The amount to reserve will depend of the specific gravity of the juice and the desired sweetness of the finished wine.<?
amespace prefix = o ns = "urn
chemas-microsoft-com
ffice
ffice" />
My approach would be to use a table that provides relationships between specific gravity, Brix and grams of sugar per liter. These tables can be found in winemaking books and on the Internet. For my example I used the table on Ben Rotter’s website.
Based on the table, a juice with a specific gravity of 1.090 will contain about 242 grams of sugar per liter. From my reading and experience a wine fermented to dry will contain about 2 grams per liter of residual sugar.
If 1.5 liter of juice was reserved, the reserve would contain 363 grams of sugar (1.5 liters * 242 grams per liter). The remaining 21.5 liters would be fermented to dry and would contain 43 grams of sugar (21.5 liters * 2 grams per liter). The final wine would be a blend of the two and would contain 406 grams of sugar and have a volume of 23 liters; this is about 18 grams of sugar per liter (406 grams / 23 liters). Looking at Ben Rotter’s table, 18 grams per liter corresponds to about 2.7° Brix.
Reserving 3 liters of juice with a SG of 1.085 (228 grams sugar per liter), would look like this:
3 liters * 228 = 684
20 liters * 2 = 40
(684 + 40) / 23 = 31, according to Ben Rotter’s table about 3.8° Brix
Remember that as the volume of the reserve increases, the percent alcohol by volume of the blended finished wine decreases.
This approach seems logical to me. Hopefully, someone will correct any faulty logic and/or provide an easier method.
I doubt that the wine will undergo MLF without the introduction of MLF bacteria; however, I would add Lysozyme to be safe. MLF and potassium sorbate can result in an unpleasant situation.