[F250Truckin]

Diesel Forum - TheDieselStop.com > Photo Gallery > F250Truckin's Profile > Albums Centrifuge Basics

Since there seems to be some misunderstanding about centrifuges, I thought I would take a shot at providing some factual information that could help others decide if they wanted a centrifuge and if so what kind.

Centrifuges are very common in our life, we just don't think of them that way. Nearly every doctor's office has one or more and a large hospital would probably have dozens. Having said that, they are mostly very small and useless for our purposes. What the medical centrifuges do is point out just how selective centrifuges can be. Anything that can seperate the red blood cells out of our blood is pretty particular.

Of the types of centrifuges commonly available for working with WMO or WVO, there are two, the open bowl type and the spinner type. There is a third type called a Sharples, but I cannot find any available so they are not worth discussion here.

Spinner type centrifuges are generally small (think flashlight with a six volt battery). They work by having the oil to be cleaned fed through tiny holes at high pressure (commonly 90 psi) that cause the centrifuge to spin at high speed. There are ridges in the cap that cause the dirt to drop out of circulation much like a snow fence or log on a sandy beach stopping wind blown grains of sand.

These centrifuges were originally designed as oil filters for large engines - battle ships in WWI and trucks. They are still available for installation in Chevy Duramax engines and Dodge Cummins engines. I know of no application available for Ford, but I'm sure it probably exists or could be made to work. They were designed to filter the crankcase oil constantly any time the engine was running. They could be cleaned every 5,000 miles or what ever service interval you wanted to set up. They do not handle water well as there isn't really any place for it to go. This type of centrifuge put into service cleaning WMO or WVO generally requires many passes through the centrifuge depending on how big a batch you are making and how dirty the product is. Oil is fed in by a high pressure pump that gives the needed operating pressure. The amount of oil processed is dependent on the time to make the necessary passes to assure clean oil. Pre-filtering is almost a requirement in order to keep from plugging the jets that power the centrifuge.

Open bowl type centrifuges are much larger with bowls a foot or more across. They are powered by their own motor and the oil to be cleaned is either gravity, air pressure (prefered methods) or pump fed. Almost any flow rate can be used. The slower the better because that gives the product more time in the centrifuge thus more "cleaning" time. Spinners would have something like two seconds in the cleanning area where open bowl would have two minutes. This design has oil flooding into the bowl, forced against the side of the bowl or some internal lip. As more oil is added, the "clean" oil is forced over the lip of the bowl or dam and out through a valve or down spout into a clean catch vessel such as a barrel. A second barrel has a seperate spout that collects sludge. Solids such as wear metals, rust particles and dust or sand build up on the wall of the bowl. These centrifuges can handle water, antifreeze and hard particles such as wear metals. Pre-filtering is prefered, but not really required. These centrifuges are capable of handling large quantities of product and can tackle very dirty product and make it into useable fuel.

The down side of all centrifuges is that they have to be cleaned. I have heard recently of a self cleaning centrifuge but have not looked into it.

Cleaning entails taking the lid or top off of the centrifuge when it is stopped and cleaning out the bowl or top that collects the dirt. It is not that big a job but a bit messy and requires some method of disposing of the greasy dirt and sludge that the centrifuge has collected from the oil.

Centrifuges are available readily on eBay and other sources and sold under various trade names. Which type would work for you depends on how much oil you need to process and how dirty it is to start with. Open bowl types are expensive but capable of processing large quantities. Spinner centrifuges are less expensive but limited in capablity.

The attached pictures are of an open bowl and a spinner including a pile of gunk that the open bowl took from 30 gallons of oil.

It would be very useful if someone running a spinner and or WVO would add to this thread.

 

[Jeffrey S Brooks]

Thanks Jerry for the useful information. I am a bit interested in how much sludge is removed from a W-80 blend that has settled for 48 hours.

