Okay, been quite some time since I've been on the scene, so I'm hardly up to the latest and the greatest, but let's talk some design.

I have a small area to work in, so a small footprint is essential
Batch sizes will be ~30-40 gallons per week. Forty until I run out of my backstock, then we go down to 30. This approximately matches my expected usages.

My little corner of the southeast is presently tottering on the edge of water restrictions, so waterless washing is desired.

Here's how I plan it, and you folks can give me suggestions, tell me I'm full of crap, or otherwise give me critiques. Politeness is always appreciated.


Processor itself is going to be a closed barrel, turned over with the bottom cut out. An open top barrel's lid with bungs will be fitted, sealed with silicone.

In each of the two 2" bungs will be threaded one of Joe's twin port adapters. The bung on the top will be pump inlets, sized so that the pump can draw from the bottom of the barrel or from a level above where the glycerin layer would settle out. The bung on the bottom will be a pair of drains, one flush with the bottom of the barrel and the other with a stand pipe sized to drain the biodiesel layer.

Heating will be accomplished by preheating and by an element inline before the pump.

Picked up an old M55HV Melling pump and disassembled it, measuring the gears and plugging them into an equation for displacement of a gear pump. The result of the calculations shows that the pump moves about 1.2 cubic inches per revolution, or approximately 11.3 GPM if driven at a speed of 2100 RPM. Accounting for some efficiency loss, I'm basing all my times on a flow rate of 11 gpm.

The intention is to cut the pedestal off, then get the shaft journal bored out to accommodate a press-in teflon seal. The pump's original inlet and outlet will be plugged, and holes appropriate for 3/4" piping drilled and tapped into the housing. The shaft will be supported on bushings pressed into plates on either side of a V-belt pulley, used to get the necessary speed increase from a ~1500 rpm TEFC motor we have in the garage.

Poor man's venturi using internally smoothed out bell reducers in black iron piping, stepping down from 3/4 to 3/8 to 1/4, then back up to 3/4 in one step on the far side. This is the part I'm not very confident about, so if someone could give me some pointers on this one I'd be grateful. The venturi will be located almost immediately off the pump outlet.

After the venturi the oil will pass through some stainless tubing lightly packed with stainless wool, to act as a random-packing static mixer. The tubing will make one pass around the lid's circumference before dumping into the smaller barrel bung.

With 11gpm from the pump, my turnover time for a 40 gallon batch would be ~3.6 minutes. 120 minute processing time would give me ~34 turnovers of the whole batch, passing through pump, venturi and mixer. A fitting on the drum return would be used to feed escaped vapors back into the mix, as well as providing a light vacuum in the drum once all of my alcohol and catalyst has been added.

Alcohol recovery will be accomplished by heat and that same venturi providing a light vacuum, captured in a repurposed 40lb propane tank (craigslist freebies are neat).

The donor open top drum will become a dry-wash tank, with alcohol-less biodiesel fed into a B100supply sourced OC-20 centrifuge, with a standard Melling M55 pump (driven at 210 RPM via another motor and belt for ~.93 gpm) with its inlet and outlet tapped rather than new holes drilled in the housing. 3/8" iron pipe would be used for both sides of the pump, and the stock pressure relief valve would be plugged. A screen lid would be constructed to keep creatures out of the barrel.

Storage (for the short term that it will be stored) will be in another drum with a hand pump (for the moment) threaded into the bung for fuel dispensing.

Questions, comments, complaints?

Aaaaaaaaanybody home? I'll even take "That's a dumb design" so long as there's something helpful in the comment...

I'm a bit confused with how you are doing you barrel design.

Not surprising, I've been called incomprehensible before.

Basically, I'm setting it up so I have bungs on both the top and bottom, by flipping over a closed barrel, cutting out what once was the bottom, and adding the top from a removable-lid drum.

Pump, pump motor, heater, static mixer and venturi will all sit atop the drum to conserve space and keep pipe runs short.

The two ports on the top bung (through the port adapter) will let me draw from two different liquid levels in the drum. The barrel lid I'll be using also has a 1" bung that I'll be using as my liquid return from the pump circuit.

The two ports in the bottom will allow me to drain from two different liquid heights.

The whole deal will sit on a stand, the barrel itself wrapped up in insulation to keep the heat in during processing.

Make more sense now?

quote:

Originally posted by TurbinePowered: My comments in red.


Processor itself is going to be a closed barrel, turned over with the bottom cut out. An open top barrel's lid with bungs will be fitted, sealed with silicone.

The open top drum lid will not seal well to the bottom of a standard drum, even with silicone. You’re much better putting a hole in the bottom of a standard drum and welding in bungs. If you can’t weld (or don’t have facilities) I have a method for mechanically installing fittings without welding.

