reflux design check

[bionicbrew]

I am building a 3" VM column. I am trying to get this right the first time so I don't have to rebuild it later. I see lots of info on 2" columns, but little detail on 3" ones. here's the plan-
15.5 gal keg boiler, propane power(I have 2 types, a turkey fryer type and a banjo type). plan is for- 3" column, 48" of packing to 2" T to 1' of 3" pipe with 8-10" reflux condenser. take off at 2" SS ball valve to elbow to 2"-1" reducer to 1" to 3/4" reducer through liebig that is 30-36" long. plan is for the A style in this diagram.


questions are-
for this diameter-
it is less than the 20-30x the diameter for packing recommendation. I can make it 5' of packing, but it really increases the cost of the build by $60 USD. (I can do that if really necessary) I don't want to have to redo this thing. but the calculator shows a .5% increase from 94.5 to 95 at 2:1 reflux. my goal for a 3" column is speed and purity of azeotrope. is 48" of packing good enough?

is the liebig product condenser good enough for the VM and a later pot still set up(on this boiler and propane)?
I know the answers are a varying degree of trade offs. just making sure I am in the right area.

[Rebel_Yell]

Go as tall as you can for purity. There is a lot of difference between 94.5 and 95 taste wise. Sorry gota go lunch!

[Harry]

Any good column design starts with MATH. That's what engineers do. It's not hit & miss. If you can handle simple math, follow along with this...


Any smallish separation device (your column) can be worked out on paper mathematically, providing you know the vapor and packing parameters.
Mike Nixon and Tony Ackland (engineers & pioneer founders of this hobby) gave us these parameters many years ago, after much research & testing. So let's use their findings and not reinvent the wheel. I have put together the following essay for the benefit of would-be designers...



Vapor parameters:

Vapor generation rate (@ 1kW power input) = 45.77 cubic inch/second. (source: M.Nixon M.Eng.)
Ideal Column speed = 12in (min) <-> 18in (max). (source: M.Nixon M.Eng.)



Packing parameters:

Typical HETPs for common packings are :

Packing HETP
S/ Steel Wool Scrubbers = 0.13 m
Marbles (10mm diameter) = 0.33 m
6mm Ceramic Raschig Rings = 0.24 m
13mm Ceramic Raschig Rings = 0.38 m

(source: Tony Ackland Ch.Eng.)

The volume of a column L inches long, with cross sectional area A square inches, is V = L x A cubic inches.
As 1 kW produces 45.77 cubic inches vapor each second, W kilowatts produces 45.77 x W cubic inches each second.
Therefore 45.77 x W = L x A (written more simply as 45.77W = LA)
L in this is, of course, the same as the speed in the column in inches/second, ie. L = Speed


So Speed S = 45.77W/A inches per second (Equation 1) In terms of the diameter of the column .... if the diameter of the column is D, then A = Pi x D x D / 4

( or A =PiDD/4)


So Speed S = 45.77W / [PiDD /4]
ie. S = 58.3W/DD (Equation 2)
or W = SDD/58.3 (Equation 3)
[Where S is in inches/second, W is in kilowatts, and D is in inches.]

Take your pick which equation you use ... they are all equivalent.
Sorry about the 'DD's etc, but I think it is perhaps a bit clearer to write
DD for D squared than D^2


(source = M. Nixon)

So, let's get down to business...

Let's work out a 2" column first, so we've got a yardstick to go by.

The standard 2" x 40" packed column requires approx 1.2KW power input to achieve good separation.
This is a known standard, however let's prove it to be absolutely sure of what we are doing...




HETP means (H)eight (E)quivalent of a (T)heoretical (P)late.

Therefore, for a 2" x 40" column, using Tony Ackland's S/S packed, we do this...

HETP = 0.13m
Change it to inches (no confusion)...


1 HETP (in inches) = 0.13 x 40
Therefore 1 HETP = 5.2 inches of S/S packing.



Thus for our 40" tall packed column, we calculate getting approx. 7 separations or distillations (40 / 5.2).

That's good separation. It's not perfect but if you want better, then just increase the column height by multiples of 5.2 inches (1 plate).


Now we need to work out the power requirements...

