For over 2000 years oak has been a fundamental component in the production, maturation and transport of wine. First used by merchants as a vessel for transporting finished wine, its capacity to transform the liquid within was in fact a serendipitous byproduct of its first purely practical use. Decisions associated with the preparation and use of oak in winemaking are arguably some of the most influential in defining style, be that regional, site-specific or that of an individual vigneron. Often under-appreciated, the journey of the oak barrel from forest to cellar reveals an intricate and complex story, one of nature, craft and science.

History

As the Roman empire expanded aggressively across the globe they encountered a number of diverse cultures, many of them possessing technologies which the Romans ’embraced’. When the Romans encountered the Gauls they found a people who were transporting beer in wooden barrels, bound together with metal hoops. The Romans themselves were aware that earlier civilisations had used palm wood barrels to transport wine, Greek historians had noted that ancient Mesopotamians used barrels made of palm wood to transport wine along the Euphrates. Despite this, until their encounter with the Gauls, amphorae had been the Roman’s transport medium of choice.

For a number of reasons oak became the preferred material used when building barrels. Oak wood is much softer, meaning that it only needed minimal toasting, and thus was easier to bend into the traditional barrel shape than palm wood. This ease of use meant that total production time was significantly reduced compared to palm wood. Additionally, oak was and is abundant in the forests of continental Europe, meaning less transport was necessary. Finally, oak, with its tight grain, offered a waterproof storage medium.

By the first century BC, the Romans were distributing wine throughout the Mediterranean. After transporting wine in barrel for extended periods of time, they began to realise that these oak barrels imparted new, pleasant qualities to the wine. They wrote that the oak contact made the wine softer and smoother, and with some wines, it also improved their taste. Over time, merchants, wine producers, and armies alike, found that the longer the wine remained inside the barrels, the more qualities the oak would impart to the wine, and thus grew the importance of oak in winemaking.

Forestry

Generally speaking there are three primary sources of oak for production of barrels for use in winemaking. These are France, America and several nations across Central and Eastern Europe, which for the purpose of this article I shall bulk together. In America the oak species used in making barrels is Quercus Alba and in much of France and Central/Eastern Europe it is Quercus Patraea with a little Quercus Robur in the forest of Limousin, France. The species and origin of the oak widely determines the physiological composition of the wood and thus the way in which it will influence a finished wine.

There are five forests in France from which oak barrels intended for winemaking are sourced; Allier, Limousin, Nevers, Tronçais and Vosges, almost all were planted in Napoleonic times. In America oak is grown primarily in the states of Virginia, Missouri, Kentucky, Oregon, Ohio, Minnesota and Wisconsin, as well as in California. In Central and Eastern Europe the nations of Croatia (mostly the region Slavonia), Slovakia, Hungary and Romania are preferred.

It takes at least 80-90 years of growth before American Oak is ready for harvest and 150-200 (300 in extreme cases) years for French and Central/Eastern European oak.

Oak trees intended winemaking are often grown in stands, this enables them to grow upward rather than outward creating tighter pores in the oak. Trees considered most suitable and more often chosen have tall straight trunks, unblemished wood, and are free of low-hanging branches. For this reason dense forests are preferential, this density forces the tree to grow straight and tall leading to higher quality lumber for barrel staves. In addition to managing planting density, undergrowth and underbrush are also often cut away meaning trees are not competing for sunlight and can grow as desired.

More often than not, particularly in Europe, there exists organisations who oversee forests. These organisations establish harvest date, determine harvest quantity and oversee the general welfare of the forest (less common in the US) An example of such an organisation is the Office National des Forêts, their purpose is to ensure the longevity of the forest, they take great care selecting trees suitable for felling. The ONF will auction chunks of forest to wood negociants who will in turn nurture their portion of the forest for a designated period of time. Prior to handing the plot back to the ONF the negociant, working with coopers, will highlight the wood they wish to fell.

