Winemakers the world over ask, "Should I base my cork purchasing decision on appearance and functionality or the budget? Do the two go hand in hand?" The fact is pricing and cork appearance do go together. To a large extent, cork pricing has been based on the visual appearance of corks. Thus, the use of visual analysis to determine both cost and quality is the generally accepted practice.

Cork grading began with a system that included six grades: from Primeira (First) to Sexta (Sixth). But as wineries demanded higher qualities and more careful selection, the grading system grew to include "Superior," then "Extra," and finally, "Flor y Nata" (Flowers and Cream!). As a way of standardizing these grades, various cork companies in Portugal, in concert with their marketing departments, developed a system of visual grading for each grade. Due to the fact that there were no industry standards, there was substan~ tial variability within each grade.

At the same time, cork producers discovered that bleaching corks disinfected them and helped eliminate or mask discoloration, thereby increasing its apparent visual quality at very little cost. Most winemakers approved of bleached corks and were willing to pay a premium for them. They quickly became the standard, and experimentation continues on cork washes.

In this system, confusion often reigned, and winemakers were frustrated. There seemed to be virtually no quantitative analysis for the visual grading of corks, and the cork industry itself seemed disinclined to change the situation. Because the visual asp.ect of the cork is one of the key elements upon which pricing is based, a more complete method was needed to help winemakers determine the value of cork lots submitted for purchase.

Several years ago, the Cork Quality Council in the U.S. developed a series of visual grades that were described in terms of qualitative analysis. This analysis provided a method to visually identify the physical characteristics of, and to place various types of corks into separate categories within a given sample. This was an important step which helped to form the foundation for some of the analysis presented here.

• Visual grading should play a key role in establishing the anticipated levels of defects in a cork lot.
• Visual grading can be used to monitor consistency of quality over time.
• Visual grading can be used to compare lots of corks from different suppliers to determine their relative value and price.
• Finally, visual grading may be used to generate specifications for minimum levels of quality categories and maximum defect levels within a cork lot.

Critical defects are those that run 75% to 100% of the length of the cork, including insect holes, channels, cracks, and heavy porosity.

Non-critical defects are mainly cosmetic in nature and alter the perception of quality. However, they can interfere with the effectiveness of bottling equipment and add dust and cork particles to the wine. This type of defect is used as the basis for the quality grading of a specific lot of cork, and includes porosity, chipping, dry years, deep holes, belly, bark, hardwood, end notching, and short cracks.

Non-critical defects may be classified as "minor" or "profound." Minor defects may be defined as those that run up to 25% the length of the cork. In some instances, a small crack originating on the end of the cork may cause slight seepage along the length of the crack. Generally, the seepage will go no further.

Profound defects may affect the performance of the cork to a greater extent. For example, a very large chip in the end of the cork may allow wine to enter the area between the cork and the bottle, with subsequent heavy seepage. Profound defects may run from 25% to 75% of the cork length. Defects arise from two different sources: those that occur naturally in the cork wood itself, and those that have been caused by faulty manufacturing processes.

Suppliers in the U.S. that regularly monitor defects are working to reduce their incidence. Education of the cork producer as to how the market perceives specific defects is key. Therefore, winemakers may want to consider investigating the relationship that their supplier has with the overseas producer.

Natural defects include those caused by growing conditions - such as cracks, dry years, porosity, hardwood, and greenwood - and those caused by parasites, such as insect larvae or molds growing in the forest.

• Cracks appear as dark brown or black fissures in the wood, and can run either vertically or horizontally through the finished cork. They are usually caused by uneven expansion of the cork bark as it grows. Depending upon the length, location, and depth of the crack, they can prevent a cork from forming a proper seal.

• Dry Years are similar to cracks, but appear as a narrow dark line down the length of the cork. This defect is caused by low moisture during the growing season which prevents the new cork bark from growing evenly into the existing layer. The result is a narrow band of hard, dry wood that is inconsistent with the rest of the cork. Creasing, and thus seepage, may occur along this dark band which has a higher resistance to compression and lack of resilience. In addition, very large dry years may crack off as they go through the corking jaws.

