Making biodiesel is so incredibly quick and simple

Raw, unwashed biodisel after just 15 minutes in the processor

The raw material: Long molecular chains of fatty acids with a glycerol (also known as glycerine) molecule attached on one end are known as glycerides. Three of these naturally bond together, forming tri-glycerides. Vegetable oil and animal fat is made of these tri-glycerides.

The chemicals: Caustic soda or potash is mixed with methanol to form sodium or potassium methoxide which will break these bonds.

The reaction using fresh vegetable oil:

Add the methoxide. It first attacks the tri-glyceride bond, releasing one glyceride, leaving a di-glyceride still bonded, then it attacks that bond and you have 3 separate glycerides.

The methoxide then breaks the glycerol molecule free from each glyceride in turn and replaces it with a methanol molecule. When complete, you will have three methyl esters (commonly known as biodiesel) and three glycerol molecules. Since these methyl esters come from vegetable oil which is made from fatty acids, the biodiesel we make is known as Fatty Acid Methyl Esters, or FAME for short.

The reaction using waste vegetable oil and animal fat:

Identical apart from the fact that used or waste vegetable oil also has fatty acid chains which have already broken away from their glycerol molecule under the heat used when cooking. These are known as free fatty acids (FFAs) and they cannot easily be reattached, so we neutralise them with extra caustic.

Agitation: To ensure that the methoxide can reach the all of the glycerides, we need to mix them together continuously until the reaction is complete.

Temperature: The reaction proceeds better above 40C and reaches a maximum rate around 60C, above which the methanol boils and evaporates (64.5C).

How Long Does It Take?

The speed of the reaction depends on a large number of factors, such as: batch size; temperature; agitation; water content; the amount of chemicals; purity of chemicals, etc., but should normally be completed in less than 30 minutes from when the methoxide is first added.

How do we know when the reaction is complete?

HM Revenue and Customs state that to qualify as biodiesel for exemption from fuel duty and to form part of your 2,500 litres allowable, the finished fuel must comprise at least 96.5% methyl esters.

We work with 4 alcohols: propanol, ethanol, methanol and glycerol which each have different dissolving powers, so we can use a simple test.

Dissolving properties:

Take a sample (many people suggest 3mls) of what you have just made and try to dissolve it in methanol (usually 27mls, hence it is called the 27/3 test). If it all dissolves, it must all be methyl esters (biodiesel) and the reaction is complete. If any won't dissolve but simply drops to the bottom of the beaker/cup/bottle then that must be vegetable oil that has not yet been turned into methyl esters. No amount of shaking will make it dissolve, nor will adding more methanol. Instant result: no drop-out = complete reaction; any drop-out = incomplete reaction.

Drop out clearly visible as a yellow band

100/10 test: We prefer larger samples to give an easier to calculate and more accurate percentage of conversion and always use 10mls of biodiesel to 100 mls of methanol. After vigorous shaking and time to settle we suck the drop-out into a 10ml or larger syringe, stand it up vertically and let it settle. Each 1ml of oil that drops to the bottom of the syringe represents 10% oil not yet reacted, while later we switch to a 1ml graduated syringe for finer accuracy, where each 0.1ml in the syringe represents 1% not reacted.

Now we just drain off the glycerol, wash it and we have pure biodiesel. See how quick and easy it is?

Easy, isn't it? So why do we have problems?

Warning: If you regularly make good quality biodiesel in less than 30 minutes then you need read no further. We do not want to change anybody using a successful method, a tried and tested processor and/or a consistent recipe. But if you don't, then read on.

The problem is us. We have different processors with different components, we use different recipes, we use different temperatures; we use different base rates for chemicals, and if one reaction is not perfect, or we see a new article or recipe on our favourite forum, we simply change something for the next one. We've all done it.

Imagine what the result would be if a baker changed the bread recipe each and every single time he/she baked bread, or if a brain surgeon used a totally new, untried method for each new operation.

No wonder we have problems. We never find the perfect processor, method and recipe and then stick to it religiously, batch after batch. So, what's the answer? What should we do? What is the right reaction temperature? What is the correct base rate? Where do we start?

The simple fact is we make so many changes at once that we never actually know what has worked and what didn't. Also, we sometimes have to change something else to compensate for the change that we originally made.

Benchmark biodiesel processor:

The only way to find out is to go back to basics and design and build a “bench mark” processor. The actual design is irrelevant, as it is just the starting point around which changes can be made, but it needs to be capable of making biodiesel, and it needs to allow total monitoring of what is happening during the reaction. It should be constructed from readily available components, which must be easy to change between reactions.

