StillDragon® Community Forum

Welcome!

Be part of our community & join our international next generation forum now!

In this Discussion

Any interest, or suggestions on, a bolt-on still automation kit?

edited August 2013 in Accessories

We're working to develop a "bolt-on" kit for still automation.

The objectives of the automation: To compile distilling protocols for different fruit mashes to automate tail-cuts and control the reflux, %abv, boiling rate, coolant control, tail-cutting etc. It will allow the distiller to focus on other issues instead of spending time on the distilling process to produce really good stuff, maintaining quality and consistency.

Off-course, once a particular mash's "ideal" distilling protocol is determined, it would be great to have a central depository from where to pull and share these protocols in a sort of "open-source" type of way - making for better product by more people...and pushing the limits beyond those achievable by big distilleries.

Any suggestions, or requirements for such a "bolt-on" still automation "package"?...or do you think it will not work?

Spirited people...loves distilling.

Best Answer

  • No problem at all. I do have my reasons.

    C adds more height. The thermal losses through stainless are much less than they would be through copper, and the bonus of the extra height is the ability to get the collection end of the product condenser to a convenient position. The extra height makes no difference to the flavour profile.

    D does indeed share the thermal load, so that you can use a smaller product condenser. It also shifts part of the weight closer to the centre point on the boiler. You don't need it, but I feel that the benefit of using a smaller (and lighter) product condenser makes it worth while.

    Two 45 degree bends gives you the option of rotating the product condenser away from vertical. Again this is mainly a height / convenience issue, and is aimed at getting the product end of the condenser into the position that you want. It is why I did not show an attached parrot.

    I assumed that B would get questioned. This might be un-neccesary, you could always pre-charge the inline thumper.
    However, coolant control at this position will give you direct control of the thumper. No coolant = no thumper - it very rapidly heats up and dries out.

    Variable coolant = the ability to change the behaviour of the still during the run. You can fine tune the product ABV over a (relatively small) range of values, or simply use it to increase heads compression and then switch it off.

    If you use power management of the boiler (a good idea in my opinion) then you might find that you don't actually need the dephlegmator B at all.

    Pot stills tend to be considered crude instruments, but add power management of the boiler and that changes things. Incorporate a thumper and you add another layer of versatility.

    They are not a sophisticated as a 3 to 5 plated column, but they are capable of a fair amount of control. At the end of the day it depends on what you want to use the pot still for.

«13456

Answers

  • Hi @Gert_B, while not targeted at distilling the brewtroller project is an open source automation project for making beer. That would seem just one step away from distillation automation.

    And BTW, welcome!

  • I am working on one at the moment, but I am held up by OZ post with deliveries from China, it took 4 days to get from China to Oz and its been 6 days since and still no deliveries.

    This is progress so far, the hole in the front of the box is where a pid controller will go, when it eventually gets here.imageimageimage I should be able to then control reflux rates at precise temperatures with the PID

    OD

    001.JPG
    640 x 480 - 87K
    002.JPG
    640 x 480 - 78K
    003.JPG
    640 x 480 - 87K
  • edited August 2013

    Hi Lloyd, Thanks. Yes, there is also The Raspberry Pi Brewing Controller. Our challenge is aimed at the last step - the still automation itself.

    Distillate flow control into different vessels has been sorted out as well as temperature corrections for altitude/ambient air pressure differences.

    The current challenge is to get an accurate/repeatable/low cost electronic alcohol meter capable of measuring up to azeotrope for interfacing to a controller. We have played around with a few experimental sensors but have not find a suitable solution yet. However, somewhere out there is the answer...just have to find it!

    There is quite noticeable differences in the distilling profile/protocol for different fruit mashes and they seem fairly consistent - ideal for automation to get consistent results - the grain based fermentations still need attention. Will get to them later on.

    Warm distilling regards, Gert

    Spirited people...loves distilling.

  • Olddog, EXCELLENT! What a neat job you did. Question: Do you base the reflux control on the column temperature, boiler temperature, distillate %abv or all together?

    Spirited people...loves distilling.

  • The probe will sit directly under the dephlegmater in the position normally occupied by a temperature gauge, I will then be able to dial in different temps for foreshots, heads, and ethanol.

    OD

  • Olddog: How is the cooling controlled by the PID? With a solenoid valve? What kind of cycle lenth would you be using?

