February 2018 Issue of Wines & Vines

Exploring French Oak and Napa Cabernet

In blind tasting, a winemaker gets help evaluating barrels with sensory and objective lab analysis

by Andrew Adams

Winemaker Andy Schweiger, whose family owns a Napa Valley vineyard and winery, conducted an unusually extensive trial exploring the effects of barrels from different cooperages and with varying toasts on the same wine. Then he asked a group of winemakers and cooper representatives to help him evaluate the barrels’ performance in a blind tasting, as well as comparing the sensory analyses with lab results.

In setting up the trial, the winemaker wanted to explore with a blind tasting how one’s sensory perception of a barrel aligns with objective, lab-based oak analysis. With the cooperation of several leading coopers, Schweiger organized a tasting of the same wine in many different types of barrels, with neutral barrels serving as a control.

Preference versus analysis
In the trial, Schweiger not only wanted to taste through different barrels, he thought it would be interesting to then compare those sensory (or “hedonistic”) results to objective oak analysis conducted by ETS Laboratories in St. Helena, Calif.

ETS was a sponsor of Wines & Vines’ 2017 oak conference and provided basic chemistry and oak aroma analysis for the event. The company also agreed to run its oak aroma analysis for the wines from the barrel trial. ETS used an average of all the samples to establish a reference line for the spider graphs.
The oak analysis by ETS determined the amounts of:

Vanillin: vanilla
Eugenol and isoeugenol: clove and spice
4-methylguaiacol and guaiacol: smoke and char
5-methylfurfural and furfural: sweet oak, butterscotch, light caramel and almond
Trans-oak lactone and cis-oak lactone: fresh oak, coconut.
All of the 15 wines in the trial came from the same lot of 2015 Cabernet Sauvignon that went into barrel Oct. 2, 2015. Each trial wine was aged in a two-barrel set from 10 different cooperages. The French oak, thin-stave barrels had different toasts. The wines underwent one racking before being put back into the same barrels and were topped from the “control” barrels. The control barrels were three-year-old, neutral barrels.

Those tasting the barrels were a mix of other winemakers, a few sales representatives from the cooperages in the trials and Schweiger.
Another of Schweiger’s goals was getting outside opinions on the barrels. “I wanted to get some more palates on this,” he said. “It was a pretty extensive trial for a small winery like ours—especially in a low-producing year like the 2015 vintage. That oak trial lot wound up being a third of my Cabernet production for the year.”
Schweiger set up all the wines in sample glass bottles. Attendees tasted at their own pace and held simple scoring sheets.

The tasters evaluated the barrels by three simple criteria: “would buy,” “maybe would buy” and “wouldn’t buy.” Schweiger asked those tasting to use the control as a point of reference to decide for their own hypothetical wine program if the barrels added something beneficial to the wine or could be a good blender for a wine. Schweiger said he adopted the idea of rating the barrels by would, maybe or would not buy from Michael Beaulac, the winemaker and general manager at Pine Ridge Vineyards in Napa Valley. Beaulac runs an extensive barrel trial every vintage to evaluate the best barrels for Cabernet Sauvignon wines from vineyards in several different Napa County AVAs.

Barrels would then be evaluated on how they matched to that particular wine, at that particular moment. The analysis by ETS provided some objective insights into the aroma profiles of each particular barrel.
After the tasting, Schweiger forwarded the tasting notes and scores as well as the completed analysis from ETS to Wines & Vines. The raw votes for “would buys,” “maybes” and “would nots” were tallied up and then assigned values. A “would buy” garnered three points, a maybe got one point, and a “would not” resulted in negative three points.
A weighted vote emphasized the difference between a definite preference to buy and a decision to not buy. A “maybe” vote still carried weight, as it indicated either indecisiveness on the part of the taster or that the barrel added something, but something best used in part of a blend.
Best barrel of the day
By those metrics, the Gamba G7 MT+ barrel came out as the top preference in the trial. The barrel garnered eight “would buy” votes and two “maybe” votes, without any “would not buys.”
In his tasting notes, Schweiger described the barrel as adding “blackberry” and “forest floor” aromas with velvety texture and tannin in the mid-palate and lingering fruit on the finish. Another taster thought the barrel “really brings the red fruit” and had “good structure and taut tannins.”
The ETS analysis found the Gamba barrel skewed dramatically to the sweet oak and butterscotch edge of the spectrum, with high amounts of 5-methylfurfural and furfural and moderate (or about average) amounts of the lactone compounds.
Close behind the Gamba barrel were the Sylvain M+T barrel with seven “would buys,” two “maybes” and one “would not,” and the Taransaud M barrel that received seven “would buys” and three “maybes.”
The Sylvain had a more balanced impact based on the ETS analysis, with a spider diagram that resembled the control but with more amounts of char and smoke flavors, clove and oak lactones. The Taransaud barrel had a much different profile than the control, with high levels of trans-oak lactone (coconut) and the clove notes of eugenol and isoeugenol. Many tasters described the barrel as having a “lighter” touch with notes of spice and a long, complex finish.
When the top three barrels’ scores were weighted, the Gamba G7 MT+ came out with 26 points, Taransaud M with 24, and the Sylvain M+T with 20. A table of the weighted scores for the top 10 barrels and the rest can be found on page 41. The five barrels with the lowest scores are identified only by toast information and the letter assigned to each in the trial.

