Industry Standards Explained for Dust & Air Filtration Media

What the Standards Actually Test — and Where They Diverge #

A cement plant in the Pearl River Delta specified “HEPA-grade” filter bags on their RFQ. Three Chinese suppliers quoted against it. Two delivered media tested to EN 1822 at H13 (≥99.95% at MPPS). One delivered media tested to the older GB/T 6165 method, which uses a different particle challenge aerosol and a different scan protocol. All three called their product “HEPA.” At incoming inspection, only one lot passed when re-tested to the buyer’s actual process requirement.

That outcome is not unusual. The core problem is that “filtration efficiency” is not a single, stable measurement. The result you get depends entirely on the particle size used in the challenge aerosol, the flow velocity across the test face, whether the test uses a scanning or integral method, and what the acceptance threshold is at the most penetrating particle size (MPPS). Change any one of those variables and you change the number. Chinese suppliers frequently cite efficiency figures from whichever standard produces the most favorable result for their media — not the standard specified by the buyer.

The practical consequence: if your RFQ says “99% efficiency” without specifying the test standard, particle size range, and flow rate, every supplier will answer “yes” using a different dataset.

The Parameters That Separate Comparable Standards #

The four standard families most relevant to dust and air filtration media procurement are ISO 16890, EN 1822, ASHRAE 52.2, and GB/T 14295 / GB/T 13554. Each tests a different application range and uses different challenge conditions.

ISO 16890, which replaced EN 779 in 2018, covers general ventilation filters and classifies media into ePM1, ePM2.5, and ePM10 groups based on fractional efficiency curves across 0.3–10 µm particle sizes. The critical test condition is a face velocity of 0.944 m/s on a 610 × 610 mm flat panel. Efficiency is measured both before and after electrostatic discharge treatment — a step that EN 779 did not require and that frequently exposes electrostatically enhanced media with poor fibre-bonding efficiency.

EN 1822 covers high-efficiency particulate air (HEPA) and ultra-low penetration air (ULPA) filters and specifically requires MPPS scanning at 0.12–0.25 µm. The H13 threshold is ≥99.95% overall and ≥99.75% at the MPPS in any single local scan point. This is the standard relevant to cleanroom and pharmaceutical inlet applications — not to general industrial dust collection, where ISO 16890 or ASHRAE 52.2 is the correct reference.

ASHRAE 52.2 is the dominant North American reference for commercial HVAC filtration. It assigns MERV ratings (1–16) using a PSL aerosol challenge across three particle size ranges: 0.3–1.0 µm, 1.0–3.0 µm, and 3.0–10.0 µm. MERV 13 roughly corresponds to ePM1≥50% under ISO 16890, but the correspondence is approximate and should not be used interchangeably in a specification.

The GB/T standards present a different challenge. GB/T 14295 covers medium-efficiency panel filters and GB/T 13554 covers high-efficiency HEPA filters. Both pre-date ISO 16890 and use different aerosol challenge conditions. A supplier reporting “98% efficiency per GB/T 14295” is not reporting the same measurement as “ePM2.5 65% per ISO 16890.” They are different tests, different particle ranges, and different acceptance thresholds. This is the specification error we encounter most often when reviewing Chinese supplier COAs against Western buyer RFQs.

The electrostatic discharge requirement in ISO 16890 and EN 1822 is the specification detail that trips up the most procurement teams. Media that achieves high efficiency through electrostatic charge loading (electrospun or resin-bonded fibre layers) can lose 20–40 percentage points of efficiency after discharge treatment. A supplier reporting pre-discharge efficiency numbers on a COA while omitting post-discharge values is not necessarily lying — they may simply be reporting the only test their lab ran. But for the buyer, the post-discharge number is the performance-relevant value, because filters in service lose electrostatic charge over time.

The standard that procurement teams most commonly conflate is ASHRAE 52.2 MERV with ISO 16890 ePM classes. They cover similar particle ranges, and rough crosswalks appear in application guides, but the test methodologies differ enough that a MERV 14 specification on a Chinese RFQ will generate inconsistent bids unless you also specify the ISO 16890 equivalent. Our internal cross-reference form (CAT-F02) maps MERV 13–16 to ISO 16890 ePM bands with specific test condition notes — the correlation breaks down below MERV 11, where the GB/T particle classification approach diverges further.

Regional Equivalence and How to Specify Cross-Regional Orders #

I’d prioritize resolving the ISO 16890 versus GB/T conflict before any other specification issue on a cross-regional order. Here’s why: the GB/T classification system does not produce an ISO 16890 efficiency class directly. A Chinese manufacturer who has only tested to GB/T 14295 cannot certify to ISO 16890 without running the full test suite — including the electrostatic discharge step — at an accredited laboratory. That typically costs RMB 8,000–15,000 per filter configuration and takes 3–6 weeks at a CNAS-accredited test lab. If your qualification timeline does not accommodate this, you will receive GB/T data presented as if it is equivalent, and it is not.

