Wondering about your experience with the small water heater as a still for methanol recovery ???



ReM

it works!

what would you like to know?

ps I've only done it a few times (to glycerol) in my system, and I'm in the process of re-doing my demonstration unit to optimise it for methanol recovery even further. More on that after I'm done with it, though, as I don't have much to report on the optimisation bit till I try it out...

the old one is at the veggieavenger.com/media forum- a topic started by Graydon Blair about my travel system. I don't have the link handy right this second...


mark

Looking for details--

Are you using vacuum ?

Measuring temps very carefully or just cooking it off until output drops ? (keeping below the boiling point of water)

Are treating it much like a water heater based BD reactor as far as valves/plumbing etc. ?

I hope I am not asking questions you have already answered but I don't remember ever seeing any details.

Tnx,

ReM

This message has been edited. Last edited by: ReM,

yes, yes, and yes...

mark

There are usually PID temperature controllers on E-bay for $10-30. They are wonderful for doing close temperature control experiments - and usually have a digital temp dispay as a bonus.

All that's needed to use one to controll a heating element is an air-conditioning contactor ($7 new) and a thermocouple with the right kind of wire to go with it.

A box might be nice, too.

The temp controllers "learn" the properties of the system they are hooked to after a few cycles up and down. They then apply just the right amount of heat to get and keep it at the setpoint temperature.

On my water heater processor, when the oil is cold, the element stays on 100 %. As it approaches the set point, it starts shutting off for a second or two at a time. Once it's fully up to temp, it only stays on for about 1 second out of 20 - just enough to counter heat loss.

Eric K

Mtushmoo,
Re:...

quote:

The temp controllers "learn" the properties of the system they are hooked to after a few cycles up and down

..

What exactly do you mean by this ? Most of the controllers that I have worked with require the set-point, ramping-up conditions etc to be programed in manually at the start. They then use these settings until you change them.
regards
dva

I've used the same sort of controllers Mtushmoo mentioned. Yes, they "learn" the best ramp rate after a couple cycles. The simplest learning algorithms adjust the heat cycle time constant to match the actual system, though most PID controllers seem to use slightly more complex ones. That prevents temperature overshoot as it approaches the setpoint. We use Omron PID controllers a lot, since they've proven more durable than A-B, and cheaper (about $100usd new). They can still have their parameters set manually, of course.

Johno,
Interesting, But surely you need to put in some initial setting so the controller knows, say, what your target temperature is. Else how does it know ? I used to use Eurotherm 2408 controllers with Pt100 platignum probes. Very versatile, but also very expensive. something like $700 by the time you have added relay and serial cards. When the company closed down I should have 'liberated a couple.
regards
dva

OK. I've got questions now.

I fixed my persistent vacuum leak that I've had since Iowa, and spent some time familiarising myself with the system which I've barely used. there are a few bugs that I'm trying to deal with.

one, I'm working with more glycerol than I'd previously used in previous batches (which makes the vacuum pump draw a REALLY good vacuum, too good for the temperature I was at...), and I developed the 'foaming over' problem. Anybody remember the recent discussion about this about Tilly's old experiments?> I'd like to read a link.

two, I'm working with more vacuum than I was able to draw previously (due to said leak) and that meant that my condensor no longer is sufficient for cooling and I'm getting quite a bit of methanol coming out of the pump';s (faraway, fortunately) vent line. Sam, what's the engineering on sizing condensors- you had a link at one point to a web page for ethanol, I think? presumably I need something that'd cover vacuum-assisted distillation, or distillation of things with lower boiling points than ethanol?

otherwise the vacuum-assist does wonders for pulling methanol out, though I"ve got a lot of problems with my small heater I'm using because I'm also circulating fluid when I do this process. the circulatino line is steel and it's a great conductor of heat. I'mworking with 1100 watts of 120V heat, which isn't enough to recover the lost heat. I think this wouldn't work for anyone who's doing it with lowpowered heaters at 120V (usually a 1500watt heater is what you'd be doing at 120V, whch isn't much more than I'm using), but there are a few ideas I'm going to try as part of my research, including inline heaters (part of what I was alluding to above in the 'optimising' bit)

thanks all,
mark

quote:

Originally posted by girl mark:
I'm working with more vacuum than I was able to draw previously (due to said leak) and that meant that my condensor no longer is sufficient for cooling and I'm getting quite a bit of methanol coming out of the pump';s (faraway, fortunately) vent line. Sam, what's the engineering on sizing condensors- you had a link at one point to a web page for ethanol, I think? presumably I need something that'd cover vacuum-assisted distillation, or distillation of things with lower boiling points than ethanol?

Here is a link to get you started:
http://homedistiller.org/cond_calc.htm

That is a handy condensor size calculator for ethanol. It makes a lot of assumptions, but fortunately for us, methanol behaves very similarly to ethanol, so those assumptions aren't too far off base.

