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

48762Re: [Distillers] Re: ⚗ Distillation Controller

Expand Messages
  • henry sangret
    Jan 11, 2013
    • 0 Attachment
      you do know you can over pump the temp and drive extra water into the vapor and drive the vapor temperature up!  I have done it and used it ocassionally in stripping.. not a good thing BTW!  Thats why in a reflux still design they suggest to you only a certain quantity of distillate at a certain rate!

      From: tgfoitwoods <zymurgybob@...>
      To: Distillers@yahoogroups.com
      Sent: Friday, January 11, 2013 4:09 PM
      Subject: [Distillers] Re: &#9879; Distillation Controller
       
      Thanks, Eddie,

      You are, of course, completely correct, but any clarification to the point I'm trying to get across, is very, very, welcome.

      I just don't want to let this wash-temperature-control thing get a foothold on these lists.

      Thanks again.

      Zymurgy Bob, a simple potstiller Making Fine Spirits

      --- In Distillers@yahoogroups.com, Eddie Hoskin wrote:
      >
      > Very good response, Bob. 
      >
      > I would add that the only thing that you, as the distiller, have control over is how much heat you are pumping into the still.  You do NOT have control over the temperature...Once you hit boiling at least.  Because we are in a state where the liquid is at boiling temperature, no matter how much heat you pump into the still, be it 1W or 4kW, you will not succeed in changing the temperature.  What you WILL succeed in doing is changing the rate of boiling.
      >
      > Now, since you are boiling off more ethanol than water, yes, the boiling temperature will change over the course of the boil.  But for short increments of time, ie a few seconds, we can say that the boiling temperature will remain effectively constant.  With the liquid not growing any warmer, all the heat input into the still will go directly into the heat of vaporization of the product.  As a result, by varying the heat input you will not affect the temperature in the slightest.  Varying the heat input is really just controlling the rate of fluid leaving the condenser.
      >
      > Now, that would be an interesting control mechanism...put a flow meter after the condenser and control the heat to maintain a constant flow rate....;)
      >
      >
      > Hope that helps,
      > Eddie
      >
      >
      >
      > ________________________________
      > From: tgfoitwoods
      > To: Distillers@yahoogroups.com
      > Sent: Friday, January 11, 2013 2:48 PM
      > Subject: [Distillers] Re: âš— Distillation Controller
      >
      >
      >  
      > I apologize for the length of the reply. I hope you can understand. Because this raises a very common question, on which I've written extensively, I'm going to start by including a previous post, which I'll color red, and then add a new section more pertinent to your process, which I'll color blue.
      >
      > Wow! I'm not sure where all this misinformation about controlling wash temperature in a still is coming from, but it's going to confuse the bejeezus out of the prospective new distillers. We have to deal with this misinformation on a regular basis, but I've  never encountered an epidemic like this before.
      >
      > The first thing you need to know is that all liquids evaporate at any temperature where they are liquid, and the rate of evaporation of a specific liquid at a specific temperature is determined entirely by its temperature and the pressure of its environment. For the sake of simplicity (and ignoring vacuum stills) I'm going to assume that all examples hereafter are made at "standard" temperature, 1 atmosphere, 760mm of mercury (760mm Hg). So, for all of my examples, rate of evaporation is set by its vapor pressure at some given temperature.
      >
      > If you set out saucers full of water, methanol, ethanol, ethyl acetate, and acetone, all common liquids found in a beer-stripping wash, in a room at "room" temperature, say 20C (68F), they will all evaporate until the saucers are dry. Your nose will give you proof that they are evaporating, because when you smell methanol, ethanol, ethyl acetate, or acetone (your nose can't smell the water), what you are actually smelling is molecules of that liquid in the air, and it got in the air by evaporating, even at the lowly temperature of 20C (69F). It's important to note that the vapor in the air will be a mixture of ALL the liquids in teh saucers; there is NO separation of vapors.
      >
      > Vapor Pressures of liquids at 20C
      > water           17mm Hg
      > ethanol         48mm Hg
      > ethyl acetate   70mm Hg
      > acetone        189mm Hg
      > methanol       200mm Hg (These numbers inferred from graphs, so they are not terribly precise)
      >
      > Because all of these liquids are at the same temperature 20C (68F) in our room, the rate of evaporation is directly proportional to the vapor pressure of each liquid. For that reason, the methanol, having the highest vapor pressure, evaporates fastest, and the acetone evaporates almost as fast, Next comes ethyl acetate, then ethanol, and finally, water. The experienced distiller will recognize that is the order in which the highest concentration of these liquids (but NOT separated fractions) come through the still as distillate.
      >
      > If you were to heat the room, as the vapor pressure of each of the liquids reached the temperature where its vapor pressure equalled the atmospheric pressure in the room, 760mm Hg, that particular liquid will boil, and evaporate quickly, while the lower vapor pressure liquids are still evaporating, but not yet boiling. Because the liquids in the saucers are pure, at least at this point, boiling of each will occur at a specific temperature, the boiling point.