 

[F250Truckin]

Jeff, I'm still playing with that. I simply centrifuge everything after the 24 hour or longer settling period when I'm sure I should be draining first. I get absolutely nothing in my sludge barrel. After 120 gallons processed, I can't squeeze a drop out of the sludge barrel. I can't eliminate it because it is the backup over flow if something goes haywire with the feed system into the centrifuge. I have a batch settling now and will drain before I fire up the centrifuge this afternoon. I will let you know what I get.

 

[RT]

That's kind of strange - I get about an inch of sludge in my mixing / settling barrel with each batch. I mix my dirty oil with gasoline and let it settle for a few days - then I draw off of that barrel from a tap about three inches from the bottom. We must have better gravity here in Texas. :icon_rolleyes:

 

[F250Truckin]

RT - I think we are mixing apples and oranges here. I get water and sludge off of the bottom of my upflow barrel and I have not been draining off the bottom of my mixing barrel before cenrifuging :nono:. My bad I know. I tried it today and there was really nothing there. I do get more grease like substance on the walls of the bowl after mixing gas so that is probably where it is going. I have work to do on this. My results are not bad as indicated by the lab report. More on that later.

 

[RT]

RT - I think we are mixing apples and oranges here.

Fruit Salad! :lol:

I don't have an upflow barrel so to speak - just a mixing barrel and a clean barrel. I mix my oil and gas in the first and let it settle. Then I pull off the side of that barrel - 3 inches from the bottom - through my centrifuge which is mounted on the top of my clean barrel. When I have transferred enough oil to the clean, I use a valve to change the input to my pump from the mix barrel to the clean barrel and let it run a few hours recirculating through my PA biodiesel centrifuge. When I clean out my mix barrel - I've got about an inch of sludge on the bottom.

 

[Jeffrey S Brooks]

Jeff, I'm still playing with that. I simply centrifuge everything after the 24 hour or longer settling period when I'm sure I should be draining first. I get absolutely nothing in my sludge barrel. After 120 gallons processed, I can't squeeze a drop out of the sludge barrel. I can't eliminate it because it is the backup over flow if something goes haywire with the feed system into the centrifuge. I have a batch settling now and will drain before I fire up the centrifuge this afternoon. I will let you know what I get.

It sounds like you are saying that your centrifuge removes nothing after you have blended and settled your oil.

How about you, RT, is your centrifuge removing anything from your blend after settling?

 

[F250Truckin]

Wrong concept Jeff. The centrifuge is removing plenty, but what it is removing is stuck to the wall of the centrifuge bowl just like it is suppose to. That is not happening that should is that there should be liquid sludge that drains away into the sludge barrel then the centrifuge is shut down. It is too heavy to escape with the good oil and not thick enough to stick to the wall of the centrifuge. That is the theory. I'm getting the thick stuff but not the thinner liquid sludge. I'm thinking that may already be gone because of the up flow barrel. I'm still working on it.

 

[thecontrolguy]

Centrifuge Crud

So, after settling (with gas added), up-flow processing and then heating and centrifuging, the pics show what came out of a 55 gallon batch circulated for 48 hours and then pumped into fuel totes for the truck. Batch consisted of old D2, used motor oil from the Subaru, the 7.3l Ford and some mystery oil from donations at work. The crud is a lot like warm tar, easily scraped off the inside of the 'fuge with a silicone spatula. I follow this with 30 and then 5 micron spin-on fuel filters. Truck runs noticeably better from the waste-oil tank than the straight D-2 tank. This is NOT the same type of crud (liquid) that accumulates at the bottom of the settling or the up-flow barrels. It is VERY concentrated. although consistency is a ???. No apparent water in the 'fuge, so maybe there was no water in any of the oils, or the heating drove it off, or it now sits at the bottom of my up-flow tank. Cheers to all.

 

[F250Truckin]

That's about what my crud looks like. You must start with cleaner oil than I do because I get a lot more than that. Sure not anything you want in your lub oil or fuel.