• Porosity is easily defined as the number and size of holes in any given cork. In many ways, this is the parameter upon which visual grading is based. Inexpensive cork may appear to have a network of larger holes and cork bark. However, even the very finest corks will have some minor porosity. Porosity as a critical defect is only seen in corks where the wood structure of an entire cork face is affected with a crumbly texture or the porosity is focused along a growth-line.

• Hardwood is seen as small sections of bark that do not have the same resiliency as the rest of the cork. Hardwood results from factors that influence the formation of the lenticels of the bark itself. Because it is not resilient, hardwood can prevent the cork from sealing properly.

• Greenwood is an unusual defect that prevents the cork wood from seasoning completely. After harvest, most cork wood is aged in the forest for six months or more before processing, then boiled and aged again for a short period. This process produces seasoned cork wood - a material that is unique for its . many physical properties. Greenwood prevents the proper seasoning of the wood. After ageing and boiling, greenwood does not achieve dimensional stability It will appear as a ruffled or bulbous area on the cork, while the remainder of the corkwood surrounding the area appears smooth and consistent.

• Insects have traditionally been the bane of cork producers. Small (1-2 mm) holes are created by insects that burrow through the cork bark leaving either open or dust-filled tunnels behind them. Insects such as Coroebus Undatus ("Bicho" or worm), Cremastogasterscutellaris Olivo ("Formiga" or ant), and Dermestes vulpinus ("Traca" or butterfly) create open tubes that - when present from end to end of the cork - cause leakage, oxidation, or both.

• Yellow stain is apparent in the raw cork wood and easily identified by its pale yellow color. It is caused by growth of the mold Armillaria mellea on the wood in the forest (personal communication with Dr. Mario Borges, Juvenal Ferreira da Silva, Lda., Santa Maria de Lamas, Portugal). It is readily detected at the cork processing facility in Europe and eliminated at that point. However, it is important to note any discoloration due to this defect as it lends a moldy flavor and aroma to the wood.

Manufactured defects are those caused by human hand during processing of the cork wood. While some have their origins in the material itself (belly and bark), most are caused by an incorrect application of the production process.

• Belly is that part of the cork that is closest to the tree. Occasionally, when the cork stopper is punched out from the cork slabs, the puncher does not avoid this belly layer, and the exterior of the cork shows the ruffled texture of the belly of the tree. Because this belly is both rough in texture and less resilient than the rest of the cork wood, this can create sealing problems. However, the more severe belly defects are sorted out in Portugal. The defects seen in the U.S. tend to be limited to small, non-critical spots.

• Bark is the outside of the tree, the part that has developed a dark brown or black appearance due to interaction with the air. The defect is a cutting error similar to belly, but the error has been made on the opposite side of the cork slab. The results are equally poor.

• Channels are purely manufacturing errors. The corks have been punched too close together, and one cork displays the cylindrical notch that was created by the previous cork. When very shallow and short, these channels may not affect performance. When deep, they can create immediate leaks.

• Chipping can be caused either by inconsistencies in the wood itself or by poor manufacturing technique. The result in either case is the same: large chunks or chips of the cork are missing. It is interesting to note that the Portuguese have two separate terms for this noncritical defect. "Falha Corpo" is a chip in the body of the cork, while "Falha Tapa" occurs on the end of the cork.

It is important to note that the results of any analysis for these defects in cork are affected by both the branding and coating. Branding the cork can introduce minute ink spots that may affect the visual appearance. Coating the cork can make the surface look more porous. The combination of the two processes tends to darken the cork. For this reason, suppliers often provide coated cork samples, which alleviates mistaking the appearance of a finished cork for discrepancies in the quality or identity of the lot itself.

Finally, while it is not a defect of a single cork per se, it is important to inspect corks for excess dust or particulate matter on their surface. This dust can affect both the seal of the cork and its visual appearance (poor printing) and should be noted on any qualitative analysis.

In addition, insertion of the cork into a standard bottle decreases the effect a defect may have upon the cork. The cork is compressed from 24mm down to 18 or 19mm, so non-critical defects are less threatening to the integrity of the wine package.

A good seal depends on several elements: the cork, the coating, the bottle, and bottling conditions. The coated cork works in combination with the interior neck of the bottle. While bottle necks do not suffer from the defects listed above, they may not conform to the same rigid criteria for dimensional consistency either. Significant variations in bottle neck diameters can introduce defects in the seal between cork and bottle.

