Winemaking is about transforming healthy and clean grapes into young, immature wine, a 3-4 month process. This is the step after growing quality grapes, an annual endeavour. The step after winemaking is cellaring, which is about maturing the young wine in barrels, bottling the wine and maturing the bottles – that takes 5-8 years.
This section is organised as follows:
On this page, we explain the general concepts and processes used. We start with a simplistic view, then describe our winemaking facility and then summarise the 15 individual process steps. Then we conclude with a summary of how we made the wine in each of the 9 past years 2009 – 2017.
On the following pages, we describe the process steps in more detail and how they apply in the last harvest (2018), what decisions we made and what we learned.
A very simplistic view
In a very simplistic view, making red wine has 4 distinct phases:
Phase 1 - from grapes to sweet must:
First we decide when we pick grape bunches in the vineyard. Then we sort out the bad bunches, destem the bunches, sort out debris and dirt and crush the grapes into sweet must. Must is a slurry of grape juice, grape skins and seeds. Phase 1 takes 6-8 weeks of monitoring grapes in the field and a few hours of picking and processing the grape bunches.
Phase 2 - from sweet must to alcoholic must
We ferment the sugar with the help of yeasts into alcohol. During this process a lot of valuable organic compounds are extracted from the skins, pulp and seeds – these compounds give the wine its characteristic odours, taste and mouthfeel. This step takes 2-3 weeks and is the most critical and difficult phase.
Phase 3 – from alcoholic must to juvenile wine
We separate the now alcoholic juice from the skins and seeds by pressing the must. This takes a few hours
To go through these four phases, we need a special-purpose facility: a winery. We built our facility on 4 levels, so we do not need to use pumps – we rely on gravity to move the product, and winches or lifts when required. The rationale is to prevent the rough physical treatment of juice, skins and seeds inside a pump. This graphic illustrates the sequence::
A brief explanation covering the entire wine-making and cellaring process:
On the Bunch Sorting Table, we sort out the damaged bunches and leaves coming in from the vineyard
The Elevator moves the sorted bunches to the mouth of the destemmer
The Destemmer separates the grapes from the stems
On the Berry Sorting Table, we pick out the "Material Other than Grapes" or MOG, mostly small stem and leaf pieces
As the berries leave the Sorting Table and fall into the Fermenter; we have the option of inserting a Crusher which breaks their skin so valuable compounds can be extracted more efficiently during fermentation
In the Fermenter, we convert Grapes into Fermented Must (i.e. sugar into alcohol). The fermenter is temperature controlled by a Glycol Cooler pumping cold or warm glycol through the walls of the Fermenter.
The Press separates the juice (i.e. wine) from the grape skins and grape seeds.
The young wine is then dropped to a Mixing Tank in the cellar and moved back and forth between Barrels until it is mature and bottled. The mixing tank and barrels can be moved up or down to allow gravity flow between them. The temperature in the cellar is kept at 55-60 dF by the Glycol Cooler pumping cold glycol through an air-conditioner.
To the left of the graphic is a picture of the physical layout. You can see the ground level outside through the window, and you can see cellar levels -1 and -2. Half the floor between level -1 and -2 is removable so that we can connect the fermenting tank with a bridge to the press. This is shown on the page explaining the press. Cellar level -3 is below. The wine is transported by a hose through holes in the floor to the mixing tank and barrels below.
Overview of the process
Here is the next level of detail: a closer look at the four phases described above. Note, this process has evolved significantly over the years; what follows is our process for the 2018 vintage and after. The flowchart on the right shows 18 steps and the decisions which link them:
1. Measure Berry Ripeness: We measure the progress towards grape maturity in the vineyard and then decide when to harvest.
2. Harvest, Sort & Destem: We pick the grape bunches, sort out the dirt, destem them, sort the berries end up with clean grape berries in a fermentation tank.
3. Crush or Stomp: We decide whether we want to break the skins of the grapes with rollers (crush) or with our feet (stomp), or not at all (i.e. Full Berry Fermentation)
4. Saignée? We decide whether we want to artificially increase the concentration of flavours in the wine. This is done by increasing the “skins & seeds”-to-“liquids” ratio through syphoning off a percentage of the liquids. The juice that is syphoned off can be used to produce rosé wine.
