Cider Keeve

I have been working with Leidel's Orchard up in La Cresent, MN.  They are looking into making hard cider, and after much discussion/research/brainstorming with the owners about the type of product they want to make, our pilot batches include a "Saison barrel", a "Natural barrel", and a "Keeve carboy".  I'll wait for another day to talk about the barrels, but I wanted to post a bit on the keeve batch.

Keeving is a old-world cider technique that has largely been forgotten, but is still done by some traditional cider makers in the UK and France. Simply put keeving releases natural pectin into the apple juice which then coagulates, removing much of the nutrients from the juice. As the yeast ferments the keeved juice, it uses the remaining nutrients, but runs out before it can ferment all of the sugar.  Traditionally, this has been done to create a naturally sweet cider.  For more information on the process, read this.

When I learned about this process and how it works, I thought about how lambic wort is brewed to have a lot of complex sugars that Saccharomyces can't eat, which leaves a portion of the sugars for the Brettanomyces.  I began to wonder if Brett would survive in a nutrient poor environment after the Saccharo died out.  So the purpose of this experiment is to see what will happen if I pitch lambic yeast into keeved cider.

However, this turned out to be a very difficult and laborious process.  I started with nearly 4 bushels (160lbs) of apples.  They had to be sent through a mill and the pulp left to sit for 24 hours (to release the pectin).  The next day we pressed the pulp, but our efficiency was about half that of a commercial cider press (and 10 times the work).  Leidel's Orchard actually spoke with a couple commercial presses keeving, but they were unable to do this process without a major rework of their set up.  I was hoping to get 8+ gallons, so I borrowed a friend's 12 gallon glass carboy to use as a keeving tank.  I only ended up with 6 gallons.

Then the keeving process starts.  The first step is to allow the pectin to coagulate in the cider.  Some modern additives help this process, but it is still a bit of an art.  The carboy was stored in my refrigerator for 10 days where it is cold enough to stall fermentation until the coagulated pectin gel has formed.

Most of the gel that formed fell to the bottom and is sitting on top of the sediment

The next step is to allow fermentation to start and the CO2 bubbles will lift the coagulated pectin to the surface making it possible to siphon out the juice in the middle.  I left it at room temperature for 5 days, which is much longer than I was told to expect, and only had a minimal amount of fermentation start.

Separation of cider and pectin gel.

Pectin gel pushed to the top by CO2 bubbles.

Only about half of the pectin gel was pushed to the top, while the rest stayed on the bottom of the jar.  I would have liked waiting until all of the pectin was pushed up, but I feared fermentation would accelerate too fast and I would be left with a cloudy cider.  Also, as the pectin gel at the top was exposed to the air, some mold was starting to grow.  I decided it was best to siphon at this point rather than risk ruining the whole batch.  As I siphoned, I also strained the keeved juice to capture any particles that might have been stirred up.

5 gallons of keeved cider with lambic blend yeast added.

In the end I was left with 5 gallons of clear, keeved, cider.  To that I added Wyeast Lambic Blend.  It will likely take 9-12 months before I know if the experiment worked, and if its anything like lambic, may take 2 years to really develop.  Apple cider is all simple sugars, and normally the Saccharo would out compete everything else.  My hope is that I've stripped out enough nutrients in this keeve to slow the Saccharo and allow the Brett/Pedio/Lacto to do their work in tandem.

Barrel Notes: Day 277

Just a quick update as I had to open the barrel to take a measurement of the bung.  While it was open I decided to snap a picture:

Lambic pellicle at day 277.

And wow!  It smells amazing.  It smells like it would be very good to drink already.  I've always seemed to be pretty good with the patience and waiting game, but this is going to be tough.

Barrel Tap

[Update: It appears the threaded tapered tap is no longer being produced.  If you would like a barrel tap, RMBC has had these fabricated and should have them in stock]

I've got plenty of time before we empty the barrel up at O'so, but its never a bad idea to get ready.  Following instructions from Vinnie at Russian River, I drilled a 15/16th inch hole at the bottom of the head and plugged it with a standard Belgian cork; the drain port.  When Russian River is ready to drain the barrels, they pull the cork and jam a 1" hose into the hole.  While I am sure that works fine, I wanted a little easier and more secure way of draining the barrel.

First I got a threaded tapered tap from winetaps.eu that look like this:

The sales rep called the one I wanted a "1/2 inch tap", which is the inside diameter of the tap.  The outside diameter of the taper goes from 22mm to 26mm (~7/8" to 1 1/16").  The drain port is 15/16", so it fits perfectly.  I tested it out on a scrap piece of wood, and the tap goes in about half way, and with a half turn, the thread lock the tap in place.

