Proof of Concept Continuous Stripper

edited January 13 in General

Hi all,

After a lot of reading here and elsewhere as well as a few offline discussions, here is my 1st stab at how i can fairly quickly develop a proof of concept for a continuous stripper.

image

My objectives are to develop a low cost proof of concept largely utilising existing equipment to learn more on building and operating a continuous stripper as I am considering one for our upcoming distillery. I am looking to get 30-40% low wines out which will be loaded into a spirit still for a batch run to produce the final product.

I am a big fan of the kiss principle, eg keep the feed rate constant at a known temp, thereby allowing only a single control variable. You will note no wash preheaters again that's part of the proof of concept / Kiss which can be added later once I have some basics locked in.

For the POC, the wash heater will be my old boiler, the steam generator will be my existing keg boiler. The PID control loops are on my brewery control panel. I will need to make the wash heater return pipe, the steam generator waste overflow pipe as well as the sections for the wash injection and bottom temp monitoring.

The diagram only shows a packed section below the wash input as a 10% input has a vapour of 55%. Further plates above the input is likely to drive up the output ABV which is not what I want. i can always add plates or more packing below the input if I find I'm loosing alcohol out the bottom of the column.

My end goal is learn from this POC to develop a system to be able to process 2000l of 7-10% wash to produce around 500l of 30-40% low wines in 7 hours. i envisage that system would have wash preheaters, as well as a more instantaneous wash heater (a rims tube ?) and have a different steam source such as the one @TheMechWarrior has detailed previously. However at the moment I'm trying to put together the concept cheaply to learn.

I look forward to hearing your comments and suggestions eg anticipated throughput as I will be running a 4" column with a 6kW element in the steam generator.

Cheers
Crozdog

continuous stripper concept.jpg
585 x 537 - 28K

Comments

  • edited January 13

    I like the idea of using the column top vapor temperature in conjunction with the bottoms boiler, that's pretty clever if you can work out the PID parameters to get it to work smoothly.

    I also think this is a pretty good "deconstructed" stripper - as you would be able to manually control a number of parameters that are not easily controllable in a continuous rig using the condensers as preheaters.

    I would say skip the bottoms tank and instead replace it with a longer section of 4" or 6" triclamp spool/pipe, and have the element vertical. This is considerably cheaper than having to use any kind of "tank" at the bottom. Also much easier to trial without fabrication. I think there is a sweet spot with regards to reboiler volume. Also much easier to just use pipe clamps to attach this thing to a wall or frame, as opposed to trying to get it to be freestanding.

    If you are going to include a dephleg, I would imagine even just a small section of packing above the input feed would make a tremendous improvement to the run speed and efficiency. I think the key here is going to be using a very light touch on the dephleg. The ratios that I've typically seen are 1/3 to 1/5.

    I do not believe it's possible to use the standard liquid/vapor chart here to estimate your abv output based on your wash input in a continuous system like this, hence my comment on the dephleg and packing.

  • I really like the feed pump with return to tank. I was so focused on a peristaltic option for feeding that this elegant and cheaper method escaped me entirely. Great way to chop off $1500+ from the cost.

    My first thought was that I'd want at least two plates above the feed point. One could always reduce the reflux if desired.

    You have mirrored my plan to reuse my older 13gal boiler as the heat source. Again great cost savings.

    My internal fight with myself is getting to the point of admitting I'll need at least a 6" column to reach my goals (almost identical to your daily run figures but I want 70%+ and 500l wash per shift). The 4" setup is just so much easier and more affordable to get dialed in.

    I'm not quite grasping grims PID comment in execution terms. I can't quite put all the pieces of info together so that I could build it myself. I get the concept tho and it seems very simple/elegant.

  • I also think that your rims idea would be better than as drawn for the prototype. Heating the whole tank is inefficient and could cook off some alcohol in the process. Plus the recirc pump will have to deal with hot wash. Better to monitor rims output one flow rate is established. Then a pid would only become a safety shut off really as not much adjustment would be needed once flow was established.

  • Anyone have pid instructions?