Throughout the time spent overseeing their plot of wood coopers will visit wood negociants and select oak, this selection will depend on a range of quality factors. Jeremy Seysess (Domaine Dujac) told Levi Dalton he once accompanied a cooper to a forest and had quickly realised it would take him a lifetime to learn this craft.

Recent advancements in technology have allowed cooperages like Seguin Moreau to more consistently select wood using objective data. Draft maps showing enological potential of oak based on quality-data gathered over 15 years of purchasing now direct their procurement policy. This data-led procurement allows them to consistently choose wood based on quality opposed to subjective characteristics such as historical significance of place.

Harvest of oak intended for winemaking takes place in the winter months when there is less sap in the trunk and once harvested a single cubic log of French oak will produce on average two 225-litre barrels, with American oak able to yield a little more.

Basic composition and chemistry of oak

Before exploring in more detail the decisions facing a vigneron when choosing both the source (and by default, the species) of their oak and the points at which to use it I would like to add context to the more technical component by establishing the fundamental chemical composition of oak.

Large, open, water-conducting vessels called xylems compose most of the wood (also present in grapevines). There are also tyloses, these are balloon-like outgrowths of cells that bulge through the pits of xylems and block water movement. American oak has a higher density of tyloses compared to French meaning it can be cut across the face of the log opposed to down the medullary ray. Coopers work with the dead, tough, heartwood to produce staves, this wood is split or sawn (depending on whether the wood is French or American) down the medullary (or vascular) rays which run outward from the centre of the tree (visible on the below diagram)

Around 40-45% of the weight of the raw material used to produce a barrel consists of cellulose, 25-35% hemicellulose and 20-25% is lignin. Phenolic compounds such as volatile phenols (such as eugenol) and gallotannins and ellagitannins make up the rest. Cellulose has little effect upon the flavours or aromas of a finished wine; however, it can provide a food source to spoilage microorganisms when toasted. Hemicellulose and lignin, otherwise known as “wood biopolymers”, are the aromatic precursors of the toasting process. During toasting their transformation under heat produces a range of aromatic molecules including furan phenol aldehydes.

Ellagitannins form a buffer to protect the sensory profile of the wine from its surroundings. They help its aromas to open up, by limiting the formation of sulfur compounds during fermentation (especially in white wines), then during ageing protect it from oxidation by fixing the oxygen.

Also present are quercus triterpenoids, ethyl-octalactone and methyl oak lactones. Quercus triterpenoids play an important role in the sensory profile of wines aged in barrels and are responsible for the sensation of “sugariness” in dry wines. Ethyl-octalactone are essential compounds for the expression of woody aromas in wine. Although naturally present the oak also contains precursors of this molecule, which can increase the lactone content of the staves during maturation, toasting and even throughout the ageing process of the wines in the barrel.

Primary compounds associated with oak flavour impact in wine are as follows:

The overall composition of oak will depend on species and origin. The way in which these molecules interact with wine will depend on a range of preferences such as toasting, size of barrel and time in barrel

Source variation

When considering their use of oak, the vigneron first must decide which species, and from which region (although the two factors are somewhat entangled) they are to opt for. This decision is one of both economics and desired style. Both the species and the source of that species will determine the chemical composition of that wood for similar reasons to the terroir in which a grape grows namely altitude, climate, annual precipitation etc.

American, Central/Eastern European (CEE) and French oak boast a number of fundamental variations which affect both the way in which they interact with wine and thus the profile of a finished wine. Predominately these are concentration of oak lactones, coarseness of grain and amount of ellagitannins. Studies have shown that wines aged in American oak display higher concentrations of oak lactones than those aged in French oak, this heightened concentration influences the intensity of associated vanilla aromas. Studies have also shown that in addition to higher concentration of oak lactones varying levels of eugenol (spice aroma) exist between French and American oak.

The difference is grain coarseness between French, CEE and American oak is vast with French oak having a much coarser grain than American oak. This tighter grain, although counterintuitive, is more permeable to oxygen. Despite this permeability American oak has a higher OTR, mostly due to how the wood is cut. As such French oak provides a more stable environment (with managed micro-oxygenation) in which to age wine contributing more to structure than flavour.

