Join Mailing List Login
Bedrock Wine Co.
Home Menu



Part of our winemaking deep-dive series, this was originally sent in our Winter 2020 Release letter.

The Boys with barrels

Joel, Morgan and JP in front of two 500 liter puncheons and three 600 gallon/23 hl foudres, and to the right: stacks of 228-L barrels.

One of the questions that most frequently comes up at tastings and wine dinners (ahh, remember those?) is about how much new oak we used on a particular wine. After hearing the percentage, and depending on whether it meets the questioner’s definition of appropriate use, they nod in approval or disapproval, and we each go on our merry ways. Sometimes this question is followed by the additional “French or American?”, but the conversation rarely proceeds from there. These truncated conversations are a bit of a shame in my mind, as the conversation around cooperage use in the winery is both remarkably important to the quality of wine and immensely complex. As I may not be able to go all the way down the cooperage rabbit hole at tastings and dinners, I thought I might like to do so here. Cooperage is not merely a flavoring component in wine but is also a key player in sculpting a wine via lees contact and oxygenation, even when adding no flavor. Thus, the type and thickness of wood and size of vessel is also quite important.

First, I am using the term cooperage here because “oak” in and of itself is a bit too simplistic. Oak, and the several species of it that are used, is the dominant wood used for aging wine, but there are places for acacia (actually black locust), cherry, almond, redwood and others (our experience with these is relatively limited).

Within oak itself, there are a number of species, each of which has an impact on quality and impact based on the physiology of the tree itself. For instance, “American Oak” is produced mainly from Quercus alba (white oak) and is physiologically quite different from the Quercus robur and Quercus sessilis that grow in Europe. Quercus alba grows with a cross-hatched grain, which both allows the wood to be sawed and accounts for its generally more aggressive fingerprint. European species grow in a single grain that requires splitting and tend to be a bit less impactful (speaking very broadly). In the cellar, it is easy to ID the American oak barrels because they are much denser than European oak and weigh substantially more.

Beyond the individual species of oak, just like with grapevines, the origin of the wood is as important as anything else. Just like how Zinfandel vines grown on well-drained, vigor-limiting soils produce higher quality wines than the same grape grown on fertile, highly irrigated soils, oak is highly dependent on its individual “terroir.” In general, the most prized oak tends to come from cool areas with soils that slow the growth of the tree, resulting in tighter grain and a more elegant wood expression. This is the reason you hear people speak reverently of the French forests of Tronçais, Jupilles, Allier, Nievre, Vosges etc. (and

more and more about the enormous potential of some Eastern European locations). It is also the reason why wood from the warmer and more humid region of Limousin, which produces a more open grain, is generally considered more suitable for Cognac and Armagnac production than wine these days. That said, I think too often the conversation, even among winemakers and barrel vendors, gets stopped here as even within forests there is enormous variability. Just like any natural setting, oak forests contain subtle contours of landscape, streams and wet areas, pockets of fertility and excessive paucity of soil, all of which mean that tree selection itself, even within the context of a “great” forest, is very important. I mean, La Tache and La Grand Rue are both “Grand Cru” vineyards in Vosne-Romanee, but I would want grapes from La Tache 100 times out of 100. This is where the skill of the cooper and their capacity to both select wood (and to navigate the complex web of how oak is sold for French oak) becomes very important.

Baby sitting on a barrel.

JP not sure about Esola in a new 228-liter Taransaud barrel.


Coopering, even after the complex process of selecting trees is done, is an enormously intricate process involving a number of variables - both controllable and not.


Coopers vary in size, from the behemoths like Seguin Moreau and the Charlois Group (owners of Ermitage, Berthomieu, Leroi, Saury and more), which produce tens of thousands of barrels each year, to tiny, artisanal coopers that make far fewer barrels, with each house tending to have its own stylistic impact. Each has its benefits and drawbacks- the larger coopers have greater R&D budgets and can work to maintain consistency in style and wood sourcing but perhaps don’t match the best small coopers in the height of quality. We have also seen coopers that attempted to grow quickly move to a much toastier style- in my opinion, as a way to cover up the use of lower quality or less seasoned wood—similar to what we see in large scale coffee roasting operations. The best coopers find the balance between consistency and quality.