5. Adjust Brix: We decide whether we need to lower the average Brix level by adding water.
6. Adjust Acidity: We decide whether we need to adjust the pH up (add carbonates) or down (add tartaric acid). We can make this adjustment now or delay until later. This adjustment can be made upfront (i.e. as step 6) or later (i.e. during fermentation or cellaring) in increments.
7. Add Stems or Oak Chips: We decide whether we want to add back some of the Stems into the must to adjust the flavour profile or add Oak Chips to adjust the phenolic extraction.
8. SO2 or native Fermentation? We decide whether we want to ferment with yeasts and bacteria native in the vineyard and winery, or with cultured yeasts purchased from external providers. If we decide to use cultured yeasts, we add SO2 to kill off all native non-saccharomyces yeasts and bacteria.
9. Enzymes? We decide whether we want to artificially increase the extraction of desirable components in the skin, pulp and seeds into the juice by adding enzymes which break down cell walls.
10. Add Fining Agents: We decide whether we want to add antimicrobial agents to bind and precipitate spoilage bacteria.
11. Cold Soak? We decide whether we want to extract desirable components of the skin and pulp into the grape juice before fermentation is converting the juice into alcohol. Again the idea is to get more aromas and flavours. We soak at a low temperature of around 50-55 oF to prevent spoilage.
12. Fermentation Phase 1: Now we raise the temperature of the must to 70 dF and decide whether to start fermenting with native yeasts living in the vineyard and the cellar, or industrial yeasts purchased from third parties. If we decide for native, we simply wait for the fermentation to start on its own, or we mix in a bucket of must which we had set aside a week or so earlier and started fermenting on its own. If we decide for industrial, we inoculate the must with cultured yeast. In either case, we consider adding nutrients for the yeast, the amount depending on the level of Yeast Available Nitrogen (YAN) in the must.
13. Fermentation Phase 2: After the fermentation accelerates and the sugar level has fallen by around a third, we have a few decisions to make. If we started with native yeasts, we might wish to decide to finish with industrial yeasts and inoculate. Also, more yeast nutrients may be required. Because fermentation releases thermal energy, we may also need to cool the tanks, so the temperature stays below 90 oF. At the same time, we need to start watching the amount of phenolics extracted from the skins and pulp. If the tannins extracted exceed the anthocyanins extracted by more than 10-20% before the fermentation is complete, we decide to press the cap separately to limit further tannin extraction while completing the primary fermentation (steps 14 & 15). Alternatively, we proceed to step 16.
14. Press Cap Separately: We scoop out the cap (mostly skins floating on top of the must) and press it, then pour the resulting juice back into the fermentation tank.
15. Primary Fermentation Phase 3: we complete the primary fermentation, i.e. all the sugars have been converted to alcohol.
16. Extended Maceration: If the fermentation has completed before tannins have reached 110% of peak anthocyanins, we decide whether to extend the time the now fermented juice is exposed to the grape skins - and, more importantly, the seeds - to extract even more phenolics (i.e. mostly tannins).
17. Press: We separate the juice from the skins and seeds by first letting the juice flow out of the fermentation tank into the settling tank (called “Free Flow”) and then pressing the remaining wet must into the same tank and other containers (called “Press Run”). The remaining, now dry, skins & seeds are carted into the field to fertilise the soil.
18. Rack into Barrels: After letting the wine settle for a few days in the mixing tank and other vessels we rack the juice into barrels and topup tanks leaving the sediment behind.
Steps 1 through 18 takes between 10 and 30 days
Up to 2015, we followed this process for a single grape variety, Cabernet Sauvignon.
In 2016 we started dealing with 4 different grape varieties (Cabernet Sauvignon, Merlot, Petit Verdot and Cabernet Franc) each possibly reaching harvest maturity at a different date. So potentially we have 4 processes running simultaneously, slightly staggered time-wise. In 2018 we had 3 separate harvests, starting with Merlot and Cabernet Franc in the upper field, followed by Cabernet Sauvignon in the lower field and finishing with Petit Verdot in the upper field.