If ideal speed = 18"/sec, then to find out what power is required, we need to start with this Mike Nixon's vapor speed equation and plug in the parameters

for our column.

Speed is 18"/sec, which is 18 = 45.77W/A.
Or, it can be expressed as 18 = 45.77W / [PiDD /4] if that is easier. They are the same thing (as Mike said).

Now plug in the parameters for the 2" column, and do one step at a time...


18 = 45.77W / [PiDD /4]
18 = 45.77W / 3.1416 x 2 x 2 /4
18 = 45.77W / 3.1416
18 = 14.57W (rounded)

So far, so good. But how do we get the figure for W (our power requirement). To continue..

As we know in math, whatever we do to one side of an equation, we must do to the other side, to isolate the unknown (in this case the unknown is W, the power

we require).

.

From the last step, we have,
18 = 14.57W
18 / 14.57 = 14.57W /14.57 (both sides divided)
1.2354152367879203843514070006863 = W

Rounded, we get
W = 1.24 Kilowatts



And now at last we have our answer.
We have PROVED what we set out to do,
ie establish the power input for a STANDARD 2" x 40" mesh-packed column.
---------------------------------------------------------------------



Using the same math , it's simple to work out power for a 3" x 40" (7 plate) column...

Plug in the 3" diameter.
18 = 45.77W / [PiDD /4]
18 = 45.77W / 3.1416 x 3 x 3 /4

Now do the math...
18 = 45.77W / 7.0686
18 = 6.5W (rounded)

Continuing...

18 / 6.5 = 6.5W / 6.5 (both sides divided)
2.7692307692307692307692307692308 = W

Rounded, we get
W = 2.78 Kilowatts


We can see clearly that the power required to run a 3" column is more than DOUBLE that of a 2" column.

[timmyjane]

Thus for our 40" tall packed column, we calculate getting approx. 7 separations or distillations (40 / 5.2).

That's good separation. It's not perfect but if you want better, then just increase the column height by multiples of 5.2 inches (1 plate).

Great answer Harry. What is considered perfect seperation? For some reason 11 is in my mind.

[Harry]

Thus for our 40" tall packed column, we calculate getting approx. 7 separations or distillations (40 / 5.2).

That's good separation. It's not perfect but if you want better, then just increase the column height by multiples of 5.2 inches (1 plate).

Great answer Harry. What is considered perfect seperation? For some reason 11 is in my mind.

The more separations you can achieve, the closer you get to true azeotrope (the practical purity limit for common fractional distilling). The paradox is, the more plates you have, the less increase (per each) you get in ethanolic strength of the product. It's very complicated, not easily explained in a couple of sentences, and not really a discussion for the "Just Getting Started" section.

Short answer, if you can achieve at least 9 HETP and desirably more, without building a ridiculously-tall unweildy still, you'll be in the ballpark of nice clean high-proof booze.

[timmyjane]

Sounds to me like I need to add a coupler and another 10.4 inches to my column! Also appears that the OP is all set with a 48in packed column. 48/5.2=9.2 plates. Does packing typically go all the way down to where the column meets the boiler or is a little lead spacing a good thing?

[Harry]

Sounds to me like I need to add a coupler and another 10.4 inches to my column! Also appears that the OP is all set with a 48in packed column. 48/5.2=9.2 plates. Does packing typically go all the way down to where the column meets the boiler or is a little lead spacing a good thing?

Depends on who builds it. Practically, it doesn't matter (lead spacing). If the boiler lid is vapor-tight (good seal) then the vapor can only follow the path of least resistance, in this case that would be the packed column (which is roughly 90% void and 10% metal solid (the packing strands)).

[bionicbrew]

great info Harry! I did read the Nixon/Stone book and read those calcs. my thought for the 3" is- yes over 2 times the power input is needed. but the 3" area is 7.07 square inches vs. 3.14 for the 2", so more than twice as much vapor available for take off and theoretically more than twice the take off rate. am I correct here? it seems it should be close from what I've been reading. my burner should be able to keep up with the heat requirements with no problem.

