Assay in 'Real Time'

October 2010
by Corey Beck, Lise Asimont, Evan Schiff and Tondi Bolkan

  • The Francis Coppola winemaking team has used the Adams-Harbertson phenolics assay for seven years.
  • The assay requires complicated lab work, but gives numbers about color and tannin that supplement tasting.
  • The assay aids both instant and long-term winemaking decisions, and it improves communication with growers.

The Adams-Harbertson phenolics assay allows a winemaker to reach a high-risk winemaking- decision quicker and with confidence. Not only does the assay give us confidence for present-day decisions, it also allows a historical look at vintage and vineyard performance to help make future winemaking decisions customized to each vineyard and upcoming vintage.

In the seven years we have been using the assay at Francis Coppola Winery in Sonoma County, Calif., we’ve termed the use of tasting alongside phenolic numbers as “real-time winemaking.” During crush, we are able to react to fermenting tanks as we taste and look at the phenolic extraction curves. For example, if a tank’s rate of extraction is potentially aggressive, we can change the extraction type (pumpover vs. punchdown), duration or numeric cycles. Yes, you can change these variables with input from tasting alone; however, it’s difficult to tell by tasting the true free tannin number at 10º Brix. Sometimes the realization of over-extraction sets in too late. “Real-time winemaking” helps us not to drain wine off the grapeskins too early or too late.

We know that in most red varieties, the anthocyanin (color) compounds are finite and the tannin compounds are infinite, all depending on the extraction regimen. Optimal color extraction with a balanced tannin level is a common goal for winemakers. During an extraction curve, we can see color compounds plateau and tannin compounds continue going up as long as you leave the wine on its skins. So to test our capabilities for adjusting wine phenolics, we tried several experiments with cold soak duration, co-fermentation and fining.


Adams-Harbertson phenolics assay


Experiments in extraction
The first experiment was to see how much extraction occurs during a cold soak. For a Zinfandel lot in Dry Creek, we were able to liberate about 47% of the grape’s color and 0% tannin during a five-day cold soak (see graph above). Liberating this much color allowed for gentler tannin extraction. Once the heating jackets turned on, the tannins came out of the skins, into the juice, and started binding with the multiple color compounds available. The binding of a tannin compound to a color compound makes a more stable wine. Imagine not liberating the color first, and then turning on the heat jackets. The tannins would come out of the skins, into the juice, but wouldn’t find a color compound to bind with, resulting in a tannic, lower color wine. This is when we say the color has crashed.
Adams-Harbertson phenolics assay
The second experiment was to see what happens in co-fermentation of varieties. We split a block of Zinfandel grapes into two fermentors. We added 10% Petite Sirah to one fermentor and left the other with 100% Zinfandel. A third fermentor was 100% Petite Sirah. We made the same additions of yeast and nutrients, and kept the extraction regime exactly the same. In the end, the tank that had 10% Petite Sirah ended up with 46% more color than the all-Zinfandel tank (see graph above), and the tannin increase was only 5% (see graph below).


Adams-Harbertson phenolics assay

After this discovery, we wanted to see what fining does to wines. So our third experiment ran the phenolics of a wine lot that we thought to be overly tannic. We made two small adds and tested the color and tannin after each. The tannin number came down 20% after the first add of gelatin, and then another 25% after the second add of gelatin, for a 45% total decrease in tannin. All the while, the color was decreased by only 12%.

From these experiments, we verified that seeing and tasting the extraction in real time helps winemakers adjust each fermentor to the target wine style vs. doing a blanket extraction protocol and then finding out later how each fermentor turned out. It’s almost like an X-ray of a wine. Once you know and are confident which lots are your A, B and C lots, you can blend similar lots together, which helps the overall picture of wine quality. It would not make sense to blend an A lot with a C lot to create more B lots.

Because we’ve been applying this assay to our winemaking for seven years now, we’ve created a database of vintage and vineyard performance. Such information is useful for predicting the type of extraction regimen needed for a particular block or type of vintage. In a late year like this, where everything will be coming in at once, tank space will be in high demand. The Adams-Harbertson Assay will be the key to this 2010 harvest.

Defining quality for growers

Defining the quality of a wine, and the implications this has on the source vineyard of that particular vintage, has long resulted in a subjective conversation lacking in measurable metrics. As a winery that sources its grapes from independent growers, the concern of grower relations at Francis Coppola Winery is threefold: 1) Identify vineyards that can deliver grapes suitable for a specific wine program; 2) Advise growers about vineyard cultural practices that are conducive to the desired wine style and quality; 3) Communicate both the desired and achieved wine style and quality to the grower.

Actually defining wine quality and style is essential to this goal-oriented approach to winemaking, a nd the use of the Adams-Harbertson assay has made this possible. A once-subjective concept based on the sensory analysis of winemakers has now been put into objective terms with measurable metrics that not only aid the winemaking team in defining the quality of wine from a vineyard but also allows us to take this process of goal-oriented winemaking into the selection of vineyard sources and winegrape growing.

