Unraveling the Mystery of O2's Effect on Wine Aging

Part one of this series, “Oxygen Status at Bottling,” which appeared in the November 2016 issue of Wines & Vines, briefly mentioned the potential benefits of oxygen but focused more on the negative aspects of excess oxygen absorption into wine. The previous article also reviewed the sources for oxygen entry into wine during vinification and bottling, when the presence of oxygen can have a critical influence on the fruitiness, freshness, color and flavor of both white and red wines. It is important for all winemakers to be aware of the crucial times when oxygen can negatively impact their wines.

The enology group at The Ohio State University’s Ohio Agricultural Research and Development Center in Wooster, Ohio, conducted a study to determine the oxygen status in Ohio commercial wines at bottling as well as the finished wines. The amount of oxygen in the headspace of bottles and the total package oxygen (the sum of dissolved oxygen and headspace oxygen) is discussed, as are analyses for sulfur dioxide and pH levels. Finally, the results and implications for the Ohio wine industry are reviewed.

Description of the Ohio study For this study, 14 commercial wineries were selected from three different regions in Ohio. Four wineries were in the north, six in the central region and four in the south. The wineries varied in size, with seven in the small category of less than 10,000 gallons in production, four wineries in the medium range, between 10,000 and 50,000 gallons, and three in the large category of more than 50,000 gallons.

The table “Description and Bottling Operations of Wineries in Study” on page 75 describes the wineries and their bottling operations.

For the filling operations, just over half of the filler types were gravity-fed hand-operated units, which were used by mostly small wineries. The other wineries employed automated systems.

Closure operations were accomplished by semi-automatic or automatic machines at 10 wineries, while the others were manual. Seven of the automated lines used some pre-evacuation process to reduce the oxygen level in the bottled wines. These bottle pre-evacuation practices included flushing with inert gas (N2 or CO2) and/or vacuum filling. Within this category, one winery failed to employ any pre-evacuation treatment.

Similar results were found for the critical headspace (HS) evacuation practice. However, just five wineries utilized an evacuation technique to remove headspace oxygen. Although several closure types are available to winemakers, synthetic cork was the most popular, with seven wineries utilizing that closure. For the remaining closures, 1+1 (twin tops), screwcaps and natural corks were used by two wineries for each type, and one winery used a technical (agglomerated) cork. It is important to keep in mind that the type of closure used may depend on the type or style of wine the winery happened to be bottling the day of the study.

Analyses conducted Commercial wines from these 14 Ohio wineries were selected for analysis. For each wine, temperature and dissolved oxygen (DO) measurements were taken at the holding tank. Then, 12 bottles were collected during the middle of the bottling process.

Several measurements were taken at various storage times: zero, seven, 21, 63 and 126 days. At each time interval, three bottles were sacrificed for analyses, except at bottling (zero time interval), when only one bottle was examined for sulfur dioxide and pH. The dissolved oxygen (DO) and headspace (HS) oxygen levels were determined by using PreSens technology with a NomaSense Trace Unit oxygen analyzer. For the free and total sulfur dioxide contents, the aeration-oxidation method was used. A pH meter, the Accumet-Basic Model AG-15 from Fisher-Scientific, was employed for all pH readings. Also, all results were recorded as averages of three replications.

Dissolved oxygen at holding tank and bottling In order to control the dissolved oxygen (DO) level in wines, every effort should be made to avoid oxygen enrichment at each step of vinification. If high O2 values are detected, measures need to be used to correct the winemaking errors. A suggested guideline level for DO at the holding tank, just prior to the filler and bottling, is less than 0.5 ppm. The table “Temperature and Average DO Levels for Wines in Holding Tank and After Bottling” on page 76 summarizes the DO results for all the Ohio wines at the holding tank location.

In breaking this down further, five wines at this site were found below the acceptable level of 0.5 ppm. The DO levels for the other nine wineries were well above the guideline value. These wines were categorized as medium (five wines) and high (four wines), averaging 1.0 ppm and 2.6 ppm, respectively. From this data, it is obvious that the majority of the wineries have difficulty limiting the O2 uptake in their wines. Later in this report, a strong relationship will be shown between the O2 level in the holding tank and the DO amount in the bottled wine. Certainly, more attention should be given to the various winemaking steps leading up to the holding tank. These include pumping, filtration, fining, blending cellar temperature and cold stabilization, just to name a few.

