Lets be clear about filtration

Introduction

Filtration of wine is normally used as an important step in the final stages of winemaking to ensure clear and stable wine. Discussion of the types and sizes of microbes that can cause spoilage are included to demonstrate one of the main reasons for employing effective filtration. Some of the aspects of filtration are discussed below in an attempt to show where problems can lie.

Why use filtration?

All wines are clarified to some degree by settling and racking. Most, but not all wines, are filtered. Whether or not to filter or is a choice made by the winemaker and depends upon the wine style they are aiming for, their personal preferences and experiences.

Filtration is a general description for a number of operations and types of equipment. We will not go into the wide variety in use, but restrict the discussion to why filtration is used and what problems can occur if it is not performed correctly.

The main object of filtering is to provide a wine that is clear and devoid of any particulate matter visible to the naked eye at the time of bottling (1,2). Stabilising wine in a number of ways, e.g. protein and tartrate stabilising, also adds to this outcome.

Another very important object of filtration is to help produce a wine that is stable over time, particularly in regard to the presence of microbes that can alter the look and taste of the wine after bottling.

Microbes and their sizes

The main groups of problem microbes in wine are yeasts and bacteria.

The yeasts include:

Brettanomyces species
Kloeckera apiculata
Saccharomyces species
Zygosaccharomyces bailii

Common bacteria include:

Acetobacter species
Lactobacillus species
Oenococcus oeni
Pediococcus species

Yeasts are typically in the range 5-10 micron (µm). Bacteria are typically smaller, for example cocci vary from 0.5 to 3 µm in diameter, bacilli diameters range from 0.2 to 2 µm (3).
Therefore the choice of filters does affect which microbes are eliminated. All yeast will be removed by 1 µm filters, however bacteria will not be. This is somewhat of a simplification, but gives a general guideline to filter selection and its impacts.

Filter pore size

If you know the size of the microbe then it should be an easy task to know which filter to use to remove the particular microbe. However, it doesn’t appear that it is quite so easy in practise. Filters for wine are usually given a notional pore size but they are not necessarily part of the product description. For example one of the most commonly used brand of filters has 3 popular products that are normally referred to as 20s, 30s or 60s. The 20s having a nominal pore size of 1.5 µm, 30s have 1.0 µm and the 60s nominal pore size is 0.6 µm. This nomenclature, although counterintuitive, seems well accepted by some users. However, it would be easier and more helpful if the pore sizes were used in the primary description by this and other manufacturers.

Sterile filtration

In microbiology the accepted definition of sterile filtration is that the filtration is done using 0.2 µm filters. At this level of filtration all bacteria and yeast are considered to be removed.

It seems that 0.2 µm filters are rarely, if ever used in the wine industry, although the use of 0.45 µm filters is quite common. Most winemakers talk about sterile filtration using 0.45 µm filters and yet this does not meet the microbiological definition. Indeed there is no guarantee that all bacteria are removed at this level of filtration. Bacteria that can cause faults in wine such as Pediococcus and Acetobacter are therefore quite likely to remain in the wine after filtration and bottling.

This is not to say that all wine should necessarily be filtered down to 0.2 µm. Merely that we face facts and don’t fool ourselves into thinking that what we in the Australian wine industry call “sterile filtration” does cause sterile conditions, as it does not necessarily follow. Of course a post bottling sterility check can be performed after bottling to confirm the wine is sterile. In fact we perform this test routinely for a number of clients.

Many, if not most, winemakers have a great concern that filtration down to these sort of levels removes much of the positive attributes from the wine and so are very reluctant to do so unless they have a specific need.

How can we limit the need for filtration?
The underlying stability of a wine can also be influenced by the wine’s composition. One of the main purposes of filtration is to remove any microbes that can cause problems. If the substrates that the microbes rely upon for growth are reduced to a low level, the potential problems are of less risk. For example the main substrates of concern are malic acid, glucose and fructose. If the levels of these compounds are kept below 0.05 g/L each in the finished wine then the microbes are very much inhibited in their ability to grow, reproduce and spoil the wine.

The pH of the wine is another very important parameter that must be managed carefully. A number of the microbes listed above grow more easily in wines with a relatively high pH – a wine of pH 3.8 has a much greater risk of microbial growth than one of pH 3.3.

Allied with the pH is the sulphur dioxide level. I have written previously about the level of this preservative that should be used. Several recommendations of around 0.6 to 0.8 ppm molecular sulphur dioxide can be quoted. This is a very effective way to reduce the potential for bacterial spoilage and in combination with the other parameters mentioned, it substantially limits the risk of spoilage. Combining these compositional criteria with effective filtration is the safest way to ensure long-term stability in bottled wine.

It is not uncommon for us to see wines that have been bottled after filtering and a deposit has formed that was not there at bottling. One serious case involved a deposit and a fizziness in supposedly still but slightly sweet wine. This was caused by the presence of Saccharomyces cerevisiae yeast that caused fermentation in the bottle due to lack of proper filtration and the absence of careful monitoring of some of the critical compositional levels. Winemakers who have experienced this nasty re-fermentation problem soon realise that Saccaromyces cerevisiae, although useful for primary fermentation, is not always the friendliest yeast in the world of winemaking.

Conclusion

Filtration can be used to remove microbes from wine that may cause faults. Knowing the size of the microbes helps in determining the correct filter that should be used. The pore size of the filter used should be known so that the effectiveness of the filtration process can be judged. Some of the compositional parameters of wine should be closely managed to also help protect against microbial spoilage if problems are encountered during the filtration operation.

References

1. Boulton, R. B; et al, Principles and Practices of Winemaking, Springer, New York, 1996, pp 293-315

2. Ough C. S; Winemaking Basics, The Haworth Press, New York, 1992, pp154-167

3. Talaro, K. P; Foundations in Microbiology, Basic Principles, 6th edition, McGraw Hill, New York, 2008, p110

Article from the Australian and New Zealand Grapegrower and Winemaker, Issue 538, 2008
Page Number(s): 108-112

Author: Greg Howell