In addition, most coopers originated making barrels for a particular region and expanded out from there. In America, most cooperages started producing barrels for bourbon and spirit aging, and in France there are generally ‘Burgundian” coopers (Francois Freres, Damy, Cadus, Gillet, and many more) and “Bordelaise” coopers (Taransaud, Boutes, Sylvain, Nadalie etc.), along with a few from the Rhone, Champagne, etc. Though these barrel houses now generally offer barrels in both traditional Bordeaux (225 Liter) and Burgundian (228 Liter) shapes, their overarching style can be somewhat related to the wines they originally made barrels for (Cabernet versus Pinot).

Once one gets past the element of large versus small coopers and the selection of trees and forests themselves, there are the elements of coopering that are the ultimate deciders of style and quality.


After splitting, it is necessary to season and dry the wood to help wash out more aggressive compounds that tend to create greenness and bitterness. Most staves are aged a minimum of 18-24 months, with the highest quality barrels seeing 36-48 months. During this process, the wood is generally stacked outside to be passively exposed to the elements for the

period of aging. This alone creates many variables. How tightly the wood is stacked or how frequently the stacks are shuffled and restacked influences the amount of sunlight, rain, etc. the wood is exposed to. Tightly stacked piles that are not shuffled often will season at a slower rate than wood that is more openly stacked. Beyond this, the location of the stave yard and the weather over the seasoning period are also critical variables. A cooper located in Australia that receives 10-12 inches of rain a year is going to have a much different seasoning profile than one located in a wet, maritime area of France. Throw a few years of drought or heavy rains and that further increases potential variability in the seasoning process. It is for this reason that we tend to work with longer seasoned wood- we feel the additional expense is well worth the mitigation in variability.


After the wood is selected, split, and then aged for 2-4 years, it is finally ready to be coopered into a barrel. Since the staves are straight, the wood must be heated and bent. There are three main sources of heat for this process- fire, water and steam. Broadly speaking, we have seen that fire-bent staves tend to play up toasty and vanillin compounds (furfural, guaiacol, syringol, etc,) in the finished barrel while staves that are water- and steam-bent tend to have a lower tannin profile, as many types of tannin are highly soluble in water. Since we tend to prefer a lower toast profile on our wines, we tend to lean towards barrels that have been water bent- though this is highly cooper-dependent.


The final process is the actual toasting of the inside of the barrel, almost always done over fire. As is the theme here, there are many different variations of how this is done. The main variables in this process is the length of toasting time and the heat of the fire as the combination of these two elements has a direct effect on the flavor profile of the wine. Some coopers opt for longer toasting times over a lower fire, which makes for a deeper but often times more subtle imprint. In contrast, a hotter fire for a shorter period might be a far more dramatic barrel, though perhaps some of the more aggressive and greener compounds are less resolved. Though almost all cooperages offer a similar spectrum of toasts – Low, Medium, Medium Plus, Heavy (or some proprietary name for each) each is very relative to the house style itself. For instance, we might find that an Ermitage “Medium toast” barrel is far toastier and impactful in profile than a Taransaud "Heavy Toast” barrel.


Joel, JP and Morgan in front of a 1600 gallon/60 hl foudre filled with 2020 Katusha's Zinfandel eventually bound for the Schmiedt Road bottling

Joel, JP and Morgan in front of a 1600 gallon/60 hl foudre filled with 2020 Katusha's Zinfandel eventually bound for the Schmiedt Road bottling

Once all of these other elements are taken into account, one then needs to look at the effect of barrel size and shape. The most commonly used barrels are around 225-228 liters in size depending on shape (Bordelaise v. Burgundian), but barrels are also available in 300 liter, 400 liter, 450 liter, 500 liter, and 600 liter iterations as well depending on the cooper. Larger barrels are generally called puncheons up to 500 liters, while 600-liter barrels are generally called demi-muids. Though all of the same variables when it comes to oak selection, aging, bending and toasting remain for these types of barrels the added size effects the ratio of oak surface area to wine. Larger barrels, when new, tend to be less impactful as the amount of oak in contact with the wine goes down. Also, because of this the amount of oxygen micro-dosed into the wine through the staves goes down as well, making for a more “reductive” aging environment. We find that wines remain fresher and more primary in the large format barrels and overall evolution during elevage slows down.

Another factor here is stave thickness. Most standard size barrels work with staves somewhere in the realm of 27 millimeters in thickness, though certain “thin stave” Bordeaux barrels made for high tannin reds can be as thin as 22 mm to accelerate oxygen ingress and resulting “softening” in the wine. Most puncheons also use 27mm staves, though it is not uncommon to start seeing 32-40mm staves on 500- and 600-liter options. As you might imagine, the thicker the stave is, the more oxygen is withheld from the wine inside, resulting in a more “reductive” environment. One of our favorite barrels for Zinfandel, Grenache and even lighter weight Syrah are these larger barrels, as they retain freshness with limited oak and oxygen impact that we feel helps preserve the fingerprint of origin in the wine better.