The back end of this piece had a BSP threaded connection.  Breweries, however, use almost exclusively Tri-Clover connections.  I figured it would be nearly impossible to find an adapter, and even if I did find it, I would be left with crevices that could hold unwanted bugs.  So, I ordered a 1" Tri-Clover ferrule from brewershardware.com:  (note: a 1" Tri-Clover still has the standard 1.5" clamp)

I then took the pieces to a machine shop to get them welded.  Its important to note that there is more than one way to weld Stainless Steel, but if you want it food-grade, you have to get someone that can TIG weld the pieces.  I also had the shop bore out the BSP threads from the original tap leaving a smooth, easy to clean, surface.

Threaded Tapered Stainless Steel Barrel Tap with 1.5" Tri-Clover Fitting.

With this final piece I will attach a shut-off valve.  Then when I pull the cork from the drain port, I can quickly push this piece in and give it a half turn to lock it in place.  Simply attach a tube to to the back of the shut-off and draining the barrels should be pretty easy.

O'so Collaboration: Day 50

I went back up to O'so yesterday to check on the barrels and install the sample port.  It has already been 50 days since we brewed the lambic!  Wow, time flies.  However, when you consider the beer will have to age close to 2 years, that isn't much time at all.  At this point in the game, it is more of a sensory check for off flavors.  I am expecting it to taste/smell essentially like a berliner weisse.  I was also curious if there would be any differences between the barrels.  There are 14 barrels, so hopefully they will develop differently and we will have some diversity to blend with.

Drill a 7/64" hole and catch a sample

As expected, they all taste/smell like a berliner weisse, which is a good sign.  We are on the right track.  There wasn't a whole lot of variety (obviously it is still very young), but there was some.  Some barrels were more lemongrass, some were more minerally, some were more brett/funky.  Aside from tasting and smelling the beer, it is a good chance to take a gravity reading.  All the barrels we tested were hovering around 5 brix (1.009 SG).

Testing gravity with a refractometer. 

All and all I am pretty happy with the progress.

Minimalist Lambic: Fermentation

We had some extra lambic wort at the end of the O'so brewday, so a few of us filled up carboys to take home.  Instead of buying a lambic blend yeast, I thought I would try something new; I wanted to try fermenting the wort as simply as possible.  The 4 main players in Lambic fermentation are Saccro, Brett, Lacto, and Pedio.  Between the two bacteria strains (Lacto & Pedio), I have read that Pedio does the lion's share of creating the acidity in Lambic, so I chose that one.  When Pedio is fermenting, it gives off diacetyl, which is a buttery off flavor.  Brett however eats the diacetyl compound and therefor should always be used in conjunction with Pedio.  So, between the two yeast options (Saccro & Brett), I really don't have a choice but to use Brett, although that is what I would have picked anyway.  The strain of Brett I picked was Brett Brux, as I like the barnyard funk flavors and thought it would complement fruit additions later on should I choose to add any.

I pitched the Brett & Pedio into my 6 gallon Better Bottle full of Lambic wort on Sunday.  On Tuesday noticeable fermentation started.  Here is what it looks like:

Lambic fermentation with only Brett B & Pedio

Once the primary settles down, I'll go put this next to my pilot Lambic barrel to sit for at least a year.  I'm looking forward to watching the pellicle and potential ropy stage through this carboy.  It will be much easier to see and document than oak barrels have been.  I'm also looking forward to seeing how Marc decided to ferment his portion of lambic.

O'so Collaboration: Brewday

Jump to:
Installing a Drain Port
Filling the Barrels

We were at the O'so brewery at 6:00am, and didn't finish until 11:00pm.  It was a long brew-day to say the least. Marc and his head brewer, Mike, had planned and prepped the turbid mash schedule as best as one could, now it was time to actually go through the steps and learn/adjust along the way.  It was incredible watching these two brew an unfamiliar style, yet hit every target number along the way.  We had some troubles with the grain mill at the start of the day, and the initial dough in was very challenging, but after that, the process went nearly perfect.

Richard adding rice hulls to the Mash Tun.

I tried to take pictures throughout the day, but many of them didn't turn out because of the steam coming out of the kettle.  There was a student there filming and photographing the process, so I may get some of his pictures soon.

Protein cake.

I believe the above picture was taken during the final transfer of the turbid mash.  As you can see, a large layer of protein has developed on top of the grain bed (even separating away from the wall).  The next picture is the wort transferring over to the boil kettle.  They were very happy with how clean it was!

