Understanding filtration › 6. Operating and purchasing

CHAPTER 6.2

Setting up a comparative trial

"This cloth lasts twice as long": only a trial on your filter, with your product, can verify it. A well-designed trial costs little; a poorly designed one proves nothing or, worse, leads to the wrong conclusion.

Nicolas Ract · 8 min read · October 2026

Why an on-site trial

Laboratory tests, such as the leaf test or the small bench-top filter press, are used for an initial screening: they measure cake formation time, cake thickness and moisture, filtrate clarity and cake discharge. A technical source points out, however, that laboratory pressure filters only give comparative results, and only if slurry concentration, pressure and filtration time are kept constant [1].

On the industrial filter, other factors come into play: mechanical wear, progressive blinding, plate condition, washing. A study on the selection of filter press cloths in mining thus lists, among the parameters to monitor, final cake moisture, cycle time, solids in the filtrate, number of cycles, and the condition of cloths and plates at the end of the cycle [2]. Only an on-site trial lets you observe everything at once.

1. Set a measurable objective

Before fitting the first cloth, write down what the trial must demonstrate, with a success criterion. For example:

• "the new cloth lasts at least as many cycles as the current cloth";
• "cake dry solids content gains at least one percentage point, at equal cycle time";
• "the filtrate stays within the site's clarity limits from the very first cycles".

The objective follows from what costs the site money (see chapter 6.3) and from the indicator already tracked in the history (see chapter 4.5).

2. A single variable

During the trial, only the cloth changes. Same product, same conditioning (flocculant, coagulant, filter aid), same pressure and same filling programme, same membrane squeeze, same washing. If a process parameter has to change along the way, record it: it may invalidate the comparison.

Field experience: the most common pitfall is to use the trial to "improve" something else at the same time, a pump setting or a new cleaning product. The result then becomes impossible to attribute.

3. Choose the plates and the reference cloths

Two set-ups are possible:

• A few trial plates in a pack of current cloths. The neighbouring cloths act as references, under the same conditions, at the same time. This is the simplest and least risky set-up.
• A complete pack compared against the history of the same filter. More representative for dry solids content and cycle time, which are measured across the whole filter, but more of a commitment and sensitive to product variations from one period to the next.

For positions, avoid the ends of the pack and plates known to fail earlier, unless that is precisely the problem to be solved. Place the reference cloths in comparable positions, ideally right next to the trial cloths. Check the condition of the plates concerned before fitting (see chapter 6.1): a damaged edge would distort the service life.

Field experience: dry solids content and cycle time are hard to read on two or three plates in the middle of a pack, because the filter operates as a whole. For these criteria, measure the cakes from the trial chambers separately or move to a complete pack.

4. Define the duration

The duration depends on the objective. For clarity and cake discharge, a few cycles give a first indication. For blinding and service life, you need to go at least to the end of life of the reference cloths, or to a number of cycles fixed in advance. Express the duration in the process unit (cycles, brews, tonnes) rather than in days.

5. What to measure

For dry solids content, ISO 11465:2025 describes the determination of the dry residue of sludges and other solid matrices by drying at 105 °C [3]; according to AFNOR (the French standards body), its French version replaces the former NF EN 12880 standard on sludges [4]. What matters is that the same method is used for the trial and for the references.

6. Write the protocol and keep the log sheet

The protocol fits on one page: objective and success criterion, cloths compared (full reference, see chapter 6.4), plates and positions, fixed process conditions, measurements and frequencies, duration, responsibilities, criteria for early termination (cloudy leak, premature failure).

Field experience: the log sheet is filled in by the team running the filter, at the time of the reading. A trial whose data are reconstructed after the fact loses most of its value.

7. Draw conclusions

At the end, compare the results against the criterion set at the outset, and against that alone. There are three possible outcomes: the new cloth meets the criterion, it does not, or the trial is inconclusive (confounding variable, missing data, a difference too small relative to process variability). In the last case, it is better to say so and extend or repeat the trial than to draw a conclusion by default.

Finally, translate the result into cost per dry tonne (see chapter 6.3) and add the data to the history (see chapter 4.5).