Filtration is often understood as a binary state: either a substance is removed or it is not. This view is technically incorrect. In practice, filtration works gradually. Retention rates change over time, depending on the medium, flow rate, contact time, and loading. To realistically assess filtration, one must understand this dynamic.
In process engineering, the performance of a filter is not described by a yes-no criterion, but by separation or retention rates. These indicate what proportion of a substance is retained under defined conditions. Even highly efficient filters do not show complete elimination, but rather a probability distribution of separation across particle sizes or substance classes [World Health Organization, Water safety and treatment processes, https://www.who.int/publications/i/item/WHO-FWC-WSH-17.05 ]
A central concept is breakthrough. With increasing use, active surfaces or pore spaces become saturated. Filter performance does not decrease abruptly but gradually deteriorates. In water treatment, this behavior is described by breakthrough curves, which show at what point a substance increasingly appears in the filtrate [US Environmental Protection Agency, Granular activated carbon treatment, https://www.epa.gov/water-research/granular-activated-carbon-treatment]
This dynamic explains why a filter may appear to "work" even though its protective effect is already reduced. The water flow remains stable, as it primarily depends on hydraulic resistance. Separation, however, is a question of available active surfaces and sufficient contact time. If these conditions are no longer met, the bypass gradually increases.
This principle is particularly relevant for activated carbon and adsorptive media. Studies show that organic trace substances are initially removed very efficiently, while the retention rate continuously decreases with increasing loading. This effect is highly substance-dependent and can vary greatly depending on molecular size and polarity [European Commission, Best available techniques for water treatment, https://eippcb.jrc.ec.europa.eu/reference ]
For everyday life, this means: statements like "filters X percent" are always snapshots under certain conditions. Without knowledge of the breakthrough dynamics and the real load, such values cannot be transferred to the entire service life. Filtration is therefore not a static state, but a time-dependent process.
A realistic understanding of filtration protects against false expectations. The decisive factor is not whether a filter "still lets water through," but how stable its retention rates remain over time. This is precisely the difference between nominal performance and actual protection.
Filtration is not a switch.
It's a process.
More on sydros.de
