A Brief Gloss on Smoke Taint

Ξ August 17th, 2009 | → 18 Comments | ∇ A Day at a Time, Technology, Wine News |

What follows is a brief overview on the matter of smoke taint. Recent fires in California have again made urgent the discussion of the topic, especially with the Crush looming. What follows is meant to be a simple primer for the more casual wine industry follower. It covers the matter of vineyards alone. With respect to minimizing smoke taint during vinification and associated ’scalping’ technologies recently developed, those matters will not be discussed.
Smoke taint, the contamination of grapes owing to exposure to fires, whether wild or controlled, is of considerable significance to winemakers and grape growers. Here in the Santa Cruz Mountains the latest blaze, the Lockheed Fire, though 65% contained, with sufficient resources on site to finish the job, the possible effects of smoke on vineyards remain. As with last year’s fires in Northern California and, again, in the Santa Cruz Mountains, there is some evidence, anecdotal in the main, of smoke taint. But the facts are difficult to establish, not only because of the potentially enormous financial impact to growers and contracted wineries and negociants, but also because of the varied tasting thresholds, whether it is a professional or a consumer performing the evaluation.
Smoke taint flavors and odors have been variously described as ‘ashy’, ‘burnt bacon’ and ‘wet ashtray’. Anyone who has wandered through the aftermath of a house or forest fire can readily understand these terms. The effect is all-inclusive, permeating and, in the case of a house fire, often resulting in the total loss of all possessions, clothes and furniture.
Not surprisingly, some research suggests the disagreeable qualities of the smoke taint may depend on the kinds of fuel the fire consumed. Bush fires, conifer forests, chaparral etc. may well have differing chemical signatures. Indeed, the research following upon the chemistry of smoke taint offers potentially interesting insights into discussions of terroir. If smoke may be readily absorbed by the vine and grapes, then perhaps the thriving flora of a given wine region may also leave an imprint not only on the grapes but on the finished wine as well. But that is another topic.
It is also important to mention that the toasting level of a wine barrel, the flavor and odor intensity of which is selectively pursued by winemakers, such toasting levels impart some of the same compounds to a finished wine as does fire smoke, obviously in more modest percentages and chemical combinations.
Though research is on-going, with the Australian Wine Research Institute (AWRI) taking the lead, a few scientific parameters are now fairly well established.
1) Smoke taint is the result of the absorption of principally guaiacol and 4 methyl guaiacol, (there are other compounds) into the grape skins, vines and leaves. Research indicates it is the grape skin absorption itself that is of the greatest concern. The first efforts to systematically understand smoke taint were undertaken by AWRI after winegrowers reported off-flavors after devastating bush fires in Australia in 2003. It was then discovered that smoke taint was not simply the consequence of ash and smaller fire particulates falling on the grapes, but was more systemic. Simply washing the grapes, or more aggressively, removing the waxy surface bloom of the grape, had no effect on the presence of smoke taint. The penetration of fire-generated compounds into the skin, leaves and vines was then established.
2) Absorption levels are grape-specific. For reasons not yet understood, there are specific grape varieties with a stronger affinity for guaiacol and 4 methyl guaiacol absorption. From the Department of Primary Industries (DPI) paper:
Anecdotal information gathered at the industry meetings suggested that there was variation in smoke taint between grape varieties. Varieties classed as having high susceptibility by industry were Pinot Noir, Sangiovese and
Cabernet Sauvignon, and lower susceptibility was found with Merlot and Shiraz.

3) Smoke exposure duration is of special significance. From a second DPI paper:
Although research is ongoing in this area, it is interesting to note that a single heavy exposure of smoke (30% obscuration/m) to grapevines for 30 min is sufficient to result in smoke taint in wine (assuming that the exposure occurs during a period of sensitivity).
4) The timing of smoke exposure is critical, with 7 days post veraison to harvest showing the greatest susceptibility. Again, from the second DPI paper:
There are three distinct periods of grapevine sensitivity to smoke exposure. The first period (P1) is characterized by a low potential for smoke uptake early in the growing season when shoots are 10 cm in length and at flowering. The potential for smoke uptake is variable during the second period (P2) from when berries are pea size through to 3 days post veraison. Grapevines show a low to medium sensitivity to smoke uptake during P2. During the third identified period (P3) from 7 days post veraison through to harvest grapevines have a high potential for the uptake of smoke compounds.
5) Harvesting, whether by hand or by mechanical means. This is hardly surprising, given that absorption of smoke occurs in the grape skin itself. From the first quoted DPI paper:
The concentration of ‘total guaiacol’ in the machine harvested grapes ranged up to 4.5 times that of the hand harvested grapes in the press fractions. Based on the original concentration in the grapes and the concentrations and
volumes of juice collected, around 40% of the ‘total guaiacol’ in the grapes came through in the press fractions and lees with hand harvesting and around 98% of the ‘total guaiacol’ in the grapes came through in the press fractions and lees in the machine harvested sample.