  • edited January 13

    The other factor is wash distribution into the packing - it might be a good idea to inject the wash onto a perf plate.

    Would serve double duty - distribution and improved stripping efficiency.

    And this sucker is going to really need to scale up to gulp down a gallon of wash a minute.

  • edited January 14

    thanks for the positive feedback guys.

    @grim as i mentioned, I'm trying to do this on the cheap hence the use of my existing keg boiler under the column. When I get this sorted, I want to progress to using a steam generator like @telluride designed here

    The perf plate is a good idea to distribute the wash. regarding scaling up, i have talked with 1 guys who thinks a 4" will be able to do it - easier with a 6". I need to get back in touch with him to get some more details. ;)

    I drew the dephled in, but don't know if i'll need it. thought it may be necessary for initial stabilisation. dunno. probably want to keep reflux ration down so as to not push up the output ABV.

    @Fiji_Spirits I deliberately didn't add plates above the input as a few things i read showed that they pushed the abv up beyond what i want. The PID comment is all about how to get it dialled in - that's the big challenge as i see it. hoping autotune will help start the journey. I agree the rims tube preheater would be more efficient, however at this point I'm not worried, simply trying to work the basics out before spending coin. Instead of a rims, I'm also thinking of using an continuous instantaneous heater similar to @telluride's steam generator.

    Please keep the ideas flowing guys.

  • edited January 14

    2000l in an 8 hour shift with a 4"?

    No way, more like 6-8" x 20' - probably a 4"x1m or 6"x1m condenser.

    You are talking like 4-5 liters a minute feed rate and approximately 50-70 liters of distillate an hour (1 gallon minute feed, 15-20 gallons of distillate an hour).

  • edited January 14

    I talked to the guy that had the first generation ASD stripping column, and I believe he said it was 8" by about 20' total height (including dephleg and superstructure), and the speed wasn't really up to what he needed. The new design is very different, and actually very close to your design. I don't believe it has a dephleg or plates above the feed. I think it has about 20 plates below.

  • edited January 14

    And you don't need steam injection unless you are stripping wash with high solids content. In which case, the packing isn't going to fly anyway. Why bother with the additional complexity and phase change if the goal is just to add energy into the bottoms boiler? If your wash is semi-clean, using steam injection means a slightly lower overall efficiency as you are diluting the bottoms product.

    Just throwing out ideas.

  • cheers grim appreciate your thoughts.

    I'm just saying what he told me. he's done a bucket load of calks to back it up. I'm keen to see em. me, i'm more a try and see then scale from there.

    yeah the volumes are pretty big. agree that the action for my requirements is below the feed point. how much is needed is the question.

    Looking to use this with rum and clarified whisky wash, so solids shouldn't be an issue. fair point re the steam injection - also saves on the need to have to get the authorities to certify it which I imagine would be a major PITA.

  • I want to do a 4" to start, then do an 8"... but just too much other crap going on... just looking at stripping without having to buy and provide power to 2000L boiler

    My plan was to have all three condensers heating the wash, the product condenser, dephleg, and bottoms drain/overflow... the bottoms and product will do a good job of keeping those outputs at room temp, and the dephleg you would not use much...

    Might try needle valves at first, but definitely want to do the metered pumps...

    I added a tab to my 'cooling water required' document for the 3 condenser math, and started a visio drawing, but have not touched in 5 months..

    Cooling water required (XLSX)

    continuous still 2016 (PDF)

    xlsx
    xlsx
    Cooling water required.xlsx
    15K
    pdf
    pdf
    continuous still 2016.pdf
    113K
  • @crozdog would you mind uploading a larger file of your drawing? I'd love some more detail, but can't seem to get the text legible

  • pdf
    pdf
    continuous still thoughts.pdf
    103K
  • @crozdog won't it be difficult to keep your reflux ratio down if you feed your fresh wash at room temperature? The big temperature difference between the wash and the vapor will cause a big increase in the reflux ratio due to the effect of subcooling. The cold wash will knock down a lot of the vapor. With that effect you might not need such a big packed section, because it will increase the separation significantly.

