Oak conversion table: chips, staves, tannins

How much chips, how many staves, what equivalent in finishing tannin to reach the same level of oak. This page gives the raw table our winemakers work with, the measurement method behind it, the minimum contact times, and what actually happens inside the wood during that time.

It is a working document, not a brochure. It is meant to be opened in the cellar, with a trial glass in hand.

Print the cellar sheet

The table

Doses are given per hectoliter of wine. The white and red columns are not read on the same line: at an equal dose, a white tastes oakier than a red.

The working range, 100 to 600 g/hL, is 1 to 6 g/L. It sits inside the 2 to 10 g/L range documented by the IFV, the French vine and wine institute, for oak pieces in winemaking. The equivalent of a new barrel is reached around 350 g/hL on whites and 500 g/hL on reds.

Contact time: minimums, not averages

• Fine chips: 1 month minimum, 2 to 3 months to reach the extraction plateau.
• Staves 7 mm: 4 months minimum.
• Staves 18 mm: 7 months minimum.

These are thresholds, not recommended durations. Below them, extraction is not incomplete in a neutral way: it is unbalanced. What comes out first are the most soluble and most volatile compounds, the ones that give the surface note. The base, the ellagitannins and the compounds that round the wine out, arrives later. Oak treatment stopped too early gives a wine that smells of wood without having its structure.

Shortening the time forces the dose up, and raising the dose does not replace time. You get a more prominent, less integrated oak, built on the same molecules. The reverse holds too: a long contact at a moderate dose gives a more integrated oak than a short contact at a high dose.

When chips are added during alcoholic fermentation, the clock does not stop at pressing: the chips must stay in contact with the wine after fermentation to reach the month. The five to ten days of a fermentation are not enough.

Oak uptake continues after the wood is removed

Pulling the chips or the staves out does not stop the oak treatment. It stops the extraction, which is not the same thing. What has already moved into the wine keeps working: tannins polymerise, ellagitannins hydrolyse and bind to anthocyanins, aldehydes evolve, and part of the volatile compounds redistribute between the phases of the wine. A wine whose wood came out three weeks ago does not have the oak profile it had on the day of removal, and more often than not it reads as more marked, not less.

This is the first cause of overdosing we see in the cellar. You taste on the day of removal, you find it is not enough, you put wood back in, and two months later the wine is oaked beyond the target with no way back.

• Aim slightly below the target at the moment of removal, never above it.
• Allow 4 to 6 weeks after removal before judging, on a homogenised wine brought back to tasting temperature.
• Record the removal date as carefully as the addition date: it is the reference point for the control tasting.
• If an adjustment is still needed, make it with a finishing tannin rather than by putting wood back: the dose is finer, and the effect reads faster.

Finishing tannins: the dilution procedure

A badly incorporated tannin is a wasted tannin, and it costs exactly the same as a well incorporated one. There is nothing spectacular about the procedure below, but it is what separates a dose that works from a dose that ends up at the bottom of the tank.

1. Weigh, do not estimate. At these doses a twenty percent error is tasted. Take the value from the table, or from the dosage simulator, and weigh it.
2. Dilute, never pour dry. Dissolve the tannin in about ten times its weight of warm water, between 35 and 45 °C. Pour the tannin into the water and not the other way round, under continuous stirring. A tannin poured dry into a tank forms lumps that never redissolve and settle to the bottom: the dose is paid for, it is not in the wine.
3. Let it hydrate for 15 to 30 minutes, then stir again. The solution must be clear and free of deposit before it goes to the tank.
4. Incorporate in a steady stream during a movement. Pump-over, stirring or racking: pour steadily across the whole duration of the movement, so that the tannin sees the entire volume and not the first ten hectoliters.
5. Wait before you judge. A tannin works over 8 to 10 days. A tasting the next day finds a closed wine and wrongly concludes that more is needed. That is the second cause of overdosing.
6. Do not stack treatments. No protein fining, gelatin, albumin or casein, in the days that follow: protein and tannin find each other and precipitate together, and you lose both. The same caution applies to bentonite and to a tight filtration. Let the reaction have its time.
7. Liquid tannins: do the conversion. Our liquid tannins are at 20 % concentration. One liter of liquid is the equivalent of 200 g of powder. Divide the g/hL figure from the table accordingly before dosing.
8. Always run a trial before the tank. The drop by drop trial gives you the answer on your wine, with your pH and your structure, within the hour. No table will do that work for you.

Where these numbers come from

We publish this table because we could not find it published anywhere else, and because a figure without its origin is worth very little. So here is the origin.

The doses come from our own technical sheets, established on our own oak: French oak air dried for 36 months, toasted by convection, then tasted by our winemakers on whites and on reds. They are averages of what we observe across campaigns, not the result of a single trial. They are cross-checked against the feedback of our partner winemakers and against published oenological research.

Every equivalence is anchored on one reference, the new barrel. A barrel releases about 60 % of its oenological capital in the first year, and that is what the last line of the table is calibrated against.

The finishing tannin column is expressed in powder. Our liquid tannins are at a 20 % concentration, which means 1 L of liquid is the equivalent of 200 g of powder.

Updated August 2026: contact times have been revised. The figure we gave for fine chips during fermentation was too short and produced incomplete extraction; stave durations are now stated as minimums, which is what they are.

Going further

• How much oak replaces a new barrel. Where the 350 and the 500 g/hL come from, and what the equivalence does not contain.
• g/hL, units, milliliters: converting without mistakes. The three conversions, and the one that costs a factor of five.
• Whites and reds are not read on the same line. The matrix effect, the pH, and why there is no single coefficient.

What this table does not tell you

• It was established on wine. On spirits, on beer and on non-alcoholic drinks, extraction does not follow the same rules and these figures do not transfer.
• It knows nothing about your cellar temperature, your maceration length or the real condition of your tanks. What measures well on a bench measures poorly in a 300 hL tank in October.
• It gives an intensity, not a profile. The character, vanilla, caramel, spice or roasted, comes from the toasting profile, not from the dose.
• It does not replace a trial. It tells you where to start, the trial tells you where to stop.

Open AMEDEE Lab See chips, staves and inserts