As part of my chem-eng day job, I’ve got access to a chemical process simulator (Hysys). I’ve been looking at adding methanol recovery to my processor, and was using the simulator to calculate the boiling points of various concentration of the methanol with both biodiesel and glycerine. Pure methanol boils at 148F, but as a solution the boiling point will be higher. Thought I’d post it on here as it varies quite a bit with the temperatures that people seem to use in methanol recovery. Any ideas why? The agitation or spraying maybe?

I used 100% m-linoleate for the biodiesel composition. I believe this ester is the main constituent of soybean oil based biodiesel. The calculations are done at 14.696 psia using Antoine.

Biodiesel / Methanol Solution
% MeOH--- Boiling Point, F
5.00 ---------- 209
4.00 ---------- 219
3.00 ---------- 232
2.00 ---------- 254
1.00 ---------- 298
0.50 ---------- 351
0.25 ---------- 411
0.00 ---------- 532

Glycerol / Methanol Solution
% MeOH --- Boiling Point, F
45.00 --------- 172
40.00 --------- 177
35.00 --------- 182
30.00 --------- 189
25.00 --------- 197
20.00 --------- 207
15.00 --------- 222
10.00 --------- 244
5.00 ---------- 288
1.00 ---------- 416
0.00 ---------- 550

You BEAUTY!!

I have had some problems lately with exactly this sort of thing and can concur that if Graham is correct in assuming that a glyc solution utilising a 5% prewash is approx 40% meth, the point at which you will start to see reclaimed meth coming from your condenser is indeed about 80*C or 177*F.

Th every interesting point is that the more meth is reclaimed, the higher the liquid temp needs to be to get it out. I would LOVE to see how these figures are affected by circulating the solution or vigourously agitating it.

Hi csp97 and Tripi,

Thanks for that - very useful! I just found a nicely written website which explains in clear terms about the behaviours of ideal liquid mixtures.

See
http://www.chemguide.co.uk/physical/phaseeqia/idealpd.html

I'm pretty certain I'm getting fairly productive distillation at less than the inferred boiling point of the mixture, but only when I circulate the mixture and pump from the bottom, to the top, spraying into the vapour zone.

I'm adding energy with pump - could that be where the extra energy for vapourisation is being supplied from? The added localised kinetic
energy? Certainly if I stop pumping, my condenser output stops or reduces dramatically, yet the bulk temperature stays the same.

All interesting stuff. Thanks again.

Hey csp,
Any way yu could model the effets of higher pressure differentials?
Perhaps like spraying the bio/meth at 80psi into 26"Hg chamber?

When I was recovering without vacuum in an Appleseed style processor the only way I could get good recovery rates was to heat the mix much higher than the BP of methanol and to circulate with the pump while recovering . I also found that it seemed to go a bit better if I allowed the system to build pressure by closing the valve and then letting the pressure out all at once an then repeating . I never did any real tests though , it just seemed to work better . The one thing that became very clear was that without circulation it was very difficult to recover the Methanol without extreme heat . This is probably fine for the glycerine byproduct , but for the fuel I did not want to overheat it too much . Too much heat is know to polymerise (sp?) the bioidiesel and cause quality issues . But this is really great information to have available . Thanks for the data .

quote:

Originally posted by GrahamLaming:
spraying into the vapour zone.

Graham - you say that you are spraying? Can you post some details of your "exit point"? I am just pumping through an open pipe but wondered about picking up a high volume spray head from Graydon.

quote:

Originally posted by Tripitaka:

Graham - you say that you are spraying? Can you post some details of your "exit point"? I am just pumping through an open pipe but wondered about picking up a high volume spray head from Graydon.

Not trying to do Graydon out of any business but you can get both Spraying Systems and Bete nozzles in the UK, it would save you time and the shipping costs.
www.spray-uk.com
www.beteuk.com

Hi Tripi.

It's just an open pipe for me too, but with a ball valve right at the entrance to the tank.

If I adjust the ballvalve so the BD hits the far end of of the tank with a 'splat' sound, I get good output.

A smooth fast flow is slow and boring for meth recovery.

A restricted spurt is good.

I don't think this would work if the valve were more than an inch or so from the tank wall. It's the restriction caused by the semi-shut valve which speeds up the flow.

It's the top valve on the left - pic taken during build - looks 'a little used' now.

quote:

Originally posted by vegenergy:
Hey csp,
Any way yu could model the effets of higher pressure differentials?
Perhaps like spraying the bio/meth at 80psi into 26"Hg chamber?

Yeah, I'll run the numbers at 26" vacuum on Monday. The pressure differential won't make any difference on how the simulator calculates. It just looks at the final temp and press.

Rather than heat up the entire bulk to the optimum temperature, how about an 'afterburner' or auxiliary heater just before the injection point?

