Measurement of cold stability of wine

Grape juice naturally contains potassium and tartaric acid, which associate together to form the salt potassium hydrogen tartrate (KHT). KHT is soluble in grape juice but less so in ethanol. After fermentation, wine becomes saturated with KHT and it precipitates out of solution. KHT solubility decreases even further at low temperatures and thus if an unstable wine is bottled and then chilled, crystals can form in the bottle. A test for cold stability gives an indication of the likelihood that precipitation of KHT will occur after bottling.

This page gives a brief summary of some commonly used techniques to test cold stability in wines and methods for cold stabilisation in the cellar.

Refrigeration/brine test

This test involves holding a sample of filtered wine at -4°C for three days and then inspecting for any sign of crystalline tartrate precipitation. If crystalline deposits are observed to redissolve upon warming to ambient temperature, the wine is generally interpreted as being cold stable, but the presence of any persistent crystalline precipitate is taken to indicate that the wine is cold unstable. The results should be expressed as a ‘pass’ if there is no permanent crystalline deposit after refrigeration, or ‘fail’ if crystals are present.

The Affinity Labs cold stability test results report three different levels of ‘fail’, including:

Fail – Level 1: Borderline fail (< 10 small crystals)

Fail – Level 2: Bad fail (> 10 small and/or larger crystals)

Fail – Level 3: Crystals visible to the naked eye

Note that some phenolic material, such as amorphous pigments and tannins, might form a transient precipitate in some red wines. Allowing the sample to warm up to ambient temperature usually results in their resolubilisation, which differentiates these precipitates from potassium bitartrate crystals. This test gives a picture of a wine’s current stability.

Equipment: Water bath containing brine (-4°C), membrane filtration apparatus
Reagents: None
Services: Wash up area, power
Space required: Bench space for water bath


This method is a variant on the refrigeration test, but is believed by some industry practitioners to provide less precise results (Wilkes, personal communication, 2004). A sample of filtered wine is frozen for a set period of time (e.g. overnight) and then thawed before being inspected for any sign of crystalline tartrate precipitation. If crystalline deposits are observed to redissolve upon warming to ambient temperature, the wine is generally interpreted as being cold stable, but the presence of any persistent crystalline precipitate is taken to indicate that the wine is cold unstable. It should be noted that many industry practitioners feel that this test is too harsh and have found that freezing until a slurry is formed and then thawing is more appropriate (Wilkes, personal communication, 2004). The AWRI does not recommend using this test as a final test of cold stability.

Equipment: Freezer (domestic is suitable), membrane filtration apparatus
Reagents: None
Services: Wash up area, power
Space required: Minimal bench space

Conductivity (contact process)

A stirred sample of filtered wine is held at a specific temperature (0°C for white and 5°C for red wines) and the change in conductivity is observed following addition of potassium bitartrate seed crystals (1 g/L) ; (Zoecklein et al. 1989). If the change in conductivity of the sample is 5% or more (due to precipitation of KHT and a resulting decrease in conductivity), the sample is considered to be unstable with respect to KHT. It should be noted that some industry practitioners vary the temperature at which this test is conducted (e.g. within the range of -4 to 5°C) to reflect what they believe is more relevant to their wine styles (Wilkes, personal communication, 2004). The chosen test temperature should be validated to improve confidence in the results.

Equipment: Conductivity meter (100-1000 mS/cm), water bath, magnetic stirrer, membrane filtration apparatus, thermometer
Reagents: Potassium bitartrate (cream of tartar)
Services: Wash up area
Space required: Bench space for water bath and conductivity meter

Concentration product

The stability of wine with respect to KHT precipitation can be predicted on the basis of the concentration product (CP). This is calculated from the concentrations of potassium, tartaric acid (total tartrate), pH and alcohol in the wine, using the following formula:
CPKHT = ((Potassium, mol/L) x (Tartaric acid, mol/L) x (% bitartrate ion))/100
= ((Potassium, mg/L)/39.1 x 1000) x ((Tartaric acid, g/L)/150.1) x ((% bitartrate ion)/100)

The % bitartrate ion, i.e. the percent of tartaric acid present in the bitartrate form, can be obtained from the pH and alcoholic strength of the wine, using the tables derived by Berg and Keefer (1958). The CP can also be calculated from potassium, tartaric acid, pH and alcoholic strength using a ‘Miller Dial’ sliding scale KHT stability calculator. The CPKHT value thus obtained is compared with historical data for the appropriate wine type. Some alternative criteria for establishing tartrate stability have been proposed:

  • DeSoto and Yamada (1963) suggest safe CPKHT values at 0°C for nine different wine types; if the CPKHT of a sample exceeds the recommended safe level for that wine type, then the sample can be considered to be cold unstable.
  • Berg and Akiyoshi (1971) suggest a maximum CPKHT value of 9.4 x 10-5 for white wines, and a maximum CPKHT value of 17.6 x 10-5 for red wines; wines with a CPKHT value greater than these maxima can be considered unstable.
  • Leske et al. (1996) suggest a maximum CPKHT value of 8.0 x 10-5 for white wines, and a maximum CPKHT value of 18.0 x 10-5 for red wines; wines with a CPKHT value greater than these maxima can be considered to be unstable.

