Hail The Mighty Thumper

Distillation methods and improvements.

Moderator: Site Moderator

Mr Sippy
Bootlegger
Posts: 117
Joined: Wed Dec 28, 2016 3:54 pm

Re: Hail The Mighty Thumper

Post by Mr Sippy »

Hiya haggy

I think your proposition to have a pot/thumper calculator validated by user data is an excellent one. For me, the pot still calculator on the parent site is a handy format. I hope that you are provided additional run data as requested.

It is clear you've invested considerable effort. To reduce all those underlying equations to the succinct form provided is elegant, I think, though I don't quite see how you got there. I trust it doesn't sound patronizing that I hope my mind is as facile at a similar age.

I have made study and done some programming of the underlying equations for my own amusement. I see their implementation a kind of 'dashboard of state' usefulness. I'm sure I'm not the only one that would like a peek at your excel model.
Thank for your work. I hope it gets some traction here.

Regards
Be Kind, Tell the Truth, Sip Good Whiskey
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Thank you for the encouragement Mr Sippy.

The pot still calculator on the parent site is most useful and my calculations for the pot still agree with it very closely. Maybe we will get to the point where we can add a pot/thumper model to that site. My calculation format is a little different, I find the time to get to set incremental decreases in pot alcohol, rather than the decrease for a set time. Power is set initially in both cases. So should I change? I may look into that later.

The Kareltje data set is very complete and useful and has helped tremendously to gain confidence in the model. I found some data for larger pot/thumper combinations from several past posts but that data is not as complete. Got some good agreement there also. I should post some of those results.

Thanks again
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

Ha, I missed the most recent posts, must have been hiding on the second page of the active threads.
The fitting of your model is encouraging! Well done! :thumbup:

I now have a challenge for your model: a run with 10 L water in the boiler and 10 L fruitpulp of about 8 %ABV sloe.
In this run I lowered the fire under the kettle to prevent boiling dry and still create some turbulence in the fruitpulp. So I could light the fire under the thumper. In my model I could simulate that by setting the heat loss in the thumper at a negative value.
So we now have two strange settings: a kettle without alcohol and a fire under the thumper.
It proved to look uncanny for a setting with thumper: it looked as a simple stripping run. But on second thought: that it was, of course: just a simple stripping run. Only a strange way of heating.
Praktijk20-12-11-bijvat.xls
(90.5 KiB) Downloaded 61 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Here are the next two runs in the Kareltje data set.

I have updated the pot/thumper model and found a little bug. The previous calculations of cumulative %ABV from the thumper were all too high. Fixing that calculation gave closer values to the data. Run 9 is a conventional run with a larger pot, it was made in 2019, a later date than the other runs. Run 10 is similar, it was made in 2017. The model results were ok for both runs. Maybe product made late in the run needs to be updated in the model, we will check this out. Other runs have been ok.

Run 9 4/12/2019
This run had the pot start at 8 liter 10%ABV mash and the thumper at 2.25 liter 10%ABV mash. My watts calculation from the run time and 58% pot usage gave 510 watts and that was used for the power input. The thumper model results are in the next graphs. The tracking to the data was good except the model distillate product made towards the end was higher, and the %ABV a little high at the start. One low product sample volume could make the rest of the following data low also since it is cumulative.

The pot, thumper and distillate data at the end when added up came to 9.56 liters, not the 10.2 liters charged. So maybe there is some product not accounted for. The pot ended up with 4 liters, model had 3.5 liters. The thumper ended up with 3.4 liters, 3.4 liters in the model.

image.png
image.png
Run 10 10/6/2017
A different thumper was probably used in this run. The pot had 7 liters and 11% mash to start and the thumper had 2 liters and 11% mash also. A power of 600 watts fit the run time and was used. The model tracked the %ABV data well but gave a little higher distillate amount. One low distillate collection data during the run ( maybe at 120 minutes) could throw off the total collected the rest of the run. Both the data and model ended up with 3.3 liter thumper volume and the pot remaining volume was 3 liter data and 2.7 liter for the model. Cumulative %ABV is tracked better with the new calculation.
image.png
image.png
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Kareltje,

I just saw your last post. Fire under the thumper. I will look at a way to get power to the thumper. So both some steam injection and fire added. I have an idea and will try it

haggy
User avatar
LWTCS
Site Mod
Posts: 12837
Joined: Mon Jan 26, 2009 6:04 pm
Location: North Palm Beach

Re: Hail The Mighty Thumper

Post by LWTCS »

I think you should make a dynamic run speed calculator and donate it to the community.
You will be the hero of heroes.

