Clean Room Projects & Equipment: A 9-Step Ordering Checklist (Built From $41,000 in My Own Mistakes)

This checklist is for anyone ordering clean room projects & equipments — whether you're a facility manager signing off on a new ISO 7 suite, a lab manager swapping out a vertical laminar air flow bench, or a contractor who has to hand over a validated room to a client. It runs through nine steps, in order, from the day someone emails you a floor plan to the day you sign the acceptance report.

I've been handling clean room construction and equipment orders for about nine years. I've personally made (and documented) five significant mistakes on those orders, totaling roughly $41,000 in wasted budget — wrong filter efficiency classes, a pressure cascade we had to rebuild, three clean benches that should have been biosafety cabinets. This is the list I keep taped inside my cabinet door now.

Step 1: Fix the ISO class (and the occupancy state) before anyone quotes you

"Clean room" is not a specification. ISO 14644-1:2015 defines classes by particle concentration, and the numbers are not interchangeable with the older Federal Standard 209E names people still use in conversation.

ISO 14644-1:2015 sets limits for particles ≥0.5 µm per cubic meter. ISO Class 5 = 3,520. ISO Class 7 = 352,000. ISO Class 8 = 3,520,000. The old "Class 100 / 10,000 / 100,000" labels map roughly onto these, but "roughly" is doing a lot of work in that sentence.

The part most people miss: ISO 14644 defines three occupancy states — as-built, at-rest, and in-operation. A room that hits ISO 7 at rest may only hold ISO 8 with four people inside and a centrifuge running. Get the required state written into the spec, not assumed.

In my first year (2018) I quoted a suite to ISO 8 and the client actually needed ISO 7 at rest. The filter order came back unusable. $4,800, straight to the trash.

Step 2: Map the pressure cascade, not just the room

Air flows from clean to less clean, and the cascade is what keeps contamination out. Most industry guidance lands on 10–15 Pa (about 0.04–0.06 in. w.g.) differential between adjacent spaces, with the cleanest room at the highest pressure — though this varies by application and I'd verify against your own design standard rather than trusting a number from a blog post (including this one).

Draw the cascade as a diagram before you approve mechanical drawings: which doors swing which way, where the pass-through hatches sit, what happens when the airlock door is open. The cascade doesn't care about your floor plan — it only obeys physics.

I had mixed feelings about a rush approval on the mechanical drawings in September 2021. On one hand, the GC was threatening a schedule slip. On the other, I knew we hadn't modeled the interlock sequence. I signed anyway. We rebuilt the pressure control panel for $12,600.

Step 3: Specify the purifier HEPA filter by test standard, not by nickname

"HEPA" gets thrown around as if it means one thing. It doesn't.

• US historical definition: 99.97% removal of particles at 0.3 µm (originally MIL-STD-282).
• EN 1822: rates efficiency at MPPS (most penetrating particle size — typically somewhere in the 0.1–0.2 µm range for glass fiber media). H13 ≥99.95%; H14 ≥99.995%.
• IEST-RP-CC001: the US filter spec that governs Type C, Type D, etc.

So when you write "purifier HEPA filter" on a PO, ask yourself: which test standard? Which efficiency class? Which gasket (gel-sealed or knife-edge)? What frame material, what face guard, what sealant?

The mistake I keep seeing — and made once myself — is specifying "H13" and receiving something that's H13 by EN 1822 but only H10-equivalent by the manufacturer's own independent test. Always ask for the test report, not just the label.

Step 4: Match the fan filter unit to the actual ceiling grid (the step everyone skips)

This is the one I'd bet most first-timers skip, and it's the one that turns into a change order.

A clean room HEPA fan filter unit has to land in a ceiling grid that already exists. Common nominal sizes are 2'×2', 2'×4', and 4'×4' — but imported FFUs are frequently metric, so you get 1175×575 mm sitting next to a 1200×600 mm T-bar opening on a drawing, and somebody decides they're "close enough" (they're not).