Equipment: Analytical capability for the determinations of potassium, tartaric acid, pH, and alcohol
Reagents: As required for the individual analyses
Services: As required for the individual analyses
Space required: As required for the individual analyses

Saturation point (Tsat)

The methods outlined above test current wine stability. As a wine ages, the nature and amount of compounds naturally present in wine that inhibit crystallisation can change, which also changes the wine’s ability to resist the precipitation of tartrates. Consequently, a wine which may have been formally stable by the ‘brine test’ at one point in its life can become unstable at a later date. This is not a reflection of the reliability of the test, but rather the changing nature of wine. Such changes are more prominent in young whites and red wines, but they can still happen in mature white wines.

A newer cold stability analysis called the saturation point test (Tsat) may be used to define if a wine is likely to become unstable in the future (Ribereau-Gayon et al. 2006). This method essentially uses conductivity to measure the amount of tartrate that can be absorbed by a wine at room temperature. The change in conductivity is then used to determine the theoretical temperature at which the wine will precipitate tartrates (Erbsloh EasyKristaTest). A low saturation temperature suggests a more stable wine (i.e. it has to be brought down to a low temperature to precipitate tartrates) while a high saturation temperature suggests that precipitation will happen much more easily. A fail by the Tsat test does not guarantee the wine will throw a deposit under normal conditions, rather, it demonstrates that the wine has the potential to be unstable. The table below summarises the interpretation of the Tsat test.

KHT stability White and rosé wine Red wine
Stable <12 <15
Unstable 12-16 15-20
Very unstable 16-20 20
Extremely unstable 20

Cold stabilisation in the cellar

Wine can be cold stabilised by keeping a wine chilled for several weeks. This is quite slow and energy-intensive and thus wineries often use the same process but with added seed crystals, called the contact seeding process. This encourages rapid KHT drop-out in hours to a few days, with the massive quantities and large surface area of smaller KHT crystals added providing multiple crystallisation nuclei sites.

A common error in cold stabilisation is that the wine is allowed to warm up before it is racked or filtered off the precipitated tartrate crystals. This allows some KHT to re-dissolve into the wine, making the wine only stable to the temperature it was filtered at. If any additions, acid adjustment or blending occurs after cold stabilisation, then the stability will be affected and the wine will need to be rechecked for cold stability. Ideally the final blend going to bottling should be checked just before being bottled.

Some winemakers use electrodialysis to remove potassium and tartrate ions from wine as a method of cold stabilisation.

Crystallisation inhibitors such as metatartaric acid, mannoproteins or carboxylmethylcellulose (CMC) can also be used to prevent crystal formation These are rarely used without some form of prior cold stabilisation, and are often used as an extra guarantee of stability for higher risk wines.

References and further reading

  • Amerine, M.A., Ough, C.S. 1980. Methods for analysis of musts and wines. New York: Wiley-Interscience: 341 p.
  • Berg, H.W. 1960. Stabilisation studies on Spanish sherry and on factors influencing KHT precipitation. J. Enol. Vitic. 11: 123-128.
  • Berg, H.W., Akiyoshi, M. 1971. The utility of potassium bitartrate concentration product values in wine processing. J. Enol. Vitic. 22(3): 127-134.
  • Berg, H.W., Keefer, R.M. 1958. Analytical determination of tartrate stability in wine. 1. Potassium bitartrate. J. Enol.9: 180-183.
  • Bertrand, G.L., Carroll, W.R., Foltyn, E.M. 1978. Tartrate stability of wines. 1. Potassium complexes with pigments, sulphate and tartrate ions. J. Enol. Vitic. 29(1): 25-29.
  • De Soto, R.T., Yamada, H. 1963. Relationship of solubility products to long range tartrate stability. J. Enol. 14: 43-51.
  • Ewart, A.J.W. 1984. A study of cold stability of Australian white table wines. Grapegrower Winemaker 244: 104-107.
  • Iland, P., Ewart, A., Sitters, J., Markides, A., Bruer, N. 2000. Techniques for chemical analysis and quality monitoring during winemaking. Campbelltown, SA: Patrick Iland Wine Promotions: 111 p.
  • Leske, P.A., Bruer, N.G.C., Coulter, A.D. 1996. Potassium tartrate – how stable is stable? Stockley, C.S., Sas, A.N., Johnstone, R.S., Lee, T.H. (eds) Proceedings of the ninth Australian wine industry technical conference; 16-19 July 1995. Adelaide, SA: Winetitles: 39-45.
  • Pilone, B.F., Berg, H.W. 1965. Some factors affecting tartrate stability in wine. J. Enol. Vitic. 16(4): 195-211.
  • Rankine, B.C. 1995. Making good wine: a manual of winemaking practice for Australia and New Zealand. Sydney: Pan Macmillan Australia Pty Ltd: 374 p.
  • Ribéreau-Gayon, P., Glories, Y., Maujean, A, Dubourdieu, D. 2006 Handbook of Enology Second Edition Volume 2: The Chemistry of Wine Stabilisation and Treatments. Chichester, UK: John Wiley & Sons Ltd: 404 p.
  • Wilkes, E. 2004. Group Manager – Commercial Services, AWRI, personal communication.
  • Zoecklein, B.W., Fugelsang, K.C., Gump, B.H., Nury, F.S. 1989. Production Wine Analysis. New York: AVI Van Nostrand Reinhold: 291-305.
  • Zoecklein, B.W., Fugelsang, K.C., Gump, B.H., Nury, F.S. 1995. Wine analysis and production. New York: Chapman & Hall: 621 p.