Instead of using a bulkhead fitting, get a steel bung from McMaster and weld it in, one in the side of each barrel

I searched McMaster.com for 'steel bung' but wasn't sure what item you may have meant, please be kind enough to supply McMaster's catalog number for the bung you meant.

It has been a long time but try p/n 12555K85

It would be interesting to try welding my plastic barrels...since I'm using plastic, I will have to use bulkhead fittings. I saw some for $10 at mcmaster.

I have success using a 1/2 NPT tap on plastic barrels, I use polyurethane construction adhesive for pipe dope. I would think youd' have no problems w/ 0 head on one.

Yes, I suppose I will have to get some new barrels. The ones I have now are covered with WVO-sticky, and should probably be replaced eventually anyway.

quote:

Originally posted by john galt:
.... ... Without a closed barrel, then one would have to pour the oil in very slowly, a rather tedious process ...

Dana's notch cut idea might work for this. You cut a small notch (triangular?) out of the threaded neck of the cubbie. How much your unscrew the cap controls the rate of oil flow out of the cubbie.

So this means you end up with.
. A tilted table that holds the cubie on its side, Neck down hill, and above height of upflow settle drum
. Funnel below neck of cubbie that feeds cubbie oil into plumbing that fills upflow drum from the bottom.
. Control rate of flow by how much you do/don't unscrew the cap of cubbie.

Another idea is to use an old picnic cooler that has the water drain on it. (lot's of new ones don't). Plumb a hose to the cooler drain with a valve on it. Pour cubbie into the cooler and set flow rate with the valve.

Would there be any benefit to make the drain from the 20 micron filter go toward the bottom of the second drum? Similar to the 30" down pipe in the first barrel. I have been building my own version of a system like this and intended to do so until I saw this thread.

I forgot to mention that I also was going to shorten my outlet pipe so that the pickup tube was much higher in the barrel. The inlet toward the bottom, the pickup 10" or so down from the top. It would limit what could be pumped out each time, but may produce cleaner oil. Any thoughts?

I haven't been following this too closely, but it appears that what you're discussing is pretty close to a water-washing method for biodiesel discussed here And a sketch is hereIt worked quite well for washing BD, and ought to work in similar fashion for SVO. I used a Facet fuel pump to push the BD into the bottom of a water column. The BD dribbled upwards through the water, losing contaminants along the way, and pool at the top, exiting through a simple overflow spout. Clean water was admitted through a needle valve at the top, gradually moving down to the bottom of the column, below the BD inlet, then exiting through a raised outlet pipe that regulated the column height. Heated water worked better than cold. A 5-foot tall, 6-inch diameter pipe, worked best, but still took two passes to adequately wash biodiesel. I would expect similar performance with WVO. The biggest problem, as also discussed here, is a pump or feed system that works perfectly. Facet (and most automotive fuel pumps) needs very clean liquid to work properly.
I hope this helps.
Cheers,
JohnO

one thing that i just noticed. the cubies that sat out side for a year in temps from 90F to -5F are a lot better dewaterd then anything else i have tried. dont know if it is the freezing and thawing that was mentioned earlier. but there has to be 75% less water then when i picked it up.
i started a new thread looking for an accurate way to test for water. i would like to test these cubies weekly, and see how much the water level drops, and if freezing temps has an effect. and just how much more water the upflow system will remove.

Hi John just to say thanks for the drawing for this system, I finished my upflow settling filter in November 07 220 gallon Ibc, and yesterday had my first 25 ltr of oil. did a pan test no water at all cant see why people would want to make it more complicated as it works so well.Thanks again John

William, good feedback, thanks. Everyone I know who actually tries the system has similar results to yours. I'm glad it works for you too. IMHO simple is almost always better.

quote:

jeepin, moggin Jessup (coachgeo):
If you have not done so already please pan test a few samples of the dirty oil your using after going thru the 100 mesh screen but before any upflow settling. This is needed to have a good comparison to analyze results against.

Dirty oil failed. Batch A passed after 5 days, B passed after 1 day, C was a total failure. Salt water looks promising with an oil introduction rate of 1/10th gallon per hour into a pail. Will be scaling up with a triplet of 30 gallon drums. First (raw) drum with strainer for dirty oil, elevated with mid-mount bulkhead fitting and ball valve to slowly introduce oil into bottom of 2nd (upflow-wash) drum via descending pipe. 2nd drum half full of salt water with near-top bulkhead fitting through another ball valve to 3rd drum. 3rd (collection) drum descent pipe fitted with 5 micron filter bag. Feed rate regulated to about 1/4 gallon per hour or about 1x 30 gallon drum a week. Possibility of chaining/cascading more 3rd stage drums to extend dwell time of oil while maintaining or increasing throughput rate. If successful, will scale up process by parallelization and start stockpiling drums of cleaned oil.

quote:

Originally posted by Burbarian:

quote:

jeepin, moggin Jessup (coachgeo):
If you have not done so already please pan test a few samples of the dirty oil....

Dirty oil failed. Batch A (w/distilled water) passed after 5 days, B (with salted distilled water) passed after 1 day, C(w/alcohol-distilled water blend) was a total failure....

Thank you. The results so far are how I assumed they would be... not that thats worth anything.

Question that remains is how well do the best of these three methods stack up against nothing being added to the up flow container. This can only realy be tested for comparison to JGalts setup; I assume, in a larger drum to equal time and height oil must rise to what JGalt's does.

From what has already been reported by JGalts and others, I have no reason to doubt that the JGalts standard upflow process works splendidly as-is. The objective of my experimentation was to see if said process could be modified to work with raw (vs pre-settled and top-decanted) WVO, and also to see if process time could be reduced. JGalts uses the non-decanted residue as stove fuel. I use solar supplemented by a stratified downdraft wood-gas stove for heat, so would like to use as much of the WVO as possible for my menagerie of diesels.

Good comments, thanks for sharing your research on this. These experiments are very helpful in getting a better understanding of why the upflow process works as well as it does. Right now my processing shed is at -5°F and I've had to put my research on hold for a few months till it warms up a bit. It's good that you guys in the warmer climes can move things ahead and continue to keep us posted.

My thanks to everyone who is contributing to making this one of the most useful discussions of VO cleaning and dewatering on the SVO forum. Keep up the good work, it just gets better.

John,
Re your use of a 2" exhaust pipe section: I am guessing that the use of 2" exhaust pipe allows for a "dip tube" that fits inside of a 2" PVC male adapter that is screwed into the drum's 2" threaded bung. Is this correct?

Also, I too am getting my WVO in cubies from a single restaurant source. What is a "cubie spout", and where can I find one?

Thanks, Allan

Welcome to the discussion Allan.

I'm using a 4" long nominal 2" steel pipe nipple to connect the two barrels. The 2" o.d. exhaust tube fits nicely inside the nipple and a slight flair on one end keeps it from dropping through the nipple.

There's a link to a description of the cubie spout in the intro message on page one of this discussion. Just click on the highlighted words "cubie spout".

quote:

Originally posted by Burbarian:
... and also to see if process time could be reduced. ...

and that is partially my point. You need a control with an upflow of nothing in it to see how fast it works. You might find it is faster or you mightr find it is the same.

Another one to try is a bakng soda water mix?