In each of the two 2" bungs will be threaded one of Joe's twin port adapters. Wise choice The bung on the top will be pump inlets, sized so that the pump can draw from the bottom of the barrel or from a level above where the glycerin layer would settle out. The bung on the bottom will be a pair of drains, one flush with the bottom of the barrel and the other with a stand pipe sized to drain the biodiesel layer.
OK

Heating will be accomplished by preheating and by an element inline before the pump.

What size elements will you be using? 110 or 220?

Picked up an old M55HV Melling pump and disassembled it, measuring the gears and plugging them into an equation for displacement of a gear pump. The result of the calculations shows that the pump moves about 1.2 cubic inches per revolution, or approximately 11.3 GPM if driven at a speed of 2100 RPM. Accounting for some efficiency loss, I'm basing all my times on a flow rate of 11 gpm.

Melling high volume pump is a good choice, should pump more than that though. A much easier way to measure volume of positive displacement pump is to fill the pump with oil, have a reservoir on top (inlet side)remove all air, turn the pump manually, say 10 times, while catching the output. How many ounces or mL you get divided by 10 is the pump’s volume/revolution. Multiply that by the speed of the motor and you get how many gpm.

The intention is to cut the pedestal off, then get the shaft journal bored out to accommodate a press-in teflon seal. Good Idea The pump's original inlet and outlet will be plugged, and holes appropriate for 3/4" piping drilled and tapped into the housing. Plugging the original outlet is fine. The inlet is large enough diameter to handle the flow. I drilled mine out a little more. You will find the housing not to be thick enough to thread in a pipe fitting. I welded mine. I have pics if you’d like to see them. The shaft will be supported on bushings pressed into plates on either side of a V-belt pulley, used to get the necessary speed increase from a ~1500 rpm TEFC motor we have in the garage.
Plate with bushings and a pulley is quite a bit of unnecessary complication, in my opinion. You will find the pump flows at quite sufficient volume directly coupled to a motor with a lovejoy coupling. The motor you have should be 1725 rpm if it’s a standard 60 cycle motor.

Poor man's venturi using internally smoothed out bell reducers in black iron piping, stepping down from 3/4 to 3/8 to 1/4, then back up to 3/4 in one step on the far side. This is the part I'm not very confident about, so if someone could give me some pointers on this one I'd be grateful. The venturi will be located almost immediately off the pump outlet.
Venturi for what? Sucking in methoxide or better mixing? Placed properly, you’ll find the methoxide tank will drain sufficiently from just the suction of the pump.

After the venturi the oil will pass through some stainless tubing lightly packed with stainless wool, to act as a random-packing static mixer. The tubing will make one pass around the lid's circumference before dumping into the smaller barrel bung.
Stainless steel wool packed tubes are not an efficient static mixer and quite unnecessary, I would say.

With 11gpm from the pump, my turnover time for a 40 gallon batch would be ~3.6 minutes. 120 minute processing time would give me ~34 turnovers of the whole batch, passing through pump, venturi and mixer. A fitting on the drum return would be used to feed escaped vapors back into the mix, as well as providing a light vacuum in the drum once all of my alcohol and catalyst has been added.
The pump size is quite sufficient for your batch size.

Alcohol recovery will be accomplished by heat and that same venturi providing a light vacuum, captured in a repurposed 40lb propane tank (craigslist freebies are neat).
Alcohol recovery takes up a lot of space, you said you’re limited on space, for small batch sizes, is it cost efficient to do this?

The donor open top drum will become a dry-wash tank, with alcohol-less biodiesel fed into a B100supply sourced OC-20 centrifuge, with a standard Melling M55 pump (driven at 210 RPM via another motor and belt for ~.93 gpm) with its inlet and outlet tapped rather than new holes drilled in the housing. 3/8" iron pipe would be used for both sides of the pump, and the stock pressure relief valve would be plugged. A screen lid would be constructed to keep creatures out of the barrel.
Centrifuge will not sufficiently remove contaminants. That’s not what dry washing is. Dry washing uses an ion exchange resin packed into a column that requires specific dimensions and flows. Usually the resin bed height is 3 times the diameter.

Storage (for the short term that it will be stored) will be in another drum with a hand pump (for the moment) threaded into the bung for fuel dispensing.

Questions, comments, complaints?

I am welding-enabled (but not a welder), but only so far as stick is concerned so that leaves welding anything to sheetmetal out of the equation. I'd be interested in your other method, though.