For industrial process dust collection (baghouses, cartridge collectors, pulse-jet systems), the relevant standard family is usually ISO 11057, which covers test methods for filtration fabrics under pulsed flow conditions, and EN 13180 or the VDI 3926 guideline for filter media in dust collection systems. These are distinct from the ISO 16890 / EN 1822 framework, which applies to air filtration panels and media for HVAC and cleanroom applications.

The practical crosswalk for specifying equivalent standards across regions:

If your drawing or specification references EN 1822 H13, write: “EN 1822:2019 H13, tested to MPPS at ≥99.95% overall, ≥99.75% local scan; GB/T 13554:2020 Class B or C may be submitted for equivalency review, subject to independent third-party test confirmation.” That language forces the supplier to either produce EN 1822 test data or explicitly acknowledge that their GB/T data requires equivalency verification.

If your specification references MERV 14 per ASHRAE 52.2, and you are sourcing from China, add: “or demonstrated ePM1 ≥75% and ePM2.5 ≥90% per ISO 16890:2016 post-electrostatic discharge.” This gives Chinese labs a path to compliance using the ISO test infrastructure that is far more common in China than ASHRAE 52.2 testing.

For industrial dust collection media (not HVAC panel filters), specify ISO 11057:2011 test conditions for pulse-jet cleaning resistance and pair them with dimensional and seam strength requirements. Efficiency class alone is not sufficient for baghouse applications because cake formation and pulse-jet cleaning dynamics dominate long-term performance more than initial filtration efficiency.

One area where opinions differ: whether to specify the ISO/EN standard alone, the GB/T equivalent alone, or both in parallel. Some procurement teams insist on EN 1822 or ISO 16890 exclusively to avoid ambiguity. Others accept GB/T test reports from CNAS-accredited labs on the grounds that the test infrastructure is equivalent even if the classification system differs. Our practice is to require the buyer’s target standard exclusively for critical applications (pharmaceutical, food processing, cleanroom supply) and to allow GB/T plus third-party verification for industrial dust collection, where the stakes of a 5–8% efficiency variance are lower than in controlled environments.

For filter fabrics and textile-based filtration media, the relevant physical property standards are different again — tensile strength, elongation, and air permeability are typically tested to ISO 9237 or GB/T 5453, and efficiency data is secondary to mechanical performance in high-temperature baghouse applications.

Decision Framework — How to Specify the Right Standard in Your RFQ #

If your application is commercial HVAC or building ventilation in North America: specify MERV per ASHRAE 52.2 and request ISO 16890 ePM equivalent data as a secondary reference for any Chinese-sourced media. Accept GB/T data only with third-party ISO 16890 confirmation.

If your application is cleanroom supply air, pharmaceutical manufacturing, or electronics assembly: EN 1822 H13 or H14 is the required standard. Do not accept GB/T 13554 Class B or C as a substitute without independent re-test. The scan protocol requirements in EN 1822 are substantially more rigorous than GB/T 13554, and the difference matters at these efficiency levels.

If your application is industrial process dust collection (pulse-jet baghouse, cartridge collector, reverse-air fabric filter): ISO 11057 for pulse-cleaning resistance, combined with your process-specific temperature rating, chemical resistance class, and seam strength threshold. Efficiency class matters less here than fabric integrity over cleaning cycles. I’d put far more weight on air permeability retention after 5,000 pulse cycles than on initial filtration efficiency for this application.

If your application crosses regions — a European plant buying from a Chinese supplier for a North American end-use — request: (a) the buyer’s target standard test report from an accredited lab, (b) a copy of the supplier’s CNAS accreditation certificate showing scope coverage for the relevant test methods, and (c) three consecutive batch COAs to verify consistency. The consistency requirement matters more than most teams acknowledge. We have seen initial sample submissions that pass EN 1822 H13, followed by production batches where the fibre blend ratio shifted enough to drop post-discharge efficiency below threshold — without any change to the supplier’s COA template.

For dust and air filtration media procurement at volume, the specification investment pays for itself. A misspecified standard costs more in re-testing, qualification delays, and potential line shutdown than the per-unit price difference between compliant and non-compliant media.

The non-obvious boundary condition here: for very low-efficiency pre-filtration applications (G3/G4 coarse filters under the old EN 779 classification, now roughly ISO Coarse ≥50% per ISO 16890), the standard reference matters less than dimensional tolerances and media weight consistency. At this efficiency level, the procurement variable that drives total cost is service life, not particle capture efficiency.

