Cooling Water Temperatures

[Quinn_the_Eskimo609]

Hello all,

I am a new distiller (Chemical Engineer straight out of college) and am looking to figure out some of the details in our process. A big question that has come up is the temperature of our cooling water. How cold is too cold? This is a utility question as well as the strain on stainless pipes. If alcohol vapor is hitting the condenser at ~170-200F, and cooling water is 50F will I risk cracking a tube?

Thank you!

[Southernhighlander]

I'm not an engineer but I can tell you that you would not risk cracking a tube with that temperature differential, as long as everything is well made.  It would not bother me a bit to have greater temperature differential.  I have mash cookers out there that have been in operation for almost 10 years that utilize a single jacket for both steam heat and cooling.  The steam temps are around 243 F and the coolant temps are as low as 36 F.  It never worried me a bit to design them that way because I had already designed, and had my employees build, steam injection boilers with temperature differentials of over 1400 degrees, between the fire box and the inner boiler with no issues.   My current mash tun design has an internal crash cooling coil.  I only changed the design because the internal coil has a great deal more thermal exchange surface area than the inner part of the jacket did.    

[Silk City Distillers]

Some folks have had operating problems running reflux condensers with a very cold input coolant directly.  This can sometimes cause some odd behaviors like cycling output  and column flooding (subcooled reflux).

Some stills temper (warm) the reflux condenser coolant by running it through the product condenser first (sequential condensers) to avoid this.

If running the condensers independently (for better control), the colder the coolant, the more sensitive the reflux condenser will be to you making changes.  Ran a still once that had the RC on the cold glycol loop, all manual controls.  Crack that valve just a little too much and you’d throw the column into full reflux.  Incredibly frustrating hair trigger.

[Alex_Sor]

11 hours ago, Quinn_the_Eskimo609 said:

If alcohol vapor is hitting the condenser at ~170-200F, and cooling water is 50F will I risk cracking a tube?

You shouldn't worry about it.

AISI 304 steel has excellent mechanical properties, and if it is welded correctly between its parts, then intergranular chipping (metal fatigue) is not a problem for you. Your temperatures are too low for problems

Example: Calculate the linear elongation of a 0.5 meter long stainless steel beam when its temperature rises from 27 Celsius to 1027 Celsius.

According to the table, the average coefficient of linear expansion of steel in the temperature range of 27...1027C is equal to 22.3 · 10-6 deg-1.

That is: an increase in size of 0.0000223 meters by 1 degree Celsius.

Let's calculate: 22.3 * 10-6 * (1027-27) * 0.5 = 0.0111 meter.

We get the value of the linear elongation of the beam 0.0111 meters.

For the case of a moonshine still, the temperature difference will not be higher than 100 degrees Celsius, then:

22.3 * 10-6 * (100) * 0.5 = 0.001115 meters, this is 1.115 millimeters. This is a very small value

[Southernhighlander]

23 hours ago, Silk City Distillers said:

Some folks have had operating problems running reflux condensers with a very cold input coolant directly.  This can sometimes cause some odd behaviors like cycling output  and column flooding (subcooled reflux).

Some stills temper (warm) the reflux condenser coolant by running it through the product condenser first (sequential condensers) to avoid this.

If running the condensers independently (for better control), the colder the coolant, the more sensitive the reflux condenser will be to you making changes.  Ran a still once that had the RC on the cold glycol loop, all manual controls.  Crack that valve just a little too much and you’d throw the column into full reflux.  Incredibly frustrating hair trigger.

Knowing what I know now.  I believe that it is always best to pass the coolant through the final condenser before it passes through the dephlegmator or dephlegmators.  This gives you a much more stable output and better control.  It also allows you to run lower coolant temps to the final condenser with no issues.  

[Alex_Sor]

9 hours ago, Southernhighlander said:

Knowing what I know now.  I believe that it is always best to pass the coolant through the final condenser before it passes through the dephlegmator or dephlegmators.

Everything is correct.

The role of the condenser-alcohol cooler and the reflux condenser is different.

It is important for the condenser to turn all the alcohol vapor into liquid and cool the alcohol. For a reflux condenser, it is important not to condense all the alcohol, but to do just partial condensation.

