Photos of Reflux Stills
Theres quite a variety of general reflux stills about.
I have constructed my own offset head still.|
The unit breaks down for easy storage, cleaning and transport.
I have plans for a bigger boiler as this one is less than 19L.
|Tim J's Still
|Ken & Mikes Project
||We get 2.7L from a 23 L, 10.7% mash. It is 91%|
Mike : We polished with activated charcoal and used
caramelized sugar for colour and oak toasted oak chips are in the diluted
(52%) product for the vanillans.
I insulated the boiler to make it
more efficient (around the pot and a round piece on top of the pot,as per picture).I also insulated the colomn with 2
pieces of insulation that are fit for a 1" pice of copper pipe.(one piece per side...look at the picture).By the way...I use
7 brass scrubbers from Zehrs market in my column and they
impart no extra flavours!!! When rolled out between the
hands...the scrubbers are about 5" long and 2" in diameter.
Put them into the column slowly and carefully!!!
(also see Mikes recipe for Capt Morgans Spiced Rum)
Pot Temperature: Measured on the lid for a better idea of the steam temp as
this is the portion that is most critical. By keeping the temperature
between 198-206 degrees F (92-96C), we are able to tell when the wert is getting too
hot and giving us that 'Brackish' smell that can be calmed with carbon
filtering but never quite goes away. Now, Mike and I have slightly
different opinions on the tails aspect but we do agree that the burning of
the wert is what was making the first few batches of rum a little on the
unpalatable side. We bit the bullet and drank it anyway but with the new
numbers, we have made a very flavourful rum product. Final flavouring with
cloves and cinnamon will be the defining bit. When the lid temp reaches
208F (98C), shut down the still. I know, there is more to come but I found a +4
degree variation between the lid and the boiler body. As soon as the boiler
body hits 212 (100), the wert started to burn and that... smell... came through
into the product.|
Variac: Because we are using 2/1500 watt heating elements from a water tank
for heat up, we get to temperature on 23L in about 45 minutes. At that
point, we switch to one heater only (all that is needed to maintain the
temp now) on variac control at 105-107 vac. Kinda like using a stove
rheostat at 9 instead of 10. This slows the heating of the pot to get a
longer run and as the alcohol is boiled off, the temperature climbs. We
don't do a full boil all the way through, just a slight rolling.
We also found that we only needed 2 scrubbies (teased until they cried) in
the column you see in our picture. We had added some more but it took quite a
bit of heat to get a flow. We clean distilled water with nothing in the
column and decided that less is more. To help the reflux happen quicker
without losing too much in percentage, ( about 5% ) we didn't have to run
quite as warm and ended up with a smoother result. We will add 1/2 of a
scrubby for the next boil but we are on the right track for us. Keep in
mind that these scrubbies have been pulled almost to the length of the
column and allowed to coil back on themselves. This, it would seem, gives
us the same area as the rashiq rings.
||Reima tells me ..."there is a neat still at http://ponu.haisee.com/only
it is in my old lingo so I have translated it hoping the owner
does not mind.
It should tickle the minds of those who like to play with thumpers,
cause it seems to have one built in."|
Lower colon is 300-400mm high. Top one is 250mm. Red tube of 8mm
and blue of 10mm copper.
This does a double distilling in one go, like a built in thumper,
and is self tuning. Only critical tuning needed is cooling water flow.
Steam goes through "A" and condenses on its way back to lower colon coil
made of 8mm by about 1m length copper tube. In this coil, being in the path
of about 90C vapours a new boiling of distillate occurs. At the bottom
of this coil there must be a 50mm liquid lock (a bend up). The lighter
ethanol in this coil boils and travels up to the T piece and to the
condenser, the heavier stuff runs back to the boiler through the liquid lock.
Both colons are of 60mm outside diameter SS-tube
Lower colon is 350-400mm and top one is 250mm length.