In playing with it you'll find that there are a few ways to increase the cooling power of your condensor, and that all of the variables work together to give you the final output.

1. First things first, is the temperature of your coolant input lower than the boiling point of your methanol? If not, do not pass go, do not collect 200mL. Normally this isn't a problem, but when you pull a vacuum this can be a reality. The boiling point of a liquid is the temperature at which it's vapor pressure exceeds the pressure of the air around it. Imagine the liquid 'pushing up', and the air 'pushing down', and when the liquid can 'push up' harder than the air 'pushes down', then bam, it boils.

Check out this link: http://www.s-ohe.com/Methanol_cal.html

It is a shnixy calculator. You input a temperature, and it tells you the vapor pressure at that temperature. Its backwards though, you want to be able to input a temp, and get a pressure. But you can input a few random temps and get an idea of the pressures at which methanol will hit those boiling points.

For instance, the tap water in my house is a brisk 9 degrees C. The vapor pressure is 52.4mmHg at that temp, meaning that if that is the pressure, then my methanol is boiling at 9 C. However, that equates to about 27 inches Hg of vacuum (remember that vacuum measurements are negative numbers with atmospheric pressure being 0, and these vapor pressure values are positive numbers with absolute vacuum being 0), which is more vac than you can probably pull, hence it would be OK. But run the numbers for YOUR coolant, and see if it has what it takes.

2. Next, you have to make sure your condensor has either enough surface area, or enough coolant flow. You can trade one for the other.

If you had a geyser flowing your coolant your condensor could be 6" long. But since you probably don't have a geyeser, but a water pump, your condensor needs to be big enough to prevent you from having too much coolant flow to be practical.

For instance, if you can pull a 25" vacuum on your rig (giving you a methanol vapor temp of about 24C), and you have input water a 9C (like my tap), and can push a gallon a minute through a 1/2" condensor, then you'll need about 87" of condensor. Thats long!

However, if you only pull 14" of vacuum (giving your vapor a temp of about 49C), then the same tap water, the same pump, and the same 1/2" pipe can now be sufficient at only 30" long.

So you can lower your vacuum, make the condensor longer, lower the power of the heating element or increase your flow rate. Tweak the variables like a mad woman until you find something that is practical.

Remember folks, when talking about distillation, the size of your boiler is irrelevant, it could be 1 gallon or 1000 gallons, the relevant variable is the power of your heating element.

Also, and this is perhaps an 'optimizing' sort of thing, it is best for a distillation apparatus's heating element to be on all the time. If it is using a thermostat to click on and off it will through everything out of equilibrium. It should still work, but won't be as effective or predictable as if you can wire your heating element to be 'always on' through the distillation.

quote:

otherwise the vacuum-assist does wonders for pulling methanol out, though I"ve got a lot of problems with my small heater I'm using because I'm also circulating fluid when I do this process. the circulatino line is steel and it's a great conductor of heat. I'mworking with 1100 watts of 120V heat, which isn't enough to recover the lost heat. I think this wouldn't work for anyone who's doing it with lowpowered heaters at 120V (usually a 1500watt heater is what you'd be doing at 120V, whch isn't much more than I'm using), but there are a few ideas I'm going to try as part of my research, including inline heaters (part of what I was alluding to above in the 'optimising' bit)

Can you just insulate the pipe more? I'd be surprised if the pipe is loosing 1100W of heat... I'm thinking that 1500W is actually a pretty good number for distillation. Try running some numbers for a 4500W element, you need some serious cooling power for something like that!

Also, another note, that calculator is for simple jacketed condensors or simple 'coil in tub' condensors. I like jacketed instead of coils because they don't vapor lock, they are easier to make, and can have very effective cooling rates if the coolant flow is high enough.

peace,
sam

Sam,

By 'Jacketed' do you mean concentric tube type, with the coolant in an annulus and the methanol in the core, presumably in a counterflow arrangement?

Was thinking more about this and was wondering if an alternative control scheme might be to heat up at a pressure near 1 atm, then regulate vacuum to control the vaporization rate. If things are evaporating too fast (as in Mark's case), reduce the vacuum on the fly to slow the process down.

I found a decent pressure-enthalpy diagram for methanol but I haven't yet tried to post images.

Also, about how much methanol can one expect to recover. For example if I use 25% methoxide to make boid, 12.5% is used in the process, the other 12.5% ends up in the byproduct. How much of that can one expect to get back?

Horn

Yes, I've been reducing the vacuum to make it work better until I have it figured out better. Previously, it reduced itself through a vacuum leak it developed, a casualty of this being a trailer-mounted rig that suffered a lot of jolting in the last 2,000 miles with the vacuum system that I'd installed on my way home this fall.

Thanks, Sam, that's the info I was looking for. I'm using a bag (or two) of ice in a tub of water as coolant assist, but didn't have a cold enough thermometer to test what temps that was. I also don't have a temp gauge in the condensor , so I'm really limited in the monitoring. Soon to be fixed.

