A bid tabulation with three HEPA or ULPA classes quoted against three different test bases is not a comparison yet. It is a set of labels waiting to be normalized, and the normalization decision — not the headline efficiency number — determines whether the award is defensible.

The Standard and Edition Behind Each Quoted Class

A quoted filter class only means something in relation to the standard and edition that defines it. Classification systems assign their class designations through specific test procedures, particle size ranges, and acceptance criteria, and those definitions change between editions of the same standard and differ structurally between standards published by different bodies. When a bidder writes a class designation on a line item without naming the standard and edition behind it, the reader has a label without a definition. Two bidders can submit the same-looking class designation while meaning different test bases, different particle size assumptions, or different sampling approaches, because the designation itself is only shorthand for a specific document.

This matters most where a bid tabulation aggregates quotes from multiple suppliers who may be working from different catalogs, different regional defaults, or different editions of the same standard adopted at different times. A supplier’s internal catalog might still reference an older edition even where the organization has since adopted a newer one, or a regional distributor might quote against whatever edition its last project used. None of this is visible from the class designation alone. The reader has to request the named standard and the specific edition for every quoted class before placing bids side by side, because the alternative — assuming similarity of label means similarity of substance — removes the basis for comparison entirely.

Where a project is specifying against ISO 29463, for example, the classification, performance testing, and marking requirements are defined in ISO 29463-1:2024, and the edition identified matters because classification frameworks are revised. A bid that cites “ISO 29463” without an edition has not yet supplied enough information to compare against a competing bid that does cite an edition, even if both appear to quote the same class. This is a bidder-disclosure problem before it is a technical problem: the fix is a question to each bidder, not an assumption resolved at the reader’s desk.

The same logic applies to filter product families offered against a quoted class. A Mini Pleat HEPA/ULPA air filter, a Separator HEPA air filter, and a V-Bank HEPA air filter can each be offered against the same nominal class, but the class claim for each still needs its own named standard and edition rather than inheriting one from a different product line in the same bid package.

Overall Efficiency, Local Leakage, and Installed Integrity Are Separate Results

A single quoted class designation can imply, misleadingly, that one test produced one number that answers every acceptance question a project has. In practice, high-efficiency filtration involves at least three distinct result types, each generated by a different test method, each answering a different question, and none substitutable for another.

Overall efficiency at the most penetrating particle size describes how the filter medium performs in aggregate at the particle size where penetration is highest — a filter-element characteristic tested under a defined method such as that described in ISO 29463-5:2022, which covers filter-element efficiency testing at the most penetrating particle size. This result tells the reader about the element’s bulk filtration performance. It does not tell the reader anything about whether an individual filter element has a local flaw, a pinhole, or a gasket defect that would let unfiltered air bypass the media in one location while the aggregate efficiency remains excellent.

Local scan leakage answers that different question. A scan method, such as the one specified in ISO 29463-4:2011 for determining leakage of filter elements, tests across the face of the filter for local defects rather than aggregate performance. A filter element can pass an overall efficiency test comfortably while still carrying a local leak path that only a scan test would catch, and conversely a scan-leakage pass says nothing about the element’s aggregate efficiency number.

Installed-system integrity is a third and separate result again. It addresses whether the filter, once installed in its housing with its seals, frame, and mounting in place, performs as an integrated system — which is a different acceptance question from either filter-element result, because installation introduces bypass paths, seal compression issues, and housing fit conditions that no filter-element test can anticipate. A bid that offers only a filter-element efficiency certificate has not addressed installed-system integrity at all, and a reader who treats the efficiency number as covering installation has conflated two acceptance results that the relevant test methods never intended to combine. Where a project’s acceptance plan requires evidence at more than one of these levels, each bid needs to be checked for which level or levels its attached reports actually cover.

Test Conditions That Must Match Before Values Are Compared

Even where two bids cite the same standard, the same edition, and the same result type, the headline numbers are not yet comparable until the underlying test conditions match. A test result is only valid for the conditions under which it was generated, and several of those conditions can shift the reported value without changing the filter’s actual field performance, or can mask a real performance difference by comparing numbers taken at incompatible conditions.

The filter configuration tested needs to match the filter configuration offered. A result generated on one pleat depth, one frame construction, or one media arrangement does not automatically transfer to a different configuration carrying the same nominal class, because the physical construction that was tested is part of what produced the number.

The tested airflow matters for the same reason efficiency and leakage results are airflow-dependent: a filter tested at one airflow and offered for service at a different airflow is not guaranteed to retain the same reported performance, since flow rate affects particle capture mechanisms and pressure conditions across the media.

The test stage — whether a result represents an initial condition, a loaded condition, or another defined stage in the test sequence — changes what the number represents. A result taken at one stage answers a different acceptance question than the same numeric value taken at another stage, even when the units and the class label look identical on the page.

The aerosol challenge used and the specific method followed complete the set of conditions that must align. Different aerosol generation approaches and different procedural methods can produce different reported values for filtration performance that is, in practice, equivalent — or can produce similar-looking values for performance that differs. Where any one of these conditions is unstated for a quoted value, that value is not yet comparable to a competing bid’s value, regardless of how similar the class designations appear. The resolution is to request the condition explicitly from each bidder or to treat the comparison as unresolved until it is supplied.

Filter Identity and Configuration Checks for Attached Reports

A test report attached to a bid is only useful if it is traceable to the filter actually being offered. This traceability check is separate from the standard-and-edition check and separate from the test-condition check; it asks whether the document in the bid package actually covers the product on the line item, which is not guaranteed simply because a report is present.

The first check is filter identity and configuration: does the report identify the specific filter configuration it covers, and does that configuration match what the bidder is proposing to supply? A generic certificate describing a product family in general terms, without identifying the configuration tested, leaves open whether the offered item is the one the report describes.

