New Condenser Calculator

[haggy]

I have been researching condensers in HD for over a year now and have developed a New Condenser Calculator that will apply to many different home distilling condenser applications.

In my past chemical industry experience, I developed calculation models for different types of heat exchangers. I applied those models to the condensers that we use here in home distilling and determined they were applicable. They were modified to fit home still condenser run data well.

The New Condenser Calculator has several options. Either (1) a Shotgun or a Dimroth Coil condenser made of (2) copper or stainless steel can be chosen for the calculator. They have different calculation models.

The calculator can run (3) a stripping run or (4) a plate or packed reflux column run. It can run either (5) a RC reflux condenser or a PC product condenser.

You make the five decisions in the Input Data to determine what type of condenser and what stilling system you want for the condenser operation.

And then there is the question of what cooling water flow rate is required for the system. That is solved for a CM column by an optional search procedure to find the cw flow rate that makes the Heat Transfer Required about equal to the Condenser Heat Transfer. More on this in the example below.

The Calculator can be used to evaluate different condenser designs to find a good design to make or buy. It can tell you if your design is adequate for the job or not. You can test the effect of different set points. Use it for your current still to see if you should adjust the cw flow rate or even try out a new design.


Input Data

The user tells the calculator what type of condenser and still is to be studied with a series of decisions. Input data for two examples of the decision tree and their run input conditions follow:

Here is input data for pot still stripping with a Dimroth coil condenser. It comes from an OtisT post about his 4" reflux column and an 8 coil Dimroth in the thread "3" CCVM" to knock down 5500 watts with about 4 L/min cw flow.. The calculator estimates the distillate rate and abv of the stripping run and uses that to evaluate the condenser design and find the cw flow rate required.

Next is input data for a case with a shotgun reflux condenser ( deflegmater ) in a 2" CM packed reflux column. It is from a PoolGuy 2" reflux column run and described in his post in the thread "First Distillation" by muse. You need to include the product rate to be run and the expected product abv. In this case I did a search to find the required cw flow rate to run.

Results

For the first case of a strip run with a Dimroth condenser, here are the results of the calculations. See that the initial product vapor was estimated to be about 13.4 Lhr at 57% abv. ( if no heat losses ) The condenser adequately condensed the outlet product vapor and even subcooled it a little. Product temperature is estimated at about 68 C.

For the second case of the 2" CM defleg shotgun condenser, here are the results of the calculations:

Explaining the Second Case Results:

A lot of info is given in the results. The top box applies to the column vapor entering the condenser. The vapor temperature in and the temperature out, the vapor abv and how much vapor is condensed, the reflux ratio and most important is the Heat Transfer Required to condense that vapor, 1101 watts.


The bottom box has to do with the condenser operation. You see the surface area, the LMTD, the U ( the calculated heat transfer coefficient ) and the Condenser Heat Transfer, 1101 watts. Then the cw temperature out of the condenser and the cw flow rate input data.

The search option was specified. The program will try different input cw flow rates and compare the two heats until a cw flow rate is found where they are about equal. ( 1101 watts and 1101 watts ). A 552 ml/min cw flow rate gave this result and is the cw flow rate required to condense the above amount of vapor and get the product flow that was specified ( 1.0 L/hr ).

Making the two heats about equal is for a CM reflux still where some of the vapor passes through the defleg RC condenser and goes over to the PC. For a VM or CCVM reflux still, the vapor split over to the PC has happened before ( or at the same time ) the vapor hits the condenser, so the condenser can condense all the vapor it sees and also subcool it. The Condenser Heat Transfer watts can be greater ( 10-15% greater is good ) than the Heat Transfer Required watts for a VM or CCVM column.


------------------------------------------------------------------------------------------

In some future posts, I will show several different applications of the Calculator. It is written in JavaScript so it can be added to the HD Calculator section for all to use.

I have tested this Calculator with many Shotgun applications and it gives good results there. Several Dimroth applications have also been studied and are ok, but not as well established as the many shotgun applications. A Liebig condenser can be run as a one tube shotgun.

Looking over the Calculator:

The Input Data is based on the set points and still that you want to run. The main unknown is the cw flow rate, and a search procedure is included to help you find the correct cw flow rate requirement. The Results give you a lot of information so you can better understand what is going on. It is not as complex as it might seem.

Is this thing any good?,,,,,, Hell yes it is!

I can add it to the HD calculator section so you can use it and test it.

Comments / improvements / thoughts are welcome.

haggy

[haggy]

A critical part of this condenser calculator is the vapor rate out the pot or out the top of the reflux column and into the condenser. How does the calculator predict this?

