Building a Continuous Still for Stripping a Sugar Wash

[badbird]

he must be adding Plates based solely on the ABV change

Flyingdutchman, I assume that's whats happening,
It all a bit strange though as the packed column above the mid point wash input is well insulated so you would expect reflux to be minimal and according to rads wash / vapor ABV graphs injecting wash at the top of the stripper should give ~55% ABV product. I got nothing close to that.

Lester, it will be interesting to see if you get similar results with different packing etc.

Drop about a foot length of the solder wire you are using into 30cc of HCl

Will have to try the dissolved solder flux. I have previously used just straight HCL with reasonable results.
You are right, the key to soldering SS is tinning and keeping the heat away from the SS, Being used to copper i was surprised just how badly it conducts heat on the first couple of tries.

[Lester]

Will have to try the dissolved solder flux.

I have first tried electroplating the SS surface with a battery + solder (tin) anode. This works well. Then I realized that the 2 dissimilar metals plus acid is already a battery so I laid some solder on the SS and placed a few drops of HCl (it's now a shorted battery). This works well also. Since the tin dissolves in HCl then there will be a lot of free tin ions floating in the liquid so I realized I just need to dissolve some tin in HCl and use it.

Wait about 60 seconds before you apply heat.

I'm gonna get some SS scrubbies today so I can test this stripper and see what numbers I get.

Cleaning runs will be done after I have arrived at the final configuration, so it's quick and dirty for now.

[rad14701]

Flyingdutchman, I assume that's whats happening,
It all a bit strange though as the packed column above the mid point wash input is well insulated so you would expect reflux to be minimal and according to rads wash / vapor ABV graphs injecting wash at the top of the stripper should give ~55% ABV product. I got nothing close to that.

If you aren't getting the right amount of alcohol out the top then it's going out the waste port at the bottom, obviously... But it's more likely that you're getting a lot more water in the output due to the top feed... I don't think I'd go higher than 1/3 - 1/2 up from the bottom... The suspended solids in the raw wash will gradually contaminate the structured packing but if it enters closer to the bottom the greater amount of water will help slow the process... Food for thought...

[Lester]

I got my scrubbies and so I ran the stripper today, 3 liters of wash.

Good news: The column is working and I got 200ml distillate with SG =0.913 = 60% ABV

Bad news: Most of the wash ended up in the drain. The boiler overflow puts out too much overflow. With the scrubbies in place the boiler is now slightly pressurized and this pressure pushes the wash out the drain, almost running my boiler dry. I could not maintain the boiler temp yet, I still need to work on the overflow. Perhaps a valve is needed in place of the passive overflow that I used, so I can do periodic dumping of the spent wash. I did put an anti-siphon vent on the overflow so this is not siphon effect.

There's no point in recording temperatures at this time since I have not yet attained a stable run.

More work ahead but I have to be out of town for a couple of days. Will continue when I get back. Meanwhile, any bright ideas from you guys?

[Jacksonbrown]

Perhaps your packing is a bit tight?!? There shouldn’t be excessive pressure building up I would think.
What about a normally-closed float valve to regulate liquid level? You could build up lots of pressure then (if you wanted to).

I’m a bit unsure of your design though. Over pressure should vent vapour not drain liquid?!? Do you have a sketch of your overflow design?

[Lester]

Perhaps your packing is a bit tight?!? There shouldn’t be excessive pressure building up I would think.
Perhaps a normally-closed float valve to regulate liquid level. You could build up lots of pressure then (I you wanted to).

I’m a bit unsure of your design though. Over pressure should vent vapour not drain liquid?!? Do you have a sketch of your overflow design?

My packing is loose, but column is tall, and there's a possibility of a fluidized bed so.......

Here's what I'm using, exactly as drawn on this page: http://homedistiller.org/equip/cont" onclick="window.open(this.href);return false;" rel="nofollow

But the end of the overflow tube goes all the way to the bottom of the pot so it gets wash at that point. After some time I saw vapor coming out of the anti-siphon tube (short vertical piece) so I know my liquid level is very low.