A 3kW heater in a 5" diameter jacket with in and out ports. (This would also make for easier heater cleaning/replacement, especially if the heater fails during a batch)

Once the liquid is up to free-flowing viscosity, switch to this auxiliary heater only. This would give a higher injection temp than a bulk heated load.

I'm not sure if this makes any sense at all, just an idea.

Something like this simulation...

Revised for 26" Hg vacuum (2 psia) using Antoine.

Biodiesel / Methanol Solution
% MeOH--- Boiling Point, F
5.00 ---------- 114
4.00 ---------- 121
3.00 ---------- 130
2.00 ---------- 146
1.00 ---------- 176
0.50 ---------- 211
0.25 ---------- 252
0.00 ---------- 467

Glycerol / Methanol Solution
% MeOH --- Boiling Point, F
45.00 --------- 88
40.00 --------- 91
35.00 --------- 95
30.00 --------- 99
25.00 --------- 105
20.00 --------- 113
15.00 --------- 123
10.00 --------- 139
5.00 ---------- 169
1.00 ---------- 259
0.00 ---------- 436

how about water? Is it the same curve or would you have an intersection of the lines which it seems to me would be a bad thing.

Thanks so much csp97.

I guess I do not fully understand, pardon what may be a stupid question...

26" Hg is around 2psia, or a differential of roughly -12.5psi from atmospheric, correct?

Would spraying a mixture at ~30psi into a chamber at atmospheric (~14.6psia) achieve the similar results seeing as how the pressure differential would be similar to your above simulation?

quote:

Originally posted by GrahamLaming:
Rather than heat up the entire bulk to the optimum temperature, how about an 'afterburner' or auxiliary heater just before the injection point?

Clever. Intuitively, I suspect it would work well, provided you had a good spray at the (above liquid level) discharge. I would also consider adding Mcguyver's air injection to help force the methanol vapors into the condenser.

Are the blue plastic drums tolerant of 180-ish temps?

quote:

Originally posted by vegenergy:
Thanks so much csp97.

I guess I do not fully understand, pardon what may be a stupid question...

26" Hg is around 2psia, or a differential of roughly -12.5psi from atmospheric, correct?

Would spraying a mixture at ~30psi into a chamber at atmospheric (~14.6psia) achieve the similar results seeing as how the pressure differential would be similar to your above simulation?

It's the final pressure not the differential that's important. I doubt 500 psi to 14.6 psi would be much different than 30 to 14.6. Operating at a vacuum allows for lower temperature not because of the differential but because of the lower pressure. Kind of like how water boils at 180F in the mountains.

thanks csp97 for the clarification.

I am curious, are the tempatures you posted for the tempature of the solution (glyc/meth) (biod/meth) one is trying to boil methanol out of, or is it for the tempature of the vapor as it exiss the boiler? Or does it not matter?

Looking at the tempatures for atmospheric pressure, one could theoretically remove ALL of the methanol from the glycerol boils off, glycerol's boiling point being 290C or 554F according to Google. Does the same apply with the tempatures you posted for boiling points at 2psia? I would have figured this myself using the online boiling point calculator I found (http://www.trimen.pl/witek/calculators/wrzenie.html), but I can't seem to find what glycerol's heat of evaporation is.

I ask this not so much out of interest of recovering ALL of the methanol but more concerned about making the crude glycerol less toxic and/or more marketable to people who would be interested in taking the glycerol off one's hands. I've seen posts on these forums where people were trying to give their glycerol to soap makers or the like and were told that their glyc was too crude.

Temperatures are for both vapor and liquid. If your vapor is cooling as it exits the boiler, you would need higher temps for the liquid. However, it seems for others' experience that the temperature requirements are lower with agitation.

The 0% methanol temperatures are the boiling points of glycerol according to the computer. Glycerol boils at 550F at 14.7 psia and 436F at 2 psia. You would not be able to recover 100% of the methanol without going above the boiling point of glycerol and using distillation.

Heat of vaporization of glycerol is 462.2 BTU/lb at atmospheric pressure according to the simulator.

That is not correct csp97. The temperatures in the table are for the liquid in the pot. The temperature of the vapour is always 65 deg C at atmospheric pressure. The best place to measure that is at the highest point in the vapour tube on the way to the condenser. If it goes any higher water is coming off.

One would not attempt to go near the boiling point of glycerol - it decomposes if this is attempted at atmospheric pressure.

Raising the pot temperature to 130 or 140 degrees C is OK.

Being very practical about distilling methanol from saved glyserol byproduct, what temperature would I have to achieve to remove the majority of the methanol. Since I don't use any water prewash and dry every bit of the raw feedstock, would 200F to 250F be high enough to do the job?
I am only considering the economics of recovering for use in BD processing........Roy