      >
      > Boiling Points of liquids at 760mm Hg
      > water         100.0C
      > ethanol        78.4C
      > ethyl acetate  77.1C
      > methanol       64.7C
      > acetone        50.5C
      >
      > As stillers, what we are really interested in is how all these liquids behave when they are mixed together and put in our still boiler, so we can know how to deal with them. First off, the mixture will have a SINGLE boiling boint, determined by the vapor pressure of each liquid, the percentage of each liquid in the mixture, and the molecular weight of each liquid, according to something called "Roualt's Law". Actually the last two numbers are combined to get something called the "mole fraction", but we don't need to understand that in this discussion.
      >
      > As you heat this (initially cool) mixture of liquids in your still boiler, evaporation of ALL liquids in the mixture increases with the temperature, although the amount of vapor is so small that you will have essentially NO distillate coming out of your condenser. ONLY when the mixture boils, at a temperature determined ONLY by the concentration of the liquids in the mixture, and according to the law of Francoise-Marie Roualt, do you get enough vapor moving to get distillate coming out of your condenser. In addition, Francoise gets the last word, and determines exactly how much of which of the original liquids is in that distillate.
      >
      > The upshot of this is that you pour your wash (let's say 10% ABV just for kicks) into your boiler, and turn on the heat. I know from my chart that your 10% wash will boil at roughly 93C, so if you used a PID controller (or any other temperature/power controller) to hold the wash at 78C, thinking to extract only ethanol, your wash will be 15 degrees short of the boiling point, and the still will just sit there, producing so little distillate it may not show at the condenser, and the funny part is this: any distillate that may occur will have the same percentage of ethanol in it as if you had boiled the wash (at 93C) and got some real product. In any event, your first real drops will be about 54% ABV. If you do get distillate holding the wash at 78C, bring some good books and a calendar, because you're going to be staring at that still for weeks.
      >
      > So you HAVE to boil, and you have no choice at what temperature it boils at, and you can't get it hotter than its boiling point, no matter how hard you try.
      >
      > Whew! I hope this clears up this issue, at least until next time.
      >
      > tyto_negro,
      >
      > Specific to your proposed setup, and (as in the previous example) assuming a 10% ABV wash, you'll heat up the still wash (I'm assuming that's where the temperature sensor is), and because the wash boils at 93C, 23 degrees hotter than your wash currently is, it does not boil and no distillate is produced. Although the controlled temperature will increase at 0.5C per minute, we don't know that the energy input is sufficient such that the wash temperature will increase that fast.
      >
      > Let's assume your heater wattage is sufficient to raise your wash temperature at 0.5C per minute. At that rate it will take 16 minutes more to raise the wash temperature 78C, still 15 degrees below the temperature at which your wash boils. No distillate is leaving the condenser yet, and will not, in any real sense, until the wash temperature is 93C, although on some strip runs you may get a couple of drops of some really vile stuff a bit earlier.
      >
      > So now we have to raise the wash temperature from 78C to the boiling point, 93C, a 15 degree difference, and your algorithm raises the controlled temperature at 0.1 degree per minute, so in another 150 minutes, your wash will boil and distillation will start, 166 minutes after the wash reached 70C.
      >
      > But we are not done yet. It takes a lot more energy, based on latent heat of vaporization, to evaporate enough of the wash to increase the boiling point, so unless you have a very large heating element and a very small boiler capacity, your controlled temperature will increase to 93.1C while the wash is still boiling at 93.0C, and from that point on, your controlled temperature will always be higher than the wash temperature, so the heat will stay turned on exactly as if a simple switch was used for control, except for that extra 166 minutes, of course.
      >
      > I'm not sure where the notion comes from that controlling the temperature of the wash will control the output of a still, but it's continually reinforced by such "experts" as the narrator on the "Moonshiners" TV show, and other entertainers. It's proven almost impossible to stamp out, at least among folks who have yet to distill.
      >
      > Certain current practices have perhaps made it seem as if this old wive's tale is true. Those hobby distillers who use compound reflux still will speak of controlling head temperatures, but in these stills "head" temperatures are the vapor temperatures at the top of reflux columns after the vapor has been through multiple condensation and re-evaporations, which the distille can control. As for the wash temperature, even in a reflux still, all that I'm telling you is true.
      >
      > Zymurgy Bob, a simple potstiller Making Fine Spirits
      >
      > --- In Distillers@yahoogroups.com, "tyto_negro" wrote:
      > >
      > > âšâ€" Distillation Controller
      > >
      > > Dear collegue-distillers, below a description of the controller I'm making.
      > > On my blog a bit more info and a few extra pictures. Comments, suggestion, ideas etc are of course more then welcomed!
      > > http://steppewolf.cato-projects.org/?x=entry:entry130108-201913