[Edwin Croissant]

I wonder how far down the column the wash actually travels.

Edwin,
As i understand it the number of plates utilized in stripping will always depend on the power input.
Increasing the ash input volume slowwwly until the boiler vapor temp starts to drop shows where that limit is.

For a 10% feed you need only one theoretical plate to go to 1% boiler concentration. I think that a good mix between wash and vapor is crucial for the efficiency. When the wash flows down the center of the packing and the steam flows up near the walls of the column it's not going to work. Maybe the kind of packing is important, to me SPP looks like small tubes that can divert the steam and liquid in many directions. Thank you for sharing your experience, this is worthwhile information

Edwin I am not sure why, based on the collection rates vs what I saw with my column using no forced reflux, but he must be adding Plates based solely on the ABV change.

I think that when your condenser is not isolated from your column like Max_Vino did that you have some parasitic reflux due to the cooling through the column wall. Without some form of condensation I cannot understand how just packing could have any effect

For initial testing I will use an electric kettle with 1.3kw element. I can cut that power in half with a single diode, so 650w for continuous operation.

The Compleat Distiller mentiones on page 47 that a column with an internal diameter of 36 mm can handle up to 750 W. With 1.3 kW you have about 0,9 m/sec vapor flow. If you didn't put the diode in (I like these simple ideas ) this could be the cause of the excessive pressure.

[Jacksonbrown]

Perhaps your packing is a bit tight?!? There shouldn’t be excessive pressure building up I would think.
Perhaps a normally-closed float valve to regulate liquid level. You could build up lots of pressure then (I you wanted to).

I’m a bit unsure of your design though. Over pressure should vent vapour not drain liquid?!? Do you have a sketch of your overflow design?

My packing is loose, but column is tall, and there's a possibility of a fluidized bed so.......

Here's what I'm using, exactly as drawn on this page: http://homedistiller.org/equip/cont" onclick="window.open(this.href);return false;" rel="nofollow

But the end of the overflow tube goes all the way to the bottom of the pot so it gets wash at that point. After some time I saw vapor coming out of the anti-siphon tube (short vertical piece) so I know my liquid level is very low.

It sounds like the liquid in the column is adding head the boiler pressure.

If that’s the case I started thinking about putting the anti siphon vent on the inside of the boiler and adding an air trap to the bottom of the overflow outlet but this isn’t going to help you with a pressureised boiler unless that air trap has enough back pressure to combat the boiler.

The simple solution would add some sort of sight glass/level gauge to the boiler and instead of a conventional over flow just hook up a hose to the drain port and adjust the height of the nozzle to balance out the pressure and give you the require liquid level.

[flyingdutchman]

Jacksonbrown wrote

It sounds like the liquid in the column is adding head the boiler pressure.

Yes I agree this is a very possible scenario. This is what I was alluding to earlier when I mentioned the possibility of column flooding. A batch distiller has a column that has to handle the mass of the vapor and resultant condensed vapor. A continuous still has all this plus the added mass of the wash entering the column. If we keep increasing that wash input rate of wash at some point for a given column diameter...

we encounter flooding as liquid backs up the column and fills all the void space in the packing bed. Poor disengagement between vapour and liquid (back mixing) reduces the separation efficiency, and the high liquid hold-up in the bed increases the pressure drop.

Infact I believe that one traditioinal method of measuring flooding was to indeed measure the pressure drop across the column (ie higher pressure in boiler). IDK if that is the case here
As DAD points out spp gets increased efficiency in the early stages of flood or semi fluidized bed but I would wager beyond that point all bets are off.

Lester Could I ask after you..
What concentration is the HCL (ie 1N 6N etc) and after you tin the stainless steel do you use normal plumbers flux to solder the copper to the pre-tinned stainless?

Edwin,
Agreed we need a mix of vapor and liquid and for them to interact to add TP. The packing is a place for the condensed vapor to reside as you point out. Ambient temps/passive cooling/parasitic cooling should allow this to happen (much more so in an uninsulated column ). Check out Young's text he describes it in multiple column designs including unpacked spiral designs. However the faster we run the column the less time the vapor and liquid has time to interact so collection rates would be crucial. That's what I meant when I said I felt the collection rates might be too high in badbird's column for this to have had a noticeable effect.

