Technical resource
Air filtration classes
This page serves two readers: the one who needs to know what an efficiency class actually measures, and the one holding a nameplate that carries a class no longer to be found anywhere.
The four efficiency groups
| Group | Minimum required efficiency | What the group measures |
|---|---|---|
| ISO Coarse | Coarse filter: less than half of the PM10 fraction is retained. It is characterised by its gravimetric arrestance, not by an ePM efficiency. | |
| ePM10 | 50% | Retains at least half of the particles up to 10 micrometres in diameter. |
| ePM2.5 | 50% | Retains at least half of the particles up to 2.5 micrometres in diameter. |
| ePM1 | 50% | Retains at least half of the particles up to 1 micrometre in diameter — the fraction that penetrates deepest into the respiratory tract. |
The rating is expressed in steps of 5%, from 50% to 95%: an intermediate efficiency is not a class.
What the classification measures
The standard in force does not rate filters on a single test aerosol: it rates them on their efficiency at retaining real particulate matter fractions, the ones measured in ambient air and used in the health literature. Each group designates a fraction, and the efficiency quoted applies to that fraction.
One point of method decides everything else, and it is the one most often forgotten: the test is carried out after the medium has been electrostatically discharged. Many synthetic media owe a large share of their initial efficiency to an electrostatic charge, which dissipates in service — on contact with moisture, fine dusts, oily aerosols. Measuring after discharge therefore measures what the filter will do over time rather than what it does on the first day.
It is also why a class from the withdrawn standard does not convert into a current class: the two do not measure the same thing, and the second deliberately removes an effect the first took advantage of.
How to read the table above
Three readings, in this order.
The group says which fraction of particles is targeted. The finer the fraction, the more the requirement bears on particles that penetrate deep into the respiratory tract.
The percentage is a minimum required efficiency on that fraction, not a nominal performance. An element quoted at that level does at least that; it may do better, and the manufacturer’s sheet says so.
The stepped rating, finally: the scale is not continuous. An efficiency between two steps is not a class, and a manufacturer quoting it as one would be describing a measurement, not a rating.
One thing the table does not say, and it needs saying: a coarse filter is not a bad filter. It is a filter for another stage. Placed at the head of a filtration train, it retains the load that would have shortened the life of the fine stage, and that is precisely its job. Judging a class outside its stage makes no sense.
I have a nameplate with a class I cannot find
This is the commonest case, and it has an honest answer.
The standard that carried those classes is withdrawn. It no longer rates any new filter, and manufacturers no longer publish products under its designations. An element installed ten years ago therefore legitimately carries a class that no longer exists in any catalogue.
The historical correspondence table at the foot of this page lets you situate that class against the current groups. Read the caveat that goes with it: this is not a conversion. Nor is it a basis for ordering. It is an orientation for discussion, so that we speak of the same order of magnitude while waiting for the only right answer — a requalification against your process.
In practice, to replace an element carrying a class from the withdrawn standard, what matters is: the stage it sits at, the air quality expected downstream, the frame dimensions, and the airflow per element. The class on the plate is a clue, not a specification.
What this page does not say
It does not give the class your process needs, and no web page can do that honestly. That choice depends on the air quality sought, on the air quality available at the inlet, on the staging already in place, on the pressure drop the unit can take and on the replacement frequency you can accept. Two neighbouring installations on one site do not call for the same class.
Nor does it give equivalences between manufacturers: two elements quoted in the same group may differ in filter area, initial pressure drop and service life, which are the quantities you actually pay for.
Historical correspondence, indicative only
Drawn between EN 779:2012, withdrawn, and ISO 16890-1:2016, in force. The withdrawn standard no longer rates any new filter: this table is for reading an older document, never for selecting an element.
Indicative only. The former standard is withdrawn and its classes are not convertible: only a requalification against the process gives the correct ISO 16890 class.
| Class of the former standard | Indicative only |
|---|---|
| G1 | ISO Coarse |
| G2 | ISO Coarse |
| G3 | ISO Coarse |
| G4 | ISO Coarse |
| M5 | ePM10 |
| M6 | ePM10 |
| F7 | ePM1 |
| F8 | ePM1 |
| F9 | ePM1 |
Other resources
155 route des 3 Lucs à la Valentine
13011 Marseille