There appears to be a strong connection between the level of DO in the holding tank and the resultant amount in the bottled wine. An increase in DO from the holding tank to the bottled wine was observed for all wines within each of the three DO categories. For the five wines in the first category (< 0.5 ppm), the DO level was 0.3 ppm, which increased to 1.0 ppm at bottling. All other wines and their respective categories showed the same direct correlation.

The recommended maximum DO level (1.0 ppm) at bottling was exceeded for these two categories. Certainly wine temperature, number of bottling stops and wine level in the holding tank with inert gas protection are issues that need to be addressed to control wine oxygen levels at this critical stage.

The oxygen amounts in wines are highly dependent on temperature, and wines at lower temperatures contain more oxygen than at higher temperatures. The recommended wine temperature at bottling is between 60º and 70º F, but only six wines out of 14 were within these limits. In this study, wine temperatures were quite variable, ranging from 54º to 72º F. Of course, some temperature changes will occur from the holding tank to the bottled wine. Keep in mind that several important issues are influenced by temperatures, including fill height at bottling, headspace volume and headspace pressure.

After obtaining a responsible amount of oxygen (< 0.5 ppm) in the holding tank, the winemaker’s next concern is controlling O2 enrichment at bottling. The common wine bottle contains 750ml of air and exceeds 200 mg of oxygen. This presents a problem for the winemaker, because bottling involves turbulence and pumping conditions that lead to high DO values when not managed correctly. To minimize the O2 impact, an inert gas needs to be used in transferring wine and covering unfilled tanks, flushing and also using a vacuum prior to filling. Of course, the wine must be in the right condition, with proper FSO2 content, temperature and DO content at less than 0.5 ppm.

Results showed that 85% of the wines at bottling were above the industry guideline of 1.0 ppm for dissolved oxygen (see table on page 76). The analyses for DO levels are summarized in the figure “Average Concentration of DO in Wines at Bottling” on page 79. These data illustrate the DO variability among the wines at bottling (zero days storage). Just two wineries produced wines that were below the recommended DO value of <1.0 ppm. However, seven wines were slightly above the guideline, ranging from 1.1 to 1.5 ppm, while five wines were not satisfactory.

The evolution of DO for the storage intervals (zero to 126 days) is shown in the table above, “Average Concentration of Dissolved Oxygen, Headspace Oxygen and Total Package Oxygen.” The majority of the measurements indicated a rapid decrease by seven days, and most were consumed by 63 days. The rate of decrease was illustrated among the results. A higher DO level at bottling (zero days) took longer to be completely consumed, meaning more time for chemical reactions to occur.

A real pattern was found concerning bottle evacuation effects on the DO content in bottled wines. Some important points can be observed in evaluating the DO results (averages) of wines from two filling operations, with and without inert gas and/or vacuum evacuation. For each bottling operation, automation and gravity, bottle evacuation was critical in reducing DO contents. Although gravity filling without evacuation led to undesirable DO levels in six wines with an average of 2.2 ppm, two gravity-filled wines under evacuation were near the DO guideline, <1.0 ppm. Surprisingly, these wines contained less DO than the five wines from the automated bottling operations that were evacuated with inert gas and/or vacuum expressing an average of 1.4 ppm. One wine that was automated in bottling without evacuation expressed a rather high level of 3.7 ppm DO in the bottled wine. However, this wine also contained a high level of DO at the bottling tank of 2.2 ppm.

Headspace oxygen Although the importance of DO content in wine at bottling is obvious, the second factor, headspace, is considered the main source of oxygen pickup in bottled wine. This factor is often overlooked or ignored by many winemakers. Wine comes into contact with headspace gas at bottling and sealing, and oxygen dissolves into the wine. Different closure types have an effect on headspace volume and therefore impact oxygen content in the bottled wine. For example, screwcapped bottles may have three times the headspace volume as bottles finished with cylindrical closures (corks). As mentioned earlier, the use of vacuum and flushing with inert gases (CO2 or N2) is often employed to minimize HSO. The recommended level for HSO is 2.0 ppm to resist oxidation.