[badbird]

I don't think I'd go higher than 1/3 - 1/2 up from the bottom...

Rad, Halfway up the column worked for me giving ~55% ABV, was origination surprised that the alcahol / water ratio at the top of the column was not similar to the vapor space above a boiler given the same wash, any the stripping column is obviously not just a boiler. 70L of moderately cleared sugar wash is the most I have run through the column in one go, there was no noticeable fouling of the scoria, maybe a good water back flush after use is all that is needed ?

Cleaning runs will be done after I have arrived at the final configuration, so it's quick and dirty for now.

Lester, If you are not planing on consuming the product, i found using a piece of clear plastic tubing between the wash heater and the injection point to be quite revealing, at least it explained some of the erratic column temp fluctuation i was getting before dealing with the vapor locks !!! Are you using a pump to feed wash?

I have the same issue with the back pressure. Next time before i run the stripper i will move my spent wash pickup from the bottom of the boiler (same as yours i think) to 2/3 of the way up (above the level of the electric element).
This will also raise the level of the overflow weir, the start up wash level will be very close to the packing but back pressure will lower it back to something more workable, i hope.

Maybe the kind of packing is important

Edwin, Raschig rings are used for stripper by some commercial operators, maybe less back pressure? defiantly easier to clean than SSP or scrubbers or scoria for that matter. One plate just doesn't sound like enough to get to 1%, but I'm a firm believer in overkill

[badbird]

That link Edwin gave to Max_Vino metering still has a interesting diagram of a cone shaped packing support and venturi arrangement at the bottom of the column that may help with some of the flooding issues Max_Vino diagran the Pro-pak brochure is also defiantly worth a read.
manu de hanoi mentioned (in an old post on AD?? i think) removing packing from certain sections of his stripping column to reduce flooding when developing his (old)continuous column still. Presumably he meant the column bottom packing.

[Lester]

..... instead of a conventional over flow just hook up a hose to the drain port and adjust the height of the nozzle to balance out the pressure and give you the require liquid level.

Lester, If you are not planing on consuming the product, i found using a piece of clear plastic tubing between the wash heater and the injection point to be quite revealing, at least it explained some of the erratic column temp fluctuation i was getting before dealing with the vapor locks !!!

Great inputs guys!

Are you using a pump to feed wash?

I am using gravity feed with the wash container some 2 feet above the feed point. I have no problems with vapor lock.

What concentration is the HCL (ie 1N 6N etc) and after you tin the stainless steel do you use normal plumbers flux to solder the copper to the pre-tinned stainless?

I used ordinary Muriatic Acid from the hardware store. Can't remember if it was 5% or 20%, senior moments, you know. The solder that I use is for electronics applications, 1mm diameter, lead free. It has some flux in the middle and I use it as it is. I think you will get good results even with just the "HCl plus solder" as flux. A few tries will get you the results you're looking for.

====================================================================================================================

I had a 20" length of the same SS tube so rather than cut the long one I used this instead. With this shorter column I no longer have problems with the overflow so that confirms the excess pressure is indeed due to the length of the first column which is 6 feet.

With the 20" column, the feed point at 8" from the top, I can get ~40% ABV from a 10% "wash" (foreshots & heads). Overflow is consistently at >99.5 deg C. What I did notice is that the column needed some time to get to 40% ABV, about 15 minutes after I started the wash flowing. I only had plain water in the boiler at the start and this delayed increase in ABV might be due to the water. It is easy to get the correct flow rate with this 20" column.

I did not bother to measure the flow rates as I was only interested in output ABV & seeing the overflow working at this time. I think I will add a little more column length and implement some of the great ideas above.

Thanks guys!