French oak also has a significantly higher concentration of ellagitannins (extractable polyphenol content) than American oak, leading to greater structure opposed to increased aromatics. Additionally ellagitannin concentration correlates with astringency, with a higher concentration of ellagitannins resulting in a heightened sense of astringency. This may partially explain why French oak appears to suit longer periods of ageing. If I were to summarise crudely the difference in impact upon wine of these oak species it would be that French oak offers less intense vanilla characters and leans more toward subtle extraction of spice, offering enhanced mouthfeel and for the grape to express place. Meanwhile American oak imparts a much more immediate vanilla and coconut character suiting much more impactful reds like Cabernet Sauvignon, the increased OTR also softens rough tannins.

Cost is another important factor when making source decisions (cost can also determine amount of new oak of oak) Fundamentally French oak barrels are in the most part substantially more expensive than American. This increased cost is a due to a number of factors. The tighter grain and less watertight nature (less tyloses) of French oak obliges coopers to split the wood along the grain, means only 20 to 25% of the tree is available. Coopers are able to serrate or saw American oak making it at least twice as economical. The cooper may also season American oak in a kiln, further speeding up the process; however, many coopers opt for identical seasoning whether French or American. For those wanting to really explore the economics of barrels Vine Analytics have put together an awesome piece which details everything from use of new oak (used oak imparts less flavour) to considering evaporation, down to calculating cost per bottle of ageing.

GuildSomm is a great reference for exploring French Oak in a little more detail.

Production and preferences

Once the vigneron has decided upon the woods provenance the barrel must be built. Throughout the production process there are decisions to be made by the vigneron, decisions which will alter the way in which the wine interacts with the barrel. A skilled craftsman called a cooper (in some cases a master cooper) is responsible for the production of the barrel.

Barrel construction begins with staves, cut from the heartwood of the tree a stave is a long pieces of oak which forms the shell of a barrel. Stave shape is important, the staves must form a watertight seal once the cooper brings them together Modern cooperages cut staves using an EDJ laser-guided jointer, this allows the cooper to cut the wood perfectly with the wood roughly 1.5cm thicker at the bilge than it is at the ends of the staves, this is to more evenly distribute pressure and increase strength. The width of staves is not always uniform, barrel constructions employs a variety of widths, this results in less opportunities for leakage and a more perfect shape.

Once cut, prior to use, staves are stacked and seasoned, this process can take around 2-3 years. This seasoning encourages physical and chemical transformations to reveal the woods best expression. Proper seasoning requires a specific method of stacking in order to allow air to circulate freely between the pieces of oak, and successive periods of watering and more air-drying. Wood arrives with 40-70% moisture, but is not worked until moisture is around 14-16%.

Once constructed the barrel can only expand outwards, never lengthways. So as the the staves are brought together they are softened using water and fire (to reduce risk of damage when bending) and metal hoops or cables used to pull in to position. If there is too much pressure once the barrels reaches the toast fire, sugar wills not travel through the stave and out, instead they will build up and form blisters.

Heads of the barrel are either cut to shape at the cooperage or received already cut. It is worth nothing that unless the winemaker requests otherwise the heads (which make up 1/3 of the barrels interior surface area) are not usually toasted in French oak barrels but are in American oak, this is due to the American oak being stronger and more sugars are toasted out, softening the effect on a wine. Once the barrel is partially constructed it is toasted over an open flame (fuelled by ‘scrap’ oak only) the toast will act as a ‘skin’ between the wine and the barrel. The level of toast is decided upon by the winemaker with the help of a cooper and has a profound impact on the influence on a finished wine. The longer a barrel toasts over a fire the deeper the toast is penetrated, for this reason a slow fire is essential, a high fire will colour the barrel but as the wine goes past that toast level the influence is not adequate.