Following is a detailed flowchart of the process and the decisions taken for each of the 3 in 2018 harvest. The only purpose of showing this chart upfront is to illustrate how the steps and decisions described in the following pages fit together. The bold arrows indicate the decisions taken. We describe the individual steps in the pages which follow.
Wine Making Summaries
The following table summarises how we made wine during the first 7 years, 2009 – 2015
During the first 3 years, I relied heavily on Aran Healy who helped me make decisions on what equipment to buy and taught me how to use it and make wine. We took relatively few measurements, relying mostly on Aran’s experience and tasting skills. The first year was about setting a benchmark: producing the wine with minimal additions and interventions in a 100% natural fashion. In the second and third year, we started experimenting with established wine-making techniques (like using commercial enzymes and yeasts). In the third year, we were particularly challenged by a bad harvest (low volume and quality of grapes)
During the following 2 years, 2011 & 2012, after Aran left, we started to become more numbers oriented and collected data more diligently.
In 2013 David Fenyvesi joined, contributing his East European winemaking experience. We started to measure phenolics with the help of WineXray, a service that converts spectral absorbance measurements into estimates of phenolic compounds in the wine. This, in turn, allowed us to fine-tune the fermentation process. We also started to document the winemaking process with a detailed flow-chart and collected data diligently.
In 2014 we started to measure the phenolics in grapes after veraison to help to time the harvest, and we pressed the cap separately and before fermentation was finished to limit the uptake of tannins.
In 2015 we fermented the different clones separately in bins within the fermentation tank – this proved that the 337 clone was of much higher quality than the Rixford clone (in term of extractable Anthocyanin concentrations)
By 2016 the new Upper Vineyard started to produce and, because Merlot matures a month earlier, we started running two harvests and two rounds of fermentations in sequence. We introduced new smaller fermentation tanks to fit inside the large tank to handle smaller lots, and we built a small crusher because stomping in new tanks became infeasible. Nicolas Vonderheyden replaced David, adding his Bordeau winemaking experience to the mix. This, and input from UCDavis encouraged me to return to the more natural approach we had used in 2009: no enzymes, no sulfur and no commercial yeasts.
The process became more complicated. The graphic illustrates the difference between 2015 and 2016. In 2015 we had one harvest (cabernet), split the grapes into three fermentation buckets (by clone) and combined the fermented juice at press into a single barrel. In 2016 we harvested and fermented the grapes the Upper Field (Merlot, Cab Franc & Petit Verdot) first in a single tank, used only the freeflow and set the wine aside. Then we harvested the Lower Field. The long row grapes (337 clones) were saigneed and fermented in 4 separate fermentation tanks and their freeflow combined with the free flow from the Upper Field into 2 barrels. The grapes from the short rows (Rixford clone) received the saignee from the long row grapes, were fermented in 2 separate fermentation tanks, then pressed together with the remaining skins of the long row grapes and filled one barrel.
In 2017 we replaced all spreadsheets with a database. This screenshot of the “REVIEW: Vintage” layout for 2017 summarises the vineyard and winery activities; the cellar activities are incomplete because the vintage is still in barrels. In 2017 we had two harvests, both poor in volume and quality: we harvested the 4 blocks (PetV, CabF, Me2 & Me1) in the upper field on October 3, the CSLR block on October 14 and abandoned the CSSR block. We fermented the Merlot blocks separately and combined the CabF and PetV blocks. We fermented the Cabernet Sauvignon block in 4 separate fermentation tanks. All fermentations completed with indigenous yeasts.
Then we pressed and filled one barrel with Cabernet Sauvignon, and a half barrel with a mixture of Merlot (55%) Cabernet Franc (25%) and Petit Verdot (20%). The third cellar batch was tiny and used as topup. We inoculated the two barrels with malolactic bacteria but have little success in converting the remaining malic acids to lactic acids.
Last Harvest: 2018
2018 was a much better year. Harvest quantity and quality were excellent. The fermentations all completed with indigenous yeasts, and we ended up with 3 full barrels and plenty of topup wine. We decided to use, for the first time, a dedicated topup container for each barrel to eliminate a potential source of cross-contamination.