If you supply very cold water to the dephlegmator, you run the risk of disrupting the normal flow of the process. Why? because there will be a very large temperature difference between cold water and alcohol vapor.

If you supply water to the dephlegmator after the alcohol condenser-cooler, this water will be heated to 30-40 degrees Celsius, this will not cause a "condensation shock" in the dephlegmator. You will get a stable dephlegmator performance.

 

[Silk City Distillers]

Sequential condensers are not always the best approach.  For passive-cooled recirculating systems, warmer municipal water, or undersized chiller systems, split condensers can provide more flexibility and control, as you aren't trying to work within a smaller set of overlapping parameters.

Where sequential condensers often fall short is with warmer input water.  The higher output temperature from the product condenser can sometimes make achieving high reflux ratios at high power inputs challenging.  Sure, you can cool the distillate further to compensate (increase flow rate), or get a bigger chiller or oversize the reflux condenser (easier said than actually done), but for many of these systems, there isn't unlimited cooling BTU on tap.

[Alex_Sor]

53 minutes ago, Silk City Distillers said:

Sequential condensers are not always the best approach. 

I wrote what I wrote and what I wanted to write.

Regardless of your thoughts, there is Physics and Chemistry.

Physics interests us in its section "gas-liquid phase transition". So, if you have a distillation process, and your equipment is correctly calculated, then you do not need any electronic "control devices". The whole process takes place independently and accurately and harmoniously.

Phase transition gas-liquid (in the condenser) and liquid-gas (in the distillation tank) are Physical Process Stabilizers.

No electronics needed at all.

From the control, only "cooling water flow" and "heating power" (stable) are needed. The use of warmer water in the reflux condenser reduces the "thermal head" (I do not know how this concept is translated into English), which allows a smaller temperature gradient in the reflux condenser tubes, and this leads to better stability of operation.

 

[kleclerc77]

On 8/18/2021 at 2:32 PM, Quinn_the_Eskimo609 said:

A big question that has come up is the temperature of our cooling water. How cold is too cold? 

I've found there is a point where the cooling water is too cold for a spirit run. Cold as I can get it for everything else (crashing mash/stripping runs), but if the water is too cold during a spirit run it can cause some serious chugging/inconsistent flow rates. 

[Alex_Sor]

4 minutes ago, kleclerc77 said:

but if the water is too cold during a spirit run it can cause some serious chugging/inconsistent flow rates. 

I will repeat once again: you consider the distillation process as a set of disparate facts, but it is a complex but interrelated process, and in the process of distillation there are two "natural stabilizers" - two phase transitions.

It is not necessary to search for temperatures, but to control the supply and removal of heat.

If you supply too much cold energy (remove heat from the process), then you knock down the smooth flow of the process.

And you get clogging.

[kleclerc77]

@Alex_Sor You seem to dwell a lot in the theoretical. People here are looking for practical advice. Using slightly warmer cooling water prevents chugging for me and others.

[Silk City Distillers]

Are we talking theory or real-world practice?

In the real world, Ive found pushing below 50f  on a split system is challenging.

I can’t speak to theory, because I’m not sure what temperature corresponds to harmonious balance.

[Alex_Sor]

4 minutes ago, kleclerc77 said:

@Alex_Sor You seem to dwell a lot in the theoretical. People here are looking for practical advice. Using slightly warmer cooling water prevents chugging for me and others.

no need to pretend to be stupid

I believe, and I am sure, that before doing something, you need to study the Theory of "how it works".

It's not as difficult as it sounds. And this (study) does not require an advanced degree from you

the usual curiosity of the boy

Smaller temperature differences in the dephlegmator allow for smoother control.

Yes it is. But why ?

And the answer is simple: cold water can take more energy from the condensation process. And a small amount (mass) of very cold water leads to great results.

That is: if you slightly change the mass of flowing cold water, and the mass of warm water, you will get different results. Because you will remove more heat from the process if the water is very cold.

Am I making it clear?

We have a variable value: this is the "degree of coldness of water" and, there is a constant, a constant (relatively) and stable value = this is the condensation temperature of steam in the reflux condenser. This is a "phase transition" and it stabilizes its temperature.

the smaller the temperature difference, the more options for the rate of water supply to the dephlegmator, and the easier the regulation.

