StillDragon® Community Forum

Welcome!

Be part of our community & join our international next generation forum now!

In this Discussion

How many theoretical plates per bubble or ProCap plate?

edited April 2015 in General

Has anyone measured or estimated the number of theoretical plates for one physical bubble or ProCap plate?

One approach might be to compare how many physical plates it takes to get the same % ABV as with a certain amount of scrubbies or SPP (where the HETP is known).

«1

Comments

  • edited April 2015

    By definition, less than or equal to one. Equal to one would require 100% efficiency, which is highly unlikely.

    Actual trays = Theoretical trays / Actual tray efficiency

    So if you required 4 theoretical trays or stages, and your real world tray efficiency was 50%, you would require 8 actual trays.

    Keep in mind that the theoretical tray calculation takes into account the reflux ratio. So depending on the theoretical RR versus what you run in the real world, you could construct a scenario where you need fewer actual trays with higher reflux than you would need theoretical at a lower reflux (I hope that makes sense).

  • Not sure why you want to do this grim! Actual plates vs real plates? You want neutral? Use a packed column. For flavour use plates and pretend it is a sophisticated pot still. But hey, thats just my opinion. :D

  • edited April 2015

    The overall tray efficiency is a little bit of a red herring.

    In our world, a plate that can flow higher volumes of vapor and reflux at a lower overall efficiency will probably outperform a plate that has a higher efficiency, but can't pass the same volume of vapor and reflux. I'm defining performance as having a higher ABV output at a faster production rate.

    In the industrial setting, reflux ratio is important, because running higher levels of reflux ratio means higher energy costs, which are a big deal. Doubling the reflux ratio at the same product rates means twice as much energy consumed (this is a gross oversimplification, but I wanted to make it clear). Engineers would rather trade more plates for a lower overall reflux ratio (increased construction cost will provide lower operating costs).

    In our setting, the output ABV and the product quality is more important than the cost of energy, so we simply dial in our reflux ratio to hit our target output. Hell, most of us have no idea what reflux ratios we are running, we just dial it in.

  • So the reflux ratio isn't the setting on the controller knob? :))

  • edited April 2015

    There are many reasons one might want to trade plates vs. packing in hybrid columns -- available max height, power consumption, product flow rate, type of product, available materials, cost, and so on.

    I'm still hoping for some actual data. Bubble plates and ProCaps don't look like theoretical plates or trays, since they have considerable surface area and depth.

  • edited April 2015

    Yeah, and to make the comparison with theoretical plates even more complicated, anyone who is controlling their dephlegmator based on maintaining a set coolant temperature is running in Constant Product Mode, and not Constant Reflux Mode - which means the reflux ratio increases significantly from the start of the run to the end.

    This applies to packed columns as well, which means your HETP isn't a fixed number, and is getting smaller as you progress through the run and reflux ratio increases.

    Don't even bother trying to compare the results of a still running in Constant Product Mode with someone running in Constant Reflux Mode (controlling the dephlegmator based on the coolant flow rate - setting a valve and keeping it at the same place through the whole run). This makes comparing real-world a challenge.

    I understand wanting to bring height into the mix here, because that's a very real world constraint. But the math is easy.

    What's your HETP in comparison to the height between physical trays in the column? If your real-world HETP is 12 inches, you might find that trays will give you a shorter column. If your real world HETP is closer to 5 inches, than it might be moot.

  • HETP definitely has its flaws.

    Maybe I should ask the question in a different way: Using the same still and technique with regard to dephlegmator operation / technique / reflux ratio, has anyone been able to generate the same %ABV with bubble or pro-cap plates and with a packed column?

    Ultimately, I'm looking for a starting point for comparing trade-offs between bubble plates and packing. HETP is just one metric; there are many others.

  • edited April 2015

    Albright's Chemical Engineering Handbook states that typical tray efficiencies (with regards to HETP) are anywhere between 40% to 75% on average. This includes valve, dual flow, sieve, and the others. I would would argue the lower end of that is probably a bit too low for bubble caps, so say 50% to 75%.

    So, a bubble cap tray at N tray distance (TD) has a HETP anywhere from X to Y:

    3" TD =  4.0" to  6" HETP
    4" TD =  5.3" to  8" HETP
    5" TD =  6.7" to 10" HETP
    6" TD =  8.0" to 12" HETP
    7" TD =  9.3" to 14" HETP
    8" TD = 10.6" to 16" HETP
    

    Problem is, short tray heights/distances increase the potential for entrainment flooding and overall reduced efficiency due to carry over.

  • Thanks @grim; those seem like reasonable estimates. Do you have any real-world measurements to go along with them?

  • :)) Absolutely right grim. HETP is not a constant, it is a variable, and reflux ratio is the key. I may be wrong but by my understanding all the other factors are defined by the reflux ratio. That was why i was questioning the HETP value of a plate. It is variable.

  • edited April 2015

    I guess what I am trying to say is that real world, reflux ratio is a bigger factor than HETP or efficiency.

