Otis’ Vacuum Still Build

[OtisT]

Otis’ Vacuum Still Build

I’ve been thinking about enhancing my mini gin still so that it can also function as a vacuum still for low temp essential oil extraction. I want this to be an attachment module rather than a dedicated use still, so that I can remove all the vacuum components when I just want to run my mini as a simple pot or gin head still.

There are two big challenges I know I will have with this still.

The first is making this still vacuum tight. My desire for a modular design means I will have lots of temporary (union) and semi-permanent (threaded) joints. It’s hard enough making permanent components air tight, let alone threaded and union joints. I plan to start with simple/quick copper union connections and will address any vacuum loss issues as they arise. I also know from experience that the factory ferrule on my keg boiler does not always get the best seal, so that’s another point I will need to watch when attempting to maintain vacuum.

The second challenge I may have is maintaining a consistent vacuum level when distilling. I need the vacuum level consistent so that the boiling temp will also remain consistent. I think that the small volume of my mini-boiler and collection tank mean that pressure changes will come quick and be a challenge to hold steady. I will start this build with just the bare minimum components needed for the still and if this concern is realized, I have a plan B to address it. Plan B: If I can’t maintain a steady vacuum level during operation, and it’s not due to a vacuum leak, I will try adding a large buffer vacuum tank to the system.

Collection Tank
I decided to start my build by creating the collection tank. I started here because I thought this would be the toughest and most costly component of the still. After several roughed out plans I landed on a plan to use all purchased parts, no welding required.

I’m using a 3” stainless steel Tee as the tank. The useful capacity of this tank is ~1 liter.
* The center of the tee faces up, and has a 3” cap with 1/2” threads to accept collection input.
* One end of the Tee has a 3” process view sight glass that functions as the collection gauge, so I can see the amount collected from beginning to end.
* The other end of the tee is capped. This is my expansion port. If I need a larger collection tank I simply add a spool to the back end. An 8” spool would add another liter capacity to the tank.

Vacuum Tank view 1

Process view

I used two 3” pipe standoff’s and a few bits of wood to make a stable base for the tank.

Vacuum Tank side view

A big part of my decision to using these purchased parts as the tank is that I can also use them on other distillation equipment I run. That 3” Tee can be added above my 3” packed column using the 3” process view sight glass on the side. I’m super happy about that.

More work to follow

Otis

[OtisT]

And here is the plan I’m working off of.

The plan

[OtisT]

Boiler and Riser

My boiler for the vacuum system will be my new 10 liter mini keg with a 500 w controlled internal stick element. ( I hope to upgrade the element to 1000 w or 1500 w soon.)

The simple pot head on that boiler did not have a thermometer so I started my copper work there. The existing riser/head was too small to cut up and add the port, so I just made a new one that is a hair taller.

Old and New riser

Here is the boiler and riser complete and ready to use. As mentioned previously, I anticipate vacuum leak trouble from the keg’s 2” factory ferrule to threaded cap fitting joint. Right now the head connects to the boiler via a 2” to 3” conversion plate then a 3” cap with 1/2” FNPT threads. I have a 2” cap with 1/2” FNPT threads on backorder now. When I get that there will be one less connection to worry about leaking.

Complete boiler solution with temp/riser

Otis

[OtisT]

Vacuum Control Manifold

My complete vacuum distillation system will consist primarily of two large/heavy components. The boiler/riser is one and the collection Tank/manifold/condenser section is the other. Each is relatively heavy and I am using skinny 1/2 pipe, so I will need each section to stand free and hold its own weight. I will reply on a flexible connection for connecting the vapor path between the two sections. I plan to build straight up from the collection tank.

Building up from the collection tank, the next piece to put together for this system is the Vacuum Control Manifold. The manifold will allow refluxed liquid to pass through it from a union connection on top into the collection tank below, while also housing the vacuum gauge, air valve, and connector to the vacuum source to the side and elevated (to keep it dry). The manifold connects to the collection tank via a 1/2” NPT connection to a 3” ferrule cap.

Vacuum System Manifold

Plan change. My original plan showed a pressure releife valve on the manifold. I decided that the pressure relief valve would be more versatile if I placed it on the boiler or riser. By moving the PRV to the boiler, that safety device will be present for all the different distilling configurations, not just for vacuum operations.

