# Collection rate

In simple distillation or in beer-stripping you collect everything which vaporizes from the boiler, but in fractional distillation you collect only about 10% of it. The reason for this is as follows:

The efficiency of a fractionating column in separating liquids of different boiling points is dependent upon two factors. One is the length of column and the type of column packing, i.e. its physical characteristics. The second is the reflux ratio, i.e. the way in which the column is used.

The principle of fractional distillation requires that the vapours rising up the column encounter the condensed liquid running back down the column. If, in the extreme case, all the vapour rising up the column were drawn off at the top via the collection valve there would be no liquid left for flowing back down the column. So there would be no counter-current flow and very little separation. At the other extreme, if the collection valve were closed and all the condensed liquid flowed back down the column (total reflux) the separation would be excellent but no product would be obtained. Obviously there has to be a compromise and this is achieved at a reflux ratio of about 10:1.

This ratio refers to the volume of liquid flowing down the column at total reflux compared to the volume drawn off through the collection valve. Thus, if the heat input to the boiler were causing the liquid to reflux at a rate of 1000 ml per hour, 100 ml per hour of distilate could be drawn off as usable product. The balance of 900 ml per hour would be flowing back down the column to provide the multiple mini-distillations on the surfaces within the packing required for the separation. It will be appreciated that the 10:1 ratio is not critical ... 8:1 would be acceptable and 12:1 even more so. The 10:1 figure is simply a reasonable value which is known to give good results with this type of packing.

So the first step involved in determining just how much alcohol can be drawn off per minute or per hour is to find out the rate at which vapour is arriving in the stillhead, i.e. the boil-up rate. When we have this figure we divide by ten and this is the volume of 96% alcohol which can be drawn off through the collection valve. There are two ways of doing this, one by calculation from the wattage input to the boiler and the other by direct measurement. First by direct measurement using the Mexican cactus stillhead.

With a known wattage input establish steady refluxing conditions and then open the collection valve WIDE for perhaps 30 seconds. It helps to tilt the column slightly to prevent condensate from running back down into the boiler. Measure the output per minute, either in terms of volume using a graduated cylinder or, more accurately, by weight using a sensitive scale. Take as small a sample as you can reasonably measure so as not to unbalance the system. You may wish to repeat with other wattage inputs. The same procedure can be used with either the metal or the glass "hatstand" design of still.

You can also calculate the rate at which a 750 watt (or any other wattage) heater will boil an alcohol/water mixture. The method of making this calculation is thoroughly explored in Appendix II, which also explains a little more about the mechanism of fractional distillation. Or, you might wish to accept the figures we provide below.

We found that with 750 watts input to the boiler the rate of reflux was about 45 ml per minute. Other wattage inputs gave proportional volumes. This means that with 750 watts input and a reflux ratio of 10:1 we can draw off 41/2 ml of 96% ethanol per minute. In practice we draw off about 4 ml to be on the safe side.

With slight variations in the construction of your column, in the way you have packed it, the amount of insulation you have used, the true wattage of your heating element, etc. you'll probably get slightly different results from the above, so do measure the rate of reflux for yourself. It's simple and informative.

It is not very convenient to set the collection valve each time you carry out a distillation by using the volume which flows out in one minute. It is too cumbersome. A better method is to laboriously find a valve-setting which does deliver 4 ml per minute and then count drops using a stopwatch. Thus, 4 ml per minute might represent, say, 30 drops in 10 seconds. Knowing this you can quickly adjust the collection valve to the right setting by counting drops with a stopwatch.

Collect at least 250 ml of this first distilate and put to one side for future processing and then start to collect the purest alcohol in a clean receiv er. Throughout this early phase test the distilate with your nose to see if you can detect any trace of heads.

The 250 ml or so of early distilate which have been put aside may be perfectly pure but the nose and the palate are extremely sensitive organs, particularly the palate, (and particularly your wife's palate!), and you would quickly detect an off-flavour if it got through into your final drink. Even commercial producers, with a laboratory full of sophisticated analytical equipment such as gas chromatographs, rely on taste panels to judge the quality of their product. It is called "organoleptic" testing and is the ultimate in testing for palatability. Play it safe, therefore, and put aside a generous portion of the initial distilate, even as much as 500 ml. It will not be wasted because, in a few weeks time, when a number of distillations have been completed and several litres of doubtful distilate accumulated, it can all be redistilled and really pure alcohol recovered from it. It will amount to a triple distillation and be exceptionally pure.

When all the ethyl alcohol has distilled over, which may take as long as 20 hours, the temperature will start to rise as the higher boiling point "tails" appear. Experience will tell you when to expect this to happen and you should start switching receivers well ahead of this point so that only a small volume of alcohol will be contaminated. The last receiver containing a trace of tails can be added to the discard bottle for later purification. There is no point in collecting extra tails for redistillation because they contain negligible amounts of ethanol.

When the fractional distillation is complete the packing in the column will be flooded with tails. These should be thoroughly washed from the column by pouring generous quantities of hot water down from the top.

When carrying out a fractional distillation for the first time the rate of production of pure alcohol will seem to be extremely slow. At a few drops per second one can believe that it will take forever to produce a reasonable amount and there will be a tendency to open the collection valve a little wider to increase the flow. Resist this temptation and be patient. The apparatus requires no attention and it is surprising how much alcohol is produced at a flow rate of 2 or 3 drops per second for several hours. Thus, at 750 watts input to the boiler and a draw-off rate of 240 ml. per hour, about 3 litres of pure, 96% alcohol will be obtained in a 12 hour day. It can even be left running overnight. This, when diluted to 40% with water will provide over 71/2 litres of vodka