Dephlegmator Control via Vapor Temperature?

[Johnnywhiskey]

I'm noodling through a concept that I recently saw at a micro-distillery that I visited. Their hybrid column was set up so the coolant to the dephlegmator was controlled by the temperature of the vapor at the top of the column. They had a controller box that they could set the threshold temperature to turn on the coolant. I asked whether they were controlling the temperature of the coolant, and the distiller said no, they were monitoring the temperature of the vapor at the top of the still, not the coolant.

This was new to me, I had heard of controlling the coolant by the temperature of the coolant exiting the dephlegmator, reflux ratio or rate of production. But vapor temperature kind of makes sense--you set the controller for the vapor temperature of the abv you are targeting and the controller will cycle the coolant valve to maintain that temperature / abv. The system will crank away at that abv until it cannot hit that abv based on your theoretical plates and remaining abv in the boiler.

Seems like this could be useful in a couple of ways. My focus is making bourbon, but with a CM column I'm always looking for ways to keep the abv on the relatively low end (around 120-130 proof), usually by dialing back the reflux using a needle valve. A high vapor temperature setting would be good for hitting a lower abv. For NGS a temp close to the azeotrope would give a high proof product. In both cases the still would run at the abv associated with the temperature until the still could not produce that abv any more. Not sure whether this would work as a good tails cut, but it would definitely give you a sign that the run was reaching an end.

Curious if anyone has seen or used a still that controls the coolant this way. Any downside to running a still this way? I did a few searches here and did not come up with anything.

Best, JW

[Monkeyman88]

I've been working on vapour temp control on an LM. Using the temp to dictate how much the needle valve opens. I have all the code written, but still working on a way to connect the stem of the valve to the stepper motor. Hasn't said much about it on here as there is quite a negative view of automation. But I was planing on doing a full wrote up once I had finished.

[googe]

Olddog and many others use solenoid valves for this, with great success. Maybe search, solenoid valves ect.

[Johnnywhiskey]

Olddog and many others use solenoid valves for this, with great success. Maybe search, solenoid valves ect.

I see Olddog floated the idea of controlling a dephleg using the vapor temperature on page 18 of the "Flute Talk" thread, but I don't see that it was actually implemented. Seems like the crowd convinced him to go with a needle valve and flow meter. Was the temperature control ever implemented and run?

I have just been thinking a little bit out of the box, sort of a brainfart.
We know that with a flute the hearts come off at around 79 degrees and we adjust the flow through the dephlagmater to maintain this.
Could a senser be connected to a motorised valve fitted to the dephlag output so the temperature will be controlled automatically. It would work a bit like the electric fan works for your car radiator, when the temp gets too high the fan cuts in to cool the radiator. Are there any tech heads out there with any ideas on this. You could then set temp stops to take forshots, heads, and then hearts.


OD

Seems like it would be fairly easy to implement.

One of these (could use this on fermentations also):

http://www.amazon.com/Inkbird-Digital-T ... controller


And a valve like this (or something similar):

http://www.amazon.com/Fafada%C2%AE-Elec ... ater+valve

Aside from implementation, the real question is whether controlling the vapor temperature would be a useful way to control distillation. I would be interested to hear any thoughts on that.

Best, JW

[sungazer]

Certainly an adjustable needle valve type device is going to be much better than a solenoid operated valve. I know that from experience just using my still all open lets out an almighty amount of liquid the temperature of the vapour will start to rise so I shut the valve and let it come back down again. This type of start stop is no way as accurate as adjusting the valve to a point that steady state can be achieved. I actually need to replace my valve I only have a ball valve as I bought all my parts from a local plumbing store and thats what they have they are not that interested in partial accurate adjustments. The current one works ok however it could be improved. Not sure how your still is set up a single take off valve and everything else gets returned or you have two valves. However you can run with just the one take off valve with everything else being returned. Then the program can also be set to close the take off at a particular Vapor temperature to make the final cut off at the ABV you wish.

Only thing you might find is that this automation will take a lot longer as it will not take into account flow rate therefore it might slow the take off to a drip a second or even slower before it reaches the shut down temp.

[cuginosgrizzo]

I've been working on vapour temp control on an LM. Using the temp to dictate how much the needle valve opens. I have all the code written, but still working on a way to connect the stem of the valve to the stepper motor. Hasn't said much about it on here as there is quite a negative view of automation. But I was planing on doing a full wrote up once I had finished.

interesting! Don't wanto to Hijack the thread, but what are you using to control the stepper motor based on the temperature output? I am a big fan of Arduino: I built my complete Acquarium controller with it (movement pumps, temperature controlled cooler, lights -neon, hqi and lunar- water level control).

