Research that was published on this topic in the early 2000s indicated that the amount of oxygen entering through a barrel depends on the age of the barrel, the grain of the wood, as well as the type of bung stopper used. According to these results, levels of oxygen entering a new barrel ranges from 28 – 45 mg/ℓ/year. High levels of oxygen also penetrate if a silicone bung is used. In older barrels these levels decrease to about 10 mg/ℓ/year. The usage of oak wood staves with a tighter grain was also found to increase levels of oxygen entering the barrels. These researchers indicated that 16% of the oxygen enters through the wood itself, 21% through the bung and 63% enters between the staves. However, the measurement of oxygen entering barrels has always been a challenge due to limited equipment sensitivity and the inability to measure oxygen diffusion in real time in a barrel. This led to these experiments being done on staves instead of barrels, as well as measurements only being done a few times. Theoretical oxygen transfer rates were then calculated. However, recently published data has challenged these figures and will be discussed in more detail.
Recent advances in measuring oxygen in a non-destructive way has led the way for more real-time measurements. This technology uses a principle where oxygen concentrations can be measured inside a bottle, tank or barrel without opening these containers. Eight of these oxygen sensors were placed at locations inside four new fine grain French, five new, fine grain American oak and four new medium grain American oak barrels which were filled with model wine. This model wine consisted of only 15% ethanol at a pH of 3.5 of which the oxygen was previously removed. This was done in order to measure the oxygen transfer more accurately, as no anti-oxidants were present in this model wine which could have reacted with the oxygen as would happen in real wine.
The researchers found oxygen levels to be higher closer to the bung hole, with levels decreasing as the distance from the bung hole increase (closer to the bottom of the barrel). Levels were sometimes up to 2x higher in the area closer to the bung hole. This illustrates the importance of obtaining representative samples from barrels when oxygen studies are conducted.
The sources of this oxygen going from the barrels to the model wine were initially due to oxygen inside the staves, between the joints between the staves and permeating through the staves. For the three types of barrels, the rate of oxygen transfer was the quickest in the first four months. The fine-grained American oak barrels for instance supplied the model wine with 7% of the total yearly oxygen additions within one week, 23% after one month and 50% after four months. The same trends were also observed for medium-grained American and the French barrels. The transfer of oxygen into wine from a new barrel is thus not linear. If one thus for instance matures wine A in a new barrel for four months and then replaces it with wine B for eight months then both these wines would only obtain about 50% of the total oxygen supplied from the barrel in a year. However, what was interesting was that American oak barrels yielded more oxygen to the wines than the French barrels. The calculated average oxygen additions for the medium-grained American barrels were 11.3 mg/ℓ, with that of the tight-grained American barrels being 11.6 mg/ℓ. The French oak barrels’ calculated average value was only 8.2 mg/ℓ. The higher amounts of oxygen entering through American oak barrels probably have to do with the physical structure of the wood. Tyloses block the rays in American oak, but in French oak the occurrence of tyloses is much lower. Wood from France thus has to be split in order for these rays to run parallel with the staves to prevent leakage. American oak can however, be sawed due to the higher occurrence of tyloses which leads to more rays in the staves not being parallel with the wine. The tyloses, although preventing wine leakage, are not impermeable to gasses and more oxygen thus comes into contact with wine in American oak barrels, compared to French oak barrels.
These results shows much lower levels of oxygen permeating into the model wine than previously thought (around 8 – 11 mg/ℓ/year, versus around 30 – 45 mg/ℓ/year). Micro-oxygenation is a process where small controlled amounts of oxygen is added to wine in tanks to simulate the amount of oxygen coming into contact with wine through a barrel. In this process up to 4 mg/ℓ/month can be added to the wine (around 48 mg/ℓ/year), but these dosages thus need some re-thinking considering the above-mentioned results.
These researchers also found more oxygen to permeate into the model wine in the first few hours of the trails, but this effect was short lived and at the end of the one-year trail the calculated oxygen transfer between the tight and medium grain American oak barrel were the same. These results thus disproves the theory that the grain plays a large role in the amount of oxygen entering a barrel. Past research only measured oxygen transfer into wine over a relatively short period of time and also found that tight grain leads to more oxygen permeating through the wood, but the recent results showed that oxygen permeation should be followed over a prolonged period of time.
The oxygen transfer rate is also to a large extent governed by the humidity of the wood. Researchers have calculated that it takes about 82 days after filling with wine for the water saturation front to occur in staves in new barrels. Oxygen travels much slower through a liquid than air and after this period the rate of oxygen diffusion will decrease and thus reaches its minimum. Toasting and bending of rough staves also reduce the oxygen transfer rate with up to 4.5-fold once they have been assembled into a barrel.
Using these results researchers were able to construct barrels with a low oxygen transfer rate and a high oxygen transfer rate. This rate was twice the amount in the high oxygen transfer rate barrel, compared to the low oxygen transfer rate barrel. The classification of staves according to how much oxygen they can transfer seems like a future option for coopers to deliver a barrel to winemakers with certain characteristics, of which one could include its oxygen transfer rate.
Du Toit, W.J., Lisjak, K., Marais, J. & Du Toit, M., 2006. The effect of micro-oxygenation on the phenolic composition, quality and aerobic wine-spoilage micro-organisms of different South African red wines. South African Journal of Enology & Viticulture 27, 57 – 67.
Nevares, I. & Del Alamo-Sanza, M., 2015. Oak stave oxygen permeation: a new tool to make barrels with different wine oxygenation potentials. Journal of Agricultural & Food Chemistry 63, 1268 – 1275.
Nevares, I. & Del Alamo-Sanza, M., 2015. Recent advances in the evaluation of the oxygen transfer rate in oak barrels. Journal of Agricultural & Food Chemistry 62, 8892 – 8899.
Vivas, N., Debèda, H., Ménil, F., Vivas de Gaulejac, N. & Nonier, M.F., 2003. Mise en évidence du passage de l’oxygène au travers des douelles constituant les barriques par l’utilisation d’un dispositif original de mesure de la porosité du bois. Premiere resultats. Sciences des Aliments 23, 655 – 678.
– For more information, contact Wessel du Toit at firstname.lastname@example.org.