Needs:
Volume / ABV / heat input = Run time
Trample the injured and hurdle the dead.
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

You mentioned the loss of product in former post. I had small losses in about every run I ran. Did not bother to find cause or reason.

But this is something that really puzzles me.
I ran a 10 L water and 10 L quincepulp one day and I got 5 L distillate, 2,9 L left in the boiler and 9,5 L left in the thumper. So I lost 2,6 L. Quite a lot, I think.
Of course: the mass balance rather than the volume balance should fit. But still: a gap of 14 % is too large.
Funny enough: subsequent runs with the same setup showed less loss. Down to about 5 %.
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Kareltje, The challenge.
Geen probleem, makkelijk.

From your spreadsheet, the run started with both pot (kettle)and thumper being fired, then at 263 minutes the fire to the pot was almost stopped and the thumper fire continued and boiled product over. At 340 minutes, the run was ended. At 263 minutes about 40% of the pot was boiled over, if some losses were put back in the pot.

The run start had 10 liters of water in the pot and 10 liters of 7.9% stuff in the thumper. So you did steam stripping and heating of the thumper at the same time up to 263 minutes. About 4.7 liters distillate resulted at 23 minutes.

Here is how I used the model to simulate this run. First calculated the power to the thumper knowing the time and % boil over. See above equations. This was 520 watts and I set that for up to 263 minutes. Then cut it drastically and sent only a very small amount of steam over to the thumper.

To model the fire power to the thumper, the following was done. Thumper outflow for each increment along the run path is obtained by calculating the latent heat (cal/gal) to condense the pot vapor times the amount transferred (gal ) for the increment and also calculate the latent heat to vaporize the expected richer alcohol outflow (cal/gal). Then the inflow calorie is divided by the outflow calorie/gal to give gal outflow maximum theoretical vapor gal that can result from the steam inflow. From examining many posted run data sets, I determined an energy efficiency factor and applied it to fit the data. That is in the existing model.

Latent heat of pot vapor condensed (cal/gal) * gal in * Efficiency factor = Latent heat vapor out (cal/gal) * gal vapor out

The thumper external heating effect was done like this. A set heating rate (calorie/min) was specified and that was mutiplied by the increment minutes and that (calorie) result was added to the incoming pot vapor ( latent heat * gal ) calorie data. So more heat was added to vaporize and get more product over. I then tried different heat amounts to add to the thumper and found that about 700 watts equivalent would result in the product amount of 4.7 liters at 263 minutes.

That was the only change made to the model up to 263 min run time. After that time, the pot watts were set near zero and the only (95%) heating of the thumper was by the specified external heating rate.

So, the outflow product rate was fit by varying the heat to the thumper. But %ABV resulted from the existing model calculations. The results are shown in the next graphs and track the product distillate %ABV very well. We have a steam stripper and a thumper external heater combined.
image.png

image.png
image.png
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

LWTCS wrote: Tue Dec 15, 2020 1:35 pm I think you should make a dynamic run speed calculator and donate it to the community.
You will be the hero of heroes.

Needs:
Volume / ABV / heat input = Run time
That is the plan. About 2/3 of the way there. I would like to publish a realistic model that covers many combinations of the pot/thumper operation. Can get that by studying and confirming more data on several different sizes and combinations.

I have reported on the model confirming most Kareltje runs and will report soon on several other larger pot and thumper size runs I have found in posts by other distillers, even though that data is not as complete. Kareltje collected several little jars (potje) during the run, got the volume and %ABV of each and recorded the starting run time of each jar.

You can help by getting your friends and yourself to send me some good run data. Initial conditions (kW if possible ) and distillate volume and ABV in little jars at several (5-10) run times. Plus pot and thumper volume at the end of the run.

Thanks for your interest
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Here are the last two runs in the Kareltje data set. They are a little different than the preceding runs. Run 11 had an external fire under the thumper the last 50 minutes and Run 12 had very high %ABV in the pot and nothing in the thumper to start. The same model parameters used in all the previous runs were used in these runs.