Well, go take a look at the Pot Still Purity Calculator in HD developed by Tony Aukland ( not sure of the person, correct me if it was not Tony ). For a pot still strip run, it gives the the ml of distillate collected for different times during the run for different inputs of pot watts and pot abv%, the two main variables.

I used the initial ( first 30 minutes ) distillate rate information from that calculator at no heat losses and with no internal passive reflux to develop the curves in the chart below. The values would be a little lower if you have heat losses and internal reflux.

Then I developed a CORRELATION to predict the stripped vapor rate values you see in the curves. The correlation is based on the pot watts and pot abv%. That correlation is in the New Condenser Calculator and used for pot stripping runs. A separate calculation is made for the stripped vapor abv% going into the condenser, it follows the standard alcohol - water equilibrium data and depends only on the pot abv%.

I tested the vapor rate values of the curves with many posts of stripping run data and they were good fits. The vapor rates also agreed with the pot vapor rates calculated in the Pot - Thumper Calculator on HD.

So, the vapor rates out of the pot for stripping runs were established.

Then a curious and unexpected thing happened.

When I examined the good and complete run and shotgun deflegmater condenser data of 13 real 2" packed reflux column runs that PoolGuy sent me, the vapor rates up the column found from that condenser data were nearly the same as the values in the curves for the pot still. Wow!

So, I use the vapor flow rate correlation for both pot still stripping and for packed reflux column condensers in my New Condenser Calculator.

That is how the vapor rates estimated in the New Condenser Calculator were obtained - and the vapor rate curves below are for both a pot still and a packed reflux column. That is what agreed with many sets of condenser data. I will show examples of different stripping and reflux column runs which fit the condenser calculations which are based on the vapor rates in the chart below.

If there are significant heat losses in the pot or column and /or some internal passive reflux, the vapor rates would be a little lower than the curve values.

haggy

[Yummyrum]

That graph looks pretty much in line with what I have found .
I initially was confused by Vapour in Litres . I was thinking Litres of vapour , but realised you mean vapour condensed back to distillate .

[Homebrewer11777]

I'd like to try the calculator. Will you be adding to site?

[haggy]

Homebrewer11777

Yes, when it is ready. Got some more work to do. Go see the latest posts in this thread.

viewtopic.php?p=7787862#p7787862

I am doing some updating of the vapor rate up the column calcs based on those last posts. The first version is not bad, it fits a lot of real run data and is ok for calculating and evaluating strip run condensing. It is not totally sound for reflux column condensers.

If you want it, send me a PM with your email address. But it comes with a price. You have to send me back some good, complete real run data so I can check out the calcs.

haggy

[Kareltje]

The other day I wondered how you were doing. Must be telepathic.
Clearly you are going strong. Nice to see that.

As I have an aircooled ss condensor of about 10 m long, your models alas are not applicable to my system.

[drmiller100]

Why not simply convert heater watts to btus. One btu is one degree change for one pound of water.

Input coolant temp difference and you are done.

[haggy]

This calculator evaluates the performance of condensers used in two different operations - pot stripping and reflux column operations.

A key factor is estimating the vapor flow rate into the condenser given a set of operating conditions.

For pot stripping, the graph above with many lines for different pot abv % is applicable and used in the calculator.

For the reflux column operation, it is a different story. The pot watts and the vapor abv ( vabv ) at the top plate ( or top of the packing ) will determine the vapor flow rate. And this can be calculated from the latent heat of vaporization of alcohol and water.

This is the calc in the browser Javascript program for the vapor flow rate. The loss term is the total heat losses in the pot and along the column.

wattsup = watts * ( 1 - loss/100 ) ;
colmvap1 = wattsup *3600/( vabv /100 *1000 *0.789*846 + (1-vabv/100 )*1000 *2256) ;

So, at the bottom is a graph of these calcs for several operating conditions.

Two top column vapor abv's are given, 95% and 80%. The top line is the vapor ( distillate ) flow rate L/hr for no heat loss in the system. The two lines in the middle are for 95% top vapor abv and for 30% and 35% heat losses. The bottom line is at 80% top vapor abv and 30% heat losses.

I studied many cases of CM reflux column run data ( that included cw flow rate ) and found that 30% to 35% heat losses fit most all the run data.

Several stillers send me their reflux column run data. higgins run data is well detailed and very reliable. Here is an example of one of higgins runs and the condenser calculator estimate of it.

He ran a spirit run using a 4" x 3 sieve plate CM reflux column using bourbon low wines at 19.2% abv in the pot. The pot power was at 2530 watts. The early in the run conditions were: Top plate vapor abv ( also distillate abv ) was 91%, distillate rate was 2.84 L/hr and cw flow was 375 ml/min. The deflagmater condenser is a Dimroth, 0.25" diam coiled copper tube about 110" long.