 

[Alex_Sor]

15 minutes ago, Silk City Distillers said:

Are we talking theory or real-world practice?

I'm talking about what I know (taught and understood the issue) and what I myself have tested in practice.

I have my own designs for distillation equipment, which I first calculated and developed, and then tested in real work.

Without the Theory of "how it works" there is no point in doing practice.

The balance of energies is what matters.

If you calculate the balance of energy input and cold water output, you will always get very good results.

[kleclerc77]

@Alex_Sor all of your answers on here are snarky and unhelpful. Wasn't it a couple months ago that you suggested barrel aging bourbon (or spirits in general) wasn't necessary, while also admitting to never making bourbon? Get a grip. Your advice is bad.

[Alex_Sor]

2 minutes ago, kleclerc77 said:

@Alex_Sor all of your answers on here are snarky and unhelpful. Wasn't it a couple months ago that you suggested barrel aging bourbon (or spirits in general) wasn't necessary? Get a grip. Your advice is bad.

What does barrels have to do with the work of a dephlegmator? You do not understand either theory or practice, but are you trying to judge me?

[Alex_Sor]

3 minutes ago, kleclerc77 said:

@Alex_Sor It was just one example of many that plainly shows that you give awful advice. Just like here. A condescending, professorial non-answer, as always.

Please indicate in this thread what I wrote incorrectly.

In this thread.

What, in your opinion, contradicts Physics.

I write my opinion and explain it.

If you personally do not understand, then this is not my problem. The laws of physics do not depend on your mood and your opinion.

I'm not going to argue with those who do not want to delve deeply into the processes.

This will be an argument between a blind man and a deaf one
 

[SlickFloss]

This is kinda silly because at some points in time y'all are saying the same shit.

4 hours ago, kleclerc77 said:

I've found there is a point where the cooling water is too cold for a spirit run. Cold as I can get it for everything else (crashing mash/stripping runs), but if the water is too cold during a spirit run it can cause some serious chugging/inconsistent flow rates. 

"If my water is too cold it causes problems

 

4 hours ago, Alex_Sor said:

I will repeat once again: you consider the distillation process as a set of disparate facts, but it is a complex but interrelated process, and in the process of distillation there are two "natural stabilizers" - two phase transitions.

It is not necessary to search for temperatures, but to control the supply and removal of heat.

If you supply too much cold energy (remove heat from the process), then you knock down the smooth flow of the process.

And you get clogging.

"If your water is too cold it causes problems"

 

 

Oh wait, this isn't silly because you're saying the same fucking thing. This is silly because we're all fucking adults!

 

Crayyyyyy

[Southernhighlander]

Hi Everyone,

The first still that I saw with athe condensers plumbed in series was a Kothe still that we did some work on.  It also had a thermostatic valve to partially automate the coolant flow.  The Ukrainians, Russians and Germans all run their coolant from their final condensers to their dephlegmators for  the reasons that  Alexander gives below.  We have tested both methods thoroughly here and the method of running coolant from the final condenser to the dephlegmator or dephlegmators before it, works the best hands down.  In the many test runs that we did we used less coolant running the condensers in series than separately.  We had better control and smoother output running in series and we were able to maintain higher proofs during Vodka runs.   We also found that we could run much colder water to the final condenser with absolutely no issues, when plumbed in series, Though I generally recommend condenser coolant be around 50 degrees F.  I see no issue running 42 F coolants through a tube in shell final condenser.  In testing we found that even lower temps are okay as long as the coolant flow is reduced so that the condenser output temp is the same.  Of course you must run at lower flow with lower temp coolant so that the condenser output temps are the same otherwise you will get chugging.

Our vacuum stills run coolant as low as -40 F through the final condensers with no issues however that is a different situation

The  German, Russian and Ukranian still builders got it right.  They all run the coolant in series from the final condenser back through the dephlegmators from the last dephlegmator to the first.   I don't know their standard recomendation but every Vendome column still that I've seen had their condensers plumbed in series as discribed above. 

