After producing biodiesel for several years now I’ve tried every type of WVO filtering setup I could find out there and concluded, THEY ALL SUCKED!

Not saying they’re all bad designs [cold upflow being an exception], most setups do exactly what they were designed to do which is to remove small particles from fairly clean oil at a specific temperature and volume. Over the years I’ve used drum filters, bucket filter, sock filters, water filters, canister filters and just about anything conceivable besides a centrifuge with limited success. Each one has it’s good and bad points, or time vs quality and cost as I see it.

What I’ve learned from all this dabbling is you need an easy way to filter large amounts of waste vegetable oil down to a usable particulate level [20 microns or so] quickly, cleanly, and cheaply without making a mess changing or cleaning the filter medium every 50 gallons or so.

I’ve always followed the rule of get it under 100 microns before filling the reactor and then let gravity and glycerin do the work for ya after the reaction. Once converted to biodiesel I then step it down to 5 microns or less before storage. I let all my collected WVO settle out for at least 30 days or more before using it, and pump only the good oil off the top. This seems to work great for removing the water and heavy stuff. So far, so good with this practice.

Most small scale biodiesel producers use some sort of setup that was designed for anything but filtering large volumes of WVO full of medium size particles. To do this with any success without a huge centrifuge you need 3 things, LARGE amounts of filter surface area, Heat, and Pressure. So I started out rethinking designs with these 3 rules.

Rule # 1, cold oil does not flow through anything fast.
Rule # 2, there is no substitute for filter surface area.
Rule # 3, gravity is ok, pressure is better, but liquids flow the best in a vacuum.

For some time now I had been using doubled up pillow cases attached to a 1 inch pipe draped over a 55 gallon drum. The oil was preheated with an inline heater as it left the 300 gallon tote and ran into the pillow case that filtered into the drum. Once the drum was full, I would pump it into the reactor. Worked ok, until the pillow cases became clogged, then you have a huge bag of oil weighing over 100lbs and no way to change it without making a mess or contaminating the clean oil. Not to mention the smell of an open top drum full of hot oil stinking up the shop.

The cleanest design out there is by far is canister filters inline with a pump pushing oil through them. Problem is, you need multiple stages of progressively smaller micron rating to work efficiently. Then there is the issue of determining what filter is clogged up and the cost of replacing them. So I started looking at large canisters with replaceable elements, also not so cheap. Then I decided to design my own canister filter from common materials anyone could find with replaceable element material that would cost less than $5. I kept coming back to my old dollar store pillow cases, and then added a new spin on the cold up flow design, heat and vacuum.

The basic design is identical to a standard canister filter, just scaled up to 55 gallon size. The unit is completely sealed with locking ring, easy to service, and contains a heating element to preheat the drum and a second inline element to maintain temp as the oil flows through. The unit uses 2 temp control points, a master on the drum that acts as a safety and controls power to the secondary unit mounted on the inline heater. Since the inline heater can only be used when oil is flowing, there is a second switch on that unit that remains off while the oil in the drum is coming up to temp. All hoses have quick couplers and I used a window AC unit 20A power cord to make a quick disconnect for the inline heater portion. This way I can quickly remove the lid to inspect and change the filter element.

Best part is, all the crud stays in the bottom of the drum and allows you to back flush the element and then drain water and or debris off from the bottom into a catch bucket. To change the filter medium, just lower the level in the drum and remove the top. No mess… The pillow cases are sealed to the output pipe by a couple hose clamps. Total filtration level can be adjusted by changing the material used if needed. Cheap pillow cases doubled up are somewhere between 20 and 150 micron.

Flow rates of 10ltr per minute is average using 10 inches of vacuum on the reactor and oil temps of 100 deg F. Higher flow rates are possible with some small design changes. I also use a 20 micron canister filter inline after this unit as a fail safe. Over 1500 gallons WVO filtered so far without replacing it. I have only changed the pillow case elements once in that same time, just to see how they looked. Unit can be back flushed with air or oil to extend filter life.

Total cost to build this unit was under $100, replacement elements cost $2 at the dollar store.

Detailed drawings and pictures of working model Click Here Comments or suggestions are always encouraged.

