Originally By Tony Ackland

## Column Diameter Properties within Reflux Still's

SS = Stainless Steel Wool Scrubbers, RR6 = 6mm Ceramic Raschig Rings,
RR13 = 13mm Ceramic Raschig Rings, M = 10mm Marbles

Instead, I scale up/down for what I know works for me ... using scrubbers for packing, a 1.5" diameter column can handle 1800W. So .. for constant vapour rate per cross-sectional area ...

Maximum Power for a Given Column Diameter

1.00" = 800 W
1.25" = 1250 W
1.50" = 1800 W
1.75" = 2450 W
2.00" = 3200 W
2.25" = 4050 W
2.50" = 5000 W
Note that these figures are roughly the maximum power that you would want to use for any given column diameter. Mike argues that you should use quite a lot less power ...
All these figures above equate to a vapor speed up the column of 119 cm/sec (47"/sec) Assuming for ease a column length of 119cm, that means that vapor will traverse the column in one second, and during that time it is busy condensing and being reboiled many times.

Tony reported that a 25mm diameter column at 1350W gave very poor results, but a 36mm diameter column handled that power well. Vapor speed through the 25mm column would have been 211 cm/sec, so it's not surprising that it didn't work well as the vapor had only 0.56 seconds to traverse the column. Increasing the column diameter to 36mm brought the speed down to 101 cm/sec, so the transit time increased to 1.2 seconds. This is on the 'safe side' of the recommended figures in the table.

I personally favor lower power settings, and the reason lies in measurements I did with column stability, measuring the temperature gradient at steady state for different power settings, and first using a simple reflux column that relied solely in internal condensation for reflux. (A not-to-scale schematic has been posted in the Photos section of Distillers. It's called 'Temperature gradients')

I used a 2" diameter column and found that at 750W the temperature gradient settled out 2/3 of the way up the column, at around 80cm. This told me that vapor moving at 28 cm/sec had fully separated after spending 2.5 seconds in the packing. Increasing the heat input raised the point at which the straight line met the curve, but when the curve reached the top of the packing, the curve switched to the straight line sharply, not asymptotically, and the temp in the top rose to match the increasing temp at the top of the packing. So using 1350W with a 25mm column would result in full separation occurring at a point right at the top of the column, with little or no leeway.

Adding imposed reflux with a compound column did little to change the temperature gradient up the column, but what it did do was add that touch more separation in the region above the top of the packing. I could detect no change in the 'asymptotic' nature of the temp gradient when the straight section still lay inside the packing, but when power was increased to raise that point to the tip of the packing then the more I increased the power a sharp 'step' began to appear. The straight line went down to the top of the packing, then quickly jumped to meet the temp in the top section of packing. This indicated to me that the composition of the cycled vapour in the void between the top of the packing and the top condenser was the result of further separation imposed by the imposed reflux operation in that region. In effect, I had two stills one on top of the other, the bottom being a simple reflux still relying on internal reflux, and a recycling still that took what the reflux still gave it and used that as its starting point. This held true until either the power was increased to a point when the curve would have settled down itself in a longer column (about a quarter extra length) or the take-off ratio was increased to a stage when the sharp step suddenly broke down and the old asymptotic curve re-asserted itself, and quality instantly dropped.

OK ... so what has all this got to do with those figures in the table? Essentially, it is that the figures in the table are good for indicating the maximum you can push a simple reflux column to and attain full separation ... just!

If consistent results are wanted, then the aim should surely be to allow some leeway and try to get that curve settling down before the top of the packing is reached. That way, the reflux column has a chance to do its job as fully as it can before either taking off product, as in a simple reflux still, or passing on the results to a secondary top section that operates with imposed reflux for that final touch of separation. My personal 'cautious old fuddy-duddy' approach would be to reduce all those wattage figures in the table to 1/4 of what they are now and regard that as a good guide for reliable operation. This sounds drastic but, when you think about it, gives much greater assurance of high quality with simple reflux stills, and greater flexibility in take-off rates with a compound still. Maybe I'm just an aging Sunday Driver, but I find that I get to where I'm going with less hassle than a Boy Racer, and both my passengers and booze samplers enjoy the ride better!

Generally, a 2" (50mm) diameter is an ideal size to use. This will happily run from 750W up to 2500W without any trouble. If in doubt, go for 2".

