A useful HEPA filter certificate links a specific filter to its test conditions and results. Read the serial number, airflow, MPPS basis, overall efficiency and leak-test result together; a large “99.995%” on the page does not do that job by itself.
The worked H14 example below uses illustrative figures. It is not a test result for a supplied filter.
First, identify what document you have
These documents answer different questions:
| Document | What it tells you | What it does not establish on its own |
|---|---|---|
| Product data sheet | The model’s declared class, dimensions, materials and operating data | The result for the particular filter delivered |
| Media test report | Properties of the flat material used to make a filter | Integrity of the completed frame, media pack and their joints |
| Individual filter test report | Recorded results associated with one identified filter element | Integrity of its later installation |
| Installed filter test report | Results for a filter system at its installed location | A replacement for the manufacturer’s original classification record |
| Management-system certificate | The scope of a certified management system | The efficiency or leakage result of an individual filter |
Scroll the table sideways to see all columns →
EN 1822-1:2019 distinguishes media documentation from the documentation normally supplied with H- and U-group filter elements. It also links the report requirements to the relevant test methods. Ask for the record that answers your question, rather than a document simply headed “Certificate.” EN 1822-1:2019, section 8
1. Match the report to the filter before reading the result
Start with the manufacturer, model and serial number. Compare them with the physical label and delivery documentation.
The serial number is especially important when a shipment contains many filters with identical dimensions. Even an H14 report with the right dimensions is the wrong record if it belongs to another unit.
Check the dimensions and configuration as well. Two filters may share width and height but have different frame depths, media areas, sealing arrangements or rated airflows. Keep one record per serial number and link it to the report and final installation location.
If the certificate’s serial number does not match, ask for the correct record. It may be a paperwork error, but it must be resolved before accepting the document as evidence for that filter.
2. Read the standard, edition and method as a set
“Tested to EN 1822” is less useful than a report identifying its classification basis and the method used. The current EN Part 1 is EN 1822-1:2019; the international classification document is ISO 29463-1:2024. BSI publication record, ISO 29463-1:2024
A report may refer to several parts of a standard. Classification, media testing, leak testing and overall-efficiency testing have different roles. EN 1822-1:2019 is used with the corresponding EN ISO 29463 test parts, which replaced EN 1822 Parts 2–5. Check the applicable editions and references; the parts do not all need to have the same publication year. EN 1822-1:2019, foreword
The method also matters. ISO 29463-4 describes the reference scan method and other leakage-test methods for specified HEPA classes. EN 1822-1 does not permit exactly the same routes: its 2019 edition excludes aerosol-photometer scan testing for its classification procedure. “DOP tested” or “leak tested” therefore does not, by itself, demonstrate the particular classification method your order requested. ISO 29463-4:2011, EN 1822-1:2019, sections 7.4–7.5
If the purchase specification requires an individual MPPS particle-counting scan, check for that evidence. Do not accept a different method merely because both reports use the word “leak.”
3. Find the airflow and the pressure drop measured at it
A pressure drop without airflow is an incomplete comparison.
Suppose a report gives 600 m³/h and 100 Pa. The pressure is tied to that test operating point. It is not a promise that the same filter will remain at 100 Pa at another airflow or after months of loading.
Similarly, the efficiency result has defined test conditions. Check whether the test flow matches the classified nominal flow and your specified duty. If the application operates outside the reported conditions, request the relevant performance data rather than extrapolating from a single number.
Keep the acceptance questions separate:
- Does the filter meet the required filtration class?
- Is its pressure drop acceptable for the ventilation design at the required flow?
A filter can satisfy the first and still be unsuitable for the fan or terminal unit.
4. Understand what MPPS tells you
MPPS is the most penetrating particle size for the relevant test conditions. It is the challenging part of the efficiency curve—not the smallest particle the filter can capture.
Find the MPPS or particle-size range used for the measurement, and how it was selected from the media test. A single number such as 0.18 µm may appear, or a report may also identify the counting channel. The laboratory should be able to explain why that size or channel represents the MPPS.
ISO 29463-5:2022 covers determining filter-element efficiency at MPPS, including provisions for sub-0.1 µm MPPS and charged synthetic media. A blanket rule that every HEPA certificate must say “0.3 µm” is therefore wrong. ISO 29463-5:2022
The aerosol name belongs to the method too. DEHS, PAO or another specified aerosol is a test material, not a filter class. An aerosol name alone does not explain particle sizing, instrument response or acceptance limits.
