Continious removal of glyc when processing, using a sedimenter?

[julianf]

This is something ive been knocking about in my head for a while, but have been put off by comments that the glyc holds the bulk of the meth.

Anyhow, the hardware would be easy to sort out -

Simply have one of the pumped pipes from the reactor connecting to the top of a sealed container, and then another pipe also coming out of the top, going back to the processor.

As glyc is heavier, the glyc would, to a large extent, fall out in the sealed container, and have nowhere to go, whilst the bio mix would continue to flow along the top.


One of the points put forward for eductors is that meth floats on the wvo in the reaction - so if this is really an issue, then the meth would pass over the top of the settling glyc also.



Has anyone tried anything like this?  Id be curious as to the result.


I have even considered it as a method for getting the glyc out post-reaction.  Rather than settling in the reactor, with the stress of glyc going hard in the pipes etc. some valves could just be set up to redirect flow over the 'glyc trap' whilst the pumps were still running.

Anyhow, its something ive considered for a while, and id be interested to hear any comments.  It may not work in the slightest, but, as i say, im not convinced (yet) that it wouldnt have some application!

[Tony]

I've also considered exactly the same thing.  I don't think you'd even need a pipe in and out, just a large bore pipe from the pumped loop - I'm pretty sure the glyc would happily fall into the trap

The problem is having an appropriate container to avoid it leaking, clearly a cubie won't cut it and I'd be nervous about a 25l drum, even if it is more robust.

Originally I thought it would be a good idea because if glyc holds the reaction back then having it kept away from the reaction might help it complete single stage.  But in my experiments with this nothing of the sort happens, presumably because of glyc's affinity for Methanol taking half the Methanol with it.

I think the floating Meth thing you mention is only when you've first introduced it, as it floats on WVO but freely mixes with biodiesel/glycerol.

[Tony]

That said I do have a stainless 55l tank with hot water coil in it, that'd make a great sedimenter as the coil could be used to heat any solids, and being stainless it's safe to leave connected all the time.  But - does it help in any way vs just draining the glyc while still liquid?

[photoman290]

i don't know if it would work on not, but having got rid of the condenser to simplify the system  isn't this another thing to possibly complicated it again.? sometimes the simplest way is the best. is there a problem to solve in the first place?

[therecklessengineer]

I've thought about this too - but using a cyclone type arrangement.

What's put me off is the need to add it back in again in order to demeth.

[Julian]

I've thought about this too - but using a cyclone type arrangement.

Me too, bottom of first post ... http://www.biopowered.co.uk/forum/index.php/topic,322.msg3619.html#msg3619

And from that thread others have had similar thoughts.  Must be about time someone gave it a try.

[Head Womble]

Form what I understand the catalyst is held in the gly (this is why gly washing reduces the acid value) so separating it to soon may cause more problems than it solves.

[Julian]

Form what I understand the catalyst is held in the gly (this is why gly washing reduces the acid value) so separating it to soon may cause more problems than it solves.

That's why I was advocating going back to a very slow introduction of Methoxide as GL originally suggested.

[julianf]

But in my experiments with this nothing of the sort happens, presumably because of glyc's affinity for Methanol taking half the Methanol with it.

Can i ask what sort of experiments you ran?

[Tony]

But in my experiments with this nothing of the sort happens, presumably because of glyc's affinity for Methanol taking half the Methanol with it.

Can i ask what sort of experiments you ran?

Sorry Julian didn't spot this question until recently.

Basically just doing a single stage but settling towards the end, dropping the glyc then carrying on pumping.  It didn't give a clear pass like two staging with the same catalyst quantities (my feedstock is very consistent in titration so pretty much all batches are the same in behaviour - when I two stage I usually always get a clear pass).

[julianf]

This also brings up somthing i have wondered about the 2-stage no-titration process for a while -

Im not that bothered about the amount of catalyst used, as thats cheap enough, but the methanol is the expensive bit.

IIRC the wiki advises 80% on s1 - going by what you say above, the glyc takes out the bulk of the methanol (somthing which has been said before lots too)

As i see it, s1 of the 2stage is mostly to clear the page, as it were - to remove all the unknowns.

Using 80% of the methanol to get to this point is possibly not the most economic way of doing things?

I should start another thread on that topic, i think.

[Tony]

I wonder, how about a sedimenter that can take 80% of total expected Glyc volume?

Then first stage of a two stage batch carries on as normal, hopefully filling sedimenter.

Then sedimenter would then be isolated from the batch via a valve, and stage two Methoxide introduced to complete 2 stage reaction.

You could then either:

- open the valve and flow through the sedimenter to do WBD, or
- after stage one drop the stage one glycerol from the sedimenter while adding the s2 Methoxide (if you want to Glyc wash your next batch).


Also, to keep the Glyc liquid in the case of NaOH you could have a coil in the sedimenter that the main reactor circuit pumps through to keep it all warm.  My stainless hot water tank is sounding more and more perfect for this job, I think I'll add it to my plan for the big processor.

Potentially any sedimenter could also double as an AAF tank, holding the Methoxide separately until the main batch is up to temp, then pump main batch through sedimenter/AAF tank for super fast Methoxide introduction/mixing.

Really got me thinking now!

[nathanrobo]

Sounds interesting. 

[Head Womble]

Tony, if you start rethinking your setup it'll never get finished.

[Tony]

It's all about adding enough BSP tees with blanking plugs to ensure all possible future additions are covered