Coils of 8mm copper tube. Tube A 10mm and B 8 mm tube.
This works on a 20 litre boiler heated with a 1 kW heater.
Condenser coil length is not critical 0.5m length is enough. Highest
pressure in this system is then 50mm water height.
When tube A is starting to warm up it is time to let in some cooling
water, only little. It should not run out as thick as a match stick, if too
much cooling there will not come out anything. Try to keep the top
colons lower part hand warm when its upper part is hot. And remember the
slower it comes out the stronger it is.
|Scott's POS Firewater Hook Column & Flamethrower Condenser
This is my column and condenser, assembled
minus the tubing. It achieves total reflux by
default, and the distillate is drawn off via the gate
valve. I've since added a needle valve below the gate
valve in order to achieve finer control.
With a 750 watt element, I've drawn 400 ml per hour of
95%+ alcohol, which isn't very fast, but I just
purchased a 1500 watt element that will probably
increase this significantly. The column is 45 in. x
1.5 in. and is packed with s.s. pot scrubbers.
The condender assembly is 44 in. long, with the
jacketed portion being 26 in. x .75 in. around a .5
in. pipe (a lot longer than necessary, I've found).|
I've only had two issues with this design.
1) The condenser screws onto the column, thus in order to
draw off alcohol, I need to wrap the column's male
fitting with PTFE tape to insure that the take-off
valve is pointing toward the ground.
assembled, the column and condenser need to be
stablized. I use a homemade wooden frame which has
|Purity : 92%
||Purity 94.5-95.0% at 40 mL/min|
The boiler is a 30 L
stainless steel vessel sold as a fermenter. It is Italian, and is
imported into the UK by Brupaks Ltd. The column is 1020mm long, of
54mm copper tubing with 20 stainless scourers, insulated in use and
without through-tubes. At the top of the column is a water-jacketed
reflux condenser (inner 22mm, outer 28mm) 300 mm long. Above this
is a thermometer port. At a tee from the port a final jacketed
condenser (inner 15mm, outer 22mm) 530mm long descends parallel to
the column and discharges into a jar.
The water-flow system is a little different from most. Cooling
water flows into the base of the final condenser's jacket and leaves
from the top. Two cocks are placed in parallel; one adjusts the
supply to the reflux condenser; the second allows water to bypass the
condenser. Water leaves the reflux condenser's top and falls from a
spout into a tundish. Waste water is collected into a single tube
where it leaves the system.|
a) reflux condenser (cool water enters at base)
b) collecting condenser
c) reflux bypass cock
d) reflux supply cock
e) thermometer port
f) water leaving system
In use, total reflux is easily maintained at a steady boil
(c closed, d opened) for, say half an hour; then c is partially opened
and the fractions are taken off. I didn't have the spout and tundish
at first, and found that air-locks occurred to the top of the reflux
condenser towards the end of operations when the water was warm.
When I put in the spout and tundish (which makes the whole thing look
quaintly Victorian) the working was made much more stable, with no
air-locks forming, and equilibration of the whole cooling system at
atmospheric pressure. The tundish allows a visual inspection of
reflux cooling water flow rates. Although the two cocks seem crude,
they only apportion the flow of water at atmospheric pressure and so
do not need to be precision-made.
I thought I would send you a picture of my new offset still-head, which replaced my old two-condenser pattern. This new one is much better; it gives far more control over reflux. The column is, as before, just over a metre of 54mm diameter copper, filled with stainless steel scrubbers. A 30L stainless drum is heated by gas.
Now I've done a few distillations I'm getting my eye in. I try to do the whole process keeping under 80 degrees at the still-head. Towards the end I turn up the reflux ratio by turning up the gas and cutting the collection. With a wide column this works really well. I give up when the returns diminish, which they do rather suddenly; it's a bit like a switch being flicked (the fractionating process must be quite efficient.) The whole process gets done in 2hrs 30mins from cold (25L.)
