Cold Upflow Settling
Written by John Galt
Dry fuel is the single most important factor for successful long term operation of any diesel fuel engine or turbine. Vegetable oil, and especially the contaminants in used VO absorb and retain moisture/water. If one takes a well mixed sample of used VO one will very likely find enough moisture present to show on a hot pan test. If one then chills the used VO and lets it settle so the water absorbing contaminants [animal fats, trans-fats, saturated fats, hydrogenated oil, food particles, burnt bits, etc] fall out of the mix, then HP tests the two fractions, where would you guess most of the moisture is?
I use cold upflow settling to remove all those contaminants listed above, along with all the water which will react to a HPT. Clear, dry, clean VO with NO contaminants gives the fewest problems when used as diesel fuel, regardless if its being used as 100%SVO, blended VO-ULSD mixes, or transmogrified into biodiesel.
Plant oils are generally too viscous to be used directly in diesel engines unless the ambient temperature is above ~80°F, and even then only in some of the more simple mechanically controlled engines. To use veg oil in all engines in temperatures below 80°F the viscosity must be reduced.
Three methods are generally used to lower the viscosity:
- Convert the oil chemically to biodiesel,
- Heat the fuel system to above 80°F and the oil to >150°F before the IP,
- Dilute the oil with solvents, after removing the oils and fats that separate out below 80°F.
Older simpler engines in warm climates where it never freezes are much more tolerant to fuel contaminated with water and other stuff. Modern engines operating in frigid climates are very sensitive to contaminated diesel fuel. Pilots in the north always drain samples from the bottom of every fuel tank and visually inspect the sample for any cloudiness indicating moisture or other contamination, before taking off. Where they work there is no margin for error.
This drawing shows a simple VO cleaning system.
The processor consists of a 10 gallon barrel with the bottom cut out, and mounted open bottom up bung-to-bung with a 2" close nipple onto a vertical 55 gallon settling barrel. A 2ft piece of 2" exhaust pipe with the top end flared, sits inside the nipple coupling and ensures that the new oil is delivered near the bottom of the settling barrel. Steel pipe is rated by inside diameter and exhaust tube is rated by outside diameter, thus they fit well together.
Primary filtration is through bugscreen and pantyhose on a simple 2x2 wood frame that sits on the rim of the 10 gallon 'funnel'. The oil settles in the barrel and every time I add oil into the settling barrel, clear oil is forced out of the 3/4" bung through a street elbow, a hose bib valve and through a clear vinyl tube to the canvas bag 'jeans leg' ~ 20 micron bag filter assembly, into the final filter barrel. I find that the 30 gal poly barrels are more convenient for the 2nd stage, and use two of them as collection/final filtering barrels.
For making biodiesel, the oil off the top of the first barrel is ready for processing.
Final filtration is by rotary hand pump, through a whole house filter assembly with a 5µ filter cartridge. The barrel pump I use is a rotating-vane, positive-displacement, self-priming type which can pull or push equally well. The 2nd barrel can be heated or the VO diluted with ULSD or kerosene, to make final filtering easier.
This system works best with no heat at temperatures below 65°F to remove saturated fats and hydrogenated oil, or low power [less than 100W] evenly distributed heat if one wants to keep them in the mix. If one does a hot pan test on the oil in a cubie, one will discover that as one goes deeper in the cubie the amount of moisture increases. In my experience most of the moisture is in the sludgy emulsion of animal fats, hydrogenated oil, and food particles in the bottom layer. Since this is the stuff that causes most of the problems with cold blend fuel systems, I use the unheated upflow settling to separate and remove the troublesome sludge and the moisture it contains.
The canola oil I use is clear and reasonably dry to start with, hasn't been mixed with animal fats, and my supplier puts it back into sealed cubies while it's still warm. The cubies sometimes settle for months and I decant the clearest portion off the top into the processor. A cubie spout makes it easier. I pour a bucket or cubie of oil into the 10 gallon barrel and let it do it's thing. The oil that comes off the top of the primary filtering settling barrel is not wet.
This VO cleaning system is based on the simple principle that water, wet oil, fats, and most other contaminants are heavier than clear dry oil. Suspended water and oil bonded with suspended water is heavier than clear dry oil, therefore it will not float to the top. Because the 'new' oil is placed at the bottom, any water, wet oil or food particles will NOT float to the top if the oil is not heated. If there are no convection currents to mix the oil then the clear dry oil comes to the top and the contaminants stay on the bottom. This cleaning process is called "Upflow Settling". This first stage gets VO so clean that very little builds up in the bag filter and the cartridge filter is good for hundreds of gallons.
Observations indicate that a couple of material properties make upflow separation work. Water molecules are more likely to be attracted to, and bond with, other water molecules, food particles and hydrogenated oils and fats, rather than bond to clear dry oil molecules. Because these contaminants are all heavier than clear dry oil, they fall to the bottom of the barrel. The 2" drop tube introduces the newest oil into the bottom 1/3 of the barrel, this gives any moisture laden particles and entrained water the opportunity to bond with the sludge. Without the drop tube the system won't work. The other factor is that the clear dry oil molecules are the lightest substances in the whole mix, and if the mix is introduced near the bottom, only the lightest molecules will float to the top. The water, moisture, entrained water, dissolved water,suspended water,...a.k.a. whatever, stays at the bottom if there are no convection currents, because it is heavier than clear dry oil. With time, water molecules, free or attached find other water molecules and bond, this process eventually makes drops of water large enough to sink to the bottom.
The slow rate of new oil introduced is controlled by the valve on the clear oil output tube to about 10gal/hr. No modifications to the 2" x 2' dip tube required. The valve can be adjusted so the oil is introduced slowly to reduce any turbulence mixing the sludge in the bottom 1/4 of the barrel. Only clear dry oil comes off the top, it couldn't be easier.
In this cold climate I don't use solid oil for motor fuel, at any time of the year. This settling process removes it as a fuel system contaminant. The thick, wet sludge is occasionally pumped off the bottom of the barrel and mixed with sawdust for woodstove fuel. That sludge is a valuable resource for me, when mixed with sawdust from chainsawing fuel wood and packed into 1/2 gallon [2 Liter] milk cartons. One of those on a wood fire first thing in the morning quickly warms the house right up to comfortable temperature, especially when it's 30 below. There's a lot of BTUs in a half gallon of transfat oil/fat soaked sawdust.
Cold processing is more effective with reasonably clear used oils that don't have a lot of hydrogenated oil and fats, and may not work with all oils. The sort of wet, goat-vomit, grey-mayonaise, hydrogenated crap that some have to contend with probably won't come out much cleaner. Cold Upflow Settling basically separates heavier crud from clear oil.
The processor is in an 8 x 8 unheated shelter tent. The processing for winter oil 50VO/50ULSD base mix is done by freeze-up in late October. Even when it's below freezing the system will work to about 0°F to process clear liquid canola. No added heat, just gravity. The system has been producing clear clean dry UVO since 2007.
This is the simplest system that requires no electricity. It can be easily assembled with a few commonly available materials. For those who want more, it can be expanded easily with additional barrels and a small electric transfer pump.
It won't produce huge volumes quickly, but it can be up and running for less than $100 while you design, develop and assemble your Ultimate-SuperMega-HyperspaceCentrifuge-VO Processor.
The basic principle of upflow settling can be successfully incorporated into other cleaning systems..