  • Hi Ben, while waiting for Olddog's response, we control the coolant water 2 ways: 1. Through servo valve on the the dephlegmator coolant to have the temperature of the distillate internally in the dephlegmator just below condensation point (to increase reflux re-evaporation efficiency) and, 2. Through servo valve on the product cooler to have the distillate at a selected temperature for quick tasting (i.e. 30 degree C) or to allow evaporation of last volatiles (i.e. 50-70 degree C)

    Spirited people...loves distilling.

  • I have ordered two different types of valves, solenoid valves and motorized ball valves, also I will have two thermocouples arrive. I want to experiment with one thermocouple directly in the coolant area in the dephlegmater, and the other in the vapor area just below the dephlegmater. It will be interesting to see which is more manageable.

    OD

  • Olddog: What about also using the two thermocouples to measure the vapor just below the dephlegmator and just above the boiler. I would think that with 0% reflux, these temperatures to be equal - and with high reflux (i.e. 80%), the top temperature to be lower than the lower temperature after some stabilization (as the ethanol content increases at the top and lowers the vapor temperature). Would be interesting the correlate the top-bottom temperature differences over a distilling run (from high volatile heads to the lower boiling point tails)....So much to learn and so little time!....

    Spirited people...loves distilling.

  • Actually tails have a higher boiling temp due to the high water content. MY Flute still stops producing ethanol automatically with the dephlegmater set @ 78c, I have to cut the water flow to the dephlegmater to be able to take tails and my temp gauge will increase to around 100c

  • Yes, I agree with the tails having a higher temperature due to the high water content. - but what if the water content in the tails REDUCE by increasing the reflux during tails?

    If we set the reflux to 100% during tails, very slowly the temperature (under the dephlagmator) DECREASES as it's water content becomes less and less (with the 100% reflux enriching the lower boiling point components). - what "compression"" is all about and is great to separate the tails into the different aroma fractions within it (on relatively short fractionating columns like a flute).

    As this is a rather sensitive/difficult process I would be greatly interested to measure the top-of-the-column and bottom-of-the-column temperatures to determine when the "compression" has peaked (i.e. max stable temp difference between top and bottom temperatures).

    I also suspect that this temperature differences be noticeable during compression of heads, harts or tails and the many sub-fractions in each of them. ... perhaps this should be a "week-end" project for me - to confirm/deny it and get to the bottom of "compression".

    Spirited people...loves distilling.

  • edited August 2013

    If and when the rest of my components arrive, the next part of the puzzle will be synching the valves. The coolant supply valve will do just that, supply coolant to the dephlegmater, the valve on the heating circuit will open and allow the coolant to bypass the dephlegmater as the column heats to the require temp, and then the cooling valve will open when the temp gets higher than require. The trick will be to set the PID so that both valve synch together. I will have to play around with the "P" Ï"and "D"settings to achieve this, I will use my low flow pump to do this so that there is not too much pressure on the system. The other way I could do this is to just use one valve for coolant to the dephlegmater and put a pressure relief valve to bypass the coolant when the valve is closed.

    OD

  • A pressure relieve valve would be easier. And if you were running cooling water from the tap you would not even need that as long as your hoses can hold the PSI of the Tap. 1 valve real simple.

  • Just received the PID SSR and thermocouple from China, it took 4 days from China to Melbourne then another 10 days from Melbourne to me, so much for Oz Post. Anyway it's all fitted now, I just have to wait for the valves to arrive to complete. image

    001.JPG
    640 x 480 - 85K
  • On the site there are nice components for making 2" or 3" pot stills. Personally if using SD components I would use a 6" keg connection, a single 4" bubble plate (inline thumper), pre-condenser and condenser. Something like this:

    image

    So given that - and getting back on topic - what are your views on automated boiler control for pot stills? Power management for the different phases of the run.

    3 inch pot.jpg
    533 x 800 - 35K
  • edited August 2013

    Hi @Myles, a few questions to educate myself.

    With reference to your configuration:

    image

    1. What would be the purpose of adding components C and D? (in my understanding C would just add height and some heat losses and D would just duplicate what E is doing?)
    2. Why have two 45 degree bends instead of one 180-degree U-bend

    Thanks for assisting me in understanding your configuration.

    marked.jpg
    299 x 775 - 18K

    Spirited people...loves distilling.