Run 11 1/23/2018
This run had 6 liters at 11% ABV starting in both the pot and thumper. At 144 minutes into the run the pot fire was stopped and the thumper external fire was started. I used 630 watts power for the pot and 800 watts for the thumper fire fit the data at the end of the run (144 to 210 minutes) . That heat added to the thumper was calculated the same way as the challenge post above. Charts of the run follow. The thumper ending volume calculated was higher than the data which saw a decrease at the end. The model tracked the distillate made and the %ABV ok, but not as close as other runs.
image.png
.
image.png

Run 12 2/20/2018
This run was very different, it had a high starting 34 %ABV in the pot and nothing in the thumper. Pot starting volume was 8.8 liters. The run was much slower with the power at 300 watts as calculated from the equation in an above post. Pot vapors condensed in the thumper and started filling it right away in the run. The run took over 7 hours to finish, ( go to < 5% product ABV).

The pot/thumper model did a very good job of predicting this run. Charts follow, I included the pot vapor %ABV , it tracked the data very well. The thumper product %ABV tracked very well and gave the right time when the ABV started to drop. This run with very high ABV starting in the pot gave constant product ABV in the mid 80's for about 4 hours into the run (2.6 liters made ). The model predicted this well. A good test of it.

We are gaining some confidence in the model. We now need run data at larger size pots and thumpers to confirm. I will document several simulations of other posted runs at larger sizes next.

image.png
image.png
image.png
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

I am very glad to find that your model runs well.
To be honest: I did not check my own model that thoroughly against my data. Just a raw fit suited me well. It helped me to understand the physics and mechanics.

As you said: more data would be nice.
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Thanks, Kareltje

I could not have done this without your data. I will report on more runs from others next. Please look those results over and give me your comments.

We will need your support if/when we put this model in the Forum Calculator section.
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

Saw this article, that might be of some interest to you. It was rather beyond my easy understanding.
https://www.academia.edu/26887304/Model ... view-paper
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Thanks, my stuff is not nearly as complicated as the article. I have some of the same equations, but only simpler and for the one stage ( thumper ) not a packed column. Article has 3 components, we only deal with alcohol and water. Maybe some conclusions are the same, I will try to digest them.

Article is a 10+ in complexity, my stuff is about a 6, but is adequate for this forum. In a case of zero reflux, closer to ours, they find an effective heating power to fit the data. I have done that also. Their sensitivity analysis is good. Says that their model can be simplified and that boiler heating and condenser cooling for a packed column have bigger effect on heart/tails cut time and volume collected but not too big on distillate strength. (figure 10). We can do sensitivity studies with my thumper model and get some valid results also.
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

I have made a format for the pot/thumper calculator that can be used in the site Calculator Section. Input data and options are made by the user, a calculator button is pressed and results in tabular form are given. An example of the proposed format is given below.

Here are the six input data, ranges, options and requirements;

Pot ------------- Gal and % ABV ----- Minimum 0.25 gal ----- %ABV 0.1% to 40%

Thumper ----- Gal and %ABV ------ Minimum 0.25 gal ----- % ABV 0% to 40%
Caution above Pot volume could result in less product recovery

Thumper ----- Heatup time ------- Normal range 5-30 min

Pot power ----- kW ---------- Usually from 10% to 40% of Pot gal

Thumper
Option Extra Reflux No or Yes - 0 or 1
Option Extra Condensing No or Yes - 0 or 1

The pot/thumper model takes the Heatup Time and finds an initial kW to get there. Then power kW from the input data is used for rest of the steady run and run time is calculated for each increment. One thing I do not do is set the time increments. The time to get to each result is given, it will not be in set multiples. It is still the correct expected time of the process. See the time in the format below.

Pot volume, vapor gal and %ABV, thumper Volume, distillate gal and %ABV and cumulative %ABV for each calculation increment are given in tabular form. Some other results are also given following the Table.

