When you're confident that you can get good results every time, even using oil from different sources, then it's time to scale up the process to provide your fuel needs. Now that you have a feel for the process and know what to expect, you'll have a much better idea of what sort of processor you want than if you'd started off building the processor (as many do) rather than learning the process first.

"Understanding of the process is vital to operate the plant." -- Prof. P.V. Pannir Selvam, Technology Center, Department of Chemical Engineering, Universidade Federal do Rio Grande do Norte (UFRN), Brazil, Biofuel mailing list, 15 Apr 2007

See Biodiesel processors.

However, one-litre test batches are not just something for beginners. It's a basic technique you'll always use. Many experienced biodiesel makers do test batches with each batch of oil. Many not only titrate the oil every time to calculate the right amount of lye to use, they also do "bracket" tests in sequence, followed by wash tests. You learn a lot that way, your fuel gets better, life gets easier.

In fact life is already easier -- people who start off making 40-gallon batches often never learn the accuracy and discipline that comes from making one-litre test batches first. Their fuel quality suffers for it, and when they encounter that inevitable "problem batch", they suffer for it too.

But if you've followed the instructions here carefully, you'll be familiar with all the variables, you'll have good methodology, and you'll be in a much better position to trouble-shoot a problem batch successfully.

Keep a Biodiesel Journal -- make notes, keep records. Get some small glass jars and keep samples of all your batches, clearly labelled and cross-referenced to the notes in your journal. You won't regret it.

When scaling up from small test-batches to a full-sized processor, be aware that the process will probably need some adjusting. All the various processing methods use averages and approximations because processors and conditions vary so widely. Blenders especially agitate much faster than any full-scale processor, a real processor will probably take longer to achieve the same result. Use the fuel quality tests to fine-tune the process to your particular processor. See Scaling up.

Removing the water

Water in the oil interferes with the lye catalyst, especially if you use too much lye, and you can end up with a batch of jelly.

Test first for water content -- heat half a litre or so of the oil in a saucepan on the stove and monitor the temperature with a thermometer. If there's water in it it will start to "snap, crackle and pop" by 50 deg C (120 deg F) or so. If it's still not crackling by 60-65 deg C (140-149 deg F) there should be no need to dewater it.

See Mike Pelly's recommendations: Removing the water.

Here's another way, from Aleks Kac -- it uses less energy and doesn't risk forming more Free Fatty Acids (see below) by overheating. Heat the oil to 60 deg C (140 deg F), maintain the temperature for 15 minutes and then pour the oil into a settling tank. Let it settle for at least 24 hours (or for a week or two). Pump the oil out from the top, leave the bottom 90% for removal later and re-settling.

Here's what Biofuel mailing list member Dale Scroggins says about water removal:

Water in vegetable oil can exist as free water, which will eventually settle to the bottom of a vessel; as suspended droplets, which may settle if the oil is heated, or the droplets are coalesced; and as water in solution with other impurities in the oil. Free water is the easiest to remove. The droplets are removed most efficiently by coalescing and draining. Suspended droplets that cannot be coalesced and water in solution are more problematic.

Boiling off the water is more difficult than it appears on the surface. Colligative properties of solutions (and some mixtures) can make removal of the last traces of water almost impossible. Water mixed with oil will not boil at the same temperature and pressure as pure water. As water is removed, more heat or lower pressure will be required to remove more water. If the oil contains salts or semi-soluble fatty acids, distillation is even more difficult.

As the percentage of water in the solution decreases (its molar fraction) its vapor pressure will continue to drop. Lowering pressure in the system alone may be insufficient to sustain vaporization when the solution becomes concentrated (the molar fraction of the solute greatly exceeds that of the solvent). Results will vary depending upon the nature of the water-soluble impurities in the oil. Few solutions are ideal, in terms of Raoult's law, and in used vegetable oil, there is no way to know what solutes are in the oil.

The important thing is how well-used, or overused, the oil is. Titration will tell you that. The higher the titration result and the more acidic the oil, the more water it's likely to contain, and the more difficult it will probably be to remove the water.

Biofuel list member Joe Street adds:

Although Dale's points about unknown solutes in waste vegetable oil and their ability to lock up water are true I have found that practically speaking oil (even terrible oil) can be reliably dried to the point of being reacted without problems by the process of heat and vacuum.