[googe]

this is one I was referring to Johnny http://www.coppercustomstillcomponents. ... 4&start=30" onclick="window.open(this.href);return false;" rel="nofollow I know there's more out there, will.look later. From memory I think most use pid control with the valves.

[Johnnywhiskey]

Yep, that is pretty close, although is sounds like Olddog is measuring the vapor temperature below the the dephelg (and Swede is measuring the temperature of the coolant exiting the dephleg). It seems like measuring the temperature of the vapor above the dephleg would be the most accurate way to get a sense of the vapor composition reaching the product condenser. Both because the condenser must be cooling the vapor some, and dripping off of the dephleg might lead to some inaccuracy in the vapor temperature.

Good idea with using the PID controller, there are a number of decently priced PID controllers for not much more than the temperature controller I linked. Sounds like the PID auto program is actually useful to maintain the target temperature. Also sounds like Olddog was able to reach a decent equilibrium by just using an on/off solenoid for the water flow. I'm still trying to figure out what the best type of water valve that I would use (suggestions? Preferably low voltage DC.). I'm thinking some combination of a 12v valve and a needle valve will give me the right balance.

Maybe Olddog and Swede will chime in with some theory and how their systems work.

Best, JW

[Monkeyman88]

I've been working on vapour temp control on an LM. Using the temp to dictate how much the needle valve opens. I have all the code written, but still working on a way to connect the stem of the valve to the stepper motor. Hasn't said much about it on here as there is quite a negative view of automation. But I was planing on doing a full wrote up once I had finished.

interesting! Don't wanto to Hijack the thread, but what are you using to control the stepper motor based on the temperature output? I am a big fan of Arduino: I built my complete Acquarium controller with it (movement pumps, temperature controlled cooler, lights -neon, hqi and lunar- water level control).

Yup. Arduino. I have a mega 2560 and an UNO

In a nut shell, I set the temp I want the take off to be ( vapour temp based on ABV I want) it has a 500ms sample rate, if the temp is above the temp I want, it takes one step closed. If the temp is lower than what I want, it takes one step open. My stepper motor has 4096 steps per revolution, so very fine adjustment.

[Jimbo]

As a matter of course now during my flute runs I increase the defleg coolant flow towards the end of the run as soon as that vapor temp starts to rise (vapor above defleg, just before product condensor). It drops the vapor temp back to hearts run temps and extends my yield quite a bit. Also bump the power in to the heat element towards the end of the run some as well. Automating both (heat in based on boiler temp, and defleg water based on vapor temp) would be slick as snot, its on my to do list.

[drmiller100]

How else would you control reflux other than temperature?

Pro tip: Put a temp gauge at the condenser for vapor.

put another temp gauge 10 percent down the packing from the top.

tune your still for max product rate and your desired condenser vapor temp.

Note the temperature of the 10 percent down gauge. It SHOULD be slightly higher than the condenser temp. If it is the same, you can pull product slightly faster.

Anyway, manage your computer controller based on the 10 percent gauge, with the goal being slightly higher temp than what you want your product at.

The last 10 percent of the packing will create a buffer to muffle the hysteresis of your computer program running the solenoid.

[Monkeyman88]

How else would you control reflux other than temperature?

By ABV. :lightbulb:

[Johnnywhiskey]

How else would you control reflux other than temperature?

By ABV. :lightbulb:

Yep, granted vapor temperature is a proxy for ABV. But in the end, ABV is what we are trying to get a handle on.

What would be cool is to be able to lock down ABV using a PID controller running on vapor temperature and cooling, and then use heating to adjust the flow rate--at least that would be easiest with a propane burner.

Not to mention that with tumbler of whiskey and a cigar, I am already short a couple of hands on my distilling nights.

Best, JW

[drmiller100]

Yep, granted vapor temperature is a proxy for ABV. But in the end, ABV is what we are trying to get a handle on.

What would be cool is to be able to lock down ABV using a PID controller running on vapor temperature and cooling, and then use heating to adjust the flow rate--at least that would be easiest with a propane burner.



Best, JW

how are you measuring ABV? the easiest, most accurate way is boiling temperature.

Why are you not just running your still at the max heat your column will handle always?

[Jimbo]

Yep, granted vapor temperature is a proxy for ABV.

It is.

What would be cool is to be able to lock down ABV using a PID controller running on vapor temperature and cooling, and then use heating to adjust the flow rate--at least that would be easiest with a propane burner.

Absolutely, I do it today, manually. Both, ABV based on defleg flow, and product flow based on power input. As I said below. Its not hard to do at all, try it.

Regarding easiest with propane.... gotta disagree there. Electric element is bone simple to control, I can give my 5500W element 2372.4 watts if need be.

[Monkeyman88]

Yep, granted vapor temperature is a proxy for ABV. But in the end, ABV is what we are trying to get a handle on.