[Lester]

I thought I'd tell you folks that I was actually able to induce a fluidized bed on the 20" column. With the top cap off I was manually feeding plain water into the top of the column, trying to see if the overflow was still working with the packing in there. At some point I dumped in too much water and immediately I could hear the boiling water inside the column itself. This went on for some time until I dumped in way too much water and the column went "whooosh!". The boiling stopped right after that.

A fluidized bed will surely increase the pressure in the boiler and run the pot dry. And it is so easy to induce: just dump in a little more wash in a single shot. Column is 35mm inside diameter, 20' length, scrubbies for packing. With the insulation on it would be very difficult to detect if this has happened.....until after the pot runs dry & the element has fried itself.

Just saying.

[DAD300]

With SPP it is very easy to get and keep a state of fluidized bed in a reflux column.

When I first noticed it, the column was sizzling. Not like boiling, but a very distinct noise, like the packing was rattling. I could lower heat input and stop the sizzle. I thought it was going to induce a puke, but it never did. Then I learned to "ride" the flood. A fluidized bed with enough thermal mass could be a great stripper. Ethanol going up and water dropping out the bottom.

The noise went away by packing the column tighter. The potential puke went away (not the fluidized bed) by adding a loose Scrubbie to the top to spread the reflux.

Running the pot dry...that is why I think it is important that the bottom reservoir be only a waste collector and the exterior of the column be heated.

[Jacksonbrown]

To reduce your start up time you could add some valving to bypass the regen on start up or add a recirc to the whole system too. A bit hard on a gravity fed set up but there are a few options there.

[badbird]

Bring the waste overflow out of the boiler above the level of the element. The worst that will happen in an overpressure is a burps from the waste outlet until you reduce the wash feed rate, the element will still be covered at all times (assuming wash is being fed in)
After researching on the Russian sites (they are the experts on this stuff) i think my issues with vapor locks/surging is actually from Co2 being expelled from the wash while heating. Now just have to find a cure

[Lester]

I ran the 35mm x 20" column again, still using watered-down fores & heads as wash. Here's the summary of the run:

At this rate of power consumption it makes sense to go for electric heating.

I still experienced the fluidized bed 2x. Perhaps a 2" column will be more tolerant of feed rates?

A longer column (~36") will give me higher ABV.

I am pleased that I lost little alcohol down the drain.

I placed a funnel about 2' above the feed point so I can monitor the feed rate. I used a manually operated valve. I did not have any trouble with vapor lock since the "wash" did not have any CO2 to begin with. Maybe I will run into problems once I feed it with real wash.

[Lester]

Here's a thought folks........

If feeding the wash at some lower point along the column (instead of at the top of the packing) improves the output ABV of a stripping column.....

.... what do you think will the same scheme do to a regular reflux column?

I think we will get to azeo faster and have azeo longer but that's just my guess.

I can easily try this with my 2" x 4' column cuz there's an alternate feed point at 18" from the top, but I don't have a wash to run ATM. Soon.....

[badbird]

Good to see it working for you Lester. No doubt 2" will handle higher throughput, that's what i use but with a 4" section at the bottom just above the boiler. I'm not 100% sure that a fluidized bed is what you want in this application, in that build thread of Max_Vino's the diagram shows a choke point at the base of the column, I think on my next run Ill make up a cone shaped shroud and cone shaped packing retainer for the base of the column then see if that reduces flooding when pushing the limits.

[rad14701]

Here's a thought folks........

If feeding the wash at some lower point along the column (instead of at the top of the packing) improves the output ABV of a stripping column.....

.... what do you think will the same scheme do to a regular reflux column?

I think we will get to azeo faster and have azeo longer but that's just my guess.

I can easily try this with my 2" x 4' column cuz there's an alternate feed point at 18" from the top, but I don't have a wash to run ATM. Soon.....

As I see it, no, returning reflux lower in a standard reflux column will not have the same effect... The difference being that with a continuous column you are having "dirty" low %ABV wash entering the column whereas with a standard reflux column you are returning already distilled and relatively clean higher proof spirits... Having already distilled spirits returned as high up in the structured packing as possible gives it more chances to reboil... Feeding wash lower in the structured packing allows it to start the boiling process down where the temperature is hotter with the "dirty" components hopefully clinging to the water and dropping rather than contaminating the upper portion of the column where only clean spirits should be...