6) There does not appear to be a ‘carry-over’ effect. Vines that have absorbed smoke do not then have the that smoke expressed in the fruit the following year.
(Update! The links have been repaired.)


18 Responses to ' A Brief Gloss on Smoke Taint '

Subscribe to comments with RSS or TrackBack to ' A Brief Gloss on Smoke Taint '.

  1. Arthur said,

    on August 17th, 2009 at 1:44 pm

    Hi Ken

    Good summary.
    Not sure if the literature has changed since last year, but I seem to recall that 4-EG and other smoke taint molecules enter the grapes via transpiration: they are taken in through stomata and transported to the berries.
    Can you comment on this?

  2. Administrator said,

    on August 17th, 2009 at 3:17 pm

    Hi, Arthur. I’m not entirely sure about the answer to your question. As you probably know, when the offending compounds were first isolated back in 2003 it was grape bunches themselves which were used for the experiments. It was sufficient to expose them to smoke in order to reveal the presence of high concentrations of guaiacol and 4- metylguaiacol. Now, inasmuch as the exposure time can be as little as 30 minutes to cause significant taint, it is not clear whether transpiration would play that great a role in any case. That’s just a guess on my part. Indeed, more current research protocols use special tents to smoke the entire vine in the field. This approach was taken to determine the most damaging exposure window to the grape’s development.
    But certainly, as a matter of basic viticulture it would be important to determine the truth of your question, no doubt. (As a side note, are you asking specifically about 4-EG transpiration or of other compounds in fire smoke as well?)
    Lastly, according to AWRI research, though the contamination of the vine persists there is no ‘carry over’ of guaiacol and 4-methylguaiacol into the next year’s crop.

    I’ll continue to look into it. And should you learn of the answer please shoot me a line.

  3. John Kelly said,

    on August 18th, 2009 at 5:27 am

    I have agreements not to disclose sources in place, but I can say that 1) taint issues are real and markers have been quantified in a number of wines, and 2) smoke compounds are transpired through the stomata, not adsorbed onto the berries or taken up through the roots.

  4. Arthur said,

    on August 18th, 2009 at 7:38 am

    I think that transpiration is believed to be the main mechanism, Ken.
    I don’t think there is anything in the physiology of the berry that would facilitate the uptake EG and MG.
    The stuff is thought to be taken up from the air by the leaves and transported through the phloem to the berries – if I recall correctly.
    Additionally, that is why washing or limited skin contact cannot free the wine of smoke taint.
    I have asked a friend to help clarify this. He will either post or email me back.

  5. Administrator said,

    on August 18th, 2009 at 8:06 am

    Thanks, John and Arthur. What puzzles me is that the initial research into smoke taint was done with whole grape clusters only. And again, in experiments done in 2006, we read in DPI’s 2007 summary Impacts of bushfire smoke on grape and wine quality – Scoping study:
    Smoke effects on grape bunches and organoleptic perceptions in wine Research was initiated to demonstrate the effect smoke exposure of grapes had on the chemical composition and sensory characteristics of wine in early 2006. Post harvest wine grape bunches were utilized in this experiment and exposed to smoke in purpose built smoking facilities at Kings Park and Botanic Gardens, Perth (WA). Two wine treatments were made from smoked and un-smoked grapes (1) free-run juice treatment (2) free-run juice on skins treatment. Quantitative chemical analysis of wines for guaiacol, 4- methylguaiacol, 4-ethylguaiacol, 4-ethylphenol, eugenol, furfural, 5-methylfurfural and vanillin has been conducted by the Australian Wine
    Research Institute (AWRI). Sensory wine analysis of difference tests and best estimate thresholds were also performed. Wines vinified from smoked grapes exhibited ‘smoky’, ‘dirt’, ‘earthy’, ‘burnt’ and ‘smoked meat’ aroma characters.