  • Thanks Myles!

    Spirited people...loves distilling.

  • @olddog said: If and when the rest of my components arrive, the next part of the puzzle will be synching the valves. The coolant supply valve will do just that, supply coolant to the dephlegmater, the valve on the heating circuit will open and allow the coolant to bypass the dephlegmater as the column heats to the require temp, and then the cooling valve will open when the temp gets higher than require. The trick will be to set the PID so that both valve synch together. I will have to play around with the "P" Ï"and "D"settings to achieve this, I will use my low flow pump to do this so that there is not too much pressure on the system. The other way I could do this is to just use one valve for coolant to the dephlegmater and put a pressure relief valve to bypass the coolant when the valve is closed.

    Option 1 - use a motorized 3-way valve?

    Option 2 - use a well type pump with pressure tank and pressure switch and then a pressure regulator to keep pressure even? (this could allow solenoid valves to actuate 3 thermostatic expansion valves each tuned for one of your product stages)

    FYI - I use a BCS-462 controller for automation on my brewery, and will implement an individual BCS-460 on each still in our distillery, the web interface, data graphing, and GUI customization work really well.

    Here are some pics of the system set up for beer:

    image image image image

    BCS GUI.jpg
    800 x 578 - 76K
    brew day 2013 06 18.jpg
    800 x 563 - 101K
    3.jpg
    800 x 598 - 119K
    4.jpg
    600 x 803 - 91K
  • @Gert_B

    I've often thought about a digital sensor so we can interface to controllers easily.

    Only one i came up with was a linear encoder as part of a alchometer floating in the output product.

    There has to be a better way!

  • @Law_of_Ohms Yes, having a linear encoder on a float is a possibility as are converting a barometric sensor, using alcohol breathalyser sensors, refractometer readings etc. all of them having some issue making it unsuitable for reliable, repeatable, easy, maintenance free operation. BUT...as you have pointed out: There has to be a better way! ...just have to find it, that's all :-) ...and the struggle continues...

    Spirited people...loves distilling.

  • I am intensely interested in this thread because computer control of the still is, in my opinion, the future of distilling and any still vendor that does not offer the solution to his customers will be left in the dust.

    @olddog and @CothermanDistilling are fast tracking this so I'd better pay close attention. Its mostly over my head but I'll try to follow along.

  • Not electronic, but more automation ideas... I have noticed Danfoss thermostatic expansion valves on every commercial still I have seen, so got one cheap on eBay and playing with it... when I get the right fittings tomorrow from grainger, will weld it up and connect to the shotgun condenser, and have the most efficient flow of cooling water possible. It should allow the exit water to always be in excess of 140F or more, which translates into a very low flow. The sensor bulb will go inside the tube, and this connects to the top port of the shotgun, then drains out the bottom. The valve body connects to the bottom, inlet port of the condenser, and will have a small trickle bypass to make sure the temp measured is due to flowing water... this may even allow dephleg water to be re-used, or maybe design a similar valve for the dephleg, but put the bulb in the vapor path....

    image

    IMG_1808.JPG
    600 x 803 - 111K
  • edited August 2013

    Here can you read more about danfoss ventiler and more.. side 106.

    Thermostatic Expansion Valves

    Cheers

  • edited August 2013

    Here is a diagram of the automation system in progress that we're working on:

    image

    Original posting available at this link.

    A short description of the various sensors and their purposes:

    FC1 – control the flow through the main reflux condenser (dephlegmater) and is the means to control reflux into the column

    FC2 – control the flow of coolant water through the product condenser. A high flow through the condenser will result in low temperature final spirits – ready to taste. A low flow through the condenser will result in high temperature spirits with more volatile fractions evaporating (to be used during low reflux distillation only) .

    TP1 – Indicates the temperature of vapor before final condensation starts and is an indication of the type of fraction (after releasing compressed vapors) or an indication of the fraction composition during reflux distillation

    TP2 – Indicates temperature just before entering reflux condenser and is an indication of the type of fraction (at the end of compression/full reflux) or the fraction composition during reflux distillation.During zero reflux operations this temperature will be equal to TP1 temperatures

    TP3 – indicates vapor composition entering the column. During compression the difference between this temperature and TP2 indicates the completeness of compression. Full compression is achieved when the difference stabilizes at its maximum.