The format is given below. It is for a run done by Midnight Brewer in the Thumper? forum on Mar 3, 2018. I added the model comparison with his results in the boxed sections.
Book1-results.xlsx
(39.94 KiB) Downloaded 101 times
Comments on the format are welcome. I will present several more model vs data run comparisons getting data from past posts with 4 to 13 gal pot runs. The model calculations seem to be realistic and ok. We can then decide if we want to put this in the Calculator Section of the Forum. I do not anticipate changing the calculation method.
Last edited by haggy on Wed Dec 23, 2020 6:58 am, edited 1 time in total.
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

My very first reaction: so there is no possibility to use a thumper larger than the pot?
I do that sometimes and not only with water in the pot and slurry in the thumper.

Dowloaded your file, will study it. ;)
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

The program will handle larger thumper volumes than the pot. I thought distillers do not like to do this. I will make a run and confirm.
Thanks for the thought.

Steam stripping as you mentioned might use larger thumpers. With a mash in the pot, one could run the pot dry and have a problem. Would you put a limit on thumper volume?
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

I ran the model for a matrix of stripping run conditions with different Thumper to Pot ratios. The model works and gives valid answers for higher Thumper volume vs the Pot volume. See the Matrix spreadsheet below for the different conditions and details of the first case.

The results show that at the same ending %ABV for similar pot and Thumper starting ABV cases, the amount of alcohol recovered is worse as the T/P ratio increases. Not too bad at 1/1 at low starting alcohol, but is much worse as total alcohol feed increases.

This confirms a reason for not having thumper volume greater than pot volume. But we will still not limit thumper input volume, just caution users about it. There may be cases that are for larger thumper volumes.


Matrix.xlsx
(15.92 KiB) Downloaded 68 times
User avatar
Kareltje
Distiller
Posts: 2176
Joined: Tue Feb 09, 2016 4:29 pm

Re: Hail The Mighty Thumper

Post by Kareltje »

I never thought of it, but you are right. I.e. I find the same effect of larger losses with larger thumpers.

On second thought this is logical. When you stop running at drops of 10 % ABV the backset still has 1 % alcohol in it. That times the volume is your loss, so a larger thumperload results in a larger loss.
This goes not for the pot, as normally you run all the alcohol from pot to thumper and for a while even transfer almost only water.

I mostly use my thumper to steamstrip grain of fruit and then I run it to lower than 10 % or even to 0 %. Not for the alcohol, for that is cheap to produce and costly to collect at such low percentages, but for the taste.
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

I have searched many posts and found several that can be used to help verify this pot/thumper model. I will document these in my next few posts.

Beerswimmer has posted some details of several of his runs. I look at two of them now that were in the Thumper Talk forum. He has 15 gal kegs for the both pot and thumper so these runs will test the model on larger volume runs.

Stripping run posted on August 27, 2019. This run had 12.5 gal and 12% ABV for both pot and thumper. It was a 3 hr run and 6 gal of 40+%ABV were collected. I calculated from the 3 hr and 70% pot boiled over a 5 kW power to use. The model at the end of 3 hrs gave similar results as the data. See the attached Beerswimmer file for details of the model results. Here is a graph of the results.
image.png

Spirit Run posted on September 12, 2019. This run had 13 and 10 gal in the pot and thumper and 40% ABV low wines in both. This high starting ABV is a good test of the model. He reported a drip rate of 1 quart in 11 minutes and that was used to find the power of 2.1 kW for the run and therefore, the time of the run. He started off at 83% ABV, model gave 84% ok. He collected 7 gal, at 81-82%ABV . Model agreed with that. From the model, 7 gal were made at 330 minutes and the instantaneous %ABV dropped just after that. See the attached Beerswimmer file for details of the model results. Here is a graph of the results.


image.png
The Excel file is:
Beerswimmer.xlsx
(15.96 KiB) Downloaded 67 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Continuing with comparing model to posted run data.

Usge has posted some complete data on the How to Determine Thumper Size Forum, Feb 23 and 28, 2015.

He gave the volume collected, ABV and graphs of the timing of each sample for four runs. I look at two runs here and will do the other two next. The first two have 5 gal pot filled with two different starting alcohol levels. The thumper was filled to just above the inlet tube with water in both cases.

