Finland is known for its alcoholism. I decided to celebrate that fact by making traditional Finnish drink “Vappukilju”, translating roughly as “strong home-made brew for first of May”.

Kilju is illegal to make here but adding some sort of fruit makes it technically wine and okay. So, with a little help from my friend, we made some “cranberry wine” with a little twist.

Quest for cheap booze

The mission was to make something that tastes good, is relatively cheap, ferments fast and most important – is strong!

With the logic “cranberry-drinks are OK, mask everything and can be fermented” I decided on making the wine out of those. Bad idea but more on that later.

Preliminary tests

Fermentation of different juices

To determine what is a good base for the drink, it was necessary to test out some juices in small batches to see how they would ferment.

Regular beer-bottles were used as containers for this experiment. They were washed properly with tap-water and disinfected by keeping them in 160°C (320°F) oven for an hour.

Test begins by pouring 3 dl (1,2 cups) of juice into the bottles with 2 grams of bakers yeast and sealing the bottles.

Experiment 1 2 3 Control
Barcode 6415131411931 6415131403950 6408620004509
Name Rybb & Deckers 1L Greippimehujuoma Marli Vital 1 l Karpalo Riitan Herkku 1l Karpalomehu Tap-water + 3 g Sucrose
Image
Notice Low sugar-content, added 2g
Gas production High, bulged cap High Medium Low
Verdict 24h Noticable yeast-smell, strong grape taste, lots of bubbles Girly partydring fuzz, pink foam, strong yeast-smell, lots of bubbles Fine foam, less acidic, smells strongly like yeast Low bubbles, mild yeast smell
Taste test 5 6,5 8 3

And thus, Riitan Herkku Cranberry juice was selected as the base for the drink.

Making the wine

Bill of Materials

Product Amount
Beer bottle 72
Bottle capper 1
Magnetic stirrer (we need this quick!) 1
25 liter barrel 1
Air-lock 2
Siphon 1

Growing the yeast

The goal of the project was to make the strongest stuff possible in the least amount of time. For this there was only one option: small sample of ALKO554 strain yeast was obtained from the Helsinki University Bio-lab.

ALKO554 is a type of brewer’s yeast. A strain that has been engineered to produce a lot of ethanol as it’s metabolic product.

The famous Finnish booze Koskenkorva is made with it’s help.

Yeast culture bottles in a shaking incubator

Yeast culture bottles in a shaking incubator

We cultured the yeast over the weekend so there was enough of it for the whole batch. Samples were taken and inspected with a microscope to verify there were no outside bacteria growing with the yeast.

After a couple of minutes in the centrifuge, most of the liquid can be just poured away from the containers, leaving a milky residue containing yeast cells in the bottom. Those containers were then washed with Milli-Q and stored in an autoclave disinfected jar.

ALKO554 yeast suspended in Milli-Q water

ALKO554 yeast suspended in Milli-Q water

Making the brew

The barrel was washed and disinfected with chlorite and vinegar was poured into the air-locks in the lid to kill any bacteria that might enter that way.

The total sugar-content was calculated to yield the maximum amount of alcohol possible and so that the yeast should be able to consume it all.

Water was heated in a large pot so the sugar would dissolve faster, all the juices and ingredients were added to the barrel and allowed to cool to 32°C (90°F) so the yeast would not die when added. After the cooling period, the barrel was lifted on top of a magnetic stirrer to try to keep everything mixed and to speed up the fermentation process.

Yeast solution was added 21.4.2015 22:30

Fermentation barrel with 2 air-locks and magnetic stirring

Fermentation barrel with 2 air-locks and magnetic stirring

Ingredients

After I made the list of what is needed, I decided the good juice from the earlier test was too expensive so I ended up making the end-product from variety of cheaper juices.

Item Amount Sugar
Water 14 liters 0 g
Rybb & Deckers Grapefruit juice 2 liters 80 g
Rainbow Cranberry juice 4 liters 480 g
Marli Vital Cranberry 4 liters 432g
Ocean Spray Cranberry 1 liters 120 g
Grapes, crushed 150 g 21 g
Cranberries, crushed 600 g 23 g
Sugar 3,5 kg 3500 g
Totals 25 liters 4656 g

Full sugar content of the solution: 186,2 g / liter.

Testing the wine

Microbiological testing

A 20 ml sample was taken from the barrel 5.5.2015 using a sterilized pipette by removing one of the air-locks.
The sample was inspected under an optical microscope and no contaminants were found.

View of ALKO554 yeast cells under optical microscope

View of ALKO554 yeast cells under optical microscope

To be sure, a culture was grown in a Petri dish in a YPD-Agar, the yeast colonies were growing as expected without anything that does not belong there. The viable count was calculated to be 3,95×10^6 cfu/ml.

Because the viable count at the beginning of the fermentation was 4×10^6 cfu/ml, we determined that the yeast did not multiply in the barrel properly. We speculated that the reason could be the benzoic acid naturally found in the berries acted as a preservative.

Note: Do not ferment cranberries

Measuring the alcoholic content

End product centrifuged (clear!) for testing

By taking the density of water solution with a 18,6% sugar content by weight from BRIX-table, we get a density of 1,07677 kg/m3.

After measuring the specific gravity from the solution to be 1,015 kg/m3, we plugged those numbers in an online-calculator, that gave us a number of 7,5% alcohol content.

The start measurement of specific gravity should have been taken, but we forgot so BRIX-table was used instead for a good estimate.

Bottling

The magnetic stirrer was turned off 9.5.2015 at 0:50 (yes I am a night-owl) and the barrel was left overnight so the yeast could settle to the bottom. Another barrel was then washed and disinfected and the contents of the first one were siphoned out to leave the berries and yeast in the first barrel.

The bottles were washed with a drill mounted brush and disinfected in the oven.

2,2 grams of sugar was added to the bottles, 3,3 dl of product was added and the bottle was sealed and left for secondary fermentation to give the drink fizzy bubbles.

I did this about half-way and quit because handling the bottles was such tedious work. I switched to larger wine-bottles.

Home made wine

Ready product with labels

If it’s worth doing, it’s worth doing right! I packaged the product in Black Tower bottles and drew a quick label in Photoshop.

Empirical testing

I found the end product to be cloudy, bright red, thick and wine-like with a low taste of yeast. After settling down for a week, the taste of yeast was almost gone and the color was even brighter.

Here is what people thought:

Not too bad, doesn’t smell like anything, reminds me of strong apple cider mixed with white wine or something with cranberry. – HR

Strong, little bit sour, smell was like actual berries.
Very delightful color to it – JD

Its good but – AG

It gives bad, completely unproportional hangover – EP

I don’t remember saying that – JD

Conclusion

The fermentation was supposed to be done in a weekend but took two weeks because of unexpected berry-related mishaps. The end-product was not that good, took way too much effort to make and cost too much.

Verdict
  1. Good science is good observation
  2. Do not try to ferment acidic berries
  3. Buy your hooch from the store, they know better.
  4. Totally not worth the effort

Special thanks to Edvard Partti for helping me with this project.

Difficulty ★★★☆☆
Time ★★★★☆
Price ★★★★☆
Share this article
Share on FacebookShare on LinkedInTweet about this on TwitterShare on VKEmail this to someoneShare on Google+Share on StumbleUponShare on RedditPin on Pinterest