Heating the oil to reaction temperature (I use 58 deg C, 136 deg F) and pulling a vacuum to 27 inches of mercury immediately before the reaction has allowed me to remove water beyond what falls out by heating alone.

I estimate that oil at 55-60 deg C can contain as much as 10,000 PPM (that's 1%) water. I have experimented with some extremely saturated oils (titrations up to 11 ml with 0.1% KOH) which require ridiculous amounts of catalyst when attempting base-only conversions. Although I cannot get complete conversions in these cases, drying the oil by heating and vacuum has allowed me to do this and still avoid problems with soap formation. (I am also very careful with my caustic and methanol.) Using a known temperature and a vacuum gauge is a very repeatable way of drying oil.

(See: Joe Street's processor)

Short of vacuum, start with heating the oil to 60 deg C and settling it, as Aleks Kac recommends, and if that doesn't give satisfactory results, try boiling the water off, as Mike Pelly recommends. Then try a small 1-litre test batch first.

If you still have difficulties, try to find a source of better-quality oil.

Or try using a glycerine pre-wash to lower the Free Fatty Acid level and dry the oil (see below, Glycerine pre-wash).

Or try this: Deacidifying WVO.

Filtering WVO

Many people filter their WVO before making biodiesel, but filtering takes time and energy, and there's really no need to filter it.

Settling the oil works just as well or better, and if it contains any water the water will settle out too.

If it's poor quality oil with a high titration level, heat it first, as in de-watering (above), and then let it settle.

If you don't have time to wait for the oil to settle, usually 1-2 weeks, it could be worth increasing the WVO supply and reserves to make the time.

If in collection you keep ahead of your processing rate, oil has a chance to settle. I have found that oil that has been sitting for several weeks is very dry if carefully decanted. Settling also results usually in oil which is spectacularly clear when observed in a glass container (you can read fine print through it) which means it is quite clean, perhaps cleaner than filtering may give you.
-- Joe Street, Biofuel mailing list, July 2006

I recently helped someone get off the ground making biodiesel. He's a tinkerer, and came up with an elaborate filtering/dewatering system. I repeatedly suggested that he trust gravity. He was away for about 10 days and when he came back he called to tell me that he couldn't distinguish the oil from the top half of an unfiltered cubie from his filtered oil. Getting rid of his filtering setup has made room for a settling tank.
-- Tom Kelly, Biofuel mailing list, April 2006

This is how Tom does it:

I allow the WVO to settle in cubies for a week. (A cubie is the 4.5 gal (17.7L) plastic container that veg oil is delivered to restaurants in.) I then pour the top 80% of each cubie into a 55 gal drum and consolidate the bottom 20% of 5 cubies into 1. Most of this will be ready for the barrel the next week. I have 4 WVO barrels. One is settled, two are settling, and one is being filled. I pump WVO out of the settled barrel from the top 3/4. This oil is very clear and requires very little drying.

We do it much the same way, settling the WVO first in the 18-litre metal cans it's supplied in here, then pouring it from the top. What's left at the bottom is re-settled.

We use a 55-gal (200-litre) steel drum for storage, but we don't pump the WVO out from the top. The drum has a bottom drain fitted with a 6"-high 3/4" standpipe (15cm-high x 1.9cm), which leaves any sediment on the bottom of the drum undisturbed.

Every now and then we drain the drum to the top of the standpipe, then remove the standpipe and drain the drum completely, sediment and all. The "bottoms" are resettled the same way, first in 18-litre metal cans.

The final sediment can be used as fire-starter, or added to the compost pile.

Simple gravity settling works well with oils titrating up to 3.5 ml NaOH solution and more.

The higher the titration level, the more water, impurities and suspensions the oil is likely to contain and the longer it will take to settle. For higher titration levels, heat the oil to 60 deg C (140 deg F), maintain the temperature for 15 minutes, then allow to cool, and let it settle.

Filtering biodiesel

There's no need to filter your biodiesel before using it either.

If you make the biodiesel properly, everything that a filter might remove will be in the by-product layer, not the biodiesel.

The biodiesel should be ready for instant consumption if it's clear and bright and without sediments.
-- Jan Warnqvist, Biofuel mailing list, Aug 2005

If the WVO has been filtered or settled to clear, any solid particles that get as far as the processor are small and won't affect the processing. During processing and settling, all unfiltered solids drop out into the glycerine by-product layer.

