macro test results.
• Welder Performance Qualifications (WPQ). One per welder per process per position, current to the project execution date (typically 6 months validity).
• Surface finish profilometer logs. One reading per spool minimum, three readings per spool on Grade A. Ra values, calibration trace, instrument serial number.
• Pickle and passivate batch records. Solution chemistry analysis, contact time, rinse confirmation, FerroxylTM or copper sulphate test results per batch.
• Leakage test reports. Per duct section, per IEST-RP-CC006 or EN 1507, with calibration trace.
• HEPA integrity test reports. Per terminal, per ISO 14644-3.
• Weld maps. Drawings showing every weld with weld number, welder ID, WPS reference and date.
• As-built isometric drawings. Updated through commissioning to reflect the as-installed geometry.
• Calibration certificates. For every test instrument used in the validation campaign — manometers, rotameters, profilometers, photometers, XRF analysers.

The documentation pack typically runs 200 to 600 A4-equivalent pages for a 1 500 m duct package, delivered as a controlled binder and as PDF. It is the deliverable the FDA inspector or EMA assessor will review first on a Pre-Approval Inspection (PAI) or a routine re-inspection. Defects in the documentation trigger Form 483 observations even when the physical installation is sound.

For broader background on weld qualification and weld procedure development see SBKJ's welding methods for HVAC duct fabrication guide.

Regulatory framework — FDA, EMA, MHRA, PIC/S

The regulatory framework applicable to a pharma cleanroom duct system depends on which markets the manufactured product is sold into. The four most consequential authorities for export-grade pharma manufacturing:

• FDA — United States Food and Drug Administration. Applies 21 CFR Part 211 (current Good Manufacturing Practice for finished pharmaceuticals) on all drug products marketed in the US. Specific HVAC scrutiny under 211.46. The FDA Aseptic Processing Guidance (2004) is a separate enforceable expectation. Inspection method is the QSIT (Quality System Inspection Technique) and the systems-based approach. Form 483 observations are the inspection finding mechanism.
• EMA — European Medicines Agency. Applies EU GMP, with Annex 1 (2022 revision) as the binding standard on sterile medicinal products manufacture. Annex 15 covers qualification and validation. Inspections are run by the national competent authorities of EU Member States with mutual recognition.
• MHRA — UK Medicines and Healthcare products Regulatory Agency. Post-Brexit applies UK GMP (substantially identical to EU GMP for sterile manufacture, including UK Annex 1). Inspections by MHRA's Inspectorate.
• PIC/S — Pharmaceutical Inspection Co-operation Scheme. Multilateral framework adopted by 50+ regulatory authorities including TGA Australia, Health Canada, Swissmedic, ANVISA Brazil, and many emerging-market regulators. PIC/S GMP is harmonised with EU GMP including Annex 1.

The duct specification implication: a facility manufacturing for the US market builds to FDA expectations including FDA Aseptic Processing Guidance. A facility manufacturing for the EU builds to Annex 1 (2022). A facility manufacturing for both — which is most export pharma plants — has to meet the more stringent of the two on every line item. In practice on duct systems the EU Annex 1 (2022) is the binding constraint on most modern projects because of its explicit contamination control strategy requirements.

EMA Annex 15 — Qualification and Validation — is the cross-cutting standard that defines what IQ, OQ and PQ have to look like and how they are documented. It applies the same way to a duct system as to a tablet press. The duct fabricator's role in Annex 15 compliance is to deliver a turnover package that allows the user's Quality Unit to execute their qualification protocol without having to re-create data the fabricator should have generated and recorded during fabrication.