After toasting is complete the hoops (truss hoops) are pressed on to form the shape of the wood, as the barrels cool down each end of the barrel is cut with a groove on the inside, where the head is inserted. A bung hole is then drilled, the largest stave is not always chosen due to the difficulty in replacing a large stave. Once constructed the external surface of the barrel is sanded, with the hoops than reapplied.

Jamie Goode has put together a helpful illustrated guide to barrel construction and Episode 13 of The Inside Winemaking Podcast will give great insight.

Fermentation in barrel

In recent years fermentation in barrel opposed to stainless steel has grown in popularity. Both the alcoholic fermentation and malolactic can be carried out in barrel and whilst this can present unique challenges to a winemaker the impact on a wines structure and mouthfeel can be stark.

As Levi Dalton pointed out to me it is important to distinguish between fermentation in wood and fermentation in barrel. For example, at Domaine de la Romanée-Conti grapes are fermented in open-top wooden tanks, this differs from fermentation taking place in a closed oak barrel. Parke Hafner talks in length about the benefits of barrel fermenting, in particular when working with Chardonnay, he had first witnessed this at Domaine des Comtes Lafon, in Meursault. The process requires substantially more labour than a single fermentation in a large tank; however, the process is said to better integrate the oak and the wine, presumedly due to the surface area, temperature and the chemical reactions taking place during fermentation encouraging integration opposed to ageing post-fermentation.

Ageing

Whether alcohol and/or malolactic fermentation has taken place in barrel, steel or some other vessel a large percentage of wines will spend a period of time post-fermentation ageing in barrel. Having already covered the chemical composition of oak and the impact of the most of these compounds on a wine the vigneron then has the decision as to how long they wish to age their wine and in what size barrel.

The length of time spent in oak will depend on a range of factors namely regional laws, desired style, quality of vintage and vigneron intuition. Some regions, such as Barolo, dictate a minimum ageing require to label a wine Barolo, others do not. Some producers may feel their wines require longer ageing based on desired style or site, a particularly vintage which leads to notably coarse tannins may encourage a producer to increase the length of time spent in barrel to ‘soften’ (polymerise) these tannins.

If you have spent much time visiting wineries, you will have seen in their cellar a whole host of different size (and shape) barrels. The size of the barrel will ultimately determine to what extent the barrel influences the finished wine, with particular regions leaning toward larger barrels and others much smaller. In Barolo, traditionalists opt for large Slavonian botti (which they reuse many times) whilst modernist opt for much smaller barrique, the former thought to better express Barolo as a place and not the character of the barrel. In the Mosel, where a percentage of Riesling are put through oak ageing, producers opt for foudre, often from the cult cooper Franz Stockinger who mostly uses wood from German and Austrian forests.

Fundamentally the size and shape of a barrel affects the surface area of wine which is both in contact with the wood and directly exposed to micro-oxygenation, thus ageing the wine, and impacting its flavour, at differing rates. A 60 gallon barrel has a surface area of 51 inches per gallon of wine, take a 132 gallon barrel and the surface area is 39 inches per gallon of wine, thus less wine is in contact with the wood and the flavours imparted will generally speaking be less prominent. The case is similar in regards to micro-oxygenation, if you use a larger barrel the wine receives proportionately less oxygen than it would if it in a smaller barrel, this is because there is a lower ratio of surface area to wine volume. Foudre are thought to age wine more slowly, this could be due to shape and thickness of staves.