Three things to check before ordering:

• Grid module. Exact opening dimensions, in mm, measured on site. Not from the architectural drawing.
• Dead load. A 4'×4' FFU with a HEPA filter can run 30–45 kg. If the ceiling grid wasn't designed for that point load, you're reinforcing after the fact.
• Face velocity. Unidirectional-flow ("laminar") applications typically target 0.3–0.5 m/s at the filter face — but always confirm the target with the design engineer, because it depends on the room and the process.

Granted, this step takes an extra half-day of measuring. It has saved me two change orders since 2022, so I'll keep doing it.

Step 5: Decide laminar flow bench vs. biosafety cabinet vs. neither

A vertical laminar air flow bench (clean bench) protects the work. It blows filtered air down over the sample and out toward the operator. It does not protect the operator from hazardous aerosols, vapors, or powders.

If you're working with anything cytotoxic, infectious, or powder-form hazardous, you need a biosafety cabinet or a powder-containment hood. This is one of those mistakes you only make once, and I made it on a $9,200 three-bench order in Q1 2019.

Rule of thumb: if you can't name what you're trying to protect (product vs. person vs. environment), you're not ready to order the bench yet.

Step 6: Don't undersize clean air ventilation makeup

Modern clean rooms recirculate heavily — often 80–90% of the air — but they still need outdoor air for pressurization, CO₂ control, and occupant comfort. Under-specify that and the cascade never holds, no matter how good the FFUs are.

ASHRAE 62.1 gives minimum outdoor-air rates by occupancy and floor area; lab and clean room applications typically target more than the minimum. I'm not 100% sure of the exact cfm-per-person figure your local code will enforce, so take that with a grain of salt — but I can tell you that the project where we cut makeup air to save on the AHU is the same project where the ISO 7 room hovered at ISO 8 for three weeks (thankfully, we caught it during commissioning and not during a customer audit).

Step 7: Buy the commissioning tests, not just the room

Accept nothing on visual inspection. The standard test set is:

1. HEPA filter leak test — IEST-RP-CC034 method.
2. Air velocity and uniformity — ISO 14644-3.
3. Particle count at your required occupancy state — ISO 14644-1.
4. Pressure differential across every boundary in the cascade.
5. Recovery test — how fast the room cleans itself up after a particle challenge.
6. Airflow visualization — the smoke pencil test. Almost nobody asks for it. It's the test that finds the leaks you can't see.

Looking back, I should have paid for a third-party commissioning agent on the 2021 project. At the time, the installer's own test reports looked fine on paper. Three filters failed the leak test on re-tested supply, and we ate $2,300 in retesting plus a two-week delay.

Step 8: Build the documentation binder before handover

As-built drawings, filter serial numbers and test reports, airflow balance reports, pressure cascade diagram, clean room log template, change log. This is not paperwork for its own sake — it's what lets the next person troubleshoot the room in three years without calling you at 11 pm.

Step 9: Set the PM schedule before the warranty expires

Pre-filters get checked every 1–3 months depending on loading. HEPA leak re-testing is commonly every 12–24 months (tighter for pharmacies under USP <797> and for aseptic operations). Pressure differentials should be logged continuously or at least daily. Gaskets and knife-edge seals get inspected at every filter change.

The warranty period is when you find out whether your PM plan is real or theoretical.

Six common mistakes, in order of how often I see them

• Ordering HEPA filters by nickname ("H13") without the test report.
• Mixing ISO class with the old 209E naming and confusing the two in writing.
• Forgetting the at-rest vs. in-operation distinction in the spec.
• Under-specifying outdoor-air makeup, then blaming the FFUs.
• Buying a clean bench where a biosafety cabinet was actually needed.
• Skipping airflow visualization because "the particle count passed."

None of this is exotic. It's just the difference between a five-minute check before the PO goes out and a five-week correction after the room is built. That trade has never once gone the other way for me.