Finally, the shape of barrel is also important. Though most barrels are pretty standardized one exception is the “cigare” barrel. Originally developed by Didier Dagueneau for his benchmark Pouilly-Fume wines, the barrels look as if one took a small puncheon and pulled it by its ends outward to lengthen it, thus the name cigare. The brilliance of this is that it is a larger barrel (320 liter), which makes for a more reductive aging environment for white wine while also cutting down on overt oak character but also creates a larger surface area for interplay between the wine and the lees. Since lees are also oxygen scavengers, this aids in making for a more reductive environment in the barrel while also allowing some of the benefits (richness of mouthfeel) of greater lees contact. Over the years we have moved to more and more of these barrels for our Sauvignon Blanc and Semillon (along with the brilliant barrels of the Austrian Stockinger cooperage).

After the different options for barrels, puncheons and demi-muids one then moves up the size ladder into foudres and cuves. Foudres are large ovals or rounds that essentially act as aging tanks for wine. They vary in size as well from roughly 1000 liters to extraordinarily large. These, unlike barrels, are considered cellar mainstays and hopefully have a working life, barring an unfortunate microbial infection, of decades in the cellar. Outside of the first couple of fills the goal of these is to be as inert as possible and to impart little oak flavor. At Bedrock our smallest foudre are 1000 liter (made by Matthias Stockinger) and our largest is 6000 liter (60 hl or 1600 gallons) with a lot of our wine aged in 22 hl (600 gallon) foundres made by the Burgundian Tonnellerie Rousseau.

Joel, JP and Morgan in front of Stockinger 1000L and 1500 Liter ovals and behind 500 liter puncheons

Joel, JP and Morgan in front of Stockinger 1000L and 1500 Liter ovals and behind 500 liter puncheons


The impulse of using too much oak in winemaking is alluring. Let’s face it, there are only so many decisions winemakers really get to make, and oak use is one of the most impactful on the finished wine. Though I have always been drawn towards high quality French oak barrels, I definitely preferred a heavier toast imprint on the earlier Bedrock wines. As we have tasted through verticals of our wines, I think that in some cases this worked quite well and in others, not quite so much. Over time, as we come to understand our sites better and also tend to value freshness a bit more than ripeness, we have moved to a bit less new oak character in the wines. On top of this, we have generally moved to a larger proportion of larger format oak and foudres where we feel like the vineyard demands it. Speaking broadly, we find that our vineyards planted on sandier soils such as Evangelho, Pato, Schmiedt Road, etc. along with those in Amador County on granite, tend to be much more sensitive to oak presence than vineyards such as Bedrock, Monte Rosso, Nervo etc. As a result, our wines from Evangelho rarely see more than 10% new oak with a higher quotient in large format barrels and foudres, while a vineyard such as Bedrock might see closer to 30-35% new oak with a smaller proportion of large format. What is clear is that it is an ongoing conversation between the vineyard, the many variables of barrel production, and our personal evolution in winemaking sensibility- something that makes for a lifetime of interest.


Whole-Cluster Use

Part of our winemaking deep-dive series, this was originally sent in our Fall 2020 Release letter.

Grapes Fermenting Grapes fermenting into wine

Once, before the development of electricity and machines that used beaters and cages and rollers and crushers, one must imagine that almost all wine was made using whole clusters— that is, grapes were fermented in a mash that included their stems (…he writes, imagining the look of a scandalized Bordelaise oenologist…). Over the last century, as technology and science have opened up a new world of winemaking possibilities, the use of whole-cluster during red fermentations became nearly abandoned outside of a few outposts of ultra-traditionalism. Though there has been a bit of resurgence in fashion for the practice in California over the last decade, it is still orthodoxy in many regions and anathema for certain varieties, according to some (“report him to Rolland for punishment,” says the scandalized Cabernet oenologist). In some ways this makes sense, as the last four decades of winemaking research have been largely about maximizing color and extraction while removing greenness—all things that whole-cluster use can be at odds with.