I figured the wool-packed tubing wasn't going to be very efficient, but I figured anything that would generate turbulence would have an affect of agitating the mixture and increasing contact area as a result. *shrug* It was a thought.

The venturi, again, was intended to add a little finer mixing to the setup while providing some lift to pull from the alcohol tank.

I have both 110v and 220v elements, and the main panel is close enough for us to add a 220 sub-panel to operate the thing on. It's what we did for the welder when we installed that piece of equipment, so we've done it recently and there is still enough capacity in the panel for another sub or two.

I'm on break at work at the moment, but I have the gear measurements for both an M55 and M55HV at the house. It's possible I did my math wrong, but it sounded in the right ballpark, to circulate the engine's entire supply of oil 10+ times per minute at high cruising RPM. I'll post the formula I found to calculate it, too, it's on the same spreadsheet.

I've read postings about some users only centrifuge-washing their finished product with good result, and my presently-low budget and the water restrictions made it a very attractive option. Perhaps I can pre-wash centrifuge the bio to extend resin life some more...

Fabricator says that centrifuging bio is a waste of time. The bio gets too hot and the soap goes back into suspension. You should use the centrifuge on the oil prior to processing to remove water and water containing crud. If you are using NaOH as your catalyst, you can run the demethed % settled bio through wood shavings to remove the residual soap and you are done. Or you can wait a couple of weeks and pull the bio off the top and use it.

quote:

Originally posted by Eurocab:
Fabricator says that centrifuging bio is a waste of time. The bio gets too hot and the soap goes back into suspension. You should use the centrifuge on the oil prior to processing to remove water and water containing crud. If you are using NaOH as your catalyst, you can run the demethed % settled bio through wood shavings to remove the residual soap and you are done. Or you can wait a couple of weeks and pull the bio off the top and use it.

I remember seeing the comment that after a while it was warm to the point of going back into suspension, but I don't recall reading his calling it a "waste of time." I thought his intention was to put together some kind of chiller.

Point me to the thread where it's all happening? I seem to have lost my link.

I have given up on fuging to remove glyc and I'm not alone, everyone who has tried it has found the heat thing to be an expensive barrier to cross, for home brewers it adds to much complexity.
I have started using naoh again and washing with wood chips, last batch I ran my 250 gallons through chips four times and had a final soap level of around 33ppm.

Hi Fab, does that mean you are not 'fugeing to remove soap as well? The chips are hardwood? Sounds simpler all round.

Paul

I still fuge my oil to clean and dry it, But no longer to get soap out, the new process with the whole batch demeth is turning out to be rediculously simple, after the demeth process let settle for two hours, drain glyc, then I let sit over night and drain any settlings in the morning, then through the chips, this morning after the initial pass through the chips it's 46ppm soap.

So what I'm getting here is that there is a problem with heat build-up in the oil as it continually loops through the centrifuge circuit?

What about putting the pump motor on a timer, to run it at equal length intervals of on and off? Take a while longer, and requires adding a timer to the electrical circuit, but that would be a bit more feasible for my footprint than adding another drum or tank to house chips (which I'd then also have to dispose of afterward, as well as finding in good supply).

At .93 GPM, which is probably a bit small for a 40-gallon batch, it'll take 35 hours to make 50 passes with the entire batch. When did the bio start getting hot enough to start dissolving the soaps back into itself? Set your timer for a little before then, let it run for that interval, then stay shut off for that interval (or less, since you aren't actively adding heat anymore? I dunno), during which you take the opportunity to clean out the rotor, then it runs again. Sure, five on/five off would make for doubling my time requirement (about three days) but considering I wouldn't have to dispose of another waste product (Wood chips or resins) or add another tank (my centrifuge barrel is also my pre-processing and dewatering barrel) and doesn't require water (which is restricted due to the drought).

I probably ought to find the money to bump up to the bigger OC-50 already, and cut that time down from the start. 1.8 GPM would bring my doubled-up time to 37 hours, or a little over a day and a half.

In my experience, and that is considerable at this point, (spent over 3K on the fuge thing this spring) what you are suggesting trying, if you make a 30 gallon batch it will take you approximately 30 days to partially fuge the soap out of your bio.

quote:

Originally posted by fabricator:
In my experience, and that is considerable at this point, (spent over 3K on the fuge thing this spring) what you are suggesting trying, if you make a 30 gallon batch it will take you approximately 30 days to partially fuge the soap out of your bio.

Thirty days? I was getting thirty hours... well, thirty six. Am I seriously missing that much here, that much off on my math?

40 gallons, 50 passes through the 'fuge, that's effectively pumping 2000 gallons through the centrifuge.