As mentioned earlier, I am not planning to include a buffer tank at this time, though I think it is possible I will need it later. My original plan showed a port on the manifold for connecting the buffer tank. I have a planned design change if I need to add the buffer tank. If needed, I will add the buffer tank between my vacuum source and the manifold, connecting the two with a soft air hose. That means no mods would be needed to this manifold to add the buffer tank.

Otis

[OtisT]

Product Condenser

I want the collection tank to support all the the weight of the condenser, so I will mount the liebig condenser vertical above the manifold and connect with a union joint for rotational/positioning control. So this will be just like my other liebig, only with a union ball added to the tip.

This little guy is 3/4” over 1/2” and has a 13” jacket. I added a piece of copper wire spiraled inside the jacket for increased turbulence. Because this will be mounted vertical, I anticipate placing some loosely packed copper mesh inside the VP.

Vertical Liebig

So that I can use the liebig w/o tank, I added a small piece of copper wire wrapped around the bottom of the condenser fashioned into a hook. The hook holds the union nut up and away from the spout when I am using the condenser w/o the tank attached.

My previous liebig was almost identical to what I made for this unit. It was bit longer and had no union ball at the tip, but the cooling connections were the same 1/4 soft copper tubing. With only two new compression nuts, I was able to quickly swap out the brass hardware from that already plumbed condenser to this one.

Otis

[OtisT]

Connection Arm

I’m not sure the official name of this piece, but this connection arm is the last piece of copper work needed before testing the system for leaks. This dual union connection arm is similar to what I use to connect a pot to a Thumper, and allows me to connect the riser to the condenser at different heights. This height control combined with the union at the base of the condenser for rotational control, gives this still a lot of flexibility on position and elevation of equipment. The downside will be managing a good seal with three union joints.

Connector arm, riser to condenser

Otis

[OtisT]

Assembling the Vacuum Still

Time to put this all together.

All of the NPT connections were hand tightened with two wrenches, and used two wraps of PTFE tape.

There are three copper union joints in the assembly. Each joint included two wraps of PTFE tape on the threads as well as over the cup.

PTFE tape on union cup

Assembly of the collection system includes the tank, manifold and condenser.

Collection Section

The boiler/riser is placed next to the collection system.

Two free standing components

Boiler output and Collection input were aligned in parallel, and the spacing adjusted until the two union joints were in alignment and tightened down.

Assembled Vacuum Still

Time to step back, mix a drink, and prepare for the vacuum test.

Otis
.

[Oldvine Zin]

Looks great


Subscribed
OVZ

[OtisT]

Vacuum Test #1

Well the first vacuum test went way better than I expected, but still below what is needed. I vacuumed the system down -15 in mercury in. which is not a lot. There was a slow leak in the system somewhere that released half the vacuum in 13 minutes.

This picture shows how the vacuum pump attaches to the manifold.

Pump attached to manifold

The dial shown is at rest. I don’t know yet how close that is to being calibrated. In this first test I vacuumed that dial down to -26 in Hg.

Vacuum gauge at rest

tyhb

I’m not sure how to find a vacuum leak. I tried looking for leaks by brushing down each joint in the vacuumed system with soapy water and listening for any inside noises from sucking water through. No luck.

I do know that soapy water trick works for a pressurized system, so my next step is to find my ferrule with the bike tire stem on it so I can put this system under pressure. Soapy water on a leaky joint should bubble away.

Otis

[Oldvine Zin]

I do know that soapy water trick works for a pressurized system, so my next step is to find my ferrule with the bike tire stem on it so I can put this system under pressure. Soapy water on a leaky joint should bubble away.

Otis

OVZ

[shadylane]

Otis, You got my attention

[OtisT]

Bubbles

I could not find my premade ferrule pressure cap, so I removed the element from the boiler and added a brass bike tire stem valve.

Valve stem added for pressure test

My vacuum gauge has a hard stop peg so I was limited to the amount of pressure I could add with a bike pump. It was enough. I brushed down the entire rig in soapy water. The only leak, other than the one in the new pressure valve, was right where I expected trouble. The 2” keg factory ferrule. Ug!

Bubbles. Damn bubbles.

I’m out of time for lookin at this system now. Maybe I’ll get back to this in a few days. Otis

[jonnys_spirit]

Wow! This is awesome!

