Filtration is a complex process. The choice made by winemakers regarding filtration requires background knowledge to ensure that the correct filtration is used for a specific purpose. This is especially important prior to the bottling of wine.
The efficiency of liquid filters is indicated by its ß-value (beta value). The ß-test measures the ability of a filter to remove particles of a certain size from the liquid. It is calculated by dividing the number of specific particles entering the filter by the number of the same particles leaving the filter. Mathematically it can be formulated as follows:
ß = Particles prior to filtration ÷ Particles after filtration
The ß-ratio can be used as follows to determine the efficiency of filtration:
(ß – 1 ÷ ß) x 100 = Efficiency
Sterile filtration is a requirement prior to the bottling of wine to prevent potential secondary fermentation and physical precipitations.
Different terminology is used by the suppliers of different filters to indicate their efficiency. This includes amongst others ʻabsoluteʼ, ʻnominalʼ and ʻlog reduction valueʼ (LRV). To qualify for ʻabsoluteʼ, a ß-value of at least 5 000 is required. The term ʻnominalʼ has however no specific meaning, except that it implies approximately. The ʻabsoluteʼ and ʻnominalʼ terminology do not really exist for membrane filters, because the beta measurement applies to specific particles and not for micro-organisms. The latter can contrary to other particles deform themselves to move through the pores of the membrane filter. The suppliers of membrane filters will usually supply technical data, which indicate the performance of a specific membrane regarding the removal of specific micro-organisms. The term ʻfully retentiveʼ (FR) is sometimes used when all the micro-organisms are filtered from a liquid, containing a certain micro-population.
The choice of filtration is mainly determined by the liquid and its composition. A filterability analysis can be done in two different ways. The first one measures the volume of wine, filtered within a certain time and the second one measures the time required to filter a certain volume of wine. The second method is easy, much filtration data are obtained, any membrane can be used and the repeatability is good. Basic equipment is used for the measurement (Theron, 2013). With the second method the filterability index (FI) is determined by recording the time (T) which is required to filter respectively 200 and 400 mL wine with the chosen membrane at a constant pressure of 2 bar. Mathematically it can be formulated as follows:
FI = T400 – 2T200
If a wine is fully filterable, meaning that no blocking of the filter will occur, the FI will be close to 0. The limitation of this measurement is however that 400 mL is a small quantity of wine, which cannot necessarily be extrapolated for bigger volumes. A third measurement of 600 can be added and the FI2 reformulated and calculated as follows:
FI2 = (T600 – T200) – 2(T400 – T200)
In many cases the FI and FI2 values will be close to each other for the same wine. This will indicate that the filterability of the wine can be maintained. If not the case factors like the membrane type, the volume of the wine to be filtered and the wine temperature must be investigated. It is important to note that the filterability index is not a replication of the real filtration prior to bottling. It is only an indication of the membrane filtration of the wine. It is rather a preliminary test to determine whether the filter will be blocked. Practically a membrane with a bigger pore size or a depth filter will be used before the final membrane filtration, which will further limit the possible blocking of the latter. It is also important that the type of membrane, which will be used in practice, is used for the test. The same temperature and a homogenous sample are other factors, which can also play a role.
From a microbiological stability point of view, the addition of concentrate for the sweetening of wine just before bottling makes sense, but it can create filtration problems. If the filterability of the wine is tested before sweetening it may lead to a wrong conclusion. Concentrate is a colloidal solution, which can block filters. Two types of colloids can occur in wine, namely associative colloids and macromolecular colloids. The first is aggregates of small molecules, which occur naturally in wines, but it can also be a result of finings. Macromolecular colloids consist of a single component like polysaccharides, proteins, mannoproteins and carboxymethylcellulose (CMC). These compounds contain a charge and can consequently stabilise the associative colloids. Colloids can cause filtration problems, seeing that its presence will not necessarily be observed by a turbidity reading (NTU), but it can be latent.
It is consequently obvious that absolute standards do not exist for filterability, seeing that different parameters exist, which can only contribute to decision making (Bowyer & Edwards, 2018).
Bowyer, P. & Edwards, G., 2018. Filtration tips: An in-depth discussion. Australian and New Zealand Grapegrower and Winemaker, November 2018: 48 – 56.
Theron, C., 2013. “Die filtreerbaarheid van wyn.” WineLand, July 2013: 76.