We have been able to collect phenolic data from the past seven vintages to create an in-depth viewpoint of sub-appellations throughout the state of California. This base of information is another tool to aid us in defining growing regions suitable for our wine programs.

For example, we can identify Alexander Valley Cabernet Sauvignon by its soil types, topography, mesoclimates, clonal choices and typical wine sensory profiles (such as dark raspberry, cherry chocolate and black tea tannins). When performing the assay, we can further identify Alexander Valley Cabernet Sauvignon by its tannin, anthocyanin and total phenol content, and use this as a tool to complement our current base of information. We can use this base to measure the impact of clonal selection, pruning methods, timing of canopy management, crop load, irrigation regime and a seemingly endless array of other cultural practices.

This becomes a powerful tool during the growing season, when we advise contracted growers regarding vineyard cultural practices to achieve our targeted wine styles. Taking the Alexander Valley Cabernet Sauvignon example a bit further, we were able to measure the impact on final wine quality of canopy management practices performed at different times in the vine’s phenology, and realized that a leaf-pull pass done at the time of berry set significantly increased the total phenolic content of the wine.

Theoretically, the results made sense, but having data from the assay allowed us to discuss this practice with our growers in an objective manner. During harvest, the data from the Adams-Harbertson assay is used as a live-time aid informing the winemaking team about phenolic trends. We discuss this information with our growers and consider this in harvest pick decisions.

How we run the Adams-Harbertson phenolics panel

The Adams-Harbertson phenolics panel or assay quantifies phenolic compounds in wine by measuring and analyzing the chemical properties of those compounds. It separates chemicals into functionally defined categories. For instance, phenols react with iron, and this reaction is measurable. This assay has been in use by our company since 2003. We collaborated with Dr. Doug Adams and Dr. Jim Harbertson (both of whom were at the University of California, Davis, at the time) and followed the most current procedures to get confirmable data for our wines.

Frequently we send our wines to outside labs to verify the validity of our data points. Currently there is an industry effort to make an openly available and accepted SOP. We are enrolled in the Gage Repeatability and Reproducibility Study for the Adams-Harbertson assay. Our staff strives to achieve quality numbers that give actionable information to our winemaking team. We take great pride in being part of the cutting edge of winemaking.

The assay demands a high degree of training and attention to detail. We have several trained technicians, and it is important that each technician uses the exact same protocol every time the assay is run. We have dedicated an area in our lab that is free from distractions and contains all the equipment necessary for the assay. Our samples are taken from fermentation tanks with red musts. The samples are representations of how the musts are changing with respect to phenolic compounds. We run the test daily for the duration of harvest.

Typically, one lab technician runs the assay for 2.5 to 3 hours depending on the number of samples needed in a particular day. The consideration of technique and exact timing of the chemical reactions during these 2.5 hours is critical to the reproducibility of the test. Equipment utilized includes a micro-centrifuge, a spectrophotometer and pipettes, along with many chemical solutions. When required, chemical solutions are pH-adjusted in order to account for pH-sensitive chemical reactions. The amount of light at specific wavelengths absorbed by working reagents mixed with samples is recorded. These absorbencies are used to calculate levels of phenolic compounds. Each sample requires multiple readings to complete a representation of its unique phenolic composition.

When the assay is finished, our winemakers taste and note the levels of total phenols, red wine pigments and tannins. The combination of tasting and awareness of chemical compounds (known to affect taste, color and mouthfeel) lead to informed winemaking decisions. Running the phenol assay daily has become part of our harvest routine because of the great value we place on wine quality.
--C.B., L.A., E.S. and T.B.

Having phenolic data has made communicating wine quality and style with growers a far less antagonistic process. Many growers have been turned off by wineries that dislike their grapes but don’t explain why or what can be done. Being able to tell a grower how their vineyard performed that vintage by taking the mystery out of a subjective wine tasting and giving them numbers to aim for has been critical to our relationships with growers.

We sit down with each grower at the end of the vintage and review the phenological and climate data, discuss cultural practices used and taste the wines from their vineyard in concert with the phenolic data we have collected during fermentation. We have a conversation about the vineyard, vintage and desired wine style while using the same language.

Growers leave their feedback meetings with a sense of how their vineyard performed, what wine style we are trying to achieve, and how their vineyards line up with our desired wine style. This becomes the jumping-off point for creating vineyard growing plans for future vintages, and it rouses further discussion about goal-oriented winemaking that starts in the vineyard and is a partnership between the grower and winery.

Corey Beck is director of winemaking for the Francis Coppola Winery, Geyserville, Calif. Collaborating with him on this article were members of his winemaking team: Lise Asimont, director of grower relations; Evan Schiff, enologist, and Tondi Bolkan, winemaker. To comment on this article, e-mail
Currently no comments posted for this article.