The first run has a 5 gal pot with 8.5% ABV mash to start and thumper with 0.5 gal water to start. I used both reflux and some initial cooling of the reflux to get a 72% start ABV, same as the data. A power of 660 watts gave the posted end run time. The first graph below gives the model compared to the data over the run time, a good fit for when the distillate ABV starts to turn down. Next graph is the ABV produced from the thumper vs the gal produced along the path and shows good agreement with the data.
image.png
image.png
The second run had the 5 gal pot filled with 40% fients and the same 0.5 gal water in the thumper. Starting ABV was 82% and 640 watts power was used in the model. With much more alcohol, the product ABV held steady over a longer time and started to drop after 5.5 hours. The same two graphs follow and indicate good results of the model fitting the data.
image.png
image.png
Details of both run results are given in the following Excel file:


Usge1.xlsx
(15.59 KiB) Downloaded 52 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

The next two Usge runs from the How to Determine Thumper Size Forum Feb 23, 2015 are here.

These two Usge runs had a larger keg rig and a larger thumper but the thumper was still only filled just above the down tube. The pot was 11.5 gal and filled with 8.5% wash in both run. The thumper was filled with water first and feints the next run.

For the first run with 11.5 gal of 8.5% wash in the pot, the model used 1.3 kW for the power to fit the time data. With 0.5 gal water in the thumper, a 72% ABV was produced initially in the model with some reflux and agreed with the data. Water in the thumper seems to influence the results somewhat and some adjustment is made in the model. The model time data fit is as good as previous data and has the right trend. The second graph of gal vs ABV for the thumper distillate has the model predicting a little lower ABV at several gal points, but not too bad.
image.png
image.png

The second Usge run had the same pot charge but had 40% feints in the 0.5 gal thumper charge. The pot power to fit the time data was 850 watts. Run time is a little longer due to lower power. The initial data %ABV was 82%, the model had about 80%. A similar trend in ABV vs time was shown by the model in the first graph below, but was a little different again in gal product vs ABV product as shown in the second graph.

The 40% thumper charge, even though the volume 0.5 gal was small, gave somewhat different results than the water only charge. This was seen in both the model and the data.
image.png
image.png
The model results are given in the following Excel file. The model was ok in all four Usge runs.
Usge2.xlsx
(14.71 KiB) Downloaded 59 times
Mr Sippy
Bootlegger
Posts: 117
Joined: Wed Dec 28, 2016 3:54 pm

Re: Hail The Mighty Thumper

Post by Mr Sippy »

haggy

Thank you for your continuing work. It is a generous gift of time and expertise to a perspective of distilling the forum rarely sees these days. I miss the (non working) potstill calculator V2 that is on the parent site. It was a good representation of the physical behavior of a still and your model seems to nicely fill that void. I think there is sometimes a bias against a quantitative view of distilling process because it can detract from the qualitative approach favored by so many. Personally, I think there is room for both without being mutually conflicting but rather as complimentary.

I hope you receive support from others for the inclusion of your model in a prominent location that can be easily found and accessed. I think your efforts represent a shining example of the sentiment expressed in Rule 13 and further hope they get recognized as such.

Best Regards

edit: comments above are in no way meant to diminish the time and efforts of the various moderators whose commitment make this forum what it is.
Be Kind, Tell the Truth, Sip Good Whiskey
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Continuing to compare the pot/thumper model to posted data.

The rest of the posted data I found does not give as much information as my previous posts. But we can still use it and compare one or two observations to the model. No time data is given, so I found a power kW to fit whatever %ABV data given. Mostly gallons made and their %ABV was given in the posted data.

Two runs are examined here. Runs by Moosemilk and Moonshiner67.

The Moosemilk post is in the Thumper? (17) June 28, 2016 forum. The pot start was 11 gal at 22% ABV of mix of low wines and mash. The thumper had 1 gal and 30% ABV from a previous strip. This was a slow spirit run. The initial %ABV was 80%, which the model with no initial reflux also calculated. He shut down at around 50% ABV coming out as tails were appearing and it was 3am. Volume collected was a little less than the pot output. The model did well in simulating this data. Here is a chart of the model ABV and results are in the file below.

image.png

I did not save and I could not find again the source of the Moonshiner67 post. The pot was 9.5 gal, the thumper at 2.5 gal, and both with 22% ABV low wines and mash at the start of this spirit run. No timing was given for the run so I set a power input reading to agree with the data. He did note two data points, the 70% ABV at one gal output and 40% ABV at two gals output. The model results were within 0.1 gal at theses two ABV points using the initial reflux option. See the chart and file below.


image.png

Here is the table of results:

TheMoos.xlsx
(15.65 KiB) Downloaded 60 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

More comparisons of the pot/thumper model with posted data. Runs by adamf and newengland are studied here. No power setting or run time data were given, I estimated the input power for the model.

