Jump to content

Operations by Temperature monitoring and control


DrDistillation

Recommended Posts

I come from BIG distillery ops where this type of thing is common, so I'm biased of course.  I'm pretty sure all the big distilleries are all computer controlled these days but I'm wondering how many of you guys use temperature monitoring/control in your distillation operations?

By this I mean manipulating boiler energy input and/or reflux cooling based on temperature of one or more probes in your column?

Of course if you use plates with multiple take off points you are doing this mechanically so you automatically fall into the yes I do this. :)

I don't care if it's computer controlled, PID for the smaller electric guys or steam heating, or if you use temp controlled valves. Don't really care if it's automated or manual operation.  What I'm wondering is how many of you guys use temps as a guide to fractions or cuts or to automate the distillation process to produce faster.  This could be predetermined cut off points for batches you run often or just a point to get a distiller involved to manually check for the cut over from heads to hearts or hearts to tails.

Of course the type of spirit dictates how you do the run as a whiskey put on oak is good with some smearing (or heads and tails) where a vodka/neutral will want tighter cuts/fractions.

So anyway without giving away any trade/proprietary secrets can you comment if you use any type of automated or manual control based on temperature?

Link to comment
Share on other sites

Mine's totally manual (gate valve on the water input to the dephlegmator) but I use the head vapor temperature as a guide for my heads to hearts cut for my whiskey. I run a hybrid still with one plate.

Link to comment
Share on other sites

I use temperature monitoring all the time. It's one of the things I look at, but not the only one. In the case of a spirit run, I'm also looking at the parrot hydrometer, the run time, and the output volume or weight. However the most important factor is taste. Recognizing the characteristic flavors of heads, hearts, and tails is the most important part of making good cuts. Flavor recognition is particularly important in experimental runs.

Now, in the case of a stripping run, you don't need to taste the output and can go by vapor temperature, the parrot hydrometer, and run time.

Temperature monitoring is also very important in mashing and making wash.

  • Thumbs up 1
  • Thanks 1
Link to comment
Share on other sites

10 minutes ago, Jedd Haas said:

I use temperature monitoring all the time. It's one of the things I look at, but not the only one. In the case of a spirit run, I'm also looking at the parrot hydrometer, the run time, and the output volume or weight. However the most important factor is taste. Recognizing the characteristic flavors of heads, hearts, and tails is the most important part of making good cuts. Flavor recognition is particularly important in experimental runs.

Now, in the case of a stripping run, you don't need to taste the output and can go by vapor temperature, the parrot hydrometer, and run time.

Temperature monitoring is also very important in mashing and making wash.

Ditto, well said.

Link to comment
Share on other sites

We pay a lot of attention to dephlegmator temperature control as a proxy for managing reflux ratio through the run.

We don't run constant dephleg temperature, we run a pattern that looks like a U - high reflux ratio at heads and tails, low reflux during hearts.  We feel this gives us the benefits of optimizing yields and to make very specific cuts (as typical with plates), but gives us the flexibility to dial it back during hearts to pull through more flavor.  We run PID control with a proportional valve to control dephleg temp.

We do cuts by taste, but honestly we run so consistent/repeatable, that we we know almost exactly when our cut points are approaching.

I almost think we could cut by weight of collected distillate if we needed to (or PG if there is some variance).

We don't like cutting based on vapor temperature, it's fine to ballpark identify the cut point (this varies by product), but always by flavor.

Link to comment
Share on other sites

I have a continuous stripping still, input flow rate controlled by maintaining a constant input temperature.

Beer or wine feed is heated by the condenser which is a counter-flow tube in tube, with some heat also from the stillage overflow.

The slower the input flow the hotter the feed is when it enters the column.

A gate valve allows the bulk of the feed then a simple PID opens and closes a small bypass valve to keep feed temperature constant.

KIS

Link to comment
Share on other sites

Yes we do and this kind of control over process is definitely key to getting the best control of your finished product.

I have no concept of the types who try to use cooling media that it not held at a rock solid consistent temperature with the " correct " sized chiller to do the job, or other fully consistent temperature cooling media that is not affected by undersized gear and chasing a batch process up the temp curve. If your target is 50F cooling media, then you need 50F for the whole run and nothing less.