Case study — vaccine fill-finish facility, 2,400 m of welded 316L duct

A representative cleanroom duct package SBKJ engineering supports — anonymised but technically representative of multiple recent projects:

• Facility. Greenfield vaccine fill-finish line, single sterile filling suite plus formulation, component prep, gowning and warehouse.
• Cleanroom inventory. 1 × Grade A laminar flow ceiling over filling line (40 m² ceiling area), 1 × Grade B background suite (180 m²), 2 × Grade C formulation rooms (each 60 m²), 1 × Grade D component warehouse (250 m²).
• Air volume. 78 000 m³/h total supply, of which 24 000 m³/h to Grade A unidirectional flow, 12 000 m³/h to Grade B background, 18 000 m³/h to Grade C, 12 000 m³/h to Grade D, 12 000 m³/h to corridors and gowning. Recirculation 70 % overall.
• Duct package. 2 400 m total. 1 600 m fully welded 316L round duct (orbital GTAW, electropolish Ra ≤ 0.5 µm internal). 600 m sealed rectangular 304L on Grade C / D supply and return. 200 m spiral lockformed 304L on building exhaust and warehouse return.
• Filters. 64 × H14 terminal HEPA on Grade A laminar flow ceiling, 18 × H14 supply terminals on Grade B, 14 × H13 on Grade C, 8 × H13 on Grade D.
• Welding scope. 1 240 transverse welds plus 1 600 m of longitudinal seam. 4 qualified orbital GTAW welders. 100 % visual inspection, 12 % radiographic on Grade A / B, 6 % radiographic on Grade C.
• Pressure cascade. Grade A ceiling under laminar flow, suite +60 Pa. Grade B background +45 Pa. Grade B-C airlock +30 Pa. Grade C +30 Pa. Grade C-D airlock +15 Pa. Grade D +15 Pa. External 0 Pa reference.
• Documentation. 412-page turnover binder including 24 mill certificates, 8 WPS, 12 PQR, 4 WPQ, 2 400 line items in the surface finish log, 16 pickle / passivate batch records, 28 leakage test reports (class D pass on every section), 104 HEPA integrity test reports.
• Schedule. 18 weeks from approved isometrics to ex-works. 6 weeks installation on site under SBKJ engineering supervision. 4 weeks IQ / OQ / PQ.

The project completed FDA Pre-Approval Inspection on first attempt with zero Form 483 observations on the HVAC system. The duct documentation was specifically referenced by the inspector as "well organised and traceable" — a comment that compresses 18 weeks of disciplined fabrication recording into one sentence in the inspection report, but is the entire point of the documentation regime.

SBKJ cleanroom-grade machinery options

For pharma duct fabricators building cleanroom-grade duct in-house, SBKJ supplies the full machinery stack. The headline lines:

• SBTF Stainless Spiral Tubeformer. 304 / 304L / 316 / 316L coil to 1 500 mm diameter, mandrel-formed flush internal seam, AC servo drive, Siemens or Mitsubishi PLC, length tolerance ± 1 mm/m. The default machine for spiral lockformed cleanroom return, exhaust and Grade C / D supply duct. See the spiral tubeformer catalogue.
• SBLD-V Auto Duct Production Line, Stainless. Re-tooled for 304L / 316L stainless coil 0.6 to 1.5 mm wall, full Pittsburgh lockform or seam-weld option, integrated TDF flange, length tolerance ± 1 mm. The right choice where rectangular sections are unavoidable. See the auto duct line catalogue.
• Orbital GTAW Welding Stations. Magnatech / Polysoude OEM-equipped or SBKJ-built equivalents, programmable orbital GTAW for 100 mm to 1 600 mm diameter, argon back-purge with closed-loop oxygen monitoring, weld parameter recording per joint. Used for transverse joint welding on fully welded 316L round duct.
• Pickle and Passivate Lines. Closed-loop chemistry tanks for nitric-hydrofluoric pickle and nitric passivate, automated rinse with deionised water, neutralised effluent treatment, batch records by spool. Sized per project from 6 m to 24 m bath length.
• Surface Finish Inspection Cell. Mitutoyo or Taylor Hobson profilometer, calibration kit, Ra logging system tied to spool serial number. Operator-trained certification by SBKJ engineering.
• Leakage Test Rigs. Calibrated fan, manometer, orifice plate, automated test sequencer per IEST-RP-CC006 / EN 1507. Includes HEPA integrity test rig with photometer and DOP / PAO generator.