5. Separate overall efficiency from the local leak result
Overall, or integral, efficiency describes the filter element’s combined removal performance in the test. Local testing looks for areas where penetration exceeds the applicable local criterion.
A good overall result can hide a small defective area. The local test checks for that problem. However, the lowest local efficiency can be below the overall minimum and still pass, because the two limits are different.
For the reference MPPS method, H14 has a maximum overall penetration of 0.005% and a maximum local penetration of 0.025%. Do not compare a local result with the overall limit by mistake. VDMA filter classification table
The relationship is simple when both values are percentages:
Efficiency (%) = 100 − penetration (%)
An efficiency of 99.997% corresponds to a penetration of 0.003%, or 0.00003 expressed as a fraction. Writing 0.00003% mixes the two forms and understates the penetration by a factor of 100.
An illustrative H14 result, read correctly
The following is a teaching example. It is not a filter release certificate and makes no claim about an actual Ulpatek product.
| Illustrative field | Illustrative entry | What you should conclude |
|---|---|---|
| Identification | EXAMPLE-001 | Must match a real filter label when used in a real record |
| Required class | H14 to EN 1822-1:2019 | Establishes the classification being assessed |
| Test airflow | 600 m³/h | Defines the operating point for the listed measurements |
| Clean pressure drop | 100 Pa | Must be compared with the agreed resistance at that flow |
| Test basis | Particle counting at an MPPS of 0.18 µm | Identifies the example’s efficiency basis |
| Overall efficiency | 99.997% | Equivalent to 0.003% penetration; below H14’s 0.005% overall maximum |
| Maximum local penetration | 0.010% | Below the reference H14 local maximum of 0.025% |
| Local result expressed as efficiency | 99.990% minimum | Consistent with 0.010% maximum local penetration |
| Scan conclusion | No local failure detected under the stated procedure | Must be supported by the actual scan coverage and test record |
Scroll the table sideways to see all columns →
The two numerical performance results are compatible with their respective H14 limits. The local efficiency of 99.990% does not fail simply because it is lower than 99.995%: that second number is the overall requirement.
The numbers show how to compare each result with the correct limit. A real certificate also needs the full test conditions, equipment information and required records.
6. Treat the scan graphic as data, not decoration
A scan report may show tracks, coordinates, counts or a map of the downstream face. Find the legend, units, acceptance threshold and orientation before interpreting the colours.
A coloured spot might show a relatively high count within an acceptable range. It might show a threshold exceedance. Without the legend and result criteria, you cannot know which.
Keep the axes, test conditions and filter identification with the scan plot. Ask for the report explaining the measurement and acceptance criteria before interpreting the graphic.
Also look for the test date, report identifier, equipment identification and the person or system authorising the result. Some traceability and calibration information may be held in supporting laboratory records rather than printed on a one-page certificate. Ask for the supporting record when needed. It does not all have to fit on the certificate’s front page.
The certificate ends where the installation begins
The factory report belongs with the filter throughout delivery and installation. Record its final location and connect it to the installed test record.
Neither document replaces the other. IEST’s current leak-test practice covers testing at manufacture, before installation and after installation precisely because these are distinct stages. IEST-RP-CC034.5
Frequently asked questions
Is 99.995% on a certificate necessarily a measured result?
No. It may be the target or minimum for H14. Find the fields identifying the measured result and acceptance limit separately. A certificate that repeats only the class minimum does not, by that number alone, show the filter’s actual measurement.
Does “leak-free” mean zero particles pass through the filter?
No. It means the filter met the leakage criterion of the stated test. HEPA classes allow finite penetration. Ask which method and acceptance limit support the conclusion; “leak-free” is not a claim of absolute zero penetration.
Can I compare a site photometer percentage directly with the factory MPPS percentage?
No. The aerosol, instrument response, sampling and acceptance basis can differ. Use each result with its own method and limits. A site result cannot be interpreted by simply subtracting it from the factory efficiency printed on the certificate.
Why does an E12 filter not have the same scan certificate?
E12 belongs to the EPA group, whose classification does not use the HEPA local scan requirement. Request its applicable classification and performance documentation rather than assuming every filter class must arrive with the same report format.