[Lester]

I think on my next run Ill make up a cone shaped shroud and cone shaped packing retainer for the base of the column then see if that reduces flooding when pushing the limits.

I think the cone-shaped shroud just by itself will do a great job of minimizing the "vena contracta" in the column. The same scheme is used on the mouth of carburetors (the velocity stack) for the same reason.

[Lester]

Here's a thought folks........

If feeding the wash at some lower point along the column (instead of at the top of the packing) improves the output ABV of a stripping column.....

.... what do you think will the same scheme do to a regular reflux column?

I think we will get to azeo faster and have azeo longer but that's just my guess.

I can easily try this with my 2" x 4' column cuz there's an alternate feed point at 18" from the top, but I don't have a wash to run ATM. Soon.....

As I see it, no, returning reflux lower in a standard reflux column will not have the same effect... The difference being that with a continuous column you are having "dirty" low %ABV wash entering the column whereas with a standard reflux column you are returning already distilled and relatively clean higher proof spirits... Having already distilled spirits returned as high up in the structured packing as possible gives it more chances to reboil... Feeding wash lower in the structured packing allows it to start the boiling process down where the temperature is hotter with the "dirty" components hopefully clinging to the water and dropping rather than contaminating the upper portion of the column where only clean spirits should be...

Right at the start of stabilization the vapor still contains a lot of water. The vapor is not yet "clean" and the vapor temperature bear witness to this. I think we will be helping the column do it's job by returning the reflux lower into the column rather than vaporizing the water again near the top of the packing. Anyway, I will try it out soon and report my observations here.

[DAD300]

Returning the "liquid" distilate to the bottom of a batch reflux column was long ago put to rest. It cause other problems...the return was outside the column and cooled (to some degree) before reinsertion. It cause power and flooding problems, as I recall.

[Lester]

Returning the "liquid" distilate to the bottom of a batch reflux column was long ago put to rest. It cause other problems...the return was outside the column and cooled (to some degree) before reinsertion. It cause power and flooding problems, as I recall.

How about 10" from the top? Has that been done also?

[Lester]

Some pics of my build. Won't be long now.

Vapor cooling via 2x liebigs 7/8" x 2.5'.

Wash preheat is done inside the liebig. 1/4" coiled tube goes inside the liebig and exits at the top.

Column is 4" x 40", SS scrubbies for packing.

[Lester]

Boiler & column. I still need to make the cooling tower.

[badbird]

Nice work Lester, like the coiled 1/4 inside the liebigs. Might try that myself one day when I acquire some more suitable copper.
Looks like it should be able to handle a lot of heat of heat, I assume by the bottle beside it means you will be running gas.

Are those air conditioning sight glasses or valves near the top of the liebigs?

[DAD300]

Lester, I praise you for putting torch to metal. Practical experience will triumph.

I have a thought on doing away with the boiler at the bottom and getting waste out.

Continuous Delph.png (11.25 KiB) Viewed 7570 times

The red portion is a common dephleg, but used as a heater rather than a cooler. The delphleg would be fed heated water at the right temp. Waste drops thru the dephleg and out.

The cascading plates are not meant to hold liquid, just slow it. This is the column close to what Bundaberg Rum uses, but the outside of the column is heated also.

I think this would be a lot simpler than heating the outside of the column.

[Lester]

thanks guys!

..... I assume by the bottle beside it means you will be running gas.

Are those air conditioning sight glasses or valves near the top of the liebigs?

Yes I will be using gas and yes those are airconditioning sight glasses. And the valve is from aircon supplies also.


Sorry dad but you can't use water to boil water. You're better off with pressurized steam but in that case no need for the dephlag just inject steam directly to the bottom of the column and you're good to go.

[Lester]

Initial trials, wash feed point at approximately the middle of the 4" x 40" column:

With wash preheat I only got 20% ABV at the output so I removed the pre-heating and routed the wash directly into the column, no other changes made. Output purity ~50% ABV.