The adamf run data was posted in Thumper? Sept 27, 2019. The pot started at 10 gal and 8% ABV, high volume and low ABV. The thumper was charged with 2 gal of 12% wash and feints mix. The model used extra cooling and reflux initially to agree with the low 1.5 gal produced at the 60% ABV mark. That was the only data point. A chart of the model distillate ABV vs gal produced follows.

image.png

The run by newengland was posted in Thumper? (8) on April 29 2013. The pot had 4 gal at 9.5% to start and the thumper was filled with water about 0.5 gal just above the down tube. The thumper was made from a Clawhammer 0.8 gal pot so extra rectification could initially happen in the column above the thumper. And this was seen by the high 83% starting ABV. The model predicted 74% since no extra plates are added.
The run was stopped when the thumper filled to 0.8 gal, this fill was seen in the model also. A thumper at 20% of the pot volume is too low. The run started back the next day. The ABV of the distillate was measured at several points along the run. After the high false initial start the model agreed with the data as shown next in the distillate ABV vs gal chart.

image.png

Here is the file for the model results along the run time for both cases.

adamf.xlsx
(16.01 KiB) Downloaded 61 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Here are model calculations and data comparison of runs made by Chauncy and by Yonder. Again, I did not have the time duration of the runs, so I set a power input for each.

Post by Chauncey » Sun Jul 12, 2020 11:57 am
Re: Max thumper charge in your experience for double keg strip setup

Here is his setup, an insulated keg thumper with one opening.
image.png
image.png (100.48 KiB) Viewed 2557 times
The pot charge was 13 gal at 8% UJSSM(1gen) wash and thumper at 6 gal of 13% UJJSM (2 gen) wash. He ran down to 20% ABV and got 3.5 gal of 43% ABV. The one hole insulated thumper probably did not have much initial reflux or cooling, so those options were chosen for the model. The model results were fairly close to the data using these options. A 69% initial ABV was calculated and about 3.7 gal at 49% cumulative ABV product. Here is graph of product ABV vs gal made, a gradual decline in ABV. See the file below for run details.

image.png



So, what if the run was made with a two opening non-insulated thumper where initial reflux and cooling is present. Something like this from Usge, in post Thumper? Dec 13, 2012:

image.png
image.png (128.5 KiB) Viewed 2557 times

We can use the model to get an idea of how this run would turn out if there were initial reflux and cooling These two options were input to the model. About 75% ABV distillate is made at the start of thumper boiling. At the 20% stop point, about 0.4 gal less is made, but at about 4% higher cumulative %ABV. Results are in the following graph and in the file below. Pot calculations have not changed with time but show different pot ABV values at the different thumper gal plotted along on the x - axis.

image.png



The run by Yonder is posted in Thumper Talk Aug 26, 2019. This run had about 10% wash in both the pot and thumper with 5,5 gal in the pot and about 1 gal in the thumper. Initial reflux and cooling were specified for the model. The starting ABV was high at 80% vs model 77% but the model agreed well with the 1 gal of product made at cumulative 63% ABV at the point when run stopped at about 35-40% ABV. A similar graph as above is next with the data point given. Model details are in the file attached.


image.png

Here is the file for all the model runs.

ChaunceyYonder.xlsx
(19.11 KiB) Downloaded 57 times
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Two items of interest in the pot/thumper model:

1. Of course, all thumpers are not the same. There are differences in heat losses, radiation, cooling, power efficiency, insulation, open volume, wind, etc. All this mainly leads to having different amounts of reflux and condensing in the thumper, especially at the start of the run before all the metal gets heated up.

That is why I have included two optional inputs, initial reflux and initial condensing. These are either on or off (0 or 1).

The reflux and condensing options are used in the first four time increments after the thumper starts boiling. Extra reflux increases the initial ABV and extra condensing reduces the output rate also and helps fill the thumper. In the input data of many of the cases I have shown above, there are different options chosen. I used the option which best fit the particular thumper being run. This would be done also when one models their own thumper to find the best fit to their data.