Like most things verboten, it turns out that whole-cluster use during fermentation can be delicious and fun! When done well, whole-cluster can amplify perfume, increase structural complexity, lower alcohol in the finished wine (stems absorb some ethanol), improve fermentation health by moderating fermentation heat, and increase a wine’s overall character. Used incorrectly, it can push wines to becoming thin and bitter, cause green character to overwhelm the wine (particularly in under-ripe fruit), cause pHs to rise due to additional potassium coming from the stems, or if used for pure carbonic maceration, potentially overwhelm the wine with the bubble gum/banana Runts character. Talking about “whole-cluster” use in a monolithic way would be like talking about oak use, elevage, or any other facet of winemaking as a single thing— and like all things, it can be used in good and bad ways. However, one only has to look at some of the finest users of the method—Dujac in Burgudy, Chateau Rayas or Henri Bonneau in Chateauneuf, and Allemand in Cornas and Jamet in Cote Rotie come to mind—to see the enormous complexity than can be imparted by the practice.

Outside of Zinfandel (and this is changing), we use whole-cluster on virtually all red varieties that come into the winery—Syrah, Carignan, Mataro, Grenache, Cinsault, even a little on Cabernet (take that, imaginary Bordelaise oenologist!)— when we feel like we want to play up perfume, spice and structure in the finished wine.

The first decision to make is how much whole-cluster to use. We make this decision based on what we are seeing in the vineyard, how ripe stems and grape tannins appear to be, and experience with a given vineyard. We find that certain varieties and sites have tipping points—up to a certain point cluster use enhances varietal character, perfume, and mouthfeel but beyond that begins to diminish those same characteristics. An example of this would be Carignan, where 30-60% cluster inclusion can be of great benefit depending on site and year but much beyond that brings about a distracting coarseness. In Northern California many grapes are grown on potassium-rich soils, which means that adding a fair amount of whole cluster can have a dramatic effect on a finished wine’s pH. Grape stems hold potassium, and as they break down during fermentation, they will buffer acidity and cause pH to rise. In this situation, what you might gain in perfume and structure, you may lose in freshness and natural acidity. Thus, for a high pH site like Hudson Vineyard, we have generally included less whole-cluster over the years, whereas a site like Bien Nacido (grown on limestone and granite with very little potassium), we can use plenty of cluster knowing the pH won’t be that affected.

To tread or not to tread? Once we have decided to include some whole-cluster in a ferment, we then have the secondary decision of whether we want some of the grapes crushed or if we want them as whole as possible. We find the latter option brings out more juiciness, aromatics, and spice in the finished wine as many of the berries will go through carbonic maceration in tank, bringing with it the characteristic lift and ebullient aromatics. However, too much carbonic in some wines makes them just smell like carbonic and not like where they came from (it can be a leveler, just like too much oak), and the amount of carbonic imprint really varies based on the variety and site. For instance, 50% carbonic on Bechthold Cinsault makes the whole winery smell like Hubba Bubba and Banana Now & Laters, while the same percentage of carbonic on Hudson Syrah merely tips the aromatics forward a little bit. We tend to trod most of our lots as we find we achieve a nice balance between the fruit that remain intact and that which breaks down more quickly in fermentation—for us this lends to wilder aromatics, perhaps a bit more spice, and overall a less confectionary imprint on the finished wine. Again though, every lot is different—100% whole-cluster, foot-trod on Evangelho Carignan will turn it into a hardened beast (trust me, we have done it), while 100% on Bien Nacido Syrah makes the aromatics soar and the structure relatively harmonious.

Like everything in winemaking, we don’t believe in absolutism, and cluster use is no different. Cluster is a bit like spice; you want it to contribute to the finished wine, but you do not want it to dominate. It is a fundamental part of our craft to optimize what we are seeing in a given year and vineyard to maximize the expression of that vintage—some years a vineyard might seem to need 30% whole-cluster and others, 80%. Though we typically have never used whole-cluster on Zinfandel and field-blended lots, even that is evolving a bit. Certain lots of Carignan and Mataro from Evangelho and Bedrock that were fermented with whole-cluster have begun to make their way into the finished Heritage wines, and this year we actually did a small tank of Old Hill Ranch with 100% whole-cluster that we were all pretty besmirched by. Though traditionally Zinfandel has been almost entirely destemmed, I do wonder why. Is it not also a larger clustered, fruit and spice-rich Mediterranean red grape similar to the others we are comfortable including clusters on? I sometimes wonder if winemakers’ manic avoidance of anything “green” in Zinfandel has precluded a world of exploration that might help create wines of more freshness and aromatic dexterity, as “green” is only one of many possible byproducts of cluster inclusion. This will definitely be one of the things we continue to experiment with.