High efficiency filters

The three families, and what separates them

The eight classes in the table fall into three families, from least to most efficient. What separates them is not a difference in kind but an order of magnitude: from one family to the next, penetration — the fraction that gets through — is divided by a substantial factor. Reasoning in penetration rather than efficiency makes the scale readable: going from ninety-nine point nine five to ninety-nine point nine nine five per cent does not “gain zero point zero four five of a point” — it divides what passes by ten.

Nor is the test point a particulate matter fraction, as it is for general ventilation: it is the most penetrating particle size. Counter-intuitively, the efficiency of a fibrous medium is not monotonic: it is high for large particles, high for very fine ones — which Brownian motion deflects onto the fibres — and it passes through a minimum in between. Classification is done at that minimum, that is, in the worst case. This is what makes these classes demanding and comparable.

Integral and local efficiency

This is the column the table carries and almost nobody explains.

Integral efficiency is an average over the whole face of the filter. Local efficiency is a minimum measured point by point, scanning the face of the element. The standard requires the second only on the most efficient classes, and the first three therefore leave that cell empty — not because it is zero, but because it is not required.

Why the distinction matters in practice: an element can hold its average perfectly and leak at one point. A bonding defect over a few centimetres of pleat, a separator that has migrated, a closed-up pleat — and the element locally passes what its average conceals. On an installation where filtration is a containment function, it is the leak point that counts, not the average.

When a class carrying a local requirement is specified, the manufacturer has scanned the element as it left production. That is information to ask for and to keep.

Installation, sealing, in-place testing

Here is the most useful fact on this page, and it is not in the table: the efficiency of a high efficiency element is lost at installation.

A faultless element, clamped into a frame whose gasket does not bear evenly, lets air pass beside the medium. The leakage flow is small in volume and catastrophic in effect: it is not filtered at all. The performance of the installation is therefore that of the installation work, never that of the element alone, and an element of a higher class fitted in a doubtful frame is worth less than one of a lower class fitted properly.

That is the purpose of the in-place leak test: it measures the assembly as fitted, frame and gasket included, where the production test measures the element. The two do not answer the same question, and the first does not replace the second.

What this implies for supply: frame geometry, gasket type and section, and the clamping system are ordering characteristics just as much as the class is. An element of the right class with the wrong gasket dimensions is not an acceptable compromise.

We supply these elements and pass on the manufacturer’s documentation that accompanies them. The in-place test belongs to the operator and to those they appoint: we claim no standing of our own here.

What this page does not say

It does not say which class you need: that depends on the function sought, on the staging upstream and on the requirements specific to your installation. Nor does it say that a higher class is a better choice — an over-fine stage without adequate pre-filtration clogs quickly, costs more and is replaced more often.