I've been racking my brains for an easy way to clean out the column; water doesn't do it very well. Upward travelling steam is ineffective, as the fusels in the column condense and fall back into the boiler and contaminate the water. Now I pass the steam downwards through the thermometer port after loosely capping the top of the condenser with a small weight (for safety.) This seems highly effective - the fusels come out quickly and the whole column smells sweet and fresh after ten minutes.
My wife and I enjoy the spirit obtained from a sugar/grape juice/orange wine. Quite a lot of flavour seems to come off in the distillate, so (after throwing away the first 50mL) I only bother to polish the first and last quarters of the product (after dilution), and mix these with the unpolished middle. 'Polishing' is one slow pass through about 10cm of activated charcoal. This spirit makes a nice vodka and a good base for gin.
|Victors Aircooled Still
I've been tinkering with the home brewing and distillation thing for
some wee time and are up to the Mk3 iteration of my distillation
Mk1 was a simple pot still.
Mk2 was a short (approx 500mm column) reflux still.
Mk3 is the one refered to here.
Only the boiling vessel has survived the three iterations.
I've not used any calculations whatsoever, but relied on a chemical
engineer (Who has built a 60 plate equivalent, 75mm diameter x 6m long
reflux column,7kW steam heated fractionating still using
calculations.) mate for rule of thumb guidance. The result I've
acheived (I believe) is outstanding.
Going through it for you from the floor up, the components are,...
1. 2kW electric heater.
2. 15 litre boiling vessel.
3. 50mm diameter by 1.5m long refluxing column, packed with stainless steel wool.
4. Refuxing controller.
5. Air cooled cooling tower, including thermometer.
Assembly of all these components takes 5 minutes.
Bringing 13 litres of wash up to refux temperature takes much longer,
2 hours, a rate limited by the size of the 2kW heater. I can bring
this time down by preheating just the pot on the stoves 3.5kW
hotplate and then assembling the reflux apparatus.
The pot is a domestic 15 litre pot. The steel band that pulls the lid and pot
together is from a 20 litre paint tin. Simply remove the bands
existing over centre lever closure assembly, adjust its length to
fit your pot, bronze on a nut to each end, drill one of these nuts,
and fit a suitable set screw (ie. A fully threaded cap screw.) to
tighten the band, and seal the pot.
The reflux tower attaches to the lid of the pot. Because of the towers
length, three issues present themselves.
1. The pot is only 0.8mm thick, so I folded a 1.2mm rib to rivet to
the lid to improve the rigidity.
2. I always rope the column up to the ceiling. I don't want the thing
3. Jointing the [pot lid/column/refux controller/cooling tower] sections together.
Something simple to fabricate, assemble, vapour tight was
required. Re: Jointing the [pot lid/column/refux controller/cooling tower]
The solution was for the tubes to slip together by flaring one of the
tubes, courtesy of an automotive muffler joint. These joints are made
vapour tight by wrapping a paper gasket onto the male end of the joint
during assembly. At the pot end, the tube is attached to the lid by a
flange silver soldered to the end of the tube.
I've run it a few times and its brilliant.
Max delivery rate is 1800ml/hr and I throttle this down to 600ml/hr
using a needle valve on the refuxing controller. This achieves a refux
rate of 3:1.
I still distill twice, the first time without the column to strip the
wash from 12% to 40% to reduce the volume. Removing the column for
this stripping operation is easy as all the tubes slip together with
paper wrap gaskets (there is no pressure in the apparatus at all) (The
tubes are flared on the ends, courtesy of a muffler joint that flared
the ends of the tubes for me).
The reflux controller I made, and its a simple cup inside the column
that collects the distillate from the cooling tower. The cup is 50mm
high to give some head to stabilize the flow rate from the needle
valve. The refux portion simply overflows the cup to percolate down
The cooling tower is a work of art.
I built this from an air conditioner evaprator.
