hist-brewing: Brandy and fortified wines

Hiram Berry burningb at burningbridges.com
Sat Sep 23 20:40:32 PDT 2000

"Sean Richens" <srichens at sprint.ca> writes:

> Jeff Renner provides a valuable service in demonstrating how the answer
> get depends on the professional biases of the "expert" you ask.
I agree.  This subject is _extremely_ interesting; from the sources Jeff
quoted it sounds like noone really knows the reason for these divergent
> Being a chemical engineer, I favour the mass transfer explanation:
> > "Yet another noted distiller suggested that the year-round humidity
> > of Scotland favored the proportionately greater evaporation rate of
> > alcohol there, whereas the relatively less humid climate of Kentucky
> > (especially in the upper floors of the warehouses in the hot summer
> > time) favors a relatively greater rate of evaporation for water.
> It's the same principle behind why the pressure of CO2 in a bottle of beer
> doesn't keep oxygen from diffusing in through permeable plastics or cap
> seals.  I don't blame the people Jeff surveyed for not thinking of it, I
> didn't think of it either until I read it.
It makes sense, but it can't be the whole story; differential diffusion
coefficients for ethanol vs water across a wood "membrane" must be a factor.
As you point out the transport processes of the water and of the ethanol can
be considered essentially independent of each other.  If in the Scottish
warehouse the outside humidity is near saturation (in air coinciding pretty
closely with the vapor pressure in vacuo), while inside the barrel vapor
pressure of water alone in the ullage space must be less than saturation
over pure water due to dilution of the water in the whisky by the ethanol,
there will be a water gradient inwardly decreasing across the wood, meaning
that water would be sucked into the barrel.  The vapor pressure of ethanol
outside the barrel would be effectively zero, while it would be considerable
inside the barrel, so if the barrel is permeable at all to ethanol there
will be net loss over time from it.  This situation is consistent with dilut
ion in the Scottish warehouse.

But it isn't consistent with a strength increase in the case of the Kentucky
product if selective permeability isn't operant.  Consider: over the range
of  probable warehouse temperatures, say 10C to 30C, the ratio of vapor
pressures between the two solvents over their pure liquid is pretty
constant, around 2.5:1 in favor of the ethanol, so for the purpose of
analysis knowledge of the precise average temperature in the Kentucky
warehouse isn't imperative.  Let's say for the sake of argument it's 25C;
also we don't know the precise strength involved, but let's guess it's
around 90 proof (volume basis, I think?, which works out to around 39 wt
percent,  conveniently close to 0.20 mole fraction ethanol.)  I realize that
whisky probably doesn't follow Raoult's Law exactly, but it won't deviate
enough to invalidate the conclustion.  Just take the extreme case, ie. a
completely arid 0% humidity external environment (which is a great
exageration over the true average hygrometric condition of anywhere in the
state of Kentucky).  Then we have decreasing vapor pressure gradients for
both solvents across the barrel interface: 19 torr -> 0 torr for water and
12 torr -> 0 torr for ethanol.  Now, if wood does not provide differential
permeability we can assume that the Fick's Law molar diffusion constants are
at least in the same ratio as for the pair of their free gas phase values
even if we don't know anything about their absolute magnitudes; this is just
the square root of the ratio of their molecular weights, or 1.60 times
faster for water given an equal molar gradient.  So for this case the molar
diffusion rate would favor water by a factor of 2.53:1; multiplying each
part by the molecular weight of its solvent gives a mass composition of the
vapor of 49.5% water and 50.5% ethanol.  But this is a higher percentage
ethanol than the barrel contents (39%), so the net result is a process
_decreasing_ the strength of the barrel contents.  And this happens
hypothesizing the extreme case of complete aridity on the outside; results
for more reasonable humidities favor more relative ethanol removal.  So
while your humidity differential thesis looks right in terms of a base cause
for the phenomenon, differential permeability in wood must be at work to
explain the Kentucky results using it.

> All of the other answers don't quite pass the BS test, in my arrogant
> opinion.  There are osmotic processes for removing alcohol from beer etc.,
> but if wood did the job there would be all sorts of folk recipes for
> applejack by aging in barrels.

I'm not sure that follows.  Would folk practice concern itself with a
process that takes a very long time (years) to achieve a very modest
increase in alcohol strength?  It seems unlikely, given that there are known
competitive methods which are much faster and easier if the point is simply
to concentrate the alcohol-- distillation by freezing comes to mind, and it
only requires a single winter season, and regular simple distillation
without a column is technologically trivial giving a product around 30%
alcohol.  Still, a recent experience I had does indicate that osmotic
processes under mild conditions can concentrate alcohol content, though it
doesn't say anything about whether wooden barrels are appropriate for this.
Around a decade ago I made a batch of braggot which wasn't very good; I put
it in some 2-liter plastic soft drink bottles and left it in a storage shed,
where its existence was forgotten for about 10 years.  Then I found it last
month, but it was very strange: while originally the bottles were under
pressure and had a significant void space, now there was no void space at
all and the bottles had caved in to around 2/3 of their full volume.  I
expected it had turned into some sort of vinegar or at least that all the
alcohol had long since evaporated, but upon opening one of the bottles found
that it was quite palatable, tasting completely unlike the original braggot:
flat, very aromatic much like a fortified wine, and  enriched in alcohol
from the original batch.   I really hadn't a clue as to what had transpired,
and was puzzled until this thread was posted.  Then it dawned on me that the
PET plastic in the soft drink bottles must have acted as a selectively
permeable membrane to CO2 and water!  So Sean, I've got to thank you, Jeff
and the others in the thread who've solved this conundrum.

If the "solvent transfer mediated by ambient humidity" mechanism is
important for aging whisky, brandy, etc., please allow me to speculate.
Typically, the humidity at a particular geographic location over a period of
time varies widely, so that during arid times there will be water vapor
diffusion outward from the barrel; during very humid times there will be
some diffusion of water vapor inward.  In a sense the barrels would
"breathe".  Is it possible that during the "inhalation" phase that the water
movement carries key flavor molecules from the wood into the liquid contents
of the barrel?  I've seen some of these barrels, and they seem to be charred
on the inside-- the lignin in hardwoods contains a lot of cross linked
phenolic compounds which would release a variety of aromatics on being
heated.  It might be that at least some periods of very high humidity are
essential for these to be carried into the whisky-- has good whisky ever
been been made in a uniformly arid environment?  It's grist for the
speculation mill anyway.

Hiram Berry

> Relative volatility of solvent pairs can
> vary with temperature, but in 99.9999% of cases, not very much.

It's definitely true of ethanol/water in the temperature range under
> Sean
> -------------------------------------------------------------------------
> To unsubscribe from this list, send email to majordomo at pbm.com containing
> the words "unsubscribe hist-brewing" (or unsubscribe hist-brewing-digest,
> you get the digest.) To contact a human about problems, send mail to
> owner-hist-brewing at pbm.com

To unsubscribe from this list, send email to majordomo at pbm.com containing
the words "unsubscribe hist-brewing" (or unsubscribe hist-brewing-digest, if
you get the digest.) To contact a human about problems, send mail to
owner-hist-brewing at pbm.com

More information about the hist-brewing mailing list