Practical Guidance for Buyers #

When sourcing dust and air filtration media from China against a specific standard, the first document to request is not the product datasheet — it is the accredited test report showing the standard, test conditions, face velocity, particle size range, and whether the efficiency measurement was taken before or after electrostatic discharge. A datasheet efficiency number without those parameters is unverifiable.

The most common risk scenario we encounter: a supplier submits a GB/T 13554 test report for a specification that requires EN 1822 H13. The numbers look comparable on the surface. The buyer accepts them. At incoming inspection six months later, independent re-testing to EN 1822 reveals the media passes overall efficiency but fails the local scan requirement (≥99.75% at any point). The entire shipment requires disposition. This happens because GB/T 13554 does not mandate a scan — only an integral test — so media with localised weak zones can pass GB/T and fail EN 1822.

Before volume commitment, insist on: three consecutive production batch samples (not three samples from the same production run) tested to your target standard at a CNAS- or DAkkS-accredited laboratory, with face velocity and discharge conditions explicitly stated on the test report. Sample size should be a minimum of five specimens per test per batch. For HEPA-grade media, this step is not optional. For industrial dust collection media, incoming air permeability spot-testing per ISO 9237 at a tolerance of ±15% from nominal is a practical ongoing control.

FAQ #

What is the difference between ISO 16890 and EN 1822, and can I use them interchangeably?
No. ISO 16890 covers general ventilation filters and classifies efficiency across PM fractions (0.3–10 µm). EN 1822 covers HEPA and ULPA filters and requires scanning at the most penetrating particle size (0.12–0.25 µm). Specifying one on an RFQ does not substitute for the other — they test different performance regimes entirely.

EN 779 is still referenced in some Chinese supplier datasheets. Is it still valid?
EN 779 was withdrawn in 2018 and replaced by ISO 16890. A supplier citing EN 779 efficiency classes (G1–F9) on a current datasheet is either using old test equipment, old reporting templates, or both. Reject that COA and request ISO 16890 data. This matters most for M5, M6, F7, F8, and F9 class media, where the post-discharge requirement in ISO 16890 produces meaningfully lower reported efficiencies than the old EN 779 gravimetric test.

How do I write the standard reference on a Chinese RFQ to avoid ambiguity?
Write the standard number, publication year, efficiency class, specific test condition (face velocity, particle range), and whether post-discharge efficiency is required. Example: “ISO 16890:2016, ePM1 ≥65%, tested at 0.944 m/s, post-electrostatic discharge per ISO 16890-3.” That gives the supplier no room to substitute a different test method.

Can MERV ratings from ASHRAE 52.2 be directly converted to ISO 16890 ePM classes?
Approximately, but not exactly — and the approximation degrades below MERV 11. MERV 13 corresponds roughly to ePM1 ≥50% under ISO 16890, but the test methods use different challenge aerosols and flow conditions. For a Chinese RFQ, I’d specify the ISO 16890 class directly rather than relying on a MERV crosswalk, because ASHRAE 52.2 testing infrastructure is uncommon in Chinese labs.

Do Chinese-made HEPA filters need to be re-tested outside China to be accepted in Europe?
It depends on the buyer’s quality system requirements. A test report from a CNAS-accredited Chinese laboratory is technically recognised under the ILAC mutual recognition arrangement, which covers DAkkS and UKAS. In practice, some European end-users require re-testing at a European-accredited lab for first qualification, then accept CNAS data for ongoing batches. For pharmaceutical and medical device applications, the buyer’s quality assurance agreement typically governs — our dataset here only covers industrial and commercial HVAC applications.

What is the MPPS and why does it matter for filter specification?
MPPS is the most penetrating particle size — the particle diameter at which a given filter medium shows minimum efficiency. For HEPA media, this falls in the 0.12–0.25 µm range. EN 1822 requires that both the overall efficiency and the point efficiency at MPPS meet the class threshold. Media that passes an integral efficiency test can still fail at MPPS — which is exactly the failure mode that GB/T 13554’s integral-only test method does not catch.

Should I specify ISO 16890 or GB/T for a Chinese supplier audit?
Specify the standard your application requires, then ask the supplier to demonstrate they can test to it — not just certify to it by reference. The distinction between testing capability and paper certification is where most sourcing risk concentrates. For a first-time supplier audit, we log the laboratory accreditation scope as the primary indicator under our MED-Q04 qualification checklist, not the product certificate.

Published by sinoraw.com Technical Team | Request a sourcing consultation