For instance, seepage may occur when the cork length extends beyond the flare point of the bottle and does not form a continuous seal. Squared off entry bores may also affect the cork, causing slight chipping to occur during compression and insertion into the bottle. For this reason, it is recommended by the authors that the winemaker set up bottling trials with a supplier when considering purchase of new glass.

It was then applied to corks which had no structural flaws, but did have visible surface anomalies. Of course, corks with perfect exterior appearance are rare and command much higher prices. Such corks also project an image of quality and integrity for the bottle of wine itself which is an important marketing aspect of corks that cannot be overlooked.

It is the defects, or lack of them, which allow one to determine the overall consistency in a lot of corks. A specific lot of corks may have a continuous spectrum of qualities, but with a simple grading method in place, the cork buyer may determine the general quality profile of the lot and identify possible problems.

Some California cork suppliers use a grading scale detailed in terms of five grades: "A" type corks have very little porosity and no defects whatsoever, "B" corks have slightly more porosity and very minor defects, "C" corks have moderate porosity and larger defects, "D" corks are extremely ugly but functional, and "E" corks have critical defects. Noncritical defects are usually placed into C and D categories, depending on whether the defect is minor or profound. The authors recommend the "Three strikes and you're out" rule: If a "B" or "C" cork is found to have three minor defects, the grade is downgraded to a "C" or "D."

In the more expensive grades, there will be very few C through E corks. In lots of lower quality corks, the number of Cs and Ds will be greater. Those wineries that grade a variety of qualities will notice that the defects in a "c" cork in an "Extra" grade will be minimal compared to a "C" cork in a "Third" quality. Differentiating between the two comes with the winery's familiarity with the product, and in working with their supplier to define the categories.

Within a lot, there will be a certain number of each cork category. Once the numbers are generated, a computergraphing program such as Excel can be used to depict the qualities of the lots received over time (see fig. 1) and between the actual shipment and the reference sample.

Because cork is a natural, agricultural product, the quality will vary slightly from lot to lot. Quantifying the categories within the lot gives the winery the advantage of determining when this variation becomes a cause for concern.

The data base can be used to develop specifications for the product. By creating specifications, the winery will be able to reduce variation and purchase corks that fit their needs.

In order to achieve this important goal, sampling procedures must be statistically sound, random, and representative of the entire lot. Several sampling procedures are available, such as the ANSI/ASQC (mil-spec 105) and the Sequential Probability Ratio Test. In addition, coming from the University of California at Davis, the Fraction Defective Sampling Plan should soon be published.

With whichever procedure you use, no fewer than 100 corks should ever be graded. Due to the range of quality within the cork lot (even within a bale of 10,000) it is difficult to determine the percentage of defects and overall quality profile when analyzing very small samples. By grading a larger sample, the winery reduces its risk of accepting cork lots with critical defects which exceed acceptable quality limits.

A minimum of 100 corks should also be used when working with a supplier to set up a reference sample. As seen in fig. I, there may be fluctuation in quality levels over time. With a reference sample that has been agreed upon by both the winery and the supplier, qualities are easier to match, thus ensuring consistent shipments throughout the year.

Visual grading facilitates the tracking of quality levels between suppliers as well. Because of confusion caused by the many different trade names and qualities used by suppliers, frequently cost is the only facet that can clearly define what the buyer is comparing. Using a program as presented above should help to make the decision-making process easier.

In order to develop an effective cork quality assurance program, the goals of the program must be clearly defined. The goals will vary depending on the type of wine, quantity of corks purchased, budget requirements, type of bottling equipment, storage methods, and winemaker preferences. By using the criteria provided above, the winemaker will be able to determine what is required, communicate those requirements to suppliers, evaluate samples and shipments, improve consistency, and reduce costs.

At JuvenaL Direct, Inc., Michelle Bowen is director of Quality Assurance, Paul Wagner is president, and Pern) Teaff is general manager. Authors can be contacted at Juvenal Direct, 120 Dodd Ct, American Canyon, CA, tel: 707/254-2000 ; fax: 707/642-2288.