A large part of getting consistent results is necessarily going to be enhanced by a very clean, very elegant mechanical system that does not require constant attention during the run. There is such a thing as doing things too much on the cheap. If you have natural skills with Electrical-Mechanical systems you will have an immediate edge. If you do not, then your prime directive is to hire someone who understands this side of things and can help you implement them. The internet is a fantastic resource for those who know how to use it. You can learn from both the success and failure of other approaches by careful analyses.

Link to comment
Share on other sites

  • 5 weeks later...

This kind of fit in here so I thought I'd post as it's both kind of funny and sad at the same time.

I also remember now why I never joined this site.  https://homedistiller.org/forum/viewtopic.php?p=7600904#p7600904 (some of those guys are pompous)

That particular forum seems to hate any talk of anything temperature or vapor related in doing cuts or letting you know when you are near a cut point so you can monitor product closely. 

I've seen a few threads like this where someone tries to explain the vapor temp relationship and how you can semi-automate the still for better product but they all end up the same way taken over usually by the same set of know it all guys who just don't understand distillation.

A few idiots seem to come out of the wood work every time a thread ventures into vapor temperature and how it can be used, stating everything from you can't use a PID to run a still, to you can't use a temperature gauge to anything else temperature or vapor related.  It seems these "experts" simply fail to understand the relationship that vapor temperature has to the product and how to use it for rough cut points or when to start watching to switch collection containers.  With a pot still your only source of control is heat and with a reflux still, you have the option of both boiler and cooling control or just cooling with your heat set at a specific point.  A few have mentioned doing exactly this here.

I'd love to hear these guys try to explain how large distilleries can use temperature monitoring/control or even smaller micro stills like Genio or iStill when they state it can't be done or the wrong way to run a still.

Link to comment
Share on other sites

I've never really had much success in trying to directly control vapor temperature.

While controlling vapor temperature would be the ideal case, I've found that the control logic would need to be much more complex than simple PID control - mostly due to dead time and the fact that the control function is likely nonlinear.  Consider that what you are really controlling is the column reflux ratio, and it takes time for the column to equalize to the new reflux ratio - and then be reflected in the vapor temperature.  

Controlling dephlegmator temperature via flow rate is far easier from a controls perspective.  I've been running this for a few years now and it gives me quite a bit of flexibility in control.  What happens when you maintain a fixed dephleg temperature, the coolant flow rate increases as the run progresses (more coolant required to maintain the same set point) - and the reflux ratio increases through the run - so the output speed falls, but output proof is more stable.  Fairly easy to run a low temp at startup, to quickly load the column, compress heads, take heads at a very high proof (usually 190 for me).  I can then increase the dephleg temperature and drop the proof dramatically, through hearts.  Finally once we get into late hearts, I set a cooler dephleg temp, and it allows me to push back tails a bit further than if I just kept a constant temperature.  I feel the higher reflux ratios - higher proof output - makes the cut points "clearer" if you are doing it organoleptically.

Back to using the vapor temperature as your process variable (PV) - the issue seemed to be dead time associated with column equalization.  I could get it working reasonably well one day, only to have it fail spectacularly the next.  Input coolant flow rate and temperature needs to be rock solid and consistent, otherwise you have to deal with variable dead time - a complete killer for process logic.  I toyed around with the idea of a kind of simple step logic system, not PID based at all - but simple logic control - if vapor temp > set point temp, open coolant by 1% and wait 10 minutes, repeat until set point reached.  Realistically, it seems far easier to just code up the control logic in a plc or little mini computer (arduino, etc) - this way you could build in all the constraints, etc.

Back to dephleg temperature control - the nice thing about this, input coolant temperature can vary without upsetting the system.  Our rig uses recirculating coolant for the condensers.  Depending on the time of year the chillers can't necessarily keep up, so the coolant increases from the start of the run to the end.   PID easily compensates for this by increasing the flow rate.  Easy Peazy.

So after thinking about vapor control for a while - I came to the realization that it would just make more sense to use a single overhead condenser.  Run a PID on the overhead condenser and coolant tank to keep a constant reflux temperature.  And use a proportional valve to vary the flow between column reflux and product.  The home distiller guys typically call this liquid management.  I believe this is the way the Genio and iStills work.  Easier to decouple the reflux condensation process from the reflux proportional control - the control logic/process becomes far more simple - even though the physical still design is a bit more complex.