The evaprator was a core fabricated as 20 x 8mm ID Cu tubes supported
in a finned aluminium layer heat exchanger. This exchanger measured
125x100mm square by 300mm long; perfect! I folded up the outer metal case to:
1) support two 240V fans;
2) protect the finning;
3) close the refux controller end of the heat exchanger;
4) support a thermometer port.
One end of the cooling tower is open to the atmosphere, and the other
end is closed by a me fabricated collection funnel which slips into
the top of the refuxing controller. This fabricated collection funnel
includes a port to hold the thermometer.
I intend to rework the reflux controller as the needle valve I made
on my lathe is far too coarse. Typically, the refux ratio changes from
zero through to 3:1 in 1/10 of a turn. It is real touchy. I've got
some pneumatic control needle valves and these should be much better.
I also intend to rework the amount of head over the valve from the existing
50mm to 125mm.
As it is the the performance is brilliant as I said before.
My chemical engineer advisor suggests that the column as it is
dimmensioned is close to optimal, and that increasing the heat input
during refluxing would cause the column to flood. The physics of the
column (I think?) is that with the packing density of the stainless
steel wool that I have corresponds to a plate distance of 75mm to
100mm. So for the 1500mm long column that I have, this corresponds to
somewhere between 20 to 15 plates, say 15 plates. Adding the
liquid/vapour interface at the wash surface in the boiling pot, gives
16 plates. With this number of plates the alcohol from the reflux controller
is 96% and almost tasteless.
My understanding, about number of plates and refluxing,...
The number of plates and reflux ratio are interdependant.
As a first order approximation, an increasing number of plates
corresponds to maximising the % alcohol acheived, and an increasing
reflux ratio corresponds to maximising the purity of the alcohol
Up till now my technique is, and I intend to refine this as,...
- Ferment sugar using yeast/nutrient mix to 15% alcohol.
I've been using the swedish turbo8 mix, which I have found very
I intend to change this to ferment sugar/brewers_yeast, and making
my own nutrient mix from a mix of ammonium sulphate, potassium
sulphate, vitamin B1.
Acheiving approx 12% alcohol at lower cost.
- I intend to filter the wash through a like swimming pool filter
(Paper element covered in a fine sand filter substance, I think the
sand is called benatite.) to remove all yeast cells, as distilling
yeast cells causes them to burst, which liberates nasty flavouring
- Strip the wash from 12 to 40% to reduce its volume.
Stripping without the column is quick for two reasons.
(a) Without the column the energy requirements to get the wash
vapour through the 16 plate mass/vapour_equlibria is reduced to
a single liquid/vapour interface at the wash surface.
(b) The size of the apparatus is reduced so that I can do it on the
stove at 3.5kW heat input.
- Refux the 40% stripped wash to 96%.
- Dilute the 96% to 80% with water.
This step is necessary for the activated carbon treatment, next
step, to remove traces of unwanted flavours. The water is necessary
(I think) because the alcohol is non-polar, whereas the water is
- Stand the 80% alcohol over activated carbon for 3 to 7 days.
- Flavour the 80% alcohol.
So far I've continued with the medium roast american white oak
chips for burbon, and intend to add a little self roasted sugar
(aka caramel) to give colour.
The current batch of chips I have are very dissappointing. Maybe
they are not roasted enough, because they impart a slightly fruity
(undesireable) flavour, whereas my earlier chip batch yielded the
(desireable) woodsmoke taste and aroma. I'll try a little
additional roasting in the oven to see how it goes. If the flavour
is unacceptable, bingo - redistill and start again!
Since then, I've done additional roasting of the chips, and the
flavour is much improved.
I've also given lemon vodka a go, using lemon rind and lemon juice.
I've also got some french oak chips for cognac, which I haven't
Another funny thing struck me as I was making myself an orange
herbal tea the other day. It suddenly occurred to me that here I
was spending a great deal of time researching flavours and this tea
bag I held on to might be worth a go.