    It appears that both mechanisms, berry absorption and transpiration, must be at work. But it remains difficult to understand how a single 30 minute exposure of a vine to a certain density of smoke, at the right moment of veraison, could be accounted for by transpiration alone. Can you provide me with research references? Thanks.

  6. Arthur said,

    on August 18th, 2009 at 10:06 am

    Well, it would make sense to analyze the clusters – after all that is what the wine is made from. It is not unreasonable to think that EG and MG settle on the outside of the skins – thus the results of the study with exposed post harvest grapes. Additionally, these compounds may somehow become fixed or chemically bound to the skins. Something about the plant (leaves and grapes) in the period surrounding veraison seems to raise susceptibility to smoke taint.
    However, other studies have indicated that washing grapes does not alleviate smoke taint. This then was connected to the transpiration mechanism.
    It would be interesting to see analysis of leaves and wood – particularly the phloem – to test/validate the transpiration theory.

  7. Administrator said,

    on August 18th, 2009 at 10:39 am

    The post-harvest clusters were specifically selected for smoke taint experiments, both in 2003 and 2006. Washing, whether with water or more aggressive approaches resulting in the out-right removal of the wax, had no ameliorating effect. This we know. Absorption by the skins has been experimentally confirmed. So the question becomes why this might be. Clearly there is something going on within the skins. So if the compounds in smoke are being, as you say, ‘chemically bound to the skins’ then that suggests, at the very least, a supplemental pathway for EG and MG contamination.
    I suggest you give the summary I’ve already cited a second look, as shall I. But it does appear that the issue is not whether grape skin absorption occurs, but how. Certainly other research avenues are being pursued, most notably involving transpiration. But grape skin absorption appears to be settled science. (Unless I am reading the literature incorrectly, of course!)

  8. Arthur said,

    on August 18th, 2009 at 10:48 am

    Hi Ken

    I’ve been reviewing the literature this morning. Agreed that its the *how* that remains elusive. Since EG and MG are phenols – could they be deposited in the skins because that is where the plant puts phenols? (studies have also indicated that EG & MG are found in the skins and not the pulps).
    If the mechanism is some sort or Chemisorption, at least initially, this would also make sense.

    One of the studies I came across indicated that wines made from machine-harvested grapes (thus those containing leaves) had higher taint – indicating some sort of uptake mechanism which involves the leaves.

  9. Administrator said,

    on August 18th, 2009 at 11:02 am

    Hi Arthur. Thanks for the reply. The matter of machine harvesting vs hand harvesting is mentioned near the end of my piece. The explanation for this difference in the article I cited involves the rougher handling, the rupturing of the grape skins during mechanical harvest and, yes, additional contamination via leaf inclusion. This becomes a significant issue if the vineyard’s exposure to smoke was not equal in all blocks.

  10. Sherman said,

    on August 18th, 2009 at 12:09 pm

    I recall reading somewhere (sorry, I don’t have more exact cite at this time) that EG & MG taint became an issue post-fermentation; that is, the grape juice wouldn’t show EG/MG contamination because the EG/MG molecules didn’t bind with the juice until after fermentation had taken place. Any indication of this in the study, or am I just mis-remembering something I read a few months ago?

  11. Administrator said,

    on August 18th, 2009 at 12:21 pm

    Sherman, you are exactly right. There is ample evidence (it might take me a while to find exact citations) that in some musts that have undergone some sort of experimental treatment to remove smoke taint prior to fermentation, nevertheless had the taint return in the finished juice. This has everything to do with vinification and associated ’scalping’ technologies, a topic I chose to set aside for another day owing to the very technical nature of the issues involved. The problem of amelioration centers on the atomic weights of the offending compounds. To remove EG & MG (and a host of other co-contaminants) risks stripping a wine of all character.

  12. Arthur said,

    on August 18th, 2009 at 5:12 pm

    I think that the literature discussed here does indicate that fermentation can “fix” or “augment” EG and MG in the wine. However, Clark Smith (vinovation fame) told me recently that you can never extract EG and MG fully – presumably from must or wine. Interestingly enough, EG and MG are also among the taint molecules plaguing wines affected by btrettanomyces.