The scoring reveals the subjective nature of barrel tasting. A barrel that received the low-weighted score of 12 also received four “would buy” votes and two “maybes” from the tasters. It was the four “would not” votes that caused it to suffer in the final tally.
That particular barrel was described by one taster as having “integrated tannins,” “good feel and fruit expression with good finish.” Yet another person described it as “slightly raw wood, unbalanced” and “thin with unpleasant tannins.”
A few months after the tasting, Schweiger said he was struck by how well the ETS spider graphs seemed to line up with the attributes of the barrels he preferred.
The tasting also confirmed his evolution of thinking on medium and medium-plus toasts. “If you look at the spider plots for the medium-plus barrels, you definitely saw a greater contribution to the wine in terms of enhancing the wine without obliterating it,” he said. “Whereas the medium barrels just had less vanillin and contributed less to the winemaking process.”
Schweiger said he “used to be afraid” of medium-plus and heavy toasts because he didn’t want any heavy-handed oak notes dominating his wine. “What I really saw was that we were just getting more complexity of tannin, complexity of the mouthfeel, so I shifted my orders. I used to do about 50 to 50 medium and medium-plus, and now I’m pretty much moving to medium-plus, just because I’m not seeing the value for me in the mediums.”
While the trial essentially confirmed his preference for the top-performing barrels, Schweiger said he is considering adding a few more to his program. That’s still somewhat of a risk, because there’s the ever-present question of consistency. “We like to say, ‘I like this barrel, I’m going to order 20 this year, 40 this year,’ but the reality is you’re getting 20 different barrels, you’re getting 40 different barrels, and you’re hoping that they are going to be similar to the one that you tasted.”
Because of that question of consistency, Schweiger said he’s considering a small trial on the Vicard Generation 7 barrels that are built with a patented system to determine tannin content in each stave as well as precise toasting with the goal to deliver the same barrel profile each vintage.
A surprisingly strong performer
“The biggest surprise was our control, which was three-year-old barrels, essentially neutral. They actually performed very eloquently. There was a lot of beauty to be found in the inert wood,” he said. “There is still some kind of contribution to be made from that older barrel, and sometimes it may be slow, gentle aeration is overlooked as an important winemaking tool.”
Eric Herve, Ph.D., a research scientist with ETS who oversaw the analysis for the trial, said another interesting finding was that the control barrels were not as neutral as one would assume. The control barrels exhibited significant amounts of oak lactones—nearly as much as the control line of the trial. Vanillin was about 50% of the average, and there was more than half the control of eugenol.
Herve said the neutral barrels were not inert at all but provided a gentle contribution similar to an extremely lightly toasted barrel. “This is actually a common observation in similar trials: Aroma compounds that are the most toast-dependent (furfurals, guaiacols) virtually disappear after a couple uses,” he said in an email to Wines & Vines after the trial. “They are mostly located on the inner surface of the barrel and easily extractable by wine, but the ones already present in untoasted oak keep being extracted much longer, because they are also present deeper inside the staves.”

Schweiger enjoyed the opportunity to compare his tasting notes to the lab results but still would like more data on the full impact of barrels on wine. That analysis only goes so far, as it’s still unclear how a barrel can affect a wine. ETS was able to provide a tannin analysis of the control wine, but there’s not yet a precise way to fully track the true phenolic impact of oak. “They still don’t have an accurate way of really determining the phenolic contributions of oak into a wine as it ages, and to me that would be the most fascinating thing to see,” Schweiger said.
Herve said oak tannins are highly reactive molecules that form complex bonds with wine phenolics increasing color, adding resistance to oxidation and modulating astringency and bitterness. “At the same time, unfortunately these complexes are difficult to measure within the already intricate mixture of wine phenolics,” he said.
How oak affects a wine’s mouthfeel and finish through the polymerization of tannins is still somewhat unclear. A winemaker must then rely on experience, craft and his or her palate—leaving some final decisions in the subjective arena of human perception. As Schweiger notes, the final analysis is often left to a powerful but unpredictable machine. “That is one of the issues we are dealing with: a very complicated electrode, our palate, hooked up to a very flawed computer, the brain of a winemaker.”

Print this page   PRINTER-FRIENDLY VERSION   »
E-mail this article   E-MAIL THIS ARTICLE   »
Currently no comments posted for this article.