Its this amount of energy that you put in which will determine the rate at which you make and collect the distillate. If collected at the condenser at say 95%, it works out roughly to the following figures. If you run a reflux ratio of 4 (e.g. return 40 mL for every 10 mL you keep - typical for SS scrubbers) - then the second figure is the flowrate you'd expect to collect at ...

1000 W = 52 mL/min (max, no reflux) or 10 mL/min (if RR=4)
1500 W = 78 mL/min (max, no reflux) or 16 mL/min (if RR=4)
2000 W = 105 mL/min (max, no reflux) or 21 mL/min (if RR=4)
2500 W = 131 mL/min (max, no reflux) or 26 mL/min (if RR=4)
3000 W = 157 mL/min (max, no reflux) or 32 mL/min (if RR=4)
3500 W = 183 mL/min (max, no reflux) or 36 mL/min (if RR=4)
4000 W = 209 mL/min (max, no reflux) or 42 mL/min (if RR=4)

Note though that you are probably going to be limited in how much power you can deliver to the still. Many homes only run 10 amp fuses in their fuseboxes. This will limit you to 240 V x 10 A = 2400 W before you have to have a safety chat with your electrician about upgrading the wiring.

The risk of making the column diameter too small is that the column will "flood", as discussed in "Chemical Engineering - June 2002" pp 60-67 by Simon Xu and Lowell Pless about flooding in distillation columns. These guys have been using "gamma scanning" to work out where abouts various distillation columns are flooding, and why. I'll quote a few paragraphs about "packed columns" for ya (they also did a fair bit on trayed columns) ....
For a given packed column, at the high end of liquid and vapour rates we encounter flooding as liquid backs up the column and fills all the void space in the packing bed. Poor disengagement between vapour and liquid (back mixing) reduces the separation efficiency, and the high liquid hold-up in the bed increases the pressure drop.

The traditional approach to analysing flooding in packed columns relies on measuring pressure drop. At low liquid rates, the open area of the packing is practically the same as for dry packing. In this regime the pressure drop is proportional to the square of the vapour flowrate. As the vapour rate continues to increase, eventually a point is reached when the vapour begins to interfere with the downward liquid flow, holding up liquid in the packing. The increase in the pressure drop is proportional to a power greater than 2.

At this point, the pressure drop starts to increase rapidly because the accumulation of liquid in the packing reduces the void area available for the vapour flow. This area is called the "loading region". As the liquid accumulation increases, a condition is reached where the liquid phase becomes continuous .....

The problem with this traditional approach is the difficulty in differentiating between the transition points of the loading or flooding in the pressure drop curve. Some suggestions for the definition of when a packed column become fully "flooded" are :
* the slope of the pressure drop curve goes to infinity
* the gas velocity is so great that efficiency goes to zero
* pressure drop reaches 2 in.H2O per foot of packing
* pressure drop rapidly increases in a region, with simultaneous loss of mass-transfer efficiency ......

There are two forms of liquid hold-up in packed columns. One is referred to as static hold-up. Static hold-up is the amount of liquid that is held onto the packing after it has been wetted, then drained - the film of liquid or droplets of liquid that adhere to the packing. This amount jointly depends upon the physical properties of the liquid and the type and material of the packing.

The second aspect is the operating or dynamic hold-up. Dynamic hold-up is the amount of liquid held in the packing by the interaction of the vapour and liquid flows. Dynamic hold-up must be measured experimentally. To measure this amount, instantaneously stop the liquid and vapour flows, then collect and measure the volume of liquid that drains from the packing. The total liquid hold-up in packing is the sum of these two forms of hold-up.....since the static hold-up is constant, the operating or dynamic hold-up changes in proportion to changes in liquid and vapour rates. The void fractions in a packed bed may change across the bed due to fouling or damage, and vapour-liquid loads may be different along the bed for different operating conditions. The peak loading could occur anywhere in a packed bed, or a liquid distributor could initiate the flooding......

An interesting phenomenon for random packing and most corrugated sheet packing is that the separation efficiency of an "initial flooding" bed could be better than a "normal" bed, because of high liquid hold-up and intimate vapour-liquid contact in the "frothing" regime..... But at the high-efficiency state it is difficult to keep the column stable, and the column could go out of control as a result of any slight process turbulence. For this reason it is always recommended to avoid designing a packed column close to the initial flooding point. In operation we would not then be overly concerned with some liquid accumulation or hold-up, as long as the column could be kept stable and under control......