I'm really interested in working with hard vacuum.

I was going to increase cooling as follows: One problem is that currently I"ve got four and a half feet of vertical space to work within - the thing can't be 87" tall or it wont fit under bridges. And i dont' want to remove the condensor for travel. Anyway, in the future i"ll have a taller setup possible, but I'm experimenting with a short one for the trailer. It's fine for cooling at atmospheric pressures, but not as good under the huge amount of vacuum I"m able to generate on this tank.



More later,
Mark

quote:

Originally posted by Horn:
By 'Jacketed' do you mean concentric tube type, with the coolant in an annulus and the methanol in the core, presumably in a counterflow arrangement?

Yup, a "Liebig" configuration. I like them because you can knock them together in a few minutes from hardware store parts. I can post a diagram if people would like.

quote:

Was thinking more about this and was wondering if an alternative control scheme might be to heat up at a pressure near 1 atm, then regulate vacuum to control the vaporization rate. If things are evaporating too fast (as in Mark's case), reduce the vacuum on the fly to slow the process down.

You could probably get away with this in this circumstance, but normally you monitor the state of your distillation by the temperature at the top of the condensor, and correlate that to the boiling points of your various distillates. Changing the pressure on the fly would make that number go all over the place, and you'd have to be pretty handy with your calculator and phase diagram to keep up with it. However, since we really only have one primary distillate it would probably be OK.

quote:

I found a decent pressure-enthalpy diagram for methanol but I haven't yet tried to post images.

Please give it a shot, I'd like to see it.

quote:

Also, about how much methanol can one expect to recover. For example if I use 25% methoxide to make boid, 12.5% is used in the process, the other 12.5% ends up in the byproduct. How much of that can one expect to get back?

I think about 2-3% is in your biodiesel, 12ish% gets used in the reaction, and the remaining 10% (from your example) is in the byproduct. If you do a 100L batch, thats 10L, and you could probably reasonably collect 6-8L of that. More if you are vigorous, but definately that much. There are a lot of variables determining how much you can pull off, especially in a 'pot stil' configuration, instead of a refluxing still.

peace,
sam

Just to clear up one thing

Tests I once performed using 200ml methanol/ litre WVO showed that 2 to 3% of unwashed biodiesel was methanol and about 30% of the glycerine layer was methanol,.
HOWEVER, because there is much more biodiesel than byproduct, it worked out to 35% of recoverable methanol in the biodiesel and 65% in the glycerine layer.

This test showed a methanol use of 14%

Rev Tilly KE (Actor)

quote:

Originally posted by dva:
Johno,
Interesting, But surely you need to put in some initial setting so the controller knows, say, what your target temperature is. Else how does it know ? I used to use Eurotherm 2408 controllers with Pt100 platignum probes. Very versatile, but also very expensive. something like $700 by the time you have added relay and serial cards. When the company closed down I should have 'liberated a couple.
regards
dva

www.grainger.com item number 3xc07 is pretty close to what I'm using (Same controller, but different options). They're usually about $20-30 used on e-bay.

You feed it a temperature that you want to hit, and after a few cycles (usually overshooting wildly the first couple times) it "learns" the effective thermal mass and heating power, and modulates how much power it dumps in.

For a person witha reflux condensor trying to control teh top end of the column, I'd think it would work pretty good. Maybe use a pair and some common sense - control the boiler temp with one to some temperature where things start to come tp life, and then switch control to the top of the column...

Sam,

Here is the pressure - enthalpy diagram I found for methanol.

Horn

methanol_ph_diagram_01.gif (5 Kb, 144 downloads) Methtanol p-h diagram

Horn
thanks for the diagram

Sam
how long does it take to complete your methanol recovery? I am presently running 55 deg C tank, 45 deg C vapour temp and about 1.5 meter condenser this can handle 18" vacuum but the flow rate is to slow, i am begining to wonder if recovery of a few liters per batch is worth the trouble/time
cheers robroy

Hi Folks,

The theoretical discussion is fascinating, but
I'd like to request that those who feel they
are successfully distilling post pictures,
diagrams, parts lists, etc. Pointers or
instructions about basic vacuum pump function
would also be helpful (like the difference
between "oil-less" and "oil-full" pumps...???).

Thanks in advance.

Bill

I am currently building my first reactor (apple seed) useing the plans from girl mark's book. I do not see any specs or information on what adaptations are needed to perform methanol recovery. I have a good vacume pump but how much vacume would I apply to the water heater? How much heat? how do you capture the methanol while maintaining your vacume? what materials are used for the cooling circuit? I read about the need to maintain a constent temp. for good recovery, how are you doing this?

I haven't written anything up for either the book or for the 'net as I rarely do this practice and dont 'think I know enough about it yet.

But youre in the right thread to find the info= look at the photos here:
http://www.veggieavenger.com/avengerboard/viewtopic.php?t=505&sid=f6d8896cb2adf224115b354466b58908



Mark