SBKJ's commercial model on cleanroom duct supports two routes for the pharma engineering customer. Route A: machinery turnkey. SBKJ supplies the complete cleanroom duct fabrication line — tubeformer, welding stations, pickle and passivate, inspection — and the customer fabricates in-house. SBKJ provides operator training, welder qualification support, a process flow for the documentation regime, and 12 months of remote engineering support. This route suits OEMs and large pharma engineering firms with their own fabrication capability. Route B: finished duct supply. SBKJ supplies the finished welded 316L duct package ex-works the SBKJ stainless duct shop, including all documentation, leakage testing and HEPA integrity testing. This route suits one-off projects and customers without in-house pharma fabrication. Both routes are supported from SBKJ's Australian office at Box Hill North VIC for English-speaking technical handover.

For the broader fabrication context see SBKJ's cleanroom duct manufacturing overview, HVAC duct fittings and fabrication guide, and the cleanroom industries page.

Request a cleanroom duct package quotation →

FAQ

Why does pharma duct need to be different from commercial HVAC duct?

Because the air contacts a sterile product. EU GMP Annex 1 (2022 revision) and FDA 21 CFR 211 treat HVAC as part of the manufacturing process, which drives material upgrades to 316L stainless, surface finish to Ra ≤ 0.5 µm by electropolish, fully welded seams in lieu of slip-and-drive, leakage class D, and a formal qualification documentation chain. A SMACNA office-grade spiral duct is acceptable for a commercial building but cannot pass an Annex 1 inspection on a sterile manufacturing line.

Where is welded round duct mandatory and where is spiral duct allowed?

Welded round 316L is the default on supply to Grade A and B sterile zones, on contained exhaust at OEB 4/5, and on any duct that is required to be cleanable in place under Annex 1 paragraph 4.10. Spiral lockformed stainless is acceptable on cleanroom return, building exhaust, dust collection on solid dose, and supply to Grade C / D where the URS does not specifically require welded construction.

What surface finish do regulators expect?

For Grade A and B sterile supply the industry consensus driven by Annex 1 and ISPE Baseline guidance is Ra ≤ 0.5 µm internal, achieved by electropolish on 316L. For Grade C and D 2B mill finish (Ra ≤ 1.0 µm) is generally accepted. Pickle and passivate is required after fabrication welding. Finish must be measured with a profilometer, certified by lot, and recorded in the turnover package.

How much air does a pharma cleanroom move?

ISO 5 / Grade A targets 250–600 ACH equivalent with unidirectional laminar flow. ISO 7 / Grade B runs 30–60 ACH. ISO 8 / Grade C runs 5–25 ACH. ISO 9 / Grade D runs 5–20 ACH. A medium-large vaccine fill-finish facility typically circulates 350 000 to 600 000 m³/h, of which 60–80 % is in stainless ductwork.

What welding qualifications are required?

Stainless cleanroom duct welds are typically qualified under ASME Section IX or AWS D18.1 with documented WPS, PQR and individual welder WPQ for each process and position. Most sterile-grade duct is orbital GTAW on argon back purge. Material is verified by 3.1 mill certificates per EN 10204 and confirmed by PMI on every heat at goods-in.

How is duct leakage tested and what classification is required?

Per IEST-RP-CC006 or EN 12237 / EN 1507 using a calibrated fan, manometer and orifice plate. The default for pharma is class C or D under EN 1507 — class D limits leakage to ≈ 0.009 L/s/m² at 1 000 Pa, an order of magnitude tighter than commercial HVAC class B.

What documentation has to go to the regulator?

3.1 mill certs per EN 10204, WPS / PQR / WPQ records, PMI reports, surface finish logs by spool, pickle and passivate batch records, leakage test reports per IEST-RP-CC006, HEPA integrity test reports per ISO 14644-3, weld maps, as-built isometrics, and calibration certificates for every test instrument. SBKJ provides all of this as standard.

What is the typical lead time for a pharma cleanroom duct package?

For a fully welded 316L duct package of 1 500–2 500 m including fittings, dampers, HEPA terminal housings and full documentation, plan 14 to 22 weeks from approved isometrics to ex-works. Critical path is mill stock availability, welder qualification, fabrication, and pickle / passivate / electropolish.