Settle it properly, separate the by-product carefully, wash, dry, and use. No need to filter it.

Badly processed biodiesel with poor conversion and too much soap might contain sediments in suspension, but if it's properly made it will be without sediments. Keep your processing fine-tuned by making test batches and using the quality tests.

People often want to "speed up" the process in the hopes of making it more efficient, and that often means taking short-cuts with settling times.

Don't do it -- life is easier with longer settling times, for the WVO, for the by-product to separate after processing, and for the wash-water to separate, especially after the final wash.

Actually what people speeding up the process usually want isn't more efficiency, it's more production. Probably they need a bigger processor, or two processors in parallel, rather than trying to make gravity hurry. Extra washing and settling tanks also help.

Centrifuges

We don't have much time for centrifuges either, for similar reasons.

Again, people want to "speed up" the process by using a centrifuge instead of settling (or instead of washing, in some cases).

We've had a number of reports that centrifuges give poor results compared with normal settling, and especially compared with washing -- washing centrifuged fuel has yielded very soapy wash-water, so obviously the centrifuging didn't work very well.

We have laboratory test results of "finished" biodiesel made here in Japan in a $70,000 commercial processor. The finished fuel was washed and dried, and then it was centrifuged, and then samples were sent to the lab for testing. But the centrifuged fuel didn't come close to the standards requirements.

Our biodiesel surpasses the standards requirements though, made in a $100 homebuilt processor, with no need for a centrifuge.

You don't need a centrifuge. As with filtering, if you want more production get a bigger processor, or run two processors in parallel, add more settling and washing tanks.

Glycerine pre-wash

From Biofuel mailing list member Chris Tan: "Good use for your glycerine cocktail", 6 Oct 2007:

Here's a good use for your glycerine cocktail before finally giving it away. My father came up with the idea that you can use the glycerine cocktail to dry your waste vegetable oil. And it works. Glycerine is hygroscopic enough to pull moisture out as it settles down so you don't have to heat or boil the oil to dry it. And as a bonus, most of the catalyst ends up in the glycerine cocktail so it neutralizes the Free Fatty Acids (FFAs) in the waste oil.

What we do is use at least 10/90 weight ratio: 10kg of glycerine for 90kg of waste oil. It is possible to bring the FFA level to zero if you use large amounts of glycerine (if you happen to have accumulated large amounts).

We use an ordinary 1/2 hp clear water pump with two inlet pipes to suck in glycerine and wvo. I adjust the inlet openings to regulate the mixing. We let it settle in a dedicated separate tank for about the same time as you would settle glycerine from biodiesel, though longer is better because of the viscosity of wvo.

We pump both glycerine cocktail and oil at the same time into a separate container. The glycerine will flow at a lower rate and the the inlet opening should be adjusted so that it finishes at the same time as the oil.

You can also mix it in the processor (just make sure to drain the water-rich glycerine and soap residue afterwards). Cycle the glycerine cocktail and oil mixture twice. The time it takes will depend on the gallons per minute rate of your pump, which is not the same for viscous oil, so measure and compute the time required for one or two cycles.

If the weather is cold, warming the oil first helps for mixing as well as settling: pre-heat to about 30 deg C (86 deg F).

-- Chris Tan

Washing

Biodiesel must be washed before use to remove soaps, excess methanol, residual lye, free glycerine and other contaminants. Some people (fewer and fewer of them) say washing isn't necessary, arguing that the small amounts of contaminants cause no engine damage.

Read what the Fuel Injection Equipment (FIE) Manufacturers (Delphi, Stanadyne, Denso, Bosch) have to say about these contaminants:
Summary -- html
Full document -- Acrobat file, 104kb

See also: Determining the Influence of Contaminants on Biodiesel Properties, Jon H. Van Gerpen et al., Iowa State University, July 31, 1996 -- 12,000-word report on contaminants and their effects. Acrobat file, 2.1Mb:
http://www.biodiesel.org/resources/reportsdatabase/reports/gen/gen014.pdf

Myth:

> I did notice that a lot of the chemistry in the book was wrong.
> His main argument seemed to be against losing the energy in
> the methanol that was washed out.

The "energy" does you no good if your particular thermodynamic cycle can't take advantage of it. What is the cetane rating of methanol?
-- Ken Provost, Biofuel mailing list, "Re: washing?"