The second check is tested airflow, which here serves a traceability purpose rather than a comparability purpose — the reader is confirming that the claimed performance for this bid actually rests on the airflow basis stated in the attached report, rather than on an assumed or extrapolated airflow.

The third check is test stage, confirming that the result in the attached report represents the stage the project’s acceptance plan is actually comparing against.

The fourth and binding check is whether the bidder has referenced a specific report for the specific claimed class, rather than attaching one generic document intended to stand in for all classes across a product catalog. A bid package that includes a certificate without a direct, stated connection between that certificate and the particular class claimed on that line item has not yet supplied supporting evidence for the claim — it has supplied a document whose relevance remains an open question. Project teams reviewing configuration details often submit exactly this kind of documentation request as part of preparing bid packages for comparison, which is the point at which a manufacturer’s configuration and quotation review process depends on the buyer having supplied enough identifying detail about the required filter configuration, airflow, and test stage to produce a traceable rather than generic response.

Bidder Crosswalks Without Assumed Class Equivalence

Bid packages regularly mix bidders who quote against different classification standards — one bidder citing EN 1822, another citing ISO 29463 — and the comparison problem this creates is structural rather than incidental. These are different classification systems with different defining documents, and a class designation under one does not carry an automatic, assumption-based equivalence to a class designation under the other, even where informal industry crosswalks are commonly discussed.

Where a bidder offers a cross-reference — stating, in effect, that its EN 1822 class corresponds to a particular ISO 29463 class — that cross-reference is a bidder statement. It is the bidder’s claim about equivalence, not an independently verified conversion, and it needs to be recorded and treated as exactly that: a claim requiring its own supporting evidence, not a settled fact that closes the comparison question. The appropriate response is to request the underlying test data behind the cross-reference rather than accepting the stated correspondence and converting the class designation by assumption on the reader’s own initiative.

This distinction matters because the consequence of getting it wrong runs in one direction: treating two classes as equivalent when they are not removes a check that the project’s acceptance plan was relying on, while treating them as provisionally unverified until data arrives costs only the time needed to request that data. Where a project’s specification requires a class under one named standard and the bid quotes a class under a different named standard, the project team needs the bidder’s data, not the bidder’s summary statement, before deciding whether the offered product actually satisfies the specified requirement. A short reference document addressing EN 1822 and ISO 29463 classification in the context of BIBO replacement HEPA filters illustrates the kind of standard-to-standard comparison question that recurs whenever a bid package mixes classification systems, and the same discipline — record the claim, request the data, do not convert by assumption — applies regardless of which two standards are involved.

Comparison Decisions to Record Before Award

Five distinct comparison questions run through a HEPA or ULPA bid review, and each needs its own resolved answer before an award decision treats the bids as actually comparable. None of the five substitutes for another, and leaving one unresolved does not average out against the others being resolved.

The first is whether each quoted class has been tied to its named standard and edition, since a class label without that basis is not a comparable data point at all.

The second is whether overall efficiency, local scan leakage, and installed-system integrity have been kept on separate comparison lines for each bid, since collapsing them into a single reported “performance” figure obscures which acceptance question has actually been answered and which remains open.

The third is whether the filter configuration, tested airflow, test stage, aerosol, and method have been identified for each reported value, since headline numbers are not directly comparable until that basis has been normalized across bids.

The fourth is whether each claimed class has been tied to an applicable filter configuration and a specific referenced report, rather than resting on an unrelated generic certificate that happens to accompany the bid package.

The fifth is whether any cross-standard crosswalk a bidder has offered has been recorded together with the underlying test data behind it, rather than accepted as a settled conversion.

Where a project’s priority is straightforward cost comparison across functionally similar offers, the temptation is to resolve these five questions quickly by assuming similarity; where the project’s priority is defensible acceptance and traceable qualification evidence, each of the five needs its own documented answer, filed against the specific line item it concerns, before the comparison is treated as complete. A related reference on cleanroom HEPA filter leak testing across factory, installation, and requalification stages addresses the evidentiary distinctions that feed directly into the second and fourth of these recorded decisions, since factory-stage filter-element reports and installation-stage integrity evidence answer different questions within the same acceptance plan. The bid comparison record itself — which standard, which edition, which result type, which test condition, which report, which crosswalk — is what a reviewing team, an auditor, or a future requalification effort will consult long after the award decision is made, which is the reason for resolving each line rather than carrying an assumption forward unexamined.

Frequently Asked Questions

Q: Can bids with different HEPA or ULPA class labels stay in the same shortlist?
A: Yes, but only as provisionally comparable bids. Record the standard and edition behind each label, then resolve any difference in test basis before treating the quoted classes or headline efficiencies as equivalent.

Q: What should we ask a bidder to provide for each claimed filter class?
A: Ask for the specific report linked to the offered filter configuration, including the tested airflow, test stage, aerosol, and method. This lets the comparison team check whether the evidence applies to the quoted product rather than relying on a generic certificate.

Q: Can a factory filter-element report define the installed-system acceptance requirement?
A: No. Keep filter-element efficiency, local scan leakage, and installed-system integrity as separate requirements, and state which result will be used at each acceptance stage before award.

Q: How should we handle a bidder’s claimed EN 1822-to-ISO 29463 class cross-reference?
A: Record it as the bidder’s stated crosswalk and request the underlying test data. Do not convert the classes by assumption; compare the reported basis and conditions before deciding whether the offers are comparable.

Q: What should happen if a key comparison basis is still missing at award review?
A: Mark that item as unresolved and avoid using it to declare one bid technically equivalent or superior. Identify the exact missing basis—such as the edition, tested airflow, test stage, aerosol, method, or report linkage—and require it to be resolved before the affected comparison decision is closed.