If I were to go down this path.  I'd use an oversized reflux condenser with a small ballast tank.  I'd use a PID to control the overhead condenser flow rate based on the (small) ballast tank temperature.

I'd probably use a solenoid based system to control reflux/product ratio, varying between flowing liquid back into the top plate of the column, vs exit as product.  I'd use an additional product cooler on the output, since I'd want to run the ballast tank temperature far hotter than I'd want my product temperature.

With this, I'd target vapor temperature as PV and control the proportional distribution.  Feel like with a fixed ballast tank temperature, and the ability to alter the reflux flow far more quickly, not to mention a consistent dead time - this would work far better.

But who has the time to play around with all of this stuff? :)

 

 

 

  • Thanks 1
Link to comment
Share on other sites

2 hours ago, Silk City Distillers said:

I've never really had much success in trying to directly control vapor temperature.

 

 

 

 

That's a really great explanation.  When I started building stills years ago we learned very quickly that directly controlling vapor temp was not the best way to go for the same reasons that you gave.  If the customer wants automated control at a simple level we control coolant temp using a reverse acting thermostatic valve. We can use one thermostatic valve to manage the final condenser and up to four dephlegmators with very little to no fluctuation in ethanol output.  

Link to comment
Share on other sites

We absolutely use the temp gauge in the top of our condenser. When we see it start to rise, after about 45 minutes, we close the manual steam valve about a rotation or two, throw the water to the condenser through a mechanical Danfoss valve, while keeping in mind that the outside holding tank of cooling water may be anywhere between 33 degrees and 75 degrees. We take a heads cut in about 15 minutes, then go do some real work for about 4.5 hours. After about 4.5 hours if the proof is dropping in the parrot, we throw the return water through the defleg for aprox and hour. 

So yeah, couldn't live without the temp probe in the top of the condenser. I wonder if I can get an app for it ?

Prost

 

 

Link to comment
Share on other sites

I'm pretty sure the vapor temp in pot still mode was simply a PID to control the upper limit of temp allowed.  So a temp could be dialed in where fores end, then heads/hearts transition.  Then another where hearts/tails come it.  Of course the rate would drop off which is the indication the transition is coming and you can watch/taste switch containers.  This makes total sense since temperatures constantly increase during the run on a pot.  It's just knowing your runs and maybe adjusting for air pressure.

For reflux you'd use cooling to control vapor temperature via dephleg and a proportion control valve.  Swede Olson has an older video of his controllers showing exactly how to do this using a PID and proportional control valve.  He does a run at 80% likely for a light whiskey or something similar.

This is what has been talked about on that site a few times but video showing how to do it get removed then admins step in with PIDs can't be used to control a still type comments.  That's what I was referring to.

Of course you can control it like in this video using a PID, using PLC or computer would give more options and combinations like an iStill/Genio type still but a simple PID surely works as well for simpler operation.  

Link to comment
Share on other sites

Just to be clear - whether you are using PID or a thermostatic valve - the control is of the dephlegmator temperature by varying the coolant flow.  While vapor temp is certainly impacted through this process, it’s not directly controlled.  This is how the controller in the video is working.  Vapor temp is just being monitored, not controlled.

Vapor temp is impacted indirectly, two steps removed:

Increased Dephlegmator Flow Rate -> Increased Column Reflux Rate -> Column re-equilibration at higher reflux ratio -> New Lower Vapor Temperature (you know this, I'm sounding preachy)

While vapor temp will be held a bit more stable than if you simply kept flow rate constant, the controller won't target the specific vapor temp and hold there.  Picking a specific dephlegmator set point temperature will not always yield the same vapor temperature at the same point in the run.  I've found that the Wash/Wines ABV is actually the wildcard variable that comes into play.  At a given dephlegmator temperature set point, if you increase the wash ABV, you will decrease the vapor temperature.  Or said another way, the vapor temperature is a function of the dephlegmator temperature and the wash/wines abv.

Just sharing my hard earned knowledge.

Link to comment
Share on other sites

23 minutes ago, DrDistillation said:

This is what has been talked about on that site a few times but video showing how to do it get removed then admins step in with PIDs can't be used to control a still type comments.  That's what I was referring to.