    Just running through a little Wolfram demonstration project - Rectifying an Alcohol Water Mixture with Variable Enthalpies - which is a simple McCabe Thiele.

    Rectifying an Alcohol-Water Mixture with Variable Enthalpies

    40% Input / Bottom Stage Composition

    92% Distillate Composition

    Reflux Ratio to Theoretical Trays (Equilibrium Stages)

    0.90 =  6.49
    0.85 =  6.99
    0.80 =  7.73
    0.75 =  8.76
    0.70 = 10.38
    0.65 = 13.10
    0.60 = 19.37
    0.55 = 66.59
    

    Now, real world, we run reflux ratios much larger than 0.90, and exactly why many people think that LM stills run better than CM stills (in reality, there is absolutely no difference when both are being run at the same reflux ratio).

  • HETP doesn't seem to change much once you get above 1.0.

    image

    th_hetp_vs_rr.gif
    381 x 300 - 4K
  • edited April 2015

    Check out this patent... might be usable by the home distiller

    An ~80% increase in efficiency (per plate w/ structured packing)

    found a comment about it on MD, while browsing a thread on plated column design

  • edited April 2015

    I've always thought that the graph you posted (from HD) is a bit odd, I think it would make more sense to show that graph in a linear scale and not with the Log10 axis (which makes it appear as if the RR is somewhat more linear than it is in reality).

    What actually happens is as reflux ratio drops, Total column height begins to get exponentially larger (see the table I posted above). If I continued that to .50, you'd see the number jump to the low 200s or so, then into the thousands. On the other side of the graph, the impact of increasing reflux ratio begins to get smaller and smaller as you approach the theoretical max - which is achieved when the still is running in 100% reflux (it would be roughly asymptotic). So, even that small difference in HETP can add up to a big difference in overall column height, especially if you are running lower reflux ratios. That left axis is measured in meters.

    (I know most of you are, but for folks that aren't familiar with McCabe Thiele, it's a way to estimate the number of theoretical stages you need to get from an input percentage to an output percentage (reverse actually, but saying it that way is confusing). It's nice because all you need to do is draw in the triangles, and each triangle represents a distillation stage. Once you run through it a few times, you'll see how brilliant those guys were. By the way, they did this 90 years ago.)

    For example, here is a typical McCabe Thiele for Ethanol and Water, it shows the 100% reflux line in red, and some level of reflux less than 100% in green. Posting this because some people look at a McCabe Thiele and think that the XY line represents typical operation, but it doesn't, it represents 100% reflux. To find the theoretical number of stages, you need to draw your operating line. It's too early to do math, so I just drew a random line representing some operating state (I cheated a little bit because it was easier to do this backwards, just for sake of discussion, not actual %):

    image

    Now, let's draw in the triangles to represent stages, in this case I'm going with 9 theoretical stages (because that's what I could easily draw) in the 100% reflux mode. That this means is the vapor leaving the top plate will correspond to about 87% liquid mole fraction. Now all of that vapor leaving the top plate needs to be condensed and sent right back down:

    image

    Ok, so thats the theoretical maximum we can get off the top. So now let's start collecting product, we dropping the reflux ratio to the green line I drew above, and start getting some distillate out. We check the parrot:

    image

    Holy crap, what happened? 9 theoretical plates was supposed to get me to 87%, I'm only collecting at 75%, what happened? God damn still is busted? No. Reflux ratio happened. If you want to increase from 75%, you'll need to increase the reflux ratio. Want to get close to 87%? You'll need to go up, and up, and up, and your still output is a tiny dribble in comparison to the monsoon in the column.

    So let's look at them together:

    image

    Now, move that green line up and down in your head and imagine the triangles shift. Line rotates up (lower reflux ratio) - you have lots of tiny triangles (less efficient separation stages) - as the line rotates down (towards the red line), the triangles get absolutely huge (theoretical max efficiency).

    All this and I don't even make neutral!

    RR1.jpg
    428 x 428 - 38K
    RR2.jpg
    428 x 428 - 38K
    RR3.jpg
    428 x 428 - 40K
    RR4.jpg
    428 x 428 - 41K
  • @grim thanks for that (and the wonderful graphs). I've never heard it explained that way -- very clear.

    So, if reflux ratio (the green line) can be kept constant between two configurations, it should be possible to use these curves to estimate the number of steps / triangles for each, based on %ABV, right?

  • edited April 2015

    Don't pay too much attention to the actual position my green line, it's technically in the wrong spot because I explained this completely backwards. I did this because it makes more sense to a distiller backwards - starting with the wash abv, and determining your output. However, an engineer would scream. Really, you need to start at your desired purity, draw the reflux line from that point, and work downwards based on how many stages you need to get to your input percentage. This would mean lots of trial and error.

    Lots of hobby websites omit the green line completely, and fail to add the assumption that using the red line assumes 100% reflux, which is nice for a theoretical discussion, but disappointing if you want something to drink.

  • @grim nice explanation. You can run short columns like the PDA 1, just very slowly with high reflux ratios. More height lets you increase speed at the same product quality.