Some runs have even higher initial %ABV than the model, indicating more reflux is taking place than the model calculates. Reflux to a smaller extent is probably occurring during the entire run. The model seems to fit the rest of the distillate ABV data and account for this.


2. The pot and thumper alcohol concentrations (ABV) when the thumper starts boiling are very important and must be determined for the start of the thumper calculations.

The model first calculates the volume (gal) of pot vapor needed to condense and heat the thumper initial contents from 30 C to about 85 C , the presumed boiling point or within a few degrees of it. The alcohol content of both the pot and thumper are then calculated after this pot vapor has been transferred. These are then the initial conditions for the start of the thumper material balance, heat balance and VLE equations. At each calculation increment, the vapor amount and composition into and out of the thumper are calculated.

It is assumed that all the pot vapor is condensed in the thumper up to and after the start of boiling. This could be modified later, if necessary, by having a specified fraction of pot vapor blow through the thumper.
haggy
Swill Maker
Posts: 167
Joined: Thu Sep 10, 2020 5:05 am
Location: Pennsylvania

Re: Hail The Mighty Thumper

Post by haggy »

Mr. Sippy made a good suggestion to me. He and probably many others like to monitor the pot and thumper temperatures during the run. Are these temperatures estimated in the model?

Yes, I already had a pot and thumper temperature correlation in the model, just did not print it out, lot of numbers there already. So, I will add it to the results printout. The temperature correlation is just based on the alcohol content of the liquid, no pressure effect. Might be slight error in the correlation, but it should be good guideline. I will check it with the Kareltje data set and update it if needed.

The attached Temps.xlsx file gives results of a typical stripping run. The initial reflux option is used. The two temperatures are in the middle of the columns so they can be compared easily. Pot results are on the left and Thumper results on the right. You can see as the run goes on the two temperatures get closer together. Also, it is easy to see when the pot almost runs out of alcohol and you are steam stripping the thumper.

Comments are welcome. I am about ready to see if, how and when we can publish this model in the Calculator section.

Temps.xlsx
(12.98 KiB) Downloaded 58 times
Mr Sippy
Bootlegger
Posts: 117
Joined: Wed Dec 28, 2016 3:54 pm

Re: Hail The Mighty Thumper

Post by Mr Sippy »

I'd like to offer a few reasons on why I advocate for haggy's pot/thumper model.

It gives an instructive view of what's happening. Veterans may yawn at that but a novice can find it enlightening. New members are often directed to the Temp/ABV charts to explain the unsteady state of a still. The model expands on that notion. It demonstrates the equilibria, that vapor doesn't come off in the same concentration as the liquid and how it changes over time. It shows that the vapor is enriched by added reflux. The amount of distillate produced and the volume of both the pot and thumper over the run time are given. It shows how temp can be used indicate the end of a run is near. It shows how take-off diminishes over time and can be made to show a power regime to maintain a constant take-off rate. In short it illustrates all the concepts explained to new distillers and not in little bits over multiple threads.

The model can also provide insight into to 'what if' scenarios. Change an input parameter(s), see the resulting changes in behavior. Again, maybe old hat to veterans but instructive to to the learning curve. It can be a simulator or provide a schedule of expectation. Likewise instructive.

And please consider this: A man with 40 years of credentialed professional experience in a discipline central to the core of our hobby. He offers a calculator useful to a comprehensive understanding of the process, based on over 30 sets of user data. This kind of contribution doesn't just fall from the sky. It's understandable that some find it unnecessary but the forum accommodates many types of users including those trying to grasp fundamental concepts of thermodynamics and still behavior.

I would encourage that this model find a permanent place somewhere in the forum. I would encourage that other members, if they see a value in the model, to speak up. And I would respectfully submit to the moderators that they let haggy know, one way or the other, where they stand. To let this die a silent death just seems, well... unfair.
Be Kind, Tell the Truth, Sip Good Whiskey
Hügelwilli
Swill Maker
Posts: 312
Joined: Fri Feb 15, 2019 5:52 am

Re: Hail The Mighty Thumper

Post by Hügelwilli »

Hi haggy,
does your thumper calculator take into account the heat loss of the thumper? And if yes, how? I see, there is perhaps something in the "Thumper Options", but I don't understand it I think. Sorry, I don't know much about excel calcs. If I would know more, probably I could answer this question myself.
Post Reply