      > >
      > > Since a year I'm the proud owner of a small Portuguese classical still - an Alambique -. Many Portuguese alambiques are wood fired. The classical Portuguese way of distilling doesn't use any instruments to monitor the process. The distiller just keeps an eye on the output of the condenser and makes sure the distillate comes out with a dripping trickle. To adjust this he(or she) closes or opens the door to the firebox or ads or removes wood from the fire. Traditional Alambiques range from 60 to 200 litres. I have a small pot still (10 ltrs) which I operate on an electric hotplate, so that is easier to regulate then a wood fire.
      > >
      > > THE PLAN
      > >
      > > As a home-distiller with an electronics background I got the idea to automate part of the distillation process. Especially the stripping would be a good thing to automate.
      > > After many hours sitting next to the still and even more hours sipping from a nice glass of Slivovitz or Eau-de-Vie-de-Peche the design took shape. The process of distilling is more then just keeping the temperature at a fixed level. During the distillation process the composition of the wine in the still changes, more alcohol evaporates than water, so the boiling point shifts up. To keep the distillate stream more or less constant the temperature needs to be adjusted, raised, until the operator decides to stop.
      > >
      > > I had in mind to make a programmable controller that would be able to raise the temperature with an adjustable rate.
      > >
      > > So, for example:
      > >
      > > Heat up the still till 70° as fast as possible.
      > > Raise the temperature with 0,5°C per minute until we reach 78°C.
      > > Wait 5 minutes, keeping temperature constant (time to take away the heads ;-)
      > > raise the temperature with 0.1°C per minute till 84°C
      > > etc.
      > >
      > >
      > >
      > > HARDWARE
      > >
      > > I decided to use a PIC16F690 micro controller with a small LCD display of 2 x 8 characters; a rotary encoder, a pushbutton and a few LED's complete the user interface.
      > > The hotplate is switched on and off with a varying duty-cycle that can change from 0 to 100% to adjust power. Switching of the hotplate is done by a triac during the zero-crossing of the sine-wave of the grid. The temperature sensor is an LM35 IC which is mounted in the top of the swan neck ("Point of no return") where my still has a fitting for mounting a normal thermometer.
      > >
      > >
      > > SOFTWARE
      > >
      > > The main program loop runs 10 times per second. Every tenth of a second a temperature sample is taken and the triac controlled. Therefore the hotplate cycle is 10 seconds, being 100% x 0.1 sec.
      > > The program calculates a running average of 32 temperature samples to further stabilize readings. The change of temperature (deltaT) is calculated every 15 seconds. The actual control is done only once per minute at the moment. An electric hotplate is veeery sloooow to react as you probably know when you ever tried to boil milk on an electric plate….
      > > This caused me major problems deciding on the control algorithm. Reading on-line about PID controllers, made me unsure if an implementation of this would be very successful. Standard PID controllers have problems with big time lags or dead time in a process. Implementation with a "Smith predictor" was something I found mathematically too complicated. All software is hand coded in assembly language.
      > >
      > >
      > >
      > > ALGORITHM
      > >
      > > I decided to make my own control algorithm and I implemented practical and theoretical background knowledge of the distillation process. The program calculates the theoretical power needed to achieve a set deltaT. Therefore the program needs to know how much liquid there is in the boiler, how much energy it takes to raise the temperature of 1 litre of water 1°C in 1 minute and the power of the hotplate. The program also makes an estimate of the losses when the boiler is at a steady state, or constant temperature. With this background data, the program only needs to regulate a relatively small bit when the temperature comes close to the set-point. Coarse "regulation" is done by calculation. To minimalize overshoot the deltaT is regulated down when temperature gets nearer the set-point.
      > >
      > >
      > >
      > > THE RESULTS
      > >
      > > After reading all this background information, you will probably be as curious as I was to see the results. I did numerous tests with pots of water on the hotplate and I was a bit in doubt if I would ever get it right. After all my different trials the temperature would still swing about 1°C around the set-point with my pot-with-water tests. Of course that was with the sensor submerged in the water and not in the vapor of an water-alcohol mixture. So I decided to go for the real thing and do a distillation of 5 litres of old wine.
      > >
      > > The results definitely were quite good. In contradiction of what I expected temperature swings were not so big as I feared. And, as a pleasant surprise, I found out that the temperature in the swan neck reacted faster than I expected on changes in the power setting. This made the control loop not only faster but also more stable!
      > > Besides the quicker response I also noticed that the process/temperature is very sensitive to draught. I'm sitting close to an outside door, and when it opens it can seriously change the temperature with 0.5 degree or more.
      > >
      > > Of course I encountered a number of bugs in my program code ;-) but these were easily resolved and after some more tweaking I'm ready for a next trial.
      > >
      >
    • Show all 16 messages in this topic