Quite so. The cetane rating of methanol is only 3, very low. Low cetane-number fuel in a diesel causes ignition delay and makes the engine knock. The high-speed diesel engines in cars and trucks are designed to use fuels with cetane numbers of about 50. The US biodiesel standard specifies a cetane number higher than 47, the EU standard specifies higher than 51. The methanol in unwashed biodiesel doesn't "make a great fuel anyway". It's also very corrosive. The EU biodiesel standard specifies less than 0.2% methanol content.

Quality biodiesel is well-washed biodiesel. Filtering it is no use, and letting it settle for a few weeks won't help much either. Anyway washing the fuel is easy.

See Washing

Using biodiesel

You don't have to convert the engine to run it on biodiesel, but you might need to make some adjustments and you should check a few things.

Petroleum diesel leaves a lot of dirt in the tank and the fuel system. Biodiesel is a good solvent -- it tends to free the dirt and clean it out. Be sure to check the fuel filters regularly at first. Start off with a new fuel filter.

If a car has been left standing for a long time with petroleum diesel fuel in the tank the inside of the tank may have rusted (water content is a common problem with petro-diesel fuel). Biodiesel will free up the rust, and it could clog the particle filter inside the tank. At worst the car simply stops, starved of fuel. It's not a very common problem, but it happens. See: Biodiesel and your vehicle -- Compatability: Filters.

A common warning is that biodiesel, especially 100% biodiesel, will rot any natural or butyl rubber parts in the fuel system, whether fuel lines or injector pump seals, and that they must first be replaced with resistant parts made of Viton. But rubber parts in diesel engine fuel systems have been rare or non-existent since the early 1980s -- it seldom happens, and when it does happen it's not catastrophic, you have plenty of warning and it's easily fixed. See: Biodiesel and your vehicle -- Compatability: Rubber.

See Biodiesel and your vehicle

Safety

Please read this whole section right to the end.

Wear proper protective gloves, apron, and eye protection and do not inhale any vapours. Methanol can cause blindness and death. Sodium hydroxide and potassium hydroxide can cause severe burns and death. Mixed with methanol they form methoxide. This is an extremely caustic chemical.

These are dangerous chemicals -- treat them with respect! Gloves should be chemical-proof with cuffs that can be pulled up over long sleeves -- no shorts or sandals. Always have running water handy. Have a bottle of vinegar handy to neutralise any lye or methoxide you may get on your skin -- rinse it off with vinegar, then rinse thoroughly with water. (If you don't have any vinegar handy, just use lots of water.) The workspace must be thoroughly ventilated. No children or pets allowed.

Organic vapor cartridge respirators are more or less useless against methanol vapors. Professional advice is not to use organic vapor cartridges for longer than a few hours maximum, or not to use them at all. Only a supplied-air system will do (SCBA -- Self-Contained Breathing Apparatus).

The best advice is not to expose yourself to the fumes in the first place. The main danger is when the methanol is hot -- when it's cold or at "room temperature" it fumes very little if at all and it's easily avoided, just keep it at arm's length whenever you open the container. Don't use "open" reactors -- biodiesel processors should be closed to the atmosphere, with no fumes escaping. All methanol containers should be kept tightly closed anyway to prevent water absorption from the air.

We transfer methanol from its container to the methoxide mixing container by pumping it, with no exposure. This is easily arranged, and an ordinary small aquarium air-pump will do. The methoxide is mixed like this -- Methoxide the easy way, which also happens to be the safe way. The mixture gets quite hot at first, but the container is kept closed and no fumes escape. When mixed, the methoxide is again pumped into the (closed) biodiesel processor with the aquarium air-pump -- there's no exposure to fumes, and it's added slowly, which is optimal for the process and also for safety. See Adding the methoxide.

Once again, making biodiesel is safe if you're careful and sensible -- nothing about life is safe if you're not careful and sensible! "Sensible" also mean not over-reacting, as some people do: "I'd like to make biodiesel but I'm frightened of all those terrible poisons." In fact they're common enough household chemicals. Lye is sold in supermarkets and hardware stores as a drain-cleaner, there's probably a can of it under the sink in most households. Methanol is the main or only ingredient in barbecue fuel or fondue fuel, sold in supermarkets and chain stores as "stove fuel" and used at the dinner table. It's also the main ingredient in the fuel kids use in their model aero engines. So get it in perspective: be careful with these chemicals -- be careful with ALL chemicals -- but there's no need to be frightened of them.