 

[Jeffrey S Brooks]

So, after settling (with gas added), up-flow processing and then heating and centrifuging, the pics show what came out of a 55 gallon batch circulated for 48 hours and then pumped into fuel totes for the truck. Batch consisted of old D2, used motor oil from the Subaru, the 7.3l Ford and some mystery oil from donations at work. The crud is a lot like warm tar, easily scraped off the inside of the 'fuge with a silicone spatula. I follow this with 30 and then 5 micron spin-on fuel filters. Truck runs noticeably better from the waste-oil tank than the straight D-2 tank. This is NOT the same type of crud (liquid) that accumulates at the bottom of the settling or the up-flow barrels. It is VERY concentrated. although consistency is a ???. No apparent water in the 'fuge, so maybe there was no water in any of the oils, or the heating drove it off, or it now sits at the bottom of my up-flow tank. Cheers to all.

There is a distinct problem with the above sequence. If you were blending waste oil with gasoline at sufficient levels, then there would be no reason to heat your blend before centrifuging. Also, up-flow processing introduces turbulence, which I do not believe is helpful. You also did not mention how long you let your blend settle, or what percentage you blended at.

So, just to clarify the working model that most of us are working with is: Blend gasoline with waste oil at 18-20%. Let settle for at least 24-48 hours, this means no upflow. Then, up-flow if you want for as long as you want. Then centrifuge, then filter. My working hypothesis is if you blend gasoline at 20% with waste oil and let settle undisturbed for 24-48 hours, then there will be very little for the centrifuge to take out.

A worthy experiment would be to work with a common batch of waste oil and process it in sub-batches. Say, start with a full 250 gallon tote of WMO. Then process 20-50 gallon batches, and increase the settling time by one-day increments followed by centrifuging for each batch, then weight the resulting matter that comes out of the centrifuge. This should tell us how long the settling period should be.

 

[thecontrolguy]

There is a distinct problem with the above sequence. If you were blending waste oil with gasoline at sufficient levels, then there would be no reason to heat your blend before centrifuging. Also, up-flow processing introduces turbulence, which I do not believe is helpful. You also did not mention how long you let your blend settle, or what percentage you blended at.

So, just to clarify the working model that most of us are working with is: Blend gasoline with waste oil at 18-20%. Let settle for at least 24-48 hours, this means no upflow. Then, up-flow if you want for as long as you want. Then centrifuge, then filter. My working hypothesis is if you blend gasoline at 20% with waste oil and let settle undisturbed for 24-48 hours, then there will be very little for the centrifuge to take out.

I added gasoline to be approximately 15 - 20% of the waste oil at the time it (the oil) was added to the settling tank. D2 went in with no adjustments. It settled for several DAYS, undisturbed. Then I recently processed a bunch of fuel (100 gallons +). Each time (twice) I cleaned the 'fuge yielded similar results as to crud volume and consistency. I heat after adding the gas, as I feel it helps the 'fuge seperate the very small fines by further decreasing the relative viscosity of the oil. It (heat) would also seem to help drive off any remaining water in the mix, if there were any. I have also not found any evidence that the gas volatilizes during this process (no gas smell from the tank vent) and so assume that, acting as a solvent, it has perhaps bound chemically in the oil mix?

Anyway, as discussed in another thread, I am moving towards a single (large) settling tank and no other settling or upflow tanks. I have already experimented with moving the oil through the primary W.W filters (30mic) into the 'clean' barrel with compressed air and it works very well; much better than rotary cranking. :thumbsup: Thanks for suggesting that.

The 'big' tank is approx 48" long by 29" so approx. 130 gallon. I am presently working up a base for it to enable access to the bottom drain and still be able to easily reach the top. Tank will be mounted horizontally. Pictures will follow.

 

[F250Truckin]

Jeff - I'm thinking that what you are seeing is not really very much coming from 50 or more gallons.