  13. Arthur said,

    on August 18th, 2009 at 5:18 pm

    Also, The issue with machine harvesting is that you include leaves and that may implicate the presence of EG and MG in the leaves – thus supporting the transpiration hypothesis. Ultimately, the grapes will be crushed and rinsing them will not reduce the taint. Some studies indicate that when you wash the leaves and the grapes, there is little EG and MG in the rinse water but the wine is still smoke tainted. When you consider that with the above-mentioned finding of taint in wines from machine-harvested fruit, you have a compelling argument for investigating the plant material for presence of EG and MG in leaves and shoots, etc.

  14. John Kelly said,

    on August 18th, 2009 at 6:40 pm

    There is actually a tremendous amount of work that will never be published being done on tainted wines.

    A company called Memstar is using an RO-related process to treat tainted wines – other service providers have followed suit. Different wines exhibit different degrees of stripping, but many wines turn out acceptably – for blending.

    Back to transpiration vs. absorption: Impact from 30 minutes smoke exposure is likely due to the rate of gas exchange in the plant, the surface area of the leaves relative to the berries, and the relatively low sensory thresholds for taint marker compounds. Taint markers were not found in aqueous methanol rinsings of clusters, and field washing did not reduce taint potential. I am extremely skeptical that adsorption is a mode of entry. One thing to recall is that grapes themselves have stomata. Some have more than others – I could hypothesize that these differences could contribute to the variance in sensitivity by variety.

    Finally EG & MG are simply markers used to assess taint. They are not the whole picture – smoke is a complex mixture. More complex molecules with higher sensory thresholds may be respired into the fruit which are not removed by RO and subsequently hydrolyze – hence taint recurrence after treatment.

  15. Administrator said,

    on August 18th, 2009 at 7:51 pm

    Of course, Arthur, leaves, grapes, the vine itself becomes affected by smoke. This is made clear in the literature I cited. Absorption of smoke compounds by the whole of the parts of the plant is not in dispute.

  16. Administrator said,

    on August 18th, 2009 at 7:58 pm

    Interesting, John. Are you suggesting Memstar’s techniques are proprietary and, therefore, their primary research is as well? I think the chemistry is solvable by a dedicated univerisity program. Can’t imagine, if I understand you correctly, the remedies, should any emerge, will be patentable.
    With respect to trans/absorp, bear in mind that some of the experiments were done with grape bunches alone.
    Lastly, yes, smoke is a complex mixture. In fact, not surprisingly, its composition varies from fuel to fuel.

  17. John Kelly said,

    on August 18th, 2009 at 9:09 pm

    Ken: 1) Yes, and work on stuff that is not already patented will not be forthcoming until after patents are in place. Also, many wineries, growers and grower organizations are reluctant to bring attention to the issue. Finally – research done in house by some of the largest and best-funded labs is NEVER published. 2) Yes one published trial suggested smoke effects on clusters alone. Would it surprise you to find that other trials did not see the same effect? And 3) fuel composition, temperature of the burn, weather during the burn, etc – should make a fascinating study on the effects on taint sensory qualities, as well as on the chance of recurrence following treatment.

  18. Administrator said,

    on August 18th, 2009 at 9:54 pm

    1) Of course wineries and others are reluctant to bring attention to the issue. But, certainly with respect to finer wines, should there be detectable smoke contamination to tasters, that reality, obviously, cannot be hidden. I suspect it is the lower end of the market, cheaper wines, that might benefit from novel taint removal technologies. Indeed, I am not sure such a tech will ever exist. Not as long as there is blending and other conning tools! Yes, research done in-house may never be published. But let’s be real. This stuff is basic chemistry. 2) The idea that there is only one study that did not just suggest but proved the absorption of smoke by grapes is not accurate. I am sure AWRI has a deep library of experimental results, as does UC Davis and Cornell, Fresno, too! It is not a credible claim, John. And should there be ‘other trials [that] did not see the same effect’, well, I can only say that should such results have been discovered the Australian wine industry would not have been in such a panic back in ‘03, ‘06, and just last year. Moreover, why would AWRI publish misleading or false or intentionally incomplete information? How does that benefit the farmer or, especially the university’s reputation as a world-class research facility?

Leave a reply

From the Vineyard to the Glass, Winemaking in an Age of High Tech


  • Recent Posts

  • Authors