Potential spoilage

A properly maintained oak barrel will offer many years of trouble free use; however, barrels are particularly good breeding grounds for bacteria and other organisms, especially where care and maintenance is insufficient. Common spoilage issues include;

• Penicillium mould is one of the the most common spoilage problems found in barrels in both wine and beer production and can be very difficult where it becomes widespread. Typically the mould will grow through stave or head joints and around the barrels bung hole.
• Acetobacter is also fairly common and causes alcohol to be converted to acetaldehyde and then acetic acid, it may also form ethyl acetate. This can occur where wine oxidises in barrels which offer headspace. This volatile acidity (VA) as it is known is in the most part considered a fault and can be detected by aromas of vinegar (acetic acid) and nail varnish (ethyl acetate) Winemakers have pointed out that proper barrel maintenance programs yield less VA.
• Spoilage yeasts are yet another source of problem associated with poorly maintained barrels. The most common of these spoilage yeasts is Brettanomyces; however, both Candida and Pichia have been identified as contaminants in the interior of wine barrels. These spoilage yeasts, as a result of volatile phenols, impart often undesirable aromas to a wine residing in barrel.
• Lactobacillis and Pediococcus otherwise know as lactic acid bacteria (LAB) have the potential to cause serious wine spoilage. Some species of LAB are responsible for conducting MLF; however, many do not. LAB are able to metabolise sugars, acids and other constituents in wine and can produce several undesirable compounds. Wooden barrels, when not properly cleaned and sanitised, can be a source of LAB where they have contained wine which has undergone, or is undergoing, MLF.
• TCA in new oak barrels is a tempestuous topic. Research (although unable to isolate the source of continuation) has suggested that the extent of the problem is grossly under-estimated by coopers. However, this claim has been rubbished by a number of coopers who suggest just 0.04% of barrels produced in the last three years were suspected of being infected with TCA. Whilst research is limited I have included TCA for your thought and consideration.

Maintenance and cleaning

New barrels ought to be properly prepared and used barrels should be cleaned, stored, maintained and prepared appropriately in the winery. In establishing proper practise it is important to draw a distinction between cleaning and sanitisation. Prior to sanitising it is important to have an effective cleaning schedule in place in order to remove dirt and debris.

At Mondavi, Russ Rosner, production enologist, has his team fill new barrels with cold water for one day, then inspect for leaks. The barrels are then emptied and sulphured, then later filled with wine. At Matanzas Creek, Dave Ramey is trialling hot water, cold water and no treatment at all. This purpose of this fill is to encourage the barrel to seal itself, avoiding future costly wine leakage.

After use barrels should first be thoroughly drained and then any debris and or tartrate build cleaned up before proceeding with sanitation of any sort in order to maximise efficacy of any agent or solution applied. Pressure washers, a barrel cleaning nozzle known as an oscillator, and steam are typical options available to winery staff when cleaning barrels.

It is not usually recommended to store used barrels dry for long periods of time, wineries can use a sulphite and citric acid solution for long-term storage. Wineries that store their barrels dry must rehydrate them prior to filling with wine. Steam or clean water (hot or cold, overnight) are adequate rehydrating agents. A fairly decent guide to maintaining barrels is provided by The Beverage People.

It is important when cleaning the winery not to use any chlorine based cleaning products. Presence of chlorine is one of the two major contributors to the production TCA, the compound responsible for cork taint.

Sanitisation

A paper from Cornell’s Viticulture and Enology Program demonstrated the efficacy of a number of treatments in sanitising barrels. The experiment used 100 “naturally contaminated” barrels from a number of different wineries and compared five treatments: sulfur dioxide, peroxyacetic acid, steam, chlorine dioxide and ozone. The paper found that sulfur dioxide, steam, ozone and PAA at higher concentrations all showed to be effective sanitisers.However, lower concentrations of PAA and chlorine dioxide did not significantly reduce the number of spoilage organisms.

An additional study sought to assess the effectiveness of aqueous steam, UV irradiation, gaseous O₃ and aqueous O₃. Steam and O₃ were the most effective treatments, eliminating about 70% of yeasts. UV appeared to be less effective, potentially due to the porous nature of wood, which protects cells from direct irradiation. Additional research is also underway exploring the use of microwave technology in removing spoilage. Additional secondary benefits to using Ozone in barrel sanitisation are its ability to combat airborne TCA in the winery. Whatever sanitisation programme is decided upon, this ought to be repeated on a regular schedule (usually every 2-3 months) for any barrels stored wet or dry without wine for an extended period of time.