What would be cool is to be able to lock down ABV using a PID controller running on vapor temperature and cooling, and then use heating to adjust the flow rate--at least that would be easiest with a propane burner.



Best, JW

how are you measuring ABV? the easiest, most accurate way is boiling temperature.

Why are you not just running your still at the max heat your column will handle always?

I can't speak for others, but I use an alchometer to measure ABV.

[Johnnywhiskey]

Yep, granted vapor temperature is a proxy for ABV.

It is.

What would be cool is to be able to lock down ABV using a PID controller running on vapor temperature and cooling, and then use heating to adjust the flow rate--at least that would be easiest with a propane burner.

Absolutely, I do it today, manually. Both, ABV based on defleg flow, and product flow based on power input. As I said below. Its not hard to do at all, try it.

Regarding easiest with propane.... gotta disagree there. Electric element is bone simple to control, I can give my 5500W element 2372.4 watts if need be.

Yea, that was not very clear. What I meant was that it would greatly simply things if all I had to do is turn my burner up or down to adjust the product flow rate. I can't speak for electric heating, I haven't gone there yet.

I can't speak for others, but I use an alchometer to measure ABV.

Yep, that is as easy as it gets. Just remember to correct for temperature because when things get flowing the product can heat up, and your hydrometer will substantially over report the ABV--take a look at TTB gauging table 1. Vapor temperature is an easy way to automate because it is a direct link to the ABV, but running manually I do not focus on it as much as my manual ABV readings.

Why are you not just running your still at the max heat your column will handle always?

Max flow rate is easier said that done when you are juggling heat input, reflux coolant and a product condenser coolant. Unfortunately my product condenser is generally the first to give up. And try to manually adjust all three of those inputs at the same time and the system gets very unstable. Plus then I have to put down my drink, my cigar, or both!

Best, JW

[Jimbo]

Unfortunately my product condenser is generally the first to give up.

Best, JW

Try putting some copper mesh in the product condenser vapor path. Makes a heap of difference. I run with just a trickle to the Liebig, even fast and hot. Product is always cold.

[woodshed]

I'm kinda late to this so I will just stay out of Jimbo's and JW's way. I got nuthin more.

[googe]

What sort of cm have you got?. Sounds like you'd need to fix the PC issues first?.

[Johnnywhiskey]

What sort of cm have you got?. Sounds like you'd need to fix the PC issues first?.

Nah, runs just fine. Just the limiting factor when an 18" liebig condenser when paired with a 200k btu burn, the condenser is going to give out first. It's a Mile-Hi 3" CM column, and I have added a 10" dephlegmator above the packing. When I do whiskey I remove the section of pile used to hold the packing and just run one roll at the bottom.

Anyway, we are getting a little off of the automation topic that initiated the post. I did register on Coppercustomstillsomponents.com.au to read Olddog's post and see the pics. What he was doing there is very close to what I am talking about, only varying by the location of the thermalcouple for the PID. Olddog said the system worked great, and that was how the microdistillery I visited was running their still also, so it cannot be all bad. Seems like it is a way to automate what many of us are doing manually--i.e., adjusting reflux to target a particular product profile.

Also I would love to know what water solenoid valves Olddog used to control his coolant, or if anyone had advice on which to use.

Best, JW

[googe]

from.memory, od uses ones from.ebay, http://m.ebay.com/itm/272066653341?_mwBanner=1" onclick="window.open(this.href);return false;" rel="nofollow http://www.coppercustomstillcomponents. ... 606#p21606" onclick="window.open(this.href);return false;" rel="nofollow https://www.stilldragon.org/discussion/ ... nts#latest" onclick="window.open(this.href);return false;" rel="nofollow . Od and Swede don't come on here much any more. Swede did the first types I think, pm them on empties site, there nice blokes and very helpful.

[cede]

I came accross the thread and I'm also looking at controlling the dephleg water flow.

I had a look at Swede's thing and I can not totally agree with him concerning the temperature sensor placement.

The coolant ( water ) sensor is used to open or close the valve to let water flow in the dephleg.
The water temp is then a function of the heat exchanged AND the inlet water temp.
For the same amount of heat exchanged, if your inlet temperature is lower, the output will be lower.

That's the fact annoying me, or the thing I do not understand in his setup.
I you are recirculating your cooling water, it might happen that the water temperature changes and then the cooling capacity in the dephleg changes too, thus modifying the reflux rate.

So by monitoring the coolant temperature, you can not determine the heat exchange rate because you are missing a variable, the inlet temperature.

I don't know it this is really a trouble as I do not own a column with a dephleg for now, but I'm a bit worried about that.

If I'd decide to go that way, I would instead monitor the difference between inlet and outlet temperatures to set the flow rate with a variable valve.