  • Hi Anro thanks for the feedback.

    the plan is to feed preheated wash into the column at around 92-93*C to minimise the chance of column instability.

    The number of plates / packing below the feed is to ensure all alcohol is stripped before the bottoms fall into the boiler and exit.

  • I gave up using a PID for controlling a RIMS tube heater, the temperature of the output would go up and down by about 10 deg C while hunting for the target. I now use the heated water out of the reflux condenser to preheat the wash. I find this is much more reliable than using the waste and it is also close to the wash input so you don't use so much insulation. There are 2 bubble plates above the input and 6 pro-cap bubble plates below. I probably need 2 more below as there is still about 2% of alcohol in the waste.
    I have insulated the column with builders aluminium foil covered foam. This makes it much more stable and requires less power.
    The amount of reflux controls the final ABV and the number of plates below determines how much alcohol is recovered from the wash.
    Once heated up and balanced it will maintain an output set around 80% ABV for hours. This also means that a lot of the flavour components remain. A re-tweak is required when the evening sets in and everything cools down.
    For anti-foaming I have found that 150ml of spa anti-foam bought from Bunnings is all that is needed in 1000 litres of wine. (BTW, it's non toxic).
    There is only 150 litres of the 1000 litres of sauvignon blanc that I started with, so hopefully the final development will be done with that.

  • What is your collection rate and estimated feed rate Ed?

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • Hi Ed, have you tried it without the 2 plates above the input? I'm curious as to the output ABV with no rectification ie you're getting 80% with those 2 plates. Would it be 30-40 without?

  • At the moment the output rate is somewhere between 3 and 4 litres per hour. The feed rate is between 15 and 16 litres per hour. This varies with input wash temperature and amount of reflux. With an extra two ProCap bubble plates beneath, making 8 in all, most of the alcohol should be recovered.

    @crozdog, if I turn off the reflux condenser the output drops to around 10% ABV.

  • Ed How much power are you putting into the boiler? If you could put more in to evaporate the remaining 2% could the column handle it? (My plan was to have about 10kw ready to roll for boiler heat and push it as hard as the column could handle to get max flow rates using perf plates)

    Your flow rates are quite nice actually. I'm quite impressed.

    Curious as I've not heard anyone try but... why not do a bit of preheating with the RIMS but control it via analog potentiometer? Seems to me that once you got your flow rates settled that this would provide a consistent heat without hassle.

    Digital automation is all good but sometimes an ammeter, a big diameter knob on a pot, plus a pencil mark for the set point can be pretty dern repeatable and fast

    I seriously believe we can push a 4" to output 8L/hr of 80%+ with 8-10 plates

  • edited May 17

    Odd that your output falls to 10% abv if you turn off reflux. Not that I know anything, I don't, but I would imagine this means your stripping section, which should not need significant reflux to operate, is not very efficient.

    This would indicate you are running in a kind of hybrid mode (not completely continuous, as high residence time in the column is necessary for stripping), which provides some insight into your low feed/output rates.

  • Are you using the reflux condenser as the water pre-heat for the Rimms boiler. If you are this then could help explain it.

  • @Fiji_Spirits I have 2 x 3.6KW elements putting about 4KW into the 20 litre boiler.
    grim the stripping section is two 4 inch standard bubble plates below the reflux condenser. I reckon the ABV drops when I turn off the reflux condenser because all vapours are going over the top, not just the lighter ones.
    richard I was using the PID on from a temperature probe set just at the output of the RIMS tube. The wine was preheated by a heat exchanger on the waste line. I currently use a heat exchanger from the hot line out of the reflux condenser. I find this to be much more stable. In the earlier setup the wine was under pressure to go through a jet. Because the temperature wasn't stable and was also getting the occasional blocked jet I decided to forego that setup and changed to the current low pressure setup. I've just bought another 2 Pro-cap plates to help recover the 2% going out the waste.
    It's all in a constant of development. It's great fun seeing how there are so many ways to skin a cat.

Sign In or Register to comment.