-Ken

Pretty impressive and ingenious set up. I am curious as to why you filter down to 20 micron before processing. I realize that particulates will hold water, but in my case with my 55 drum processor, I spray dry as I am warming up the wvo prior to processing. I use window screening for my filtration before loading into the processor.

Nice rig though!

quote:

I am curious as to why you filter down to 20 micron before processing.

On my old processor [a 55 gallon drum setup] I burned allot of heating elements up due to particles attaching to it. Titration was higher and I was having to boil off water most of the time. I was using those bucket filters with 200 and 100 micron screens when filling it, they always clogged up an made a mess. I'd end up dumping half the junk into the reactor sometimes. And having to change a heating element when the tank it full is no fun..

When I upgraded to the 120 gallon tank style I had to come up with a new way to fill it and was using a sump pump in a drum, that's when I was attempting several screening type setups. I had also installed 2 water filters on the exit to the wash tank. A 20 and 5 micron element, they would clog up every other batch $$. So I knew there was a good amount of crud still getting in and not falling out with glycerin as I expected. At this point I was still spilling oil all over the place, the floor was so sticky you couldn't fall down.

So I wanted a closed loop setup where I never had to touch the oil again from the time I filled the storage tank and would conserve the more expensive filters. The large inline 20 micron spin on canister filter was a left over from previous designs. Not knowing how the pillow cases and plastic bucket would hold up under high vacuum, I just plumed it into the line to catch anything that might break free in the drum. It remains in play and I've only changed one other post wash filter in the system since, no heater elements failures in this new rig either.

I'd say the doubled up pillow cases [70% polyester 30%cotton] are catching everything down near 20 micron. The cold up flow aspect seems to work good on the water part, I just drain off a few gallons of oil from the bottom of the drum before I start filling each batch.

I'll agree it's not required to go that fine, but it has saved me a bunch of work.

-Ken

Nice set up Ken,
I have to agree that vacuum rocks for moving liquids. I've been thinking of how to devise something similar to yours with the addition of a 15 gallon or so canister filled with wood planer shavings in place of your 20 micron filter.
You've definately given me some ideas here. Thanks.

Nice write up, drawings and pictures. Could you please give some details on the electric inline heater. - Thanks

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

rkpatt - I used a design someone I found on here a good while back [Maybe Murphy has it in his collection of designs], but can't recall the link now so if someone can refresh me on that, great. I also opened up my broken Titan instant on hot water heater and found inspiration.

It's pretty simple and fairly cheap to build, just basic pipe fittings, 240V water heater element, stitch, temp controller and a couple 1 gang electrical boxes with back cutouts to keep the hot wires covered up. Everything came from Lowes.

Only issues I had getting it to work was keeping air out of the tube. When you boil off any water in the WVO you create air bubbles, and trapped pockets of air will burn out the element right quick.. I burned up 3 in a day working out the bugs. Vertical with inlet at bottom and exit at top keeps the element submersed and seems to do the best. So keep that in mind when setting it up.

Second thing I suggest is placing the temp controller high up on the tube in the direction of oil flow for best results. I mounted mine on the tee near the threaded end of the element, when in a horizontal orientation it worked ok, but when I changed to vertical mounting it end up at the low point and never gets warm enough on that end to shut off. So is pretty useless now till I build a new one and position it higher up the line.

Last, output temperature is relative to flow rate. I first built one using 110V wiring, worthless for anything over a couple ltr per minute. A 3500W element using 240V seems to hold about 130F at 10ltr per minute flow with 70F starting oil temps. You could move up to a 4500W or 5500W and gain some flow rate, but since I have 2 elements on the same circuit I was limited by load amps. Also suggest some shielding or isolation on the pipes to avoid getting burned if using at higher temps.

Anyways here is a breakdown view of it, hope this helps.

-Ken

Ken- The link to the view doesn't seem to show up .

quote:

Anyways here is a breakdown view of it, hope this helps.

quote:

The link to the view doesn't seem to show up

Fixed



-Ken

Your design is perfect for me, although I'm just planning to burn the straight WVO in my generator. For it to work, I'm guessing the 55 gal metal drum has to be absoulutely airtight, or your vacuum pump will just suck air. That ring top drum will seal it tight enough? Also, what kind of vacuum pump are you using?

