A matter of reflux ...
- Did a small experiment with a new column today.
I've upgraded from a 1.2m x 1.5inch column to 1.5m x 2inch.
The power input is the same - 1800W
Hopefully the taller length would allow me to get the same purity, at a
Thought I'd give it a range of different reflux ratio's, and see what the
output was like.
The "drinkable" take from the session was 4.3L at 92% (yeah - final purity
affected by all the stuffing around) - this equates to 0.50 L pure alcohol
per kg of sugar used (it was an 8kg.25L turbo brew that got down to 0.990).
Its a bit less than the 0.610 L/kg theoretical maximum - so it looks like
I'm only getting about 81% of the available alcohol out (as drinkable
quality). This figure though doesnt count the 120 mL tossed as foreshots,
500 mL heads, and the approx 300 mL collected as tails.
Heat up took 74 minutes (vs 62 minutes off my calculator), so there might
be a little heat lost from the boiler.
Big surprise was how quickly it settled down to an a steady temperature
when under total reflux - only 7 minutes to drop from an initial peak of
81.4C down to 77.5C. It then sat there for another half hour without doing
anything different. Is it really worth waiting longer than that inital 7
minutes ? Do the heads concentrate up any further in the following period
? I cant see the benefit in spending longer equilibriating the column -
its allready there in 7 minutes.
Taking the heads off 1 drip/second - there was a noticable change in the
smell after 80mL - it cleaned up heaps and became much cleaner/sweeter.
But the temperature still remained low : 77.6C. It still sat there for
the next 350 mL to come off. Maybe my thermometer is in a cold portion of
the head (its not quite in the path of the vapour, but up a bit). By 77.9C
(another 200 mL later) i was fed up waiting for it to get up to 78.2C so I
started the main run anyway - the actual spirit was very clean & smelling
good by then. I only threw away the first 120 mL; kept the next 500 mL
into my "redistill" collection bottle.
I then did the rest of the brew off at different reflux ratios / collection
rates. What was quite interesting to see was how a very high collection
rate could be well sustained initially (eg still gave high purity spirits),
whereas much later in the run, when the pot was quite depleated in alcohol,
the same reflux rate / collection rate gave a dismall purity - thus
demonstrating the need to continually increase the reflux rate during the
course of the run if you want to keep the purity high.
The runs were ....
35 mL/min (RR=4.2) : 95%
75 mL/min (RR=1.9) : 93.5 %
29 mL/min (RR=5) : 95 %
60 mL/min (RR=2.3) : 92 %
67 mL/min (RR=1.7) : 80 %
30 mL/min (RR=4.8) : 94.5 %
47 mL/min (RR=2.6) : 85 %
20 mL/min (RR=7.2) : 94 %
23 mL/min (RR=5.6) : 87 %
12 mL/min (RR=12.2) :95.5 %
Just roughly, this would suggest that if I'm happy with say 94%+, I could
start out collecting at around 60-70 mL/min, then slowly decrease this to
around 30 mL/min mid run, and finish up with it at 10-15 mL/min.
Total collection time around 4 hours, but this involved a LOT of frigging
It has also shown too that there is no need of such a high reflux ratio to
start with (eg that of 8-10) as recommended by some. Around 4 is OK, and
it delivers it twice as fast. The main point is to head towards a reflux
ratio around 10-15 as the brew runs out of alcohol.
Next time I play, I think I'll do it slightly differently - target say 95%
efficiency, and measure the collection rate / reflux ratio required to keep
it there (say adjusting every 15 minutes ?). Should be a very interesting
Another wee interesting trial was fiddling with the cooling water rate ....
incoming cold was at 12C (yeah its the middle of summer down here !)
680 mL/min = 44C outgoing (warm)
625 mL/min = 48C outgoing (quite warm to touch)
525 mL/min = 53C warm to hot
425 mL/min = 60C (to hot to hold finger under for long)
Each of these were below the dewpoint of the spirit (78C), but I'd plan now
on running it at around 450-500 mL/min - quite warm to touch as a guide.
Overall - quite impressed with the new column - I think the extra height is