Here is my latest version of the Condenser Calculator ( Version 15 ) using these run inputs.

Input Data

You see that this run used 30% heat losses. Some calculator runs before that with a different heat loss did not fit the cw flow data.

Results

There is a good fit ( the Heat Required 1132 watts is about equal to the Condenser Heat Transfer 1128 watts ) at 367 ml/min cw flow rate.

And, please note the neat feature of the calculator. I estimated a cw flow of 300 ml/min in the input data. The program steps that up ( or down ) until it finds a cw flow where the heats are about equal. In this case - 367 ml/min fit the data. So, the 30% heat loss was best. Other heat losses did not give a the correct cw flow.

So, I am close now to having a good condenser calculator. The same condenser heat transfer coefficient equations are used in both the pot stripping and the reflux column operations. A few more run data example posts and I will send it to Uncle Jessie for review.

Here is the reflux column vapor ( distillate ) flow L/hr chart showing a most likely reflux column vapor flow rates at 95% abv and 30% to 35% heat loss. Different reflux column setups will certainly have different heat losses, but a good 95% abv column should have vapor flow rates not too far away from the two middle curves. Some more run data examples will show that.

haggy

[haggy]

Here is a very good example of pot stripping and the good fit of the Condenser Calculator to the run data. OtisT posted this information which includes cooling water flow data:

Otis' Pot and Thumper
Post by OtisT » Thu Mar 22, 2018 6:10 pm

Finished my metal work today on a new Pot and Thumper still.

Re: Otis' Pot and Thumper
Post by OtisT » Fri Mar 23, 2018 2:48 pm

Antler24 wrote:
Let us know how that dimroth thingy works out lol

No supprise; The dimroth works Great!

I just did a cleaning run and test with a 5 gallon, 12% wash. No Thumper, just the boiler and Dimroth PC.

5600 Watts the entire run. Collection @ 6 minutes per 1 Liter.

Cooling Flow data:
@ 1.5 lpm, some slight vapor leakage
@ 2.0 lpm, no vapor leaking, Hot water output and Hot distillate
@ 3.0 lpm, no vapor leaking, Warm water output and Warm distillate

The Condenser Calculator inputs and results for the 1.5 L/min cw flow case are next. Notice in this case that the product ( distillate ) flow of 10 L/hr and expected abv of 60% are calculated by using 30% for the Heat Losses.

The condenser calcs at 1.5 L/min cw flow give a Condenser Heat Transfer of 3743 watts and the Heat Required is 3885 watts, close together, no subcooling. There could be some slight vapor leakage like the data showed.

So,
I am satisfied with the PC condenser heat transfer calcs for pot stripping, both for the shotgun and for the dimroth condensers. Another post here will show a summary of many stripping runs that have been studied and used to determine the calcs.

Also, the reflux column condenser RC calcs have been updated and I am satisfied with them now. In a future post in the "Reflux Condenser Sizing - Packed Columns" thread I will give a summary of the reflux column runs that were updated and studied and fit the reflux condenser data.

haggy

[haggy]

This New Condenser Calculator can analyze condensers used for stripping runs and for plate or packed reflux column runs. You choose the option you want to run.

I have detailed the reflux column condenser runs that have been studied in my post on July 20 in the thread "Reflux Condenser Sizing - Packed Columns". Here is that thread:

viewtopic.php?t=92819

In this post, I discuss the condenser studies for stripping runs and show that this calculator does a good job of estimating these operations.


So, below is a summary table of most of the pot stripping runs used to evaluate the calculator. The vapor flow rate out of the pot depends on the pot watts, the pot abv% and the heat loss % of the still. There is a correlation in the calculator that predicts the vapor flow rate.

In the first five cases of the table ( Yummyrun (4) and OtisT ) the distillate rate was measured so we can get the heat loss values. The heat loss for the Yummyrum cases were about 10% and for OtisT, about 22% heat loss. A value of 10% heat loss was used for most of the other cases.

The important thing about the table is the column where the Required Heat Transfer is compared to the Calculated Condenser Heat Transfer. ie RHT=5.4 vs CCHT=5.6 for the Yummyrum first case

In all cases, the second number - Calculated Condenser Heat Transfer - is about equal or greater than the first number - Required Heat Transfer. This means the calculations adequately estimated the condenser operation and all the vapor could be condensed into distillate by the specified condenser.

Both shotgun and Dimroth condenser run data sets were studied and both gave adequate heat transfer condensing operation. Even a Liebig data set using water/steam met the data requirements.