The headspace oxygen levels at bottling were variable and ranged from 0.6 to 4.9 ppm (see “Average Concentration of HSO in Each Wine” on page 77). More than 65% of the wines exceeded the HSO guideline (2.0 ppm) for controlling oxidation and preserving sensory properties. The most striking observation was the extremely high HSO values for six wines (3.7 to 4.9 ppm). Another interesting finding was the HSO measurements within the suggested guideline. Six wines were found in this category, and three of these wines were filled with gravity fillers. Literature has reported that gravity filling systems have the tendency to deliver high oxygen level wines. However, results showed that wines D, E and F were handled properly and contained acceptable HSO values below 2 ppm.

The figure “Percentage of HSO in Total Oxygen Content” on page 77 illustrates the proportion of HSO to total oxygen amount as percentages in wines at bottling. Literature has shown that the HSO should be less than 67% to be acceptable. Three wineries (I, J and L) experienced very high HSO percentages (73% to 85%) and did not exercise headspace evacuation. It is interesting to note that one winery within this group utilized an automated filling line. In contrast, the lowest HSO percentage (27%) was from a gravity filling operation (D) with no evacuation. However, this particular wine contained a very high DO level (3.9 ppm), which reduced the percentage HSO to 27%. Overall, it is essential to use evacuation to remove oxygen from the bottle and headspace, no matter whether automated or gravity filling is used. The wineries demonstrated the success of evacuation in reducing DO and HSO oxygen from 2.4 to 1.3 ppm and 3.9 to 1.2 ppm, respectively.

Total package oxygen Total package oxygen (TPO) is the sum of dissolved oxygen (DO) and headspace oxygen (HSO) contents. Although the DO level is important, HSO measurements are equally decisive in monitoring oxygen during the winemaking process. The final TPO value after bottling should be below 3.0 ppm, according to available literature. In this study, TPO levels decreased rather rapidly during bottle storage (see the table on page 78). Most TPO readings indicated that oxygen consumption was completed in 63 days. However, some wines high in TPO extended oxygen loss to 126 days. It should be emphasized that high headspace oxygen was significant in delaying oxygen consumption.

Sulfur dioxide Free sulfur dioxide (FSO2) levels were measured at each storage interval. The results showed that most wines at bottling contained adequate amounts of FSO2. However, one wine was considered unacceptable (<2 ppm) at bottling. It should also be noted that two wines (B and J) did not follow the usual decrease during bottle storage. This inconsistency was most likely due to a sampling error or lack of thoroughly mixing the wine at SO2 addition.

This study showed that the rate for FSO2 decline over time was related to the initial oxygen content. For illustration, the TPO data were grouped into two categories (low and high) and compared to their respective FSO2 levels at each storage interval (see “Influence of Low and High TPO Levels on Free SO2 Content” on this page). Wines with the lowest TPO value (2.4 ppm) decreased to 31 ppm FSO2 in 126 days, a 28% drop. In contrast, the highest TPO amount (5.9 ppm) had a 42% decline in FSO2 at the same storage times. Overall, these results showed that TPO has a primary role in FSO2 decrease in four months.

The same conclusion was found in earlier studies, which also observed a gradual decline after four months. Their findings stated that another loss of 5 to 10 ppm will take place between four months and 12 months of storage. Much of this loss depends on the closure type, which influences the oxygen transfer rate (OTR). Keep in mind that 1 ppm O2 consumes 4 ppm SO2. This ratio (O2: SO2) can be used to estimate the decline of FSO2 during storage. For example, the lowest TPO level (2.4 ppm O2) in the figure “Influence of Low and High TPO Levels on Free SO2 Content” is multiplied by four (SO2) to calculate the drop of 10 ppm FSO2 at four months (126 days). Our actual reading (33 ppm) was close to the estimated value of 31 ppm. It is important to remember that the FSO2 should not fall below 10 to 15 ppm during bottle aging. Levels less than this will result in wine oxidation and sensory property faults in addition to potential microbial instability concerns.