 

We partially automate the coolant control of up to 3 columns and the final condenser with one Danvos thermostatic valve and we have done it many times with great success.  This has enabled us to have 300 gallon Vodka stills with 20 plates  that stand less than 10 ft tall.  None of our competitors are doing that.  Our 300 gallon Vodka stills with fit in spaces with ceilings as low as 10 ft.

 

I have over 500 of my stills in distilleries and I can say with absolute certainty that Alexander's statement below is correct. 

Below Alexanders statement is a pic of stills in one of my 3 assembly shops.  All 3 shops have stills under construction at all times.  The still in front is a 105 gallon Standard Series Vodka Still with gin basket and CIP.  The huge still pot back and to the right is part of a Paul Hall Signature Series 2,500 gallon Ultra Pro Vodka Still.  When running Whiskey it will put out up to 75 gallons per hour.  This pot is waiting to be jeweled.  My employees will put several thousand swirls in the stainless to form a fish scale pattern.  Each swirl must be done perfectly and no mistakes can be made.  It's all done using smal pnumatic drills running 2" sanding pads.  It will look amazing when it's complete and it;s condensers will be plumbed in series and will be controlled with one thermostatic valve.  It will have 2 dephlegmators and one final condenser

 

 

"The role of the condenser-alcohol cooler and the reflux condenser is different.

It is important for the condenser to turn all the alcohol vapor into liquid and cool the alcohol. For a reflux condenser, it is important not to condense all the alcohol, but to do just partial condensation.

If you supply very cold water to the dephlegmator, you run the risk of disrupting the normal flow of the process. Why? because there will be a very large temperature difference between cold water and alcohol vapor.

If you supply water to the dephlegmator after the alcohol condenser-cooler, this water will be heated to 30-40 degrees Celsius, this will not cause a "condensation shock" in the dephlegmator. You will get a stable dephlegmator performance."

[Southernhighlander]

9 hours ago, Silk City Distillers said:

Sequential condensers are not always the best approach.  For passive-cooled recirculating systems, warmer municipal water, or undersized chiller systems, split condensers can provide more flexibility and control, as you aren't trying to work within a smaller set of overlapping parameters.

Where sequential condensers often fall short is with warmer input water.  The higher output temperature from the product condenser can sometimes make achieving high reflux ratios at high power inputs challenging.  Sure, you can cool the distillate further to compensate (increase flow rate), or get a bigger chiller or oversize the reflux condenser (easier said than actually done), but for many of these systems, there isn't unlimited cooling BTU on tap.

Silk,

During testing we ran all scenarios including really cold coolant, really warm coolant etc.  Without adjustable dephlgmator bypasses we had all of the issues you would expect from the things you mention above, however once we added adjustable coolant bypass control at the dephlegmators there were no issues and everything worked perfectly as one would expect.

Warmer coolant up to 74 degrees F caused us no issues running sequential condensers.  We simply increased coolant flow to the final condenser so that the final condenser output temp was the same as with the lower temp coolant and we adjusted the dephlegmator coolant bypass so that the dephleg had the exact same flow and temperature as it did when we were running 52 F coolant.  It's that simple.  

Let me know the next time you are down this way and I will set up a prototype still in my equipment testing lab and you can see the results given above.   We'll run both kinds of coolant designs so you can see the difference.  I have lots of fun toys in the testing lab right now including a really fast vacuum stripping still and a subcritical CO2 extractor that operates at 1600psi without a pump of any sort.

[Silk City Distillers]

Yeah I've got to get back that way.  I wanted to go back to Top of the Rock in Branson, which puts me close enough.

I've tried it every way to Sunday.  Apologies for the crap picture, but this is the setup we've been running for the past 6 years.

Recirculation pumps are "in-the-loop" - they run at about 5-7 gallons per minute.  As such, my condensers exhibit almost zero temperature gradient.  This bypass mixing approach allows me to easily temper cold coolant, to running full open to the tank, if we outrun our chillers.

The 3 way valves are all high speed Belimo proportional actuators run of Omega PIDs.

[Southernhighlander]

Silk,

 

  You are welcome to visit any time, just give me a heads up a week or two in advance.

That's a  good design.  Better than most.  I bet your still runs really smoothly without swings with that set up.

