Removing water generated in processing?

[Tony]

i next rebuild my reactor, im going to have the pump slightly higher than the drain port, so i can leave the port open, and the idea was that it would drain the pump also.  But i realise this probably wont work, as the impellers will still have fluid lurking at their lower edge : (

Could give the pump a quick run after draining the glyc, then re-settle?

Even with an elevated pump above the glyc line (as I have at the moment) you get the liquid in and above the pump settle glyc into the dips in the pump, so you can't win.  I'm contemplating on the next build just putting the pumps at the bottom and giving them a quick run after dropping Glyc to clear them through (or just fitting drain taps to the pump housings).

[RichardP]

Sounds simple enough, the only caveat with glyc prewash is having to react right after otherwise pipes can get blocked as it settles overnight (even if you're super careful draining the glyc).

Not really true, if the processor is designed right then you should be able to drain the pipework completely and leave no glyc anywhere except in the reactor. Insulate the reactor well and the oil and glyc will stay hot enough so the glyc will not solidify, even if leaving a day or so.

I have to disagree.  I did a Glyc prewash in my hot water tank reactor (fully insulated) and settled and drained, then went back three times over the course of an hour and a half, thinking that would be OK, but the next morning the pipework was blocked.  I guess Glyc settles very slowly from oil vs biodiesel.  But perhaps I was just unlucky?

But if you leave your pipework empty what is going to solidify in it? If your pipework is open to the tank then of course glyc will settle out further, go into the pipework where it will cool and solidify.

Have a valve directly after the tank and one before the pump, glyc drain in between, easy to drain all the pipework and leave nothing to solidify.

[Tony]

Oh yes I see what you mean.  I do have a valve there but never thought to drain the pipework.  Wouldn't that just have given me a plug of glyc further back though?

I'm still liking the idea of a sedimenter

[Julian]

I do exactly what Richard is describing with a valve immediately at the exit from the tank.  My pump is about ten inches above the bottom of the tank.

I can drain the pipe work and leave the contents of the tank in place.  The lower half of the pump casing must collect glycerine as the pipe work above settles, but it's never caused a problem.  The Leo, not the most powerful of pumps, always seems to start the next time it's required.

[julianf]

I do exactly what Richard is describing with a valve immediately at the exit from the tank.  My pump is about ten inches above the bottom of the tank.

I can drain the pipe work and leave the contents of the tank in place.  The lower half of the pump casing must collect glycerine as the pipe work above settles, but it's never caused a problem.  The Leo, not the most powerful of pumps, always seems to start the next time it's required.

I have some leos, and some multistage grundfoss pumps.

The leos seem to be able to pump all sorts, all be it slowly.  The grundfoss are quicker when running on thin fluids, but theyre rubbish with even cold liquid oil.

[RichardP]

I always close off the reactor outlet then drain all the pipework, all it leaves is the small bit of bio in the bottom part of the pump so any glc left is minimal. The drained bio goes back into the processor and the pump side of the glyc drain is isolated.

Back to the original post, I agree with JRL, keep it simple and straight forward. Dry the oil well, good glyc wash and the use of ASM is simple enough and give excellent results. I think a molecular sieve would just slow the process down for maybe a tiny gain and at some expense.