[RedwoodHillBilly]

Otis,

While I don't normally visit this site anymore, I read your thread and I have 1 concern about your design.
I would suggest that you either cool your collection vessel or add a cold trap (reflux condenser) in your vacuum line after your collection vessel. The collection vessel will act like a spirit boiler heated by ambient temperature. The boiling point of ethanol at full vacuum is 93F. I would be worried about the vapor pressure even at 70F. I don't think that you want to send flammable vapor into your vacuum pump.

Just my $0.02

[OtisT]

Otis,

While I don't normally visit this site anymore, I read your thread and I have 1 concern about your design.
I would suggest that you either cool your collection vessel or add a cold trap (reflux condenser) in your vacuum line after your collection vessel. The collection vessel will act like a spirit boiler heated by ambient temperature. The boiling point of ethanol at full vacuum is 93F. I would be worried about the vapor pressure even at 70F. I don't think that you want to send flammable vapor into your vacuum pump.

Just my $0.02

Thanks RH for the safety warning. It’s good to hear from you.

I saw the second condenser before the vacuum source on Mano’s system and decided not to worry about it for now, thinking it was for protecting the equipment more than for safety. I’ll rethink adding one now.

Otis

[OtisT]

I’m traveling today but still thinking about my new Vacuum Still and the work that remains. On my last flight, I came up with a simple solution to the concern raised by RedwoodHillbilly. He suggested that for safety reasons I needed a second condenser to “dry” the air going from my collection tank to the vacuum pump.

I really do not want to add another condenser hard connected to the stack of parts tied to the manifold. Even adding an independent condenser is not appealing to me, dreading the need for yet more cooling hose and possibly more space on the water manifold I use for all my mashing and stilling.

I started wondering how I could use my existing condenser for both purposes and think I have a simple solution. I’m going to wrap the exterior of my existing liebig’s cooling jacket with some copper tubing and solder that in place. Each end of the newly wrapped copper tube will get a compression to air hose thread fitting. I’ll use the existing soft air hose to connect the manifold to the condenser, and the other end of the condenser will be connected to the vacuum pump,with a second soft air hose.

The only remaining element to work out is what to do with any condensed liquid from the second condenser? I don’t want any condensation from the second condenser to find it’s way back into my collection tank. I’m thinking now that the vacuum source will be connected to the bottom of the new condenser (top of condenser to manifold) and I will need a liquid trap inline just before the pump?

Otis

[Sharks_n_danger]

Really interesting collection vessel. I’m excited to see where this build takes you.

I could be kissing something, but why wouldn’t you want any condensation from the second condenser to make it back into the collection tank if it is potentially originating there? It is a closed system so as long as it’s not making its way back from the vacuum source it shouldn’t be contaminated.

What if you used the out line from your condenser to feed a Bain Marie style pot (in picturing a flake stand style container) to keep your collection vessel cooled? Because it’s a vacuum still I wouldn’t expect the water to be heated up much, depending on what vacuum pressure you’re working at, though that may make viewing the level in the vessel difficult.

[Single Malt Yinzer]

Subbed. As always thanks for doing awesome stuff.

[OtisT]

I could be kissing something, but why wouldn’t you want any condensation from the second condenser to make it back into the collection tank if it is potentially originating there? It is a closed system so as long as it’s not making its way back from the vacuum source it shouldn’t be contaminated.

great question.

The materials in a soft air hose are not suitable for alcohol so any condensation would be flowing through that hose. Also the air valve and air fittings are brass. I don’t want that getting back into the collection.

What if you used the out line from your condenser to feed a Bain Marie style pot (in picturing a flake stand style container) to keep your collection vessel cooled? Because it’s a vacuum still I wouldn’t expect the water to be heated up much, depending on what vacuum pressure you’re working at, though that may make viewing the level in the vessel difficult.

I can’t do that. The water coming out of the condenser will be warm. The cooling vessel will be at room temp adjusted for the liquid temp and volume put in it. The reflux out should be nice and cold, so the vessel should remain at room temp with no help.

Otis

[OtisT]

Vacuum Line Cooling Added

I added a second condenser to the system to address a safety issue. The new condenser is in the air line between the collection vessel and vacuum pump, to dry the air and ensure no flammable vapors find their way into the vacuum pump.

The second condenser is 30” of 1/4” soft copper tubing wound and soldered to the outside of the system’s existing liebig product condenser cooling jacket.