I couldn't find the minimum plate equivalent again with all my research and lack of taking notes so thanks Harry again for the 9 min statement. I thought I had seen 11 is a good # like timmyjane said too. the calculator I was using had the HETP at 11-12 with the reflux ratio at 2:1(and 9.2 at 1:1 as tj pointed out). so sounds like 48" may be acceptable?

as to the difference between 94.5% and 95% being noticeable taste wise, that I found surprising and interesting. is that the same for most people? or those with a great sense taste for this stuff? I have that ability for beer, but not with liquor yet.

so now does the condenser length and leibig length seems good?

Harry, I appreciate you helping out this seppo. (my engineer/driver at work and band mate is from Melbourne. he and I are making this still together).

[Kapea]

...so sounds like 48" may be acceptable?

as to the difference between 94.5% and 95% being noticeable taste wise, that I found surprising and interesting. is that the same for most people? or those with a great sense taste for this stuff? I have that ability for beer, but not with liquor yet.

FWIW
I have a 2" x 48" SS pot scrubber packed propane-fired VM column. 48" of packing works great.

Neutral flavor-wise. My distilling regime is 3X:
1X - Strip hard and fast on my potstill
2X - Low wines through the VM achieving azeotrope and making heads/tails cuts
3X - Azeotrope diluted to 35% and a second pass through the VM achieving remarkably neutral flavor.

THe second run on the VM column is something I started doing recently at the recommendation of a mentor here on this forum. I was quite pleased with 2X for years. Then I added the second VM run to discover that further neutrality is possible. Comparing the 2X to the 3X I discovered I could taste some tails in the 2X.

The nice thing about the third run is there are no heads left, so there is no need to make the heads cut on that run.

Being propane fired I calculate my power input by the change in temp of my condersers' cooling water (reflux coil and product liebig). I estimate I run my column at around 2000W.

Insulating the column is very inportant in order to insure stability. I use hot water plumbing pipe insulation I bought at the hardware store (the soft black foam rubber kind). Two pieces trimmed to match column diameter and glued together cover my 2" copper pipe column nicely.

I understand all of the math, but from driving my column (empirically) I've found this is how my column works best. Every machine runs differently - YMMV.

[bionicbrew]

the formula I've seen here is 20-30 times the column diameter, so there must be a difference between a 2" with 48" packing and a 3" with 48" packing. most 2"ers I'm seeing are around 40" it seems, and 48" is not much longer for going up to 3" diam. I don't have much of a place to run a real tall column, so I was hoping to keep it to 6' or less for just the column itself. I may not be able to.

I like the change in cooling water temp idea for calculating the input power. I will have to use that! I do plan to insulate the column, so good there. I figured that a stripping run would at minimum be a good idea. I'll look at the 3x run when I dial it in and get some experience with it. good suggestions.

[Harry]

great info Harry! I did read the Nixon/Stone book and read those calcs. my thought for the 3" is- yes over 2 times the power input is needed. but the 3" area is 7.07 square inches vs. 3.14 for the 2", so more than twice as much vapor available for take off and theoretically more than twice the take off rate. am I correct here? it seems it should be close from what I've been reading. my burner should be able to keep up with the heat requirements with no problem.

I couldn't find the minimum plate equivalent again with all my research and lack of taking notes so thanks Harry again for the 9 min statement. I thought I had seen 11 is a good # like timmyjane said too. the calculator I was using had the HETP at 11-12 with the reflux ratio at 2:1(and 9.2 at 1:1 as tj pointed out). so sounds like 48" may be acceptable?

Sounds like you've got most of it sorted.

as to the difference between 94.5% and 95% being noticeable taste wise, that I found surprising and interesting. is that the same for most people? or those with a great sense taste for this stuff? I have that ability for beer, but not with liquor yet.

Yep there's quite a bit of difference. It's the difference between what is classed as grain whisky, and what is neutral spirit. Let me quote a passage from a book I hold in my online Library (it's free for all to read).

The definition of Scotch Whisky limits distillate strength to less than 94.8 per cent by volume, so theoretically the maximum permitted alcohol concentration for spirit used for grain whisky is 94.7 per cent. However, no Scotch grain whisky still is operated to produce spirit above 94.0 per cent ABV, allowing sufficient contribution by congeners to its flavour.

Source: "Whisky, Technology Processing and Marketing" by Inge Russell, pg 194

so now does the condenser length and leibig length seems good?