Into 200ml of 70% alcohol it went for a few days, dilute to 40%,
add some sugar and it's not a bad approximation to orange liquer.
- Mature the 80% product.
- Dilute to 40% for drinking.
- Avoid alcoholism.
||Smithers says ..|
It has the standard urn element 2.4 kW and a 1380 W
Still Spirit element in it.
I heat up with both and cut down to the 1380 as temp starts to rise up the column.
It will collect at 94-95% all day everyday.
I know of 20 people who have copied this design and have had great success.
Column : 2" stainless tube x 600mm high
Packing type : stainless pot scrubbers
Distillate flowrate : approx 600-700ml/hr @ 95% purity
Pot capacity : 30 ltr urn
approx construction cost in aussie dollars ..
temp controller cost about $60 to make
still head materials (stainless) cost about $15
element (1.38kw) cost $40
one carton of beer to the welder $20
dig temp probe $18
urn $20 at an auction
The controller will switch 3.0kw so theoretically you wouldn't need to use
the 1.38kw element but I like a speedy heat up
|Construction details/tips :|
- pot scrubbers pot scrubbers pot scrubbers
- develop a way of controlling your input temp, ie controller, large rheostat,
- don't just rely on water flow for adjusting final temp, as this will lead to
(aka still spirits stills) unless you have a very well designed still (which
would be better)
- toss the first 100ml, don't be greedy
- equilise your column before you start collecting
- insulate your column, saves money
- when distilling I like to have my distillate coming out warm and let it drop
I think that this does help in final purity by evaporating some fusals on
the way down.
- The insulation is very similar to the foamy type insulation you'll find in
airconditioners ie goto industrial air con shops.
Most high density foams should work, just do a test, boil your jug (assuming
metal) and tape some foam on, if the outside stays so you can touch it, that is
cool (pun) just make sure that the other side isn't burning either, that is also pretty
important. The tape holding it on is just any type of material tape
||Column is 115cm long and 42 mm diameter packed with stainless steel scrubbers
Output is about 500 mL/hour at 94% purity
See Jan's site for heaps of details re construction and design options.
||The coil (for reflux) is made of 10mm soft copper, easy to wind but
really inefficient to use.
The coil has 12 windings and needs massive amounts of cooling water to
keep the top cool enough for sufficient reflux.
A innerside coil is much better, even pass trough pipes AT THE TOP!
are better, but i don't like these because I want my column to be
clean and empty on the inside.
That means NO pipes coils or other things in the column, now I can
experiment with lengths and so, it also makes it multi functional.
U can turn it with a short EMPTY column into a potstill etc
I'm experimenting with a new cooling thing a (sort of) drawing is
atached to this message.
I hope you get the idea....
It is not very easy (to clean). But I can remove a portholes by 4 screws. I don't
remove them very often. After a destilation I "clean" the Raschig Rings by
steaming them for 15 min by turning off the reflux cooler. I don't know if
thats very effective.
Maybe a hot water flow upstreams would be a better solution. This method is
also used for cleaning sand filters in traditional drinking water
I found the column in a copper rubish container, paying about NZD 4,50 per
kilo. The portholes were already there. I don't know the original function
of the "column". It is easily possible to turn the column, so may be it is
used to mix something ?
The connection between the pot an the base of the column is angled because
the heavy column is fastened to the wall and the stainless steel drum stands
on the ground. The "double angel tube" makes it possible to get always a
good connection between the pot and the column. If there is a difference in
level, than there is no problem.
The tube into the bottom of the pot is used to syphon out the
spent wash. I advise to use not a too small diameter for that tube.
Than some particals in the wash won't be a problem to get through.
The tub of glycerine that it sits in - is that open at the top, or does
it seal off against the pot ?
Now it is still open because the cooking-pot is made of cast iron.