For fire risks, see Hazards

More about methanol

Question: Just how dangerous is methanol?

Fact: Methanol is a poisonous chemical that can blind you or kill you, and as well as drinking it you can absorb it through the skin and breathe in the fumes.

Question: How much does it take to kill you?

Short answer: Anything from five teaspoons to more than half a pint, but nobody really knows.

Fact: Human susceptibility to the acute effects of methanol intoxication is extremely variable. The minimum dose of methanol causing permanent visual defects is unknown. The lethal dose of methanol for humans is not known for certain. The minimum lethal dose of methanol in the absence of medical treatment is put at between 0.3 and 1 g/kg.

That means it's thought to take at least 20 grams of methanol to kill an average-sized person, or 25 ml, five teaspoonsful. Or it might need more than three times as much, 66 grams, 17 teaspoonsful, or maybe more, and even then it'll only kill you if you can't reach a doctor within a day or two, and maybe it still won't kill you.

But it definitely can kill you. If you drink five teaspoonsful of pure methanol you'll need medical treatment even if it doesn't kill you. Yet people have survived doses of 10 times as much -- a quarter of a litre, half a pint -- without any permanent harm. But others haven't survived much lower doses. Getting rapid medical attention is crucial, though the poisoning effects can be slow to develop.

Authorities advise that swallowing up to 1.3 grams or 1.7 ml of methanol or inhaling methanol vapour concentrations below 200 ppm should be harmless for most people. No severe effects have been reported in humans of methanol vapour exposures well above 200 ppm.

Out of 1,601 methanol poisonings reported in the US in 1987 the death rate was 0.375%, or 1 in 267 cases. It might have been only 1 in more than a thousand cases because most cases weren't reported. Most cases were caused by drinking badly made moonshine, which is a worldwide problem.

Fiction: "Methanol is ... a very active chemical against which the human body has no means of defence. It is absorbed easily through the skin and there is no means of elimination from the body, so levels of methanol dissolved in the blood accumulate."

That's from a British website trying to sell Straight Vegetable Oil (SVO) solvent additives by frightening people with the alleged perils of biodiesel. See The SVO vs biodiesel argument

Fact: 30 litres of fruit juice will probably contain up to 20 grams of methanol, near the official minimum lethal dose. Methanol is in the food we eat, in fresh fruit and vegetables, beer and wine, diet drinks, artificial sweeteners.

Not only that, methanol occurs naturally in humans. It's a natural component of blood, urine, saliva and the air you breathe out. It's there anyway even if you've never been exposed to chemical methanol or its fumes.

Methanol is eliminated from the body as a normal matter of course via the urine and exhaled air and by metabolism. Getting rid of it takes from a few hours for low doses to a day or two for higher doses. Some proportion of a dose of methanol just goes straight through, excreted by the lungs and kidneys unchanged. The normal background-level quantities of methanol in humans are eliminated and replenished all the time as a matter of course.

Fiction: It's largely biodiesel's methanol content that's being blamed when the same British SVO website charges that biodiesel is wasteful and environmentally irresponsible.

Fact: Methanol is readily biodegradable in the environment under both aerobic and anaerobic conditions (with and without oxygen) in a wide variety of conditions.

Generally 80% of methanol in sewage systems is biodegraded within 5 days.

Methanol is a normal growth substrate for many soil microorganisms, which completely degrade methanol to carbon dioxide and water.

Methanol is of low toxicity to aquatic and terrestrial organisms and it is not bioaccumulated. (It's toxic mainly to humans and monkeys.)

Environmental effects due to exposure to methanol are unlikely. Unless released in high concentrations, methanol would not be expected to persist or bioaccumulate in the environment. Low levels of release would not be expected to result in adverse environmental effects.

Fiction: A European SVO fuel website using similar anti-biodiesel tactics claims: "Biodiesel is a chemically altered plant oil. However the process to chemically change the structure of Pure Plant Oil is a very costly operation and requires a lot of energy, as it removes the glycerine substituting it by methanol as well as adding other chemicals, making the end-product poisonous and equally hazardous as fossil diesel fuel."

Fact: There is no free methanol in washed biodiesel. All the national standards require washing. According to US EPA studies methyl esters biodiesel is less toxic than table salt and more biodegradable than sugar. It has none of the toxic or environmental hazards of fossil diesel fuel.