[Southernhighlander]

8 hours ago, Southernhighlander said:

Silk,

During testing we ran all scenarios including really cold coolant, really warm coolant etc.  Without adjustable dephlgmator bypasses we had all of the issues you would expect from the things you mention above, however once we added adjustable coolant bypass control at the dephlegmators there were no issues and everything worked perfectly as one would expect.

Warmer coolant up to 74 degrees F caused us no issues running sequential condensers.  We simply increased coolant flow to the final condenser so that the final condenser output temp was the same as with the lower temp coolant and we adjusted the dephlegmator coolant bypass so that the dephleg had the exact same flow and temperature as it did when we were running 52 F coolant.  It's that simple.  

 

I was just looking over my notes from the testing.  It appears that we did have some surging in the final condenser itself when we ran with colder coolant however the surge suppressor on the final condenser compensated for it, for the most part. 

    Best operation for vodka was with the final condenser coolant going in at 53 F and coming out of the final condenser going into the dephlegmator at 120F.  In that situation the system self balanced after the initial phase of loading the plates and compressing the heads, which was done manually using the bypasses.   This is what Alexander was talking about when he mentioned the system balancing itself when plumbed this way.  In fact it will self balance with up to 3 dephlegmators fed sequentially from last to first.  We have never tried to run more than that.

Alexander designs large commercial and Industrial continuous column stills and vacuum stills for a manufacturer in Ukraine.  He has a deep understanding of continuous column and vacuum still design.  He has worked for me on several projects, some of them very large.  He holds 2 engineering degrees and he was a Naval officer. I've seen his resume and it's very impressive.

 

[Alex_Sor]

I have a mathematical model that takes into account (and calculates) all flows of water and alcohol vapor in a condenser and reflux condenser.

Here is a picture of the model working.
It shows what temperatures are at the inlets and outlets of the alcohol refrigerator, vapor condenser, and reflux condenser.
Selection of 25 liters of pure alcohol per hour (in the picture), with a heater power of 16 kilowatts.

The model allows you to "play with numbers" and see what happens at different speeds of water and different powers.

It is a "closed model" that uses a conventional (car-type) radiator(green color) to exchange heat between ambient air and cooling water. Therefore, the temperature of the coldest water here = 35 Celsius. Main circuit water flow = 15 liters per minute, reflux condenser water flow = 7 liters per minute. Temperature at the inlet to the reflux condenser = 38.8 Celsius, at the outlet = 65.3 Celsius.

Please note that a water flow of 15 liters per minute heats up in the alcohol condenser from only 35.6 to 38.8 Celsius! At the same time, the condensation power is 3 kilowatts, which gives the condensation of alcohol 12 liters per hour in this place.
The model shows that only 3.144 degrees Celsius from the cooling water is used for this.
This is one of the mathematical models for simulating the operation of a distiller that I have.

[Silk City Distillers]

5 hours ago, Southernhighlander said:

I was just looking over my notes from the testing.  It appears that we did have some surging in the final condenser itself when we ran with colder coolant however the surge suppressor on the final condenser compensated for it, for the most part. 

 

Agree, overcooling on the final/product condenser is rarely a problem if you have a sufficient vent, will usually exhibits as an oscillation of the output speed, and in extreme situations causes the "chuffing" where the sudden vapor crash causes the oscillating vacuum condition where output stops as air is sucked back through the vent/parrot.  This "chuffing" was pretty common back in the early days, where folks were using undersized long Leibig (single tube in tube) condensers which were pushed very hard (high flow rates, cold water).

Like we've all agreed here, it's the reflux condenser that's far more sensitive to input temperature.

@Alex_SorWhat do the mathematical models say about the ideal surface area of the reflux condenser in relation to the column diameter or vapor speed?  How are you accounting for thermodynamic efficiency of the reflux condenser when moving from model to actual?

Asking because your models appear to assume 100% condenser efficiency.  13000 watts knock down, 7 liters per minute, calcs to roughly 26.6 delta T.

Suppose you could just work backwards from 'rule-of-thumb' heat transfers and hope for the best.  For example, tube in shell in condensing mode being about ~3000w per square meter surface area, and oversize from there.