Double condenser for my vacuum still

A soft air hose connects the manifold to the condenser, while the pump will connect to the other end of the new condenser.

Updated Vacuum Collection System

Otis

[Oldvine Zin]

Beautiful work there Otis

OVZ

[Sharks_n_danger]

I could be missing something, but why wouldn’t you want any condensation from the second condenser to make it back into the collection tank if it is potentially originating there? It is a closed system so as long as it’s not making its way back from the vacuum source it shouldn’t be contaminated.

great question.

The materials in a soft air hose are not suitable for alcohol so any condensation would be flowing through that hose. Also the air valve and air fittings are brass. I don’t want that getting back into the collection.

So I was missing something!

What if you used the out line from your condenser to feed a Bain Marie style pot (in picturing a flake stand style container) to keep your collection vessel cooled? Because it’s a vacuum still I wouldn’t expect the water to be heated up much, depending on what vacuum pressure you’re working at, though that may make viewing the level in the vessel difficult.

I can’t do that. The water coming out of the condenser will be warm. The cooling vessel will be at room temp adjusted for the liquid temp and volume put in it. The reflux out should be nice and cold, so the vessel should remain at room temp with no help.

Ya thinking about it more you wouldn’t be able to get a temp gradient on the Liebig if you were to keep the water in the flake cold enough to work if your ambient air temp was higher than your ethanol phase change temp.

Are you planning on devising a way to make cuts in the future with this setup?

[Chauncey]

Amazing work as always Otis. Bravo sir

[OtisT]

Sharks, you asked if I was considering an engineering change to allow for cuts.

Initially no, I intend to only have one collection vessel for oil extractions. I did consider something more complex but I decided to keep my first vacuum still simple until I learn more about Vacuum Still operations. The solution I was thinking of would be a bit spendy.

I recently started making gin and am learning that botanicals will concentrate in one section of a run. (Example, juniper is in the first few jars, Angelica comes over in the last few jars. Not a lot in the middle. If I can infer that I will see the same thing when a I try extracting various essential oils, I may want to add a mechanism for cuts after I have more experience.

This may all change by the time I get there, but my previous thoughts on how to do this would be to install a three way A/B valve after the product condenser that feeds one of two sealed collection vessels. Switch to vessel A to swap out the B vessel and visa versa. Some method of releasing and adding vacuum to the swapped vessel may be needed. I say “may” because I theorized that if my system volume was great enough, like by adding a vacuum buffer tank, adding a collection vessel at 1 atm would only have a minor impact on vacuum and it would allow me to use the existing system vacuum to readjust pressure after a swap. For collection vessels, I was thinking of two large Tees, each with a process view glass in the side, cap on the bottom, and collection input from a threaded cap on top. I would put a mason jar in each collection vessel that gets swapped out with each cut. Also, the tees would be filled with something solid that positions the jar directly in front of the sight glass and fills most of the unused space so the volume to vacuum is minimized. Maybe a wax filler so it fills most of the unused spaces and can easily be removed?

Otis

[Sharks_n_danger]

I like the idea of the tee with the mason jar inside, the wax filler is good thinking too, but what about the implosion risk of the mason jar? Or because the mason jars are designed for the vacuum of canning would a visual inspection for imperfections or compromises prior to use be enough? I’m just thinking if part of the jar is embedded in the wax, would the vacuum forces be unevenly distributed on the outside of the jar?

Also are you planning on somehow laminating or using a screen over your process view glasses in case of implosion?

[OtisT]

I like the idea of the tee with the mason jar inside, the wax filler is good thinking too, but what about the implosion risk of the mason jar? Or because the mason jars are designed for the vacuum of canning would a visual inspection for imperfections or compromises prior to use be enough? I’m just thinking if part of the jar is embedded in the wax, would the vacuum forces be unevenly distributed on the outside of the jar?

Also are you planning on somehow laminating or using a screen over your process view glasses in case of implosion?

The jar will be open at the top to collect product so there is no risk of the jar imploding (same pressure inside and out.) Also, the wax would not not be perfectly sized for the jar because I need to be able to easily swap out jars, so there is no risk of sealed air pockets between the glass jar and wax.