Condenser sizing is a subject of its own.
The reflux condenser (overhead HX) must be able to cope with TOTAL refluxing duty. That means it must be capable of removing ALL the "latent heat" from the vapor entering it, to effect phase change to liquid. If the OHX can't do this, some vapor will pass through and out the vent.

The Liebig HX (aka product HX) can be fed either liquid (an LM still) or vapor (a VM still).
In a vapor fed arrangement and duty, it must remove BOTH the "latent heat" from the vapor feed (to produce liquid), and the "sensible heat" from said liquid, to reduce the product temperature back to room temp. There's calculators on Tony Ackland's site for sizing condensers. http://homedistiller.org/legacy/cond_calc.htm

Harry, I appreciate you helping out this seppo. (my engineer/driver at work and band mate is from Melbourne. he and I are making this still together).

Did he tell you the origins of that term?

Anyways, glad to be able to help. Good luck with the build.

[bionicbrew]

...Sounds like you've got most of it sorted.

I 'think' I do. but did need some reassurance to be sure and guidance if not. I can be kinda anal and want to get things right the first time if possible.

Yep there's quite a bit of difference. It's the difference between what is classed as grain whisky, and what is neutral spirit. Let me quote a passage from a book I hold in my online Library (it's free for all to read).

The definition of Scotch Whisky limits distillate strength to less than 94.8 per cent by volume, so theoretically the maximum permitted alcohol concentration for spirit used for grain whisky is 94.7 per cent. However, no Scotch grain whisky still is operated to produce spirit above 94.0 per cent ABV, allowing sufficient contribution by congeners to its flavour.

Source: "Whisky, Technology Processing and Marketing" by Inge Russell, pg 194

that is very interesting info! I like it. I will look more at that one. (I have seen some of the books on your site I believe)

Condenser sizing is a subject of its own.
The reflux condenser (overhead HX) must be able to cope with TOTAL refluxing duty. That means it must be capable of removing ALL the "latent heat" from the vapor entering it, to effect phase change to liquid. If the OHX can't do this, some vapor will pass through and out the vent.

The Liebig HX (aka product HX) can be fed either liquid (an LM still) or vapor (a VM still).
In a vapor fed arrangement and duty, it must remove BOTH the "latent heat" from the vapor feed (to produce liquid), and the "sensible heat" from said liquid, to reduce the product temperature back to room temp. There's calculators on Tony Ackland's site for sizing condensers. http://homedistiller.org/legacy/cond_calc.htm

somehow I missed that one. thanks, I will play with it.

Did he tell you the origins of that term? ...

[/quote]

yes he did. I told him I didn't need a bogan coming around pissing in my pocket.... ok that was a few months after I learned the term and I was helped by someone else. we have an aussie dictionary started at work (we are firefighters), so far we have seppo, esky, bay marie, shifter(the wrench), and a few others.

[punkin]

You definately need 'fair dinkum' and 'streuth'. 'Fuck me dead' is another good one.

[bionicbrew]

ahhh yes fair dinkum in deed. that was on the list. I'll run streuth by him tomorrow.

[Harry]

"streuth" = shortened form of "God's Truth", an exclamation.
"Fair dinkum" = something or someone that's real or genuine.
"'Fuck me dead" = An exclamation, made when some scene or experience floors or gobsmacks you.

A lot of Aussie slang or "strine" as it's called in the dictionary (shortened "Australian") also comes from a mixture of the older English Cockney rhyming slang, and the Aussie penchant for shortening things or making them ridiculously long.
e.g "telling Porkies" = "pork pies" = "lies"
e.g. "have a butchers at this" = butchers hook = "look".

This one's hard to define to foreigners...
Blokes (males) with red hair almost always get the nickname "bluey". But a brawl or fight is a "blue".
Sheilas (females) with red hair are "rang..ers", while "ringers" can be stockmen (cowboys) or in horse-racing "ringers" or "ring-ins" are horses of good quality substituted for poor quality horses in a race to 'fix' the outcome. A famous incident in racing was when the horse "Fine Cotton" was substituted by "Bold Personality" (it won) in 1984 (see Fine Cotton Affair)