The value of cold soaks for red winemaking; the value of cold soak research for winemakers

Cold soaking seems to be an especially divisive winemaking technique, at least in the Pacific Northwest, and that’s saying something in an industry full of strong personalities. Cold soakers say that allowing crushed red grapes to rest for one to several days in an environment too cold for Saccharomyces activity, before warming everything up to yeast-pleasing temperatures and allowing fermentation to begin in earnest, deepens color and augments flavor and tannin extraction. The anti-cold soak camp claims that these benefits aren’t real and sometimes adds that cold soaks allow for the dangerous possibility — dangerous, that is, if you’re also in the anti-spontaneous ferment camp — of illicit microbial growth before winemakers inoculate commercial yeast strains at the soak’s end.

Research to date has been unhelpfully mixed. Some studies show increased phenolic (color and/or tannin) extraction, some don’t, some even show lower phenolics following cold soak, and the variables responsible for the differences haven’t yet been worked out. Adding to the confusion is the inevitable mess that follows pro-spontaneous from anti-spontaneous fermenters, since the non-Saccharomyces activity that might occur during cold soaks is a source of desirable complexity to some and unconscionable spoilage to others.

I would love to say “until now” and herald the arrival of a brilliant, conclusive paper outlining a robust explanation for how and why and where and when cold soak works. My inability to do so isn’t likely to come as a surprise. Nevertheless, there is new research and, while far from once-and-for-all conclusive, it helps, if perhaps not in the expected way. A new study from an Argentinian team* tested cold soak on cabernet sauvignon, merlot, syrah, pinot noir, malbec, and barbera d’asti, looking for differences both when the wines were pressed and after a year of bottle aging. Cold-soaked wines saw four days of 6.5-11.5ºC (44-52ºF) courtesy of periodic dry ice additions, then 10-day fermentations at 21.5-26.5ºC (71-80ºF); control wines went straight to 14-day fermentations. All varieties were made in the same way: same full twice-daily pump-overs, same twice-daily punch downs. All were inoculated with the same commercial yeast strain five hours after crush. Regrettably, the study didn’t include multiple variations on the cold soak theme — different times, temperatures, or techniques — that might have helped to suss out where any cold soak differences are happening and given much more information to winemakers. In particular, it’s important to emphasize that chilling with dry ice meant as much as a 10ºC (18ºF) difference in temperature between different parts of the tank because the dry ice clumped. Jacketed tanks would have applied a more uniform treatment.

The agglomerated results were straightforward enough. Cold soaks increased color density, but didn’t increase phenol or tannin concentrations. Cold soaking also didn’t make a statistical difference to any basic wine chemistry parameters: ethanol concentration, pH, acids, glycerol, and residual sugar. Tasters found that the most important difference between all of the wines was driven by grape variety, though that’s hardly meaningful and says nothing about cold soak. That’s the big picture.

The details in the supplemental data attached to the main paper show something more interesting. Each variety responded a bit differently to the cold soak treatment. In the barbera and the syrah, tannin concentrations actually were higher in the cold-soaked wines. The opposite was true for the pinot noir, where cold-soaked wines measured tannin concentrations statistically significantly lower than the control. Cold soaking related to increased total phenols in cabernet, decreased in pinot noir.

What this says to me is that we’re measuring the wrong construct at the wrong level of detail. Asking whether “cold soak” works seems to be the wrong question. Instead, we need to be testing out different potential cold soaking parameters in specific grape varieties to identify what precisely makes a difference and what is moot. This is the kind of data that could really help winemakers who through the lens of their communal experience are saying that cold soak sometimes makes a noticeable positive difference and sometimes doesn’t, and who might reasonably look to science to help them figure out what features separate the worthwhile instances from the useless ones. Unfortunately, if the research question continues to be “Does cold soak increase phenol concentrations?” instead of “Under what conditions does cold soak make a difference to phenol concentrations?” we’re likely to continue seeing confused yes-no-or-maybe reports instead of useful, applicable explanations of what winemakers seem to observe.


*Including Federico Casassa, who has in the past published excellent phenol-related research with James Harbertson at Washington State University, including the American Society of Enology and Viticulture’s 2014 Best Enology Paper of the Year, on the phenolic effects of extended maceration and regulated deficit irrigation, the full text of which is freely available here.

6 thoughts on “The value of cold soaks for red winemaking; the value of cold soak research for winemakers

  1. Having done detailed analyses myself I know the parameter most important to those of us in cooler climates is the degree of polymerisation of anthocyanins. Hence the noted colour difference outside any major differences in tannin or total anthocyanin concentrations.
    An interesting experiment I also did was to add a “sacrificial” inoculum of Saccharomyces at receival along with a decent dose of SO2 (good for several reasons including protecting the anthocyanins from oxidation prior to polymerisation) prior to cold soak… The yeast struggles and produces aldehyde which then releases the SO2 and anthocyanin at a point when the appropriate tannins are present to polymerise the anthocyanin. The rate of colour increase (and measured polymerisation to verify) was astounding. Interestingly though the results after six months were not statistically different which showed that neither the evolution of tannins nor the amount of anthocyanin present were limiting in our case. In another region or vintage, or by modifying the ferment regime (for speed or cost perhaps), this may very well not be the case and the technique could have enormous value.



    • Thanks for the additions, Dave. Much good research is out there, in and outside of the published literature, but it’s hard to find coherence amongst it all when most plays more with gross results than underlying principles.

  2. Good point about the wrong type of question being asked. The move to questions about what are the drivers of cold soak is a much better one. Time to break out DOE and chemometrics approaches.

    The only other thing this research highlights is the massive divergence in experimental methods. Reading Red wine papers always makes me want to shake the researcher and say “why have you done that”. It seems to me in this paper the cap management regime is on the more aggressive side of the spectrum and 10 days of 4 times daily manipulation in ethanol environment will tend to obscure many subtle results.

  3. Pingback: The value of cold soaks for red winemaking; the value of cold soak research for winemakers | The Wine-o-scope | phen•ology•vin•ology

  4. Pingback: Cold Soaks and Color Extraction: My Observations | Nova Cadamatre

Leave a Reply

Your email address will not be published. Required fields are marked *