So, when using the Condenser Calculator, the condenser design you specify will determine the Condenser Heat Transfer and if it is greater than the Required Heat Transfer, your condenser design should be ok.

------------------------------------------------------------------------------------------------

But, heat loss % is an important variable in a strip or reflux column operation and needed for these condenser calculations. A rule of thumb for the heat loss to use might be:

Use 10% heat loss for very well insulated pots and columns
Use 25% or 30% for average insulated pots and columns
Use 38% or 40% for not well insulated pots and columns
You can also run several heat loss cases to see the range of results the calculator predicts.

And --------------------------------------------------------------------------------------------
Some more heat loss examples that come from a question from Bunny that I had not answered well a (long) time ago but can answer now.

Re: LM temperature photo essay
Post by bunny » Thu Mar 30, 2023 12:27 pm

If I run 40% through an empty 1.87" riser at 1003w with zero reflux I collect less than 3l/hr.

If I run 40% through 56" manu's spp in my 1.87" column at 1003w with zero reflux I collect 70+ml/minute. (4.2l+/hr) - {edit > after the column has been stabilized with reflux}

Today, I'm wrapping up another 28l spirit run.
I've come to another one of those thing I can't explain.
I took a zero reflux volume measurement between heads and hearts.
I captured 70ml in 1 minute through the packing and boiler temp 184*f.
A few minutes ago I took another zero reflux volume measurement.
I captured 69ml in 1 minute through the packing and boiler temp 204*f.
I expected the volume to drop 20-40% with the much lower boiler alcohol content.
Could this have something to do with the fact my column is pretty much full of azeo?

Can you help me understand what I'm seeing?

OK.
First case is like a stripping of 40% abv charge from the pot. From the pot strip curves above ( July 4 ), at 1003 watts / 40% abv, the 0% heat loss vapor rate would be 4.1 L/hr. So 3/4.1 is 73% - so there is about 27% heat loss in that pot and riser still.

Second case is like a reflux column with reflux and the vapor at the top of the packing is at about 96% at all times up to the tails. The chart above ( July 14 ) for this case ( 1003 watts / 96% abv ) gives about 5 L/hr vapor flow rate at 0% heat loss. So, 4.2/5 is 84% - so there is about 16% heat loss in that reflux column.


Finally, yea
I am now satisfied with the new Condenser Calculator and will check with Uncle Jesse and try to include it in the HD Wiki Calculator Section for all to test, evaluate and use. I hope it helps with your evaluation and sizing of new and current condensers and determining what to expect from your operating set points.

haggy

[haggy]

I found a post with complete stripping run data using a shotgun product condenser. It shows that there are heat losses in these stills. Also, the run data confirms the calcs in the New Condenser Calculator. So, I want to report this. Here is the post:

Re: EZ Solder 2" Shotgun
Post by Crabmanstyle » Sun Nov 22, 2020 3:28 am

I've a 2" x 24" shotgun with 5 1/2" tubes and with 4kW power and maybe 1/2 liter per minute cooling I'm getting 1 litre ever 10 minutes on a strip and it's coming nowhere close to its full potential. I'm restricted by my power supply only.

So, at 4000 watts, he stripped a wash and got 6 L/hr product rate. It takes about 2562 watts to condense that rate. The heat losses are (4000-2562)/40000 or about 36%. Real data.

The New Condenser Calculator was run with this data and predicted close to the 6 L/hr strip rate. Also. about 535 ml/min cw flow rate to the shotgun condenser gave enough heat transfer to condense all the distillate product. Here is the calculator run:

Input Data

Results

This is a good representation of the run data. So, you might ask what the maximum watts the condenser can do. Can it do 11,000 watts? Yes. the calculator did a search and found that about a 2.6 L/min cw flow rate could knock that down. You would run the cw flow a little higher ( about 3 L/min ) to cool the condensed liquid. Here is that run:

Input Data

Results

This was one more example showing that the New Condenser Calculator is good enough for the HD wiki.

haggy

[kiwi Bruce]

Bang up job here haggy, you've put a lot of effort into this and it's appreciated!

[tommysb]

Hi Haggy,

Thanks for this, it's brilliant!

Are you able to give any reference or information about the type of model that's being used to do these calculations? How would one go about working something like this out?

Cheers

[Salt Must Flow]

Haggy, I'm a little confused about the input data in every instance in your last post. It was testing a Shotgun Condenser, but in each instance you selected 1 for Reflux in the cell titled "Decide On Product Or Reflux Condenser ... 0=Product 1=Reflux". Why is that? I assumed you select 0 for 'Product' as in a 'product condenser' or 1 for 'Reflux' as in a 'reflux condenser'. Does it make a difference either way or did I misunderstand the calculator?