    During normal reflux operations the difference between TP3 and TP2 is an indication of the amount of reflux. Under zero reflux conditions TP3 will be equal to TP2 and TP1 .

    With more plates in the column the temperature differences between TP2 and TP3 will be greater (during compression or partial reflux).

    If a temperature difference exists between TP3 and TP1 during zero coolant flow through the main reflux condenser, it indicates that zero reflux is not achieved and the column should be temperature insulated if zero reflux is required.

    TP4 – Measures the actual mash temperature and will be most useful during start-up to project “start boiling” time. The difference between TP4 and TP5 is an indication of the heat transfer potential from the jacket water to the mash. (The higher the transfer potential, the more heat can be provided, the quicker the "bring-to-boil" time - and vice versa.)

    TP5 – Measures the jacket water temperature and, together with TP4 and SP, control the heat input

    TP6 – Measures the distillate outflow temperature for control of FC2

    VH – Controls the heat input finely as determined from TP5, TP4 ad SP

    R1– Controls the heat input coarsely as determined from TP5, TP4 ad SP

    R2– Controls the heat input coarsely as determined from TP5, TP4 ad SP

    %A – measures the end product % abv and is main input to determine FC1 control according to distilling protocol settings

    AP – measures ambient pressure and temperature to adjust all protocol temperature settings (standardised at standard ambient pressure of 101.3kPa at 20 degree C)

    SP – determine when steam reaches the top of the cooling jacket and heat input is more than what can be transferred to the mash. Used to adjust heat settings first coarsely and then finely.

    FD – diverts the distillate flow into different receivers to allow automated collection of desired fractions according to the style and specific distilling protocol. Where heads and tails are normally discarded, sub-fractions of it may form part of a specific style. (i.e. grappa with higher than normal methanol fraction)

    ANY COMMENTS will be greatly appreciated!

    Still_automation_diagram.jpg
    662 x 435 - 35K

    Spirited people...loves distilling.

  • Interesting Concept. You also might look at a scaled back version. A lot of people are not using a Jacketed Boiler and thus TP5 would not be needed. Heat input would need to be controlled by TP4 temp.

    FC1 could be controlled based on Temp of TP1 and TP2. It would be Nice to have %A but based on my research I have not found an accurate fluid density meter sensor that would not be cost prohibitive.

    Also not sure I like the concept of the Flow Diverter for automated cuts. Cuts are not that exact based on a set temp and needs the human touch, smell, taste to make these decisions.

  • Great thread. I'll look for fogging on the site glass till I can get myself up to speed,,,,,,,,smiley face here.

    StillDragon North America - Your StillDragon® Distributor for North America

  • edited August 2013

    Digitally/electronically measuring %abv: The following method almost worked to get a cheap digital/electronic alcoholmeter:

    image

    As 100ml beaker was filled with distillate coming from the still and had an overflow. This beaker was placed on a cheap pocket digital scale with range of 100g with 0.1g resolution.

    The distillate %abv was measured with a range finder (0-100% abv) with temp. correction and also weighed.

    At 92% abv it measured 72.6g and at 15% abv it measured 87.2g This gave us a good range (14.6g and with the scale resolution should give us %abv values in approximately 0.5% steps - which seems reasonable. This was in theory....

    In actual operation we struggled to stabilize the reading and could only only get to accuracies with about 3% abv. To better this we required a) a bigger beaker (which the small scale couldn't handle) or b) a bigger range scale with same resolution. However, we could not "hack" into the bigger scale to get the readings electronically. ...the struggle continues....somewhere there is a solution...

    Cheap-electronic-alcoholmeter.png
    374 x 304 - 10K

    Spirited people...loves distilling.

  • edited August 2013

    @RedDoorDistillery: Thanks. As with the Still Dragon Modular concept, the automation should definitely also be modular. On the hardware part it is not difficult to achieve but the software is a bit of a nightmare.

    As to the flow diverter. Yes, temperature alone is not good enough. Readings of the temperature, %abv, data history and a few phase diagrams (in the software) seems to give workable (but not yet good enough) results. ...We're also tried to use a "lookup list" for various components that "should" be in the tails and match that to the available data. To date no success with this. :-(

    Spirited people...loves distilling.

Sign In or Register to comment.