Drew Horton, Enology Specialist
Matt Clark, Assistant Professor
No doubt about it, one of the worst things that can happen with a wine is re-fermentation in the bottle. Instead of a beautifully-colored wine with brilliant clarity, you discover a once-perfect wine has become cloudy and bubbly, or even worse, numerous "bottle bombs" are exploding in the cellar or on winery or retail shelves, or at the least, corks are pushing out of the bottle. This is a disaster.
Whenever there is any amount of unfermented sugar in a bottle of wine, the wine risks re-fermenting. This article is really about sweet white or rosé wines, but also applies to any non-dry red wine as well.
There are a variety of methods, techniques and additions that can help eliminate or reduce this risk.
Firstly, overall winery sanitation is the most important. Each and every surface; tank, pump, hose, valve, bottle-filler, or corker that comes into contact with wine must be clean and sanitized. There are many ways to sanitize, but none of them will work unless the surface is first cleaned. Cleaning is the removal of dirt or other residue from a surface, sanitizing is the killing of all microbes on a surface. A wine maker needs to do both when handling any sweet wine.
When preparing a wine for bottling, the first step after blending (if any blending is to be done) is to treat the wine with bentonite, which removes excess protein and keeps the wine heat stable. A complete and exhaustive article on the use of bentonite is available on the internet, posted by Purdue University Extension: https://www.extension.purdue.edu/extmedia/fs/fs-53-w.pdf
After the bentonite treatment, some method of achieving cold-stability must also be done. Traditional cold-contact treatments are the norm, but they do require the wine be super-chilled. This means bringing the wine close to, or just at, freezing for a period of 5 to 15 days, depending on the amount of excess tartaric acid in the wine, among other variables. It should be noted that there are new preparations of manno-proteins and cellulose that can provide almost instant cold-stability, but they lack the often-beneficial side effect of traditional cold-contact stabilization which can reduce the overall total acidity in a wine by precipitating excess tartaric acid in the form of crystals. See this link to the Laffort website for alternate cold-stabilization methods and products: http://laffort.com/en/products/stabilisation
Electrical conductivity tests can be done by a laboratory to confirm that cold-stability has been achieved. In the Midwest, Iowa State University's Midwest Grape & Wine Industry Institute can conduct this testing quickly and for a small fee, see this link to their Lab services order form: http://www.extension.iastate.edu/wine/sites/www.extension.iastate.edu/files/wine/PriceList-May%202016.pdf
The addition of a sweetener (sugar, sweet reserve juice, or concentrate) should happen late in the wine preparation process, usually between a nominal rough and polish filtration step is best. Any extra time a wine is in storage with any sugar it risks re-starting fermentation. During the aging and storage period for a wine with existing natural residual sugar, you are advised to pre-filter the wine and keep the wine at a cold-enough temperature (45F degrees or less) and with sufficient free sulfur (based on wine pH) in order to inhibit re-fermentation.
So, with this now blended, heat and cold stabilized wine, you are ready to start final filtration and any final sweetening. Keep in mind when discussing filtration that there are both "nominal" and "absolute" filtration equipment and techniques. Nominal filtrations are achieved by the use of sheet or pad filtration, sometimes used with DE (diatomaceous earth), in a plate and frame type of filter. Cross-flow filtration, though it does not require DE or filter sheets, is also to be considered a nominal filtration and not "absolute".
Nominal filtration (sometimes called depth filtration) will remove 99.9% of bacteria and yeast cells. Only absolute (aka sterile membrane) filtration will keep the wine 100% free of bacteria and yeast cells. Sterile membrane filter cartridges and housings are not inexpensive, and they must be integrity-tested both before and after use. If a wine is not pre-filtered well, membranes will plug quickly. Here is a link to the bubble-point method to test sterile cartridge membrane integrity: http://www.scottlab.com/uploads/documents/downloads/25/ScottCart%20Bubble%20Point%20Test%20.pdf
In lieu of the use of absolute sterile membrane filtration, and the use of strict chemical or heat sterilization of all bottling/capping/corking equipment, you can consider the use of a yeast fermentation inhibitor such as Sorbic acid, usually used in the form of powdered Potassium Sorbate. The big issue with sorbate is that it does not guarantee stabilization, and it must be used properly and in conjunction with proper levels of free sulfur. Also, if used in large amounts, or in a wine that may undergo spontaneous malo-lactic fermentation, there is a risk of "geranium-smell taint." Also, above a certain level, some consumers will notice the smell and taste of sorbate. Here is a link to an article on the nature and proper use of sorbate : http://www.extension.iastate.edu/wine/sites/www.extension.iastate.edu/files/wine/SorbicAcid1.pdf
If you are able to produce a sweet wine with alcohol at or above 13% and a pH below 3.5, then sorbate may be an option to help stabilize the wine without the use of absolute sterile membrane filtration, and without the risk of geranium taint. A low alcohol wine (say 11% or below) with high pH (above 3.6) and any amount of residual sugar is virtually impossible to stabilize and guarantee stability without the use of a high-speed bottling line utilizing integrity-tested absolute sterile membrane filter equipment.
Keep in mind, though, that an absolute sterile membrane is only a doorway, with dirty on one side and sterile on the other. Any downstream equipment (i.e., post-membrane), such as valves, hoses and bottle-filler that are not also completely sterile, can allow even a few bacteria or yeast cells to enter the bottle, and can lead to re-fermentation. If you have ever spent a few hours, or days, "de-corking" and attempting to re-treat and re-bottle a wine that has become unstable or is re-fermenting, you will take the preceding to heart.