Also does 'Temperature Out Of Condenser' mean the exiting product temp? That's what I assumed it means.

[haggy]

Salt Must Flow,

You asked:
I'm a little confused about the input data in every instance in your last post. It was testing a Shotgun Condenser, but in each instance you selected 1 for Reflux in the cell titled "Decide On Product Or Reflux Condenser ... 0=Product 1=Reflux". Why is that? I assumed you select 0 for 'Product' as in a 'product condenser' or 1 for 'Reflux' as in a 'reflux condenser'. Does it make a difference either way or did I misunderstand the calculator?

OK, you are asking about the Reflux Column Operation - and the numbers that follow that decision. Yes, it does make a difference either way.

First Reflux Option was Product or Reflux Condenser ( 0 or 1 ) - Why is that?
You assumed correct - There are calcs for either a Product Condenser or a Reflux ( Deflag ) Condenser. You choose which one you want to evaluate.

So, if you choose a Product Condenser, ( 0 ) the product rate specified will be used in the heat transfer calcs for a Product Condenser with the design given above.

If you choose a Reflux Condenser ( 1 ), the program calculates the vapor rate out the column and into the condenser - then subtracts the specified product rate from it to get the reflux flow rate and calcs the Heat Transfer Required to cool this reflux flow rate. So, you are evaluating a Reflux Condenser with the design given above.

And, yes that is about the Product Temp out of the condenser or close to the inlet temp if there is no subcooling. The temp will be lower, down to 25C, if the Condenser Heat Transfer is greater than the Heat Transfer Required.

And remember, for a deflag reflux condenser operation the Heat Transfer Required must about equal the Condenser Heat Transfer. The Search Option finds the cw flow rate to do this for you. If you get a NaN, the cw flow was too low, start the Search with a higher cw flow.

Haggy

[haggy]

tommysb,

Thanks for the compliment. I hope others can use it as well as you.

You asked:
Are you able to give any reference or information about the type of model that's being used to do these calculations? How would one go about working something like this out?

Use a Google browser, type in - Condenser Heat Transfer Calculation
several references there give the basic method and equations for this

Good reference websites are:
MyEngineeringTools - the equations there will scare you
The Engineers ToolBox - mostly basic info - physical properties - guidelines
You can get a lot of info from these two

The Condenser Heat Transfer is the standard equation Hc = U * A * LMDT
U - Overall Heat Transfer Coefficient BTU/hr/ft^2/"F
A - Surface area of the tubes ft^2
LMDT - log mean temperature difference of the vapor and cooling water F

I am old school and was brought up on these units, not metric units. The U is what you have to find a correlation for that fits real data.

To determine U for a tubular heat exchanger, you need correlation equations for the tube inside, hi, and shell outside, ho, heat transfer coefficients. Most shotguns in HD have condensing alcohol and water vapor inside the tubes and cooling water through the shell. A Dimroth is the opposite. My hi and ho equations are similar to those scary ones in the reference above. You combine the hi and ho together with a fouling factor and metal conductivity factor to get the Overall Heat Transfer coefficient U.

I had done some work previously in the chemical process industry on heat exchangers, so I had basic info and some hi and ho model equations to start with. I modified and adapted those equations to model both HD shotgun and Dimroth condensers and by trial and error found the best fit to a lot of real run data, both PC data and Deflag condenser data. Several stillers supplied me with their good condenser data, and I searched HD a lot and found posts with useful condenser run data.

Then there is the determination of the Heat Transfer Required - or the heat given off by the condensing vapor and some subcooling when applicable. That is the heat the condenser heat transfer calc must match for a deflagmater or be greater than for a VM reflux condenser or a PC. So, you need to calc the amount of alcohol vapor and water vapor condensed for the operation you are doing. That takes a while to figure out - you need real run data to construct equations that reliably calculate this. Several posts in this thread and my Reflux Condenser Sizing thread discuss this.

So, that is the way you work something like this out. Set aside a hell of a lot of computer time if you want to try it.

Haggy

[tommysb]

Thanks for the quick response Haggy -

I had recently been toying with something similar, but to work out how much a product condensor would heat incoming wash in a continuous.

In this case, could a simulator such as yours be used, but rather than using the heat capacity of water as the coolant, instead the heat capacity of an ethanol/water mix?

It's interesting to hear that there's what I would call a 'fudge factor' in there - e.g. we NEED some real-world data to make the model fit. Maybe this is needed if we look at using a different cooling medium that water, but then again maybe not...

Anyway, it's been really useful to me in trying to work out that portion of a continuous design!