No, I don’t plan on a blast shield of any form. I may be naive, but I don’t think there is a big risk of implosion. For starters, I’m not taking this down to a full vacuum. I believe I’m likely to see a seal leaks long before that small thick piece of glass gives way. If I make a mistake and things implode, I’ll be sure to take pictures and will post about it here.

Otis

[OtisT]

Mind The Gap

I think the right thing to do is add the 4” ferrule to my mini. That said, I wanted to try the idea of filling the gap with my TIG and grinding it flat, for a better seal. Here it is.

The dreaded gap

The gap

Filled the gap with 316 SS fill rod

SS filler in the gaps

I ground off most of the mounds, hoping to rough finish with a nice set of files this size. I did not realize the ferrule is lower than the protective edge.

File is too big

So I ended up using stuff this size. PITA.

Small file

Finished with several small sheets of 360 grit wet sand paper on a small block.

Finished. No more gap.

There we have it. No more gap. Time to test that seal.

Otis

[OtisT]

That seal works now.

Passed both pressure and vacuum test.

Here it is with the new hoses in place for the vacuum air dryer. Red for hot vapor, blue for chilled.

Passed the vacuum test. :-)

I think I’m ready for a cleaning run, under vacuum.

[Sharks_n_danger]

That is awesome. Can’t wait to hear how it runs.

Probably more trouble than it’s worth on this scale, but there is a vacuum distillery that uses a pump from the collection vessel to a second collection jar so they can make their cuts without breaking vacuum. Though based on some of the other things I’ve seen, I’m not sure they have all the synthetics out of the vapour path.

[OtisT]

Vacuum Still - It’s not finished yet

I completed a cleaning run under vacuum but I have some engineering changes to make before I run it again.

Edit: . Previous posts where I wrote vacuum pressures I measured in psi, should have said In.Hg (inches mercury). Sorry for the error. 1 In.Hg = 0.5 psi

Vacuum Leak
I had a small vacuum leak again. 1 In.Hg every 4 to 5 minutes. It’s only a slow leak but I need to fix it before I can use this for any form of extended run. I learned you can’t effectively re-vacuum the system mid run, at least not as I have this designed. More on that later.

Air Condenser Upgrade
I need to look at this again. The small condenser I added for air is not even close to adequate. I went small, and should have gone the other direction. The vacuum pump pulls vapor fast, so to condense the air before it gets to the pump I need to slow the vapor down with a much larger volume condenser, or somehow slow down the air removal process, or both.

Maiden Voyage - Cleaning run and a few lessons learned

Vacuum leak aside, the run was successful as far as a cleaning run as well as teaching me a few things about operating a vacuum still.

Vacuum was consistent (leak aside) from cold startup to first drops. I thought there would be some fluctuations, but that was not the case.

I was able to collect 200 ml before loosing too much vacuum to continue. The 30% charge started boiling at 142 F @ -15 In.Hg. As I slowly lost vacuum I recorded that -14 In.Hg = 144 F, and -12 In.Hg = 150 F. I lost the boil below -12 In.Hg, which means my charge was approx 149 F.

First collection

Important Lesson
I learned that once you loose vacuum you can’t (easily) get it back. In my case the vacuum dropped to the point that boiling had stopped. I powered the still down to stop new vapor creation and tried to increase vacuum with the pump. Vacuum increased as air was pumped out, until the vacuum was just enough to allow boiling of my charge. Well, when you hit this level of vacuum your charge begins boiling again (creating a lot of vapor) and any air you remove with the vacuum pump is offset by new vapor being created by your boiling charge. I.e. your vacuum stays put, at the level necessary to start your heated charge boiling.

Hitting this vacuum level that starts your charge boiling also means that a lot of hot vapor is flowing very fast all the way from the boiler to the vacuum pump, and all the parts in between get hot, fast, and wet vapor and liquid reflux is pulled throughout the system and into the pump. Not good.

Moving forward
* I need to find/fix that damn leak
* I need to increase vacuum in my next test to reach a lower boiling temp
* I need to re-engineer the vacuum system from air condenser to pump
* I should put a thermometer in the boiler charge for both improved operations and safety. This seems like a necessity considering that a change in vacuum can cause a boil to stop/start instantly.
* I could benefit from an instant read thermometer in the head. Likely not as important for steady operation of the system, but when making adjustments it would be great having an instant readout of vapor temp.

The journey continues......

Otis