I am making a beter version now by using an other half copper boiler instead
of the cooking-pot. It wil be soldered with an expansion tube. Than it is
possible to use water instead of the glycerine too. By using a pressure
relief valve the boiling-point of the water will decrease. This system is
mostly used nowadays by all the german stills. By the way: you can buy these
profesional home stills in germany for about NZD 10.000,--. That is a lot of
money, quite a lot of farmers in germany and austria have them for stilling
Column with Plates:|
Each "block" = 5mm. So the diameter of the column is about 100mm.
Each level has 3 standard caps of 35mm diameter intern. The cap height
intern is 26 mm. I find out myself that that the LEVEL of the overflow pipe (one next to 3
caps) is very important.
An engineer of Jacob Carl in Stuttgart (making profesional still equipment)
adviced me to make the level of these overflow pipes variable.
The U-tube on the bottom stage is necesary to force the
alcohol vapour to go up only through the "gas" pipes under the caps.
In the U-tube is always alcohol liquid, so that no vapour can go up here;
only liquid getting down.
Alambic: (for aromatic distillates)|
The cooling coil in the drum has 10 cycles of about 0.26m makes: 8,2 m.
I got plenty of 12 mm copper pipe and it is good to have some extra for the
safety. It is fixed in a 50 L wine cask.
||HuckMubb comments :|
Running the dual tubes instead of one single 1/4" OD
copper tubing increased output
to 900ML per hour.. This still is very small and not a high volume
design but it can be built easily by anyone and is a great homeade
example for a beginner not having the means to heat a larger still...
|Phil's Still (email : firstname.lastname@example.org )
5 L boiler|
1.4 m fractioning column packed with SS scrubers utilizing the cage system as
shown with scrubbers installed. This allows the easy removal from column.
then headspace of @ 250 mm with reflux cooling coil with water flowing down
the coil then exiting via vertical center pipe. This allows a progressive
cooling up the column rather than introduce cold water 250mm down and have to heat it
out of order. The fractioning column splits apart. I have the
column of 1.4m odd of 50mm ID pipe rising from the boiler, this contains the
SS Scrubers, then the top section just slids over this using a length of
54mm ID pipe has 50mm ID soldered to it at the top.
Where the 50mm meet inside the 54mm i have carefully lapped the two together
to ensure a near gas proof joint with neither solder or thread required. Then
to make it fully gas tight the outside slip joint has teflon tape wound
around the 50mm pipe at the point the 54 slides to. This has proved simply
The cooling coil is blocked of at the top by a plate forcing all vapor to
pass over the coil at some point
The digital temp gauge as you can see is inserted to measure the vapor temp as
it rolls over to head towards the condensor.
The condensor is a shotgun type with 12 6mm pipes
inside a 38mm jacket. It is 635mm long internal cooling length @ 8m cooling
length and a HUGE surface area
After a couple of small cleaning runs and a test with some plum wine a real
run was done. 95% was achieved without any problem.
Flow to the reflux coil can be adjusted independent of the shotgun
condensor. A few more runs has yeiled the following..
I run total reflux for @ 30-40 minutes or until the vapor temp stabilises
about 40C. I then slowly back off the reflux water and raise the vapor temp
till i get 77c reading on the DT (digital temp) (it reads 1.4 low) over
about 10 minutes.
I take WHATEVER the reflux ratio is and smile and patiently collect the end
product in 250-300ml batches. (empty and sterilised barcadi breezer bottles)
Late in the run i will increase reflux once only to hold the temp reading
Once the vapor temp rise 1C I stop collecting the main cut. collect some
tails till 85C then stop.
Total water usage is less than 20 litres/hour and with some better valves for
water control this should drop to about 12 l/hour.
This is due to the reflux condensor not even getting warm past the top T
piece. Just can't get fine enough control at this time.
Urn has three heat setting high (2000W) med and low. Use high to heat up wash
and Med to keep a rolling boil
Purity : 95 %