Usually, when people say this, they are responding to people who are trying to use a PID to control the kettle/wash temperature - which can't be controlled with a PID.

Link to comment
Share on other sites

Quote

So a temp could be dialed in where fores end, then heads/hearts transition.  Then another where hearts/tails come it.  Of course the rate would drop off which is the indication the transition is coming and you can watch/taste switch containers. 

Auberins makes a cool little "PID-Style" Distilling controller that a few folks have used with some success - it does what you are talking about.

https://www.auberins.com/index.php?main_page=product_info&cPath=53&products_id=559

Downside is - we're talking about electric control only here, not steam (I tried to get Auberins to produce a 4-20ma version that could be more flexible, they weren't interested).

But for something like a simple electric gin rig - you could wire up this controller up to give you a little bit more convenience - since it can control the heatup process, cut points, and shut down.  The more advanced version can control multiple elements or even control your product condenser water flow (on/off).  Running gin from neutral - once you've got it dialed in, it's as simple as filling and flipping switches.  Our gin runs are incredibly repeatable, down to a few minutes total run time.

If you are distilling all manner of whiskies, rums, washes, wines (all at different ABV) - something like this becomes significantly less useful - as the temperature set points will vary dramatically (and you won't necessarily know them before hand).

Link to comment
Share on other sites

I fully agree with you on the first post of what's actually happening and you can see the vapor temp rise during the run from about 87.1 to 93.3 for the hearts portion of the run while holding the ABV.

I also agree you certainly can't control the boiler temp nor would you want to as it has to rise as the alcohol is removed. But when you know the temp your heads and hearts end at you can monitor vapor temps in either configuration and use that info for cut over points.  Totally useful info, when you run the same batches over and over.

What I dislike in threads like those are people assuming other MUST run the same way they do and dismiss using a tool a different way then they would use it.  Knowing the vapor temp, you can for example slow the production as you get close to where tails come in on a pot.  It will just run slower and slower, then stop producing at the vapor set point. If you just run with a set power level you'll push right past the cut over points.  Just a different way of running, not better or worse, just different.

Funny that you mentioned the Auber controllers.  I just ordered the DSPR400 to play with.  It's actually more a power regulator from what I read on it with some PID functionality thrown in, but I could be wrong.  Will know how well it works in a week or two hopefully. I saw that mentioned in one of the PID threads and it caught my eye.

Have you ever played with either the 220 or 400 on a small electric setup?  Probably not to many people here using electric.

Link to comment
Share on other sites

On 4/14/2020 at 9:55 AM, Southernhighlander said:

That's a really great explanation.  When I started building stills years ago we learned very quickly that directly controlling vapor temp was not the best way to go for the same reasons that you gave.  If the customer wants automated control at a simple level we control coolant temp using a reverse acting thermostatic valve. We can use one thermostatic valve to manage the final condenser and up to four dephlegmators with very little to no fluctuation in ethanol output.  

So just to be clear.  If you are controlling cooling in an automatic fashion,  it is best to control the vapor temp indirectly by directly controlling the coolant flow.  The temp probe for a thermostatic valve should be at the output of the final condenser (from the final condenser, the coolant should go to the last dephlegmator in line and then to the one before that and so on) and the valve itself at the input of the final condenser.  If you have  your temp probe directly in the vapor path, using a PID, your spirit output will constantly fluctuate.  Also controlling the coolant flow is best done in a proportional manner using a valve that acts proportionally instead of on off.

Danfoss thermostatic valves operate in a proportional manner.  I like them better than PID control because they are simple and do not need electricity or any other possible source of ignition.  Also, thermostatic valves make controlling multiple dephlegmators simple.  I like mechanical gauges better than PID or PLC for the same reason and because there is no electrical ignition source.

My electrical department head developed a PLC controller and wrote the software for it.  It basically controls everything including heating, cooling, cuts, as well as condensate and other pumps.  It has been sitting down at my equipment development lab for almost 3 months, connected to a 2" diameter continuous column still.  We used it to control a couple of little vodka still prototypes before we reconfigured it for the continuous still for data logging etc.  It works great but I just have not seen the need to move forward with it at this time.

As far as cuts go, my grandfather always said that your best instruments for doing cuts are your taste and smell, but just like most things, there are a lot of different ways to skin that cat and none of them are wrong as long as they work for you.

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...