  • edited April 2015

    @Myles said: More height lets you increase speed at the same product quality.

    Yeah, same with plates. You can see with the graph that lowering the reflux ratio (increasing product rate) with more plates lets you connect A to B (input to output) just the same.

  • edited April 2015

    Here is a graph of a theoretical 2 plate column (3 stages) producing the exact same output % as a theoretical 3 plate column (4 stages) with the only difference being reflux ratio:

    image

    And to take it one step further, here is a theoretical 5 plate column (6 stages), producing the same output % as the theoretical 2 and 3 plate columns, only difference is the reflux ratio (well, and the theoretical 5 plater should be much faster).

    image

    So in a world where a theoretical 2 plate column could hit the same output % as a theoretical 5 plate column, you see why I call "efficiency" a red herring, especially when so many of us have no idea what reflux ratios we are actually running. It's also why trying to compare two different columns is so difficult. If someone's city water temperature is colder than yours, but they have an inferior column design, you might be fooling into thinking their design is more efficient, when in reality, their inefficiency just being masked by a higher reflux ratio.

    The other thing to keep in mind is that red line isn't always achievable, an inferior column or plate design might flood (downcomer backup flooding or flooded packing) before you can reach it.

    RR5.jpg
    428 x 428 - 39K
    RR9.jpg
    428 x 428 - 40K
  • @grim thank you for this fantastic posts, graphs and explanations. ^:)^

  • While I understand the theory better now, my original question remains.

    Although there are complexities in comparing one type of column to another, people manage to do it all the time. Surely it's possible to compare bubble plates to scrubbies or SPP. If the term "HETP" is too loaded, then perhaps "number of plates vs. column height for equivalent %ABV at the same yield rate" -- or something like that.

  • If neutral is your aim the math is simple, packing wins hands down every time.
    Like most of these questions you need to start at the end and then work your way forward. So if I can better understand your goal I can better help.

    If all you want to know is what your expected %abv output will be by adding a plate you should find a few great posts here that already have some very well detailed studies. You already have the height details for each site glass tee, so you can work out your total height per ABV from that.

    This is all just academic...until you have a desired goal in mind.

  • Neutral is not my only goal.

    My goal is a setup that's flexible over a wide range of operating profiles, while fitting in the limited height I have available.

    I would like to be able to target a specific operating profile (power, %ABV, product type, yield rate, available materials, cost, etc), then quickly come up with a mix of plates and SPP that will meet that goal.

    A table of plates vs. %ABV is helpful, but by itself isn't really enough to accurately compare to packing, due to reflux rate and other issues, as @grim has so ably described above. Would be better to compare by holding as many other parameters constant as possible.

    It's starting to sound like I'll need to develop my own data...

  • A flexible rig to do both, sounds like your ideal solution would be a 4 plate column on top of a boiler with a secondary rectification column accessible via 3-way bypass valve. A common design, see Kothe, Carl etc.

    I run a 4 plate setup for flavoured spirits and when I want to make a neutral I install an extra plate and a 510mm packed section. I do that because it's easy with what I already have not because it's the best solution. I'd prefer a secondary stand-alone rectification column, I can't be arsed climbing ladders and pulling gear apart all the damn time.

  • Sounds like a good type of setup -- I'll probably end up heading in that direction at some stage.

    I have reasons for wanting to explore a much wider design space than just two configs at the moment, though. It's a straightforward thing to do by trial-and-error, of course -- just time-consuming.

    Not unlike finding just the best wash recipe, I guess.

  • @TheMechWarrior said: A flexible rig to do both, sounds like your ideal solution would be a 4 plate column on top of a boiler with a secondary rectification column accessible via 3-way bypass valve. A common design, see Kothe, Carl etc.

    I run a 4 plate setup for flavoured spirits and when I want to make a neutral I install an extra plate and a 510mm packed section. I do that because it's easy with what I already have not because it's the best solution. I'd prefer a secondary stand-alone rectification column, I can't be arsed climbing ladders and pulling gear apart all the damn time.

    This is the exact type of rig that I will be setting up. Least amount of changes for different runs.

  • edited April 2015

    Has anyone ever tried using TC pipe spools instead of the sightglass tees to reduce column height? Maybe only use a sightglass tee at the top and bottom.

    You should be able to decrease the bubble cap HETP to an extent. I suspect most could probably squeeze in an additional plate or two before they hit the roof. Probably save a couple bucks too.

    Won't look as pretty, but a packed column typically don't look pretty anyway.

    On a 4" column, maybe a 4" high spool, you could probably fit 10 plates on a keg before you hit the ceiling.

  • edited April 2015

    Depends on the ceiling...

    image

    Woodinville Whiskey

    image.jpg
    800 x 600 - 70K

    I'm more like I am now than I was before.

  • The biggest issue with consistency is not knowing your RR.

    Ace, with that VM head you posted elsewhere you could make up an RR meter for under 30 bucks using 3 temp sensors and a small arduino. You just need to change the plumbing to series. I even did up an equation to take into account product sub cooling at az although the effects are minimal.

Sign In or Register to comment.