@ 1.8 GPM, you have 2000/1.8 = 1111.1 minutes, or so. Divide that by 60 minutes in an hour, and you get approximately 18.5 hours to pump that volume through the centrifuge at that rate. Divide that into, say, six even sections of three hours apiece, add a couple minutes to take out the slop factors from rounding, then give yourself an identical length window between each. So that means you have 11 increments of 3 hours each, discounting the three hour period after your final 'fuge run. 3 x 11 = 33 hours of 'fuging.

I'm not really sure where the thirty days came from?

quote:

Originally posted by TurbinePowered:

quote:

Originally posted by fabricator:
In my experience, and that is considerable at this point, (spent over 3K on the fuge thing this spring) what you are suggesting trying, if you make a 30 gallon batch it will take you approximately 30 days to partially fuge the soap out of your bio.

Thirty days? I was getting thirty hours... well, thirty six. Am I seriously missing that much here, that much off on my math?

40 gallons, 50 passes through the 'fuge, that's effectively pumping 2000 gallons through the centrifuge.

@ 1.8 GPM, you have 2000/1.8 = 1111.1 minutes, or so. Divide that by 60 minutes in an hour, and you get approximately 18.5 hours to pump that volume through the centrifuge at that rate. Divide that into, say, six even sections of three hours apiece, add a couple minutes to take out the slop factors from rounding, then give yourself an identical length window between each. So that means you have 11 increments of 3 hours each, discounting the three hour period after your final 'fuge run. 3 x 11 = 33 hours of 'fuging.

I'm not really sure where the thirty days came from?

It comes from expensive and painful experience, but if you want to reinvent the wheel and use a lot of time and treasure doing it help yourself, remeber, math does not always reflect the real world, in this case it is not even close.

My experience with fuge washing is limited, but I can believe 30 days. I gave up on the concept after a couple of days. The soap can not be mechanically removed until after sufficient methanol has been removed. I don't really recommend fuge washing because it gradually flashes off the methanol while it's running. 30 gallons of biodiesel will have a couple gallons of methanol in it. That's all released into the air above the fuge.

FYI, A pressure driven fuge will be more effective at flashing the methanol than a motor driven one.

quote:

Originally posted by TurbinePowered:
I am welding-enabled (but not a welder), but only so far as stick is concerned so that leaves welding anything to sheetmetal out of the equation. I'd be interested in your other method, though.
.

OK, as they say, a picture is worth a thousand words. I came up with this idea when I was too scared to weld a bung on an old propane tank that I turned into a super sucker. You make a hole in the tank (drum, whatever) that is a little larger than the pipe size you are going to use. Also make the six holes for the bolts. The two half flanges fit through the hole and the bolts come out the bolt holes. You can install this w/o access to the other side of the tank wall.





You can see how the tank wall would be sandwiched between the plates. Put a gasket over the bolts before putting on the threaded bung and it is leakproof. My super sucker will hold a vacuum for days with this flange. The one pictured is 2 inch pipe.


I didn't have a piece of 2 inch pipe to screw into it, so I used one of my two pipe ports. (it's 2 inch NPT)

Those are really. really nice looking units Joe.

quote:

Originally posted by RickDaTech:
My experience with fuge washing is limited, but I can believe 30 days. I gave up on the concept after a couple of days. The soap can not be mechanically removed until after sufficient methanol has been removed. I don't really recommend fuge washing because it gradually flashes off the methanol while it's running. 30 gallons of biodiesel will have a couple gallons of methanol in it. That's all released into the air above the fuge.

FYI, A pressure driven fuge will be more effective at flashing the methanol than a motor driven one.

Especially with KOH, if there is one single ml of methanol left in 200 gallons soap will NOT come out of suspension.

quote:

Originally posted by RickDaTech:
My experience with fuge washing is limited, but I can believe 30 days. I gave up on the concept after a couple of days. The soap can not be mechanically removed until after sufficient methanol has been removed. I don't really recommend fuge washing because it gradually flashes off the methanol while it's running. 30 gallons of biodiesel will have a couple gallons of methanol in it. That's all released into the air above the fuge.

FYI, A pressure driven fuge will be more effective at flashing the methanol than a motor driven one.

So even if you've distilled/boiled/vacuumed off the methanol you're going to have issues with it? That's heavily discouraging...

I wasn't intending to centrifuge raw bio, just the de-methed stuff.

Trying to keep my "Waste" products down and my continuing costs down as well, which was why I was steering clear of the resins (recurring cost + disposal) and the chips (Unfamiliar, but disposal). Water washing is a method of last resort because of the issues of drought and restrictions on usage.

Would it be feasible to do a combination of resin and 'fuge, to extend the life of the resin?