I'm planning to set up a mini experiment using a 5 gallon jug. Mix 15 gallons of WMO so you are experimenting with the same base, then run three 5 gallon samples using different % or gasoline, other solvents, different settling times and any other differences you can think up and check your results.

 

[JR3342]

Here's most of what I know about centriifuging wvo. Tried cold upflow got clean wet oil. Heat and settle, frybridstill, rock salt and baking soda, salt collum. None of these could pass hot pan test for suspend water. Got a spinner cf 60 gph and pump from pa bio supply. From all the failed methods and 8 mos of settling this was well filtered oil. Set up cf in open top insulated barrel with 1500w immersion heater. Put in first batch heated to 140 deg. Started cf ran batch for 4 hrs cleaning rotor every 1 hr. Lots of gunk in rotor and visable water. From 160 to 185 deg steam was coming from the rotor. Oil passed hpt perfectly no bubbles from water at all. All the bag filters and other mess is gone all the settling barrels gone. Now I pour oil into cf barrel through window screen or sewn shut jeans legs to pre filter and cf it and pump into ibc for storage. 3000 mi I change fuel filter on car no restrictions showing on gauge. I still pump oil thru 2 mic cat spinon before storagand while filling car just for insurance.bigest expense now is paper towels to clean rotor. Biggest tip I learned is put a cup under rotor of cf to catch output when it is slowing down to keep the crap in rotor from falling into clean oil.

 

[Jeffrey S Brooks]

I added gasoline to be approximately 15 - 20% of the waste oil at the time it (the oil) was added to the settling tank. D2 went in with no adjustments. It settled for several DAYS, undisturbed. Then I recently processed a bunch of fuel (100 gallons +). Each time (twice) I cleaned the 'fuge yielded similar results as to crud volume and consistency.

Thanks, thecontrolguy, for the clearer description of what you are doing.

I heat after adding the gas, as I feel it helps the 'fuge seperate the very small fines by further decreasing the relative viscosity of the oil. It (heat) would also seem to help drive off any remaining water in the mix, if there were any. I have also not found any evidence that the gas volatilizes during this process (no gas smell from the tank vent) and so assume that, acting as a solvent, it has perhaps bound chemically in the oil mix?

Since water has a higher boiling point than gasoline, and gasoline volatilizes at low temperatures, and boils at 160F, and it has a low flash point, it seems unsafe to heat a blend containing gasoline. Further, there seems to be no reason to heat a waste oil blend containing gasoline at 15% or more. What temperature did you heat your blend to?

Anyway, as discussed in another thread, I am moving towards a single (large) settling tank and no other settling or upflow tanks. I have already experimented with moving the oil through the primary W.W filters (30mic) into the 'clean' barrel with compressed air and it works very well; much better than rotary cranking. :thumbsup: Thanks for suggesting that.

The 'big' tank is approx 48" long by 29" so approx. 130 gallon. I am presently working up a base for it to enable access to the bottom drain and still be able to easily reach the top. Tank will be mounted horizontally. Pictures will follow.

I look forward to seeing your photos, and you are welcome. I am glad some of my suggestions have helped you.

 

[Jeffrey S Brooks]

Jeff - I'm thinking that what you are seeing is not really very much coming from 50 or more gallons.

I'm planning to set up a mini experiment using a 5 gallon jug. Mix 15 gallons of WMO so you are experimenting with the same base, then run three 5 gallon samples using different % or gasoline, other solvents, different settling times and any other differences you can think up and check your results.

I think some experiments are certainly in order. So, if I had a centrifuge I would do those experiments, so I am glad you are going to do them.

It is not my agenda to disprove the usefulness of the centrifuge; however, I am interested in proving to myself that it deserves the investment, because I am very happy with my process of blending at 20%, let settle for 24-48 hours, then I drain off the sludge, which is often about 10% of the blend, then I filter down to 1-micron. I end up with no precipitates coming from my fuel samples, so I know my process works well enough to justify not purchasing a centrifuge.