That said, I'm still try to understand things on paper for now

[Johnnywhiskey]

The efficiency of the dephleg will also change during the run. If heat is kept constant, early in the run more vapor will be running through the dephleg (more vapor speed). Hence the fast moving vapor will not be cooled as efficiently. Later in the run as the boiler is depleted the vapor speed will slow down and the vapor will be cooled more efficiently and will be closer to the exiting coolant temperature. On YouTube there is a video with the PID hooked up to the coolant and the ABV went up slightly through the run as the product slowed--i.e. the vapor was being cooled more as the run when on and the vapor speed slowed.

That said, in the end it is what works for you, and it seems like the monitoring the coolant temperature gives a very reproducible run. And the coolant temperature is a decent proxy for the vapor temperature. But I'm setting my still up to measure the vapor temperature, which should be one step closer to the actual ABV. That said, practically it might be tougher to get a good reading on the vapor temperature because the vapor has a lot less thermal mass than a liquid, and it might fluctuate more. But I have a port at the top of my still, so that is what I am going to go with.

I'm collecting parts for the PID controller and should have everything together by the end of the week. In total I'll have about $120 into the project, with the most expensive parts being the PID controller (and SSR (a relay)), the water valve and the box to build it in. It definitely could be done for less, but I'm adding a few extra lights and switches, and I bought most of the electronics on Amazon (easy but not always the cheapest). BTW after researching I decided to go with the 110v AC water valve. I would have rather run 12v DC over to the water valve for safety, but I might have needed a bunch of separate parts for a 12v source to run. And with 110v AC I'll be able to use the PID controller for other 110v ac temperature control uses (like fermentations), and I read the 110v valves are much faster than the 12v.

[cede]

Johnny, I would also go vapor path for the temp probe as vapor temp is related to ABV.

Getting the temp right should not be a trouble with a thin wall thermowell.
I bought a bunch at brewershardware for my brewery and put a bunch of thermal paste before fitting the gauges.

I do not know a lot about solenoid valves in 110V. I mainly use 24V ones, direct or semi direct acting. You should be able to get opening and closing time for valves in the datasheets.
Using 110V for valves, you should use a GFI and make sure the valve is properly grounded. Well, chances are small that a shortcut would happen, but I would not take any chance !

I must start the still construction first before thinking about putting a heating regulation on it, but I'll add spare ports for thermowells for sure.

[Johnnywhiskey]

Here is a little preview of what I have been up to. I still need to finalize some of the wiring and get the water valve hooked up. But I have the controller cycling and I was investigating the autotune feature.

BTW anyone familiar with tuning a PID controller? I'm hoping there is a way to keep the cycle time something reasonable for the water valve that I have. I don't want it to be cycle on-off every 2 seconds. I think that is something that can be adjusted by the "P" Band variable.

[cede]

I'm not familiar with this pid. On my temp controler I can adjust the dead band or hysteresis and time between on and off as it's controlling the compressor of a freezer.
The porportional band is where the pid is controlling the output from 0 to 100%. If temp is above or below that band, it's either 100% power or 0%.

[Johnnywhiskey]

What type of PID do you use? I'm looking for a variable to set the equivalent of a compressor delay on simple thermostats. Maybe 20 or 40 seconds so the valve is not continually pinging on and off. My PID does have a hysteresis range I can set, but I think that just sets how much the actual temperature (PV) can vary from the set temperature (SV).

The "P" band is a variable that sets one of the three algorithms of the controller--Proportional–Integral–Derivative. The P band determines how much the algorithm looks at the difference between the PV and SV. The output is still and on/off relay.

These are the instructions--not the greatest.

http://www.sanyoutech.com/UploadProduct ... 749860.pdf

[Johnnywhiskey]

Ok, figured it out. The "CtL" value is the control cycle value (in seconds I think). The instructions are a bit confusing, and I thought that was a value to set the output for SSR (it says "CtL=1-3 means SSR control output").

Anyway I have it set to cycle every 20 seconds (turns on no more than once every 20 seconds), which seems reasonable for the valve--maybe I'll set it for more. So in theory the controller and valve should be ready to control the vapor temperature, hence the ABV output.

Seems like a good time for a quote about theory v. reality.

"In theory there is no difference between theory and practice. In practice there is." --Yogi Berra

I'll see how practice goes next weekend when a run a batch of blue corn mash through it.

[cede]

You found it

My controler is not a pid, it's a basic temp controler with a dead band setting and on/off time.

I have some pids for the brewery but I made them myself using microcontrolers, but that's another story.

they say in your doc : In normally, the control cycle of the heating equipment
should be 20-30 seconds. For huge capacity heating
equipments, the value should be 30-120 seconds, in
order to longer the use life of the relay.

So I think you're right.

Theory and reality are sometimes way apart !