I'd love to start making biodiesel, but gettin g methanol here in the islands might be a littl costly.

What are you using for a vacuum pump? I've tried an old compressor from a fridge, but it didn't like to run on a continuous duty cycle long enough to fill up my 50 gallon processor.

Hello. I wondered if anyone is familiar with a super effective WVO filter that is being made in Asheville? I've heard it's been testing for a year and can handle almost anything poured into it. Thanks. Claire

quote:

Originally posted by clairek:
Hello. I wondered if anyone is familiar with a super effective WVO filter that is being made in Asheville? I've heard it's been testing for a year and can handle almost anything poured into it. Thanks. Claire

Apparently not, but keep trying to get somebody to ask you about it, you may sell one someday.

The third party stuff doesn't get by the BS detector.

Norman

quote:

Originally posted by Jim Sloan:
What are you using for a vacuum pump? I've tried an old compressor from a fridge, but it didn't like to run on a continuous duty cycle long enough to fill up my 50 gallon processor.

Sorry for the lag Jim, been super busy an not checking in on the forum.

Your correct the fridge pumps get hot pretty if ran continuously. My partner found a couple old air cooled rotary compressors / vacuum pumps at a farm sale "cheap" and we have one of them connected to the reactor. If you have sealed up all the air leaks in the system, try pulling the reactor down to about 15 inches and then let the pump cool off for a while before opening the valve to the filter. It should continue to flow down to about 5 inches before you need to flip the pump back on again. Keep a watch on eBay for good deals on vacuum pumps also, any small oil bath unit will work fine also.

-Ken

hey guys, i just picked up a universal style 5500 watt replacement water heater element and the thread on it doesnt seem to be NPT.. it will go in maybe a single thread before binding up. anyone know what thread these are? thats the only part of the inline heater diagram im not sure of .. the "weld box to hex after assembly" part..

thanks in advance
mike

All the water heater elements that I have worked with have 1 inch STRAIGHT pipe threads, same pitch as normal pipe thread but not tapered. I usually have to run a pipe tap into the fitting to inlarge the threads so the element will screw in several threads, it depends on the fitting though. If screwing the element into a tapered pipe thread you will need to use LOTS of thread sealent to keep it from leaking. Best would be to buy a straight 1 inch pipe tap and pass it through the fitting and then machine a recess for the rubber washer to keep it sealed. Alternative to maching the recess is to use a couple machine bushings (big washers), one just fitting over the threads of the heating element to compress on the rubber washer, the other fitting over the outside of the rubber washer to keep it from squishing out.

thanks for the info. this was a 5500W universal replacement unit i picked up at lowes for $9.50 or so.. its definitely straight thread, closer to 1.25 OD though. i'll keep looking around.

Don't go by the direct measurment, it is 1 inch straight pipe, measure it against a 1 inch pipe nipple.

Thanks for the great plans. This does seem like a great idea.
The only limitation I see is acquisition of a vacuum pump.

The recommendation of using a tap to create the proper threads for the heating element was a great suggestion.

My question is how long does the rubber seal typically last? Would Viton be a good selection?

Here are Viton Seals for hotwater heater elements.

http://www.bayareabiodieselsup...iewProd&productId=82

But the 1" NPS pipe taps are pricey for what is most likely a one time use ......

Edit -

Screwing the NPS WH element into the NPT fitting with pipe dope has worked for me ( just lucky?). It is probably a better idea to follow the advice of Sunwizard - http://biodiesel.infopop.cc/ev...9605551/m/2001011761 and others and use the washers and o-ring .

Also, I used !" pipe (and tees) for my inline heater assembly ( for a Dieselcraft CF setup ). The clearance between the element and the the inside of the pipe is minimal but it has hae been working fine for many months.

quote:

est would be to buy a straight 1 inch pipe tap and pass it through the fitting and then machine a

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

quote:

Originally posted by rkpatt:
But the 1" NPS pipe times are pricey for whatis most likely a one time use ......

You're not kidding.

http://www.mcmaster.com/#2526a36/=46yrfe