Transferring to boil kettle.

As the kettle was being filling, they started the boil.  It took nearly an hour to fill the kettle with 35 barrels of wort, and it was boiling the entire time.  During that time we added the hops.  We used 40 pounds of 5 year old Liberty hops.  They smelled like a hockey locker room.

5 year old hops

Some people assume that Lambics don't have much hops in them, but there are IPA's that use less hops than Lambics.  The only difference is that they have been aged, which removes the bitter flavor of the hops, but retains its antimicrobial power.  The boil went for 4 hours.  Then the wort was transferred over to the whirlpool and left the hops behind.  Wow do those hops ever expand!

Hops at the bottom of the Boil Kettle

At this point that hops mass is about 8" deep and absorbing nearly 3 barrels of the wort.  Marc cleaned this tank the following day and told me they had puffed up and it was about 2 feet deep!

Installing the Drain Port:

While much of the brewing process was going on, I was further prepping the barrels.  We had to install the drain port prior to filling these barrels.  I talked to Vinnie at Russian River and he sent me over the instructions on how to do it.  Its always nerve racking drilling a whole into the barrel, so it was nice to know the exact method.

Drilling a 15/16" hole into the barrel head.

The hole is about 2 fingers up from the edge.

Insert a standard Belgian cork into the hole.

It is quite simple.  The French Oak is pretty hard and can be a bit of a pain to drill through.  We tried a couple different ways of inserting the cork on our practice 2x4.  It is possible to hammer the cork in, but if you use a simple floor model corker, you can pull the handle part way so it just compresses the cork.  The cork has a bit of memory at that compressed state giving you enough time to easily slide the cork into the hole.  It will then quickly expand back to its normal state creating a very tight seal in the barrel.

Filling the Barrels:

So with the barrels prepped and the wort ready, it was time to fill the barrels.  Initially our plan was to short fill the barrels and then after the primary had finished, to rack over the contents of two of the barrels to top off the other 12.  As it happened, we had yielded more wort that estimated, and so we filled all 14 barrels to the top (stopping half way to add the yeast).

Filling the barrel with lambic wort.

Since this is the first use for these barrels, we needed to add a blend of yeast/bacteria to the barrel, which we did half way through filling each (1 liter of wort per barrel).  Future uses of these barrels will not require yeast to be pitched.  We used the same East Coast Yeast blend that I used for my pilot barrel.  I can't stress enough how happy I have been with this yeast.  Thank you Al!

Adding East Coast Yeast lambic blend to the barrel.

And so, 17 hours after we started, we have 14 barrels of lambic sitting over the drain waiting for fermentation to start.

I figured it would be 2 days before the fermentation kicked off, but Marc sent me this picture the next day:

Fermentation in less than 12 hours!

O'so Collaboration

Over the weekend I went up to start the preparation work for what will become a collaborative O'so/Funk Factory lambic project.

(click picture to visit the O'so website)

I arrived Friday evening to 14, French Oak, used wine barrels delivered and waiting for me.

While I had the barrels on end, I decided to quick apply a stencil I made up for them.  Its a simplified version of the FF logo, and creates 4 spaces to record the fill date.  With that done, it was time to line up the barrels and fill them with water.

Soaking the barrel overnight was done for two reasons.  First, it will swell the barrel ensuring there are no leaks when it comes time to fill it with lambic wort.  Secondly, it will extract some of the wine flavor from the barrel.  In brewing lambic, you want as neutral of a barrel as possible.  When I arrived to the very last barrel and filled it up, I noticed it wasn't holding water.  On the barrel head opposite to me, it was gushing out water along the bottom croze.  I left the hose valve 1/3 of the way open, and it was losing water faster than the hose filled it.  Unfortunately, I think that barrel will need to be replaced even though it did swell up overnight.  I just don't have any confidence in a barrel that struggles that much to hold water that it will limit O2 permeation sufficiently.

The next morning I had to empty the barrels.  Moving around a 60 gallon filled barrel is no easy task, and in fact took me most the morning to empty all 14 barrels.  It wasn't until nearly noon that I started steaming the barrels.

This is the same steamer I used before.  Behind the barrel in the picture being steamed are 4 that I had already completed.  You can see the bung is in and seals the barrel.  As the steam cools down, it creates a vacuum that will extract more wine from the wood.  Once I finished with all the barrels, I gave them a final hot water rinse and sealed them again.  The air inside was hot from the water, so they will be vacuum sealed until the day we go to fill them with wort.