In general, we must not overlook the importance of TSO2 in controlling oxidation. In this study, there was a noticeable trend among the TSO2 levels with storage time. Although some inconsistency was observed, results indicated a decrease during storage (zero to 126 days). An explanation for this variability may relate to the forms of SO2 in wine, particularly the bound species. These species include, of course FSO2, but also two bound forms with degrees of binding. One form has a lasting bond while the other dissociates and releases FSO2. This occurs when the initial wine FSO2 drops to a certain level and lasts until the supply disappears. All forms depend upon the wine type and composition.

A recent article from the Australian Wine Research Institute (AWRI) stated that SO2 addition to wines should yield about 35% to 40% FSO2. Failure to reach these percentages is related to small SO2 additions during vinification. It is best to treat wines with large amounts, but less frequently. This practice yields higher levels of FSO2, because a high TSO2 ratio to bound was established.

AWRI also mentioned that many problem wines were related to small SO2 additions. Their FSO2 percentages to the total SO2 were less than 25%. This outcome was evident due to oxidation and spoilage products binding with FSO2. For this study, calculating the percentage of FSO2 concentration to TSO2 revealed that five Ohio wines were near 25%. Another important aspect of TSO2 involves its influence upon wine sensory properties. Past literature has suggested that wines approaching 200 ppm TSO2 possess certain sensory faults.

pH values The pH of wines has an important influence on many phases of the winemaking process. With regard to this study, low pH values are less susceptible to oxidation and more responsive to SO2. Wine pH levels between 3.1 and 3.6 are generally recommended.

The majority of the pH values at bottling were within the acceptable range. However, wines from wineries B and D were somewhat low, 3.05 and 2.98, respectively. In contrast, winery N was slightly above the suggested limit at 3.72. Also, a trend was observed for pH levels at the various storage intervals. These data showed a modest increase in pH from bottling to the last storage period, 126 days. An average pH rise near 0.2 with a range between 0.4 and 0.19 occurred in 13 wines of this study. Explanations for this pH increase may include: sulfur dioxide loss and extent of potassium exchange with wine acids during bottle storage.

Also, the slight pH increase may suggest to the winemakers to add an extra amount of SO2 before bottling. However, most Ohio wineries are small and usually store their wines for a short time, three to four months. Thus, the SO2 results of this study indicate that SO2 levels were adequate for brief storage times. On the other hand, prolonged bottle storage may experience a significant FSO2 decrease with levels approaching the minimum guideline of 10 ppm.

Conclusion Management of DO can affect wine quality during bottle aging. Wines in this study varied considerably with respect to their DO levels in the holding tank and at bottling. Most wines at these locations were above industry guidelines for DO. For example, only two wines of 14 measured were below the recommended level of 1.0 ppm.

The same trend was found for HSO, with only six wines below the proposed level of 2.0 ppm. Also, the HSO percentage in the total wine oxygen content indicated that HS was an important source of oxygen in the wines. Most wines showed HSO percentages were above 50%, which may have a significant impact on the wine shelf life.

It is evident that most wineries in this study need to employ techniques to better manage their oxygen levels. One specific issue that needs special attention is to improve the utilization of inert gas to remove wine oxygen at various steps of the vinification process.

Another meaningful observation in this study showed that the FSO2 decrease during wine storage was related to the oxygen content at bottling. Wines with high O2 amounts experienced the greatest loss in FSO2. With this in mind, a future study may answer the question, “Can winemakers compensate for high oxygen wines with FSO2 without diminishing wine quality during bottle aging?”

We believe that the results of this study would correlate with a majority of the wines from the Ohio commercial wine industry and throughout much of the country. The study’s results provide a stimulus for winemakers to review their winemaking and bottling protocols in providing best management practices to limit unwanted oxygen pickup in addition to appropriate sulfur dioxide use at critical times. If winemakers do that, it will help put the wine industry on the right track toward producing premium quality wines with longer shelf life potential on a more consistent basis.