A reasonable series of experiments could be easily conducted by blending a large quantity of WMO at 20%, and leave it to settle, then after the first day of settling draw off the top a batch, say 5-20-gallons, then run it through a centrifuge, then weight the amount of sludge extracted. Then the next day do the same thing with the same size sample from the same batch. And, do that every day until the amount of sludge removed by the centrifuge is the same as the previous day's sample. At that point one would know how long to settle the WMO.

 

[thecontrolguy]

..SNIP..

It is not my agenda to disprove the usefulness of the centrifuge; however, I am interested in proving to myself that it deserves the investment, because I am very happy with my process of blending at 20%, let settle for 24-48 hours, then I drain off the sludge, which is often about 10% of the blend, then I filter down to 1-micron. I end up with no precipitates coming from my fuel samples, so I know my process works well enough to justify not purchasing a centrifuge.

A reasonable series of experiments ...

I think that if you settled for whatever interval is normal for you, filtered as you normally do, and THEN the fuel that you would normally pour into your van / truck was centrifuged, it would be interesting to see what, if anything, the centrifuge pulled out of your 'clean' fuel. Is this the minimum filtration that is required for your type of vehicle and fuel system or would this work in all of our vehicles? I am fairly concerned about the health of my expensive HEUI injectors and feel that I need to process the fuel to whatever best standard I can get so it would be cool to see who's 'clean' fuel was actually clean to the minimal standards of pump D2. (which, around here, is not so clean actually) Or, better, I guess?

I would actually have tried this myself but I am a few days away from having a new setup, without an up-flow, and my 'plant' is in pieces again.

Should we have a new series of threads / stickies documenting each of the different types of processes and the inherent characteristics / issues associated with each as documented by the actual users.

Cheers, Jim

 

[Jeffrey S Brooks]

Should we have a new series of threads / stickies documenting each of the different types of processes and the inherent characteristics / issues associated with each as documented by the actual users.

Cheers, Jim

That sounds like an interesting idea.

 

[RT]

Jeff - your proposal above is basically what I do - and you are right on. I don't drain the sludge, rather I drain the oil from a bung 4 inches above the bottom of the barrel through my centrifuge into the clean barrel. This is after mixing to 18% and settling for two days to a week. The centrifuge (PA Biodiesel running at 95 psi) pulls out about a cup of uber-fine grain sludge after running for about 4 hours. There is about an inch of sludge in my settling barrel (probably the equivalent of a gallon) when I go to clean it. So the settling is taking about 95% of the crap out, the centrifuge the remaining 5% - I then pull it through a 2 micron filter before putting it in my truck.

 

[thecontrolguy]

Since water has a higher boiling point than gasoline, and gasoline volatilizes at low temperatures, and boils at 160F, and it has a low flash point, it seems unsafe to heat a blend containing gasoline. Further, there seems to be no reason to heat a waste oil blend containing gasoline at 15% or more. What temperature did you heat your blend to?

I am heating to a MAX of 120F; the small hot water tank cycles between about 105-120F. Usually, it sits around 110F or less. The system is closed-loop - that is, no open vents unless I am pumping out (clean) processed fuel, then an auto vacuum-breaker opens to allow air in so the clean drum doesn't collapse. When I am pushing oil in via pneumatic pressure on the settling tank, I open the sludge/overflow drain on the 'fuge to release air so the clean drum doesn't bulge like Oprah W. :lol: This is the ONLY time air is released from the system and it is always during a cold-fill cycle.

So, the REAL question I have, and was planning to post as a new thread - IS GASOLINE MISCIBLE in Oil? I ask this as I have not (yet) smelled any gas vapor at any time except when I pour a bunch into the settling tank and the gas jug and funnel are open to the atmosphere. I expect that the gas is chemically bound into the oil, or vice-versa, and therefore its vapor-pressure, boil temps, and other characteristics are altered. Perhaps someone's Uber-Smart Chemical Engineer Son could shed some light on this? :icon_wink: Anyone?