Loading ...
Sorry, an error occurred while loading the content.

Your parallel design

Expand Messages
  • Mike Nixon
    Hi Alex, I apologise for not getting back to you (and others) sooner on this. I haven t had much time to myself recently. Your design I m talking about is
    Message 1 of 1 , Sep 22, 2002
      Hi Alex,
      I apologise for not getting back to you (and others) sooner on this.  I haven't had much time to myself recently.
      Your design I'm talking about is this: The problem with it lies in the management of the liquid reflux.  You would have to ensure that all the take-off points were each aligned exactly horizontally, and for all columns.  Quite a difficult task as they are widely separated.  Failure to do this would result in unequal reflux flows, and the problem is further complicated by surface tension and any ripples in the surface of the reservoir as you have the reflux flowing over the lips of the take-off tubes.  Adding extra overflow paths for each column would make matters worse if the columns were not aligned precisely.
      This is the crux of the whole parallel operating system.  Levelling, and thereby ensuring equal distribution of reflux.  Others have also come up with ingenious designs, some of them very elegant indeed, but I have seen none that offer an easy means of ensuring equal distribution.
      The first thing to bear in mind is that you are very unlikely to be able to make any system that delivers precisely accurate amounts of reflux to each column.  So, with this in mind, the next best thing is to try and reduce the errors that might lead to unequal flows.  One way of achieving this (as you have done in your design) is by having just one reservoir for the reflux underneath the main top condenser.  This gives you a common starting point.  This would have one tube leading down to a needle valve to control product take-off, which could be situated in the middle.  The next thing is to provide as many submerged take-off points in the base of the reservoir as there are columns, and have each with exactly the same sized drilled aperture (or accurately sized needle jet such as used in carburettors).  Of greatest impotance, they should also be as close to each other as you can get them to minimise the effects of any tilt.  In this manner, for any given level of reflux in the reservoir, each take-off point would have the same head of liquid above it and, being of equal size, each should then deliver the same amount of reflux in a given time.  Of course, in a 'real' system. you will never quite achieve that ideal, but with this arrangement you should be able to come pretty close, and it's a darn sight easier to adjust one small assembly than try and level the whole rig.
      The next thing to ensure is that the delivery tubes to each column do not contribute to the fluid head that you have tried so hard to make equal.  This can most easily be done by having an air space between each take-off point and the tube it is feeding ... something like a small cup or funnel on the collection end of each delivery tube.  By this means, it doesn't matter if the delivery tubes vary in length, diameter, or connection point to their respective columns.  All that is then needed are small U-bend vapor locks to ensure that vapor doesn't leave the columns at these connection points.
      This is what I'm getting at.  The region highlighted in pink is the central distributor, and is shown connected to the top condenser shroud as one means of getting fine levelling adjustment by means of shimming gaskets.  You just have to ensure that all the feed jets are level.
      By varying the rate at which you draw off product by the needle valve, you simultaneously vary the reflux to each column by almost exactly the same amount, and with just one control.
      Now, all this is a lot of trouble to go to when, in most of our applications, increasing the diameter of a single main column would be all that is needed to increase production rate.  However, if considering medium-scale production, for ethanol fuel say, then multiple columns have their attraction.  Also, as I said earlier, someone out there may have a better idea of how to achieve equal reflux flows ... this is just the one I offer as a 'solution' to the 'challenge' I originally posed about designing a multiple column still. It is an interesting problem!   If anyone has a simpler, more elegant solution ... let's hear about it!  Always eager to learn!!!
      All the best,
      Mike Nixon
    Your message has been successfully submitted and would be delivered to recipients shortly.