We chose a plastic HDPE cone of 220 litre capacity as it had a nice big screw top hatch so we could open it up and see exactly what was happening during the reaction. Similarly, we chose 1” ID clear plastic wire reinforced tubing to connect the system, as, again, this allowed full viewing of the mixture as it reacted. The pump we connected was the Tam 105 rated at 40 lpm (our's is, but some seem to quote 33 or 38 lpm), a trusty little pump, widely recommended. For getting fluids in and out, we fitted an inlet/outlet tube, connected through a 1” ball valve to a 'T' piece on the inlet (suction ) side of the pump. With the pump running, we could use this to suck in WVO, water, methoxide, and with the pump off, it could be used, being the lowest point of the system, as a drain valve for water, glycerol, and biodiesel. For heating we installed a 3Kw stainless steel 27” immersion heater controlled by a PID with a range of 0°C to 200°C and accurate to 0.1°C with a built in 16A mains relay to directly control the heater.

Once we had the design, we then needed to eliminate the natural variations that occur, namely different batches of WVO and different chemicals. The easiest way was to bulk buy. We bought an IBC full of 1,000 litres waste vegetable oil from a single source and thoroughly mixed up the contents, before allowing it to settle for 7 days to eliminate water. We titrated the oil several times and took the average reading of 4.5. We ordered a fresh IBC of methanol at 99.9% purity, a fresh 25Kg sack of Potassium hydroxide flakes at 90% purity and a fresh 25 Kg sack of sodium hydroxide pearls at 99.9% purity. Finally, we needed to go back to basics with the recipe to use. From “Journey to Forever” we chose their simplest “all in one” process, being a 100 litre batch carried out at 50°C, and elected to use their lowest chemical base rates, being 20% methanol and 3.5g per litre for sodium hydroxide and 5.45g per litre for potassium hydroxide (3.5 x 1.4 / 90%).

Once all these are set, we then carried out a “bench mark” reaction and noted the time it took to reach completion, defined as reaching 96.5% methyl esters content in line with HMRC specifications, confirmed by the 10/100 methanol test (a modified 27/3 test), which gave very accurate methyl ester percentages. This time taken was 2 hours and 15 minutes.

Once we had our bench mark, all that was required was to select one and only one variable to change for the next reaction from a relatively short list of possibles.

Variables to be tested were:

The reaction vessel; copper, steel, plastic, cone, cylinder, etc

The pump; 40 lpm, 80 lpm, 110 lpm centrifugal and self priming

Temperature controller; immersion thermostat, PID

Base rate caustic; 3.5g, 4g 5g 6g, 8g, 10g per litre (and equivalent for KOH)

Basic methanol; 15%, 17%, 20%, 25%

Reaction temperature; 30°C, 40°C, 50°C, 60°C, 65°C

Methoxide introduction; venturi, all in, 80/20

WVO water content; water removal, water addition

While there are only 8 variables, each one can have many different values, meaning that they can probably be combined to give many thousands of combinations altogether.

We carried out one reaction after another, changing just one, and only one, variable each time, and measured the time taken to completion, or 3 hours which was classed as a failure. After each reaction, we reset the variable and chose another variable or setting. We quickly found that some variables made almost no impact at all on the reaction time, while others showed fantastic leaps, faster or slower. Variables that showed a reaction time difference less than 15 minutes faster or slower were disregarded. Variables that made large differences were explored in more detail.

Over the next few weeks we will be releasing all of the findings, but we start with the single biggest variable, and definitely the most surprising in that it is barely mentioned on any of the web sites or forums that we have ever found. It made a difference in reaction time that was simply staggering.

Agitation energy:

By far the biggest drop in reaction time was when we changed the pump. The original Tam 105 rated at 40 lpm gave a reaction time of 2 hours 15 minutes, yet changing to a 80 lpm centrifugal pump reduced that to just under 30 minutes. And changing to a 110 lpm centrifugal pump dropped even that quick time to just over 12 minutes for a 100 litre batch. We were astounded and set about testing the impact of pump size and design, and when we ran out of pumps (the largest we had was the 110 lpm), we just dropped the batch size (processing just 25 litres meant that each litre received over 4 times the agitation compared to processing 100 litres with the same pump).

Our table of results shows average time taken to reach 96.5% completion (we decided that testing the 40 lpm pump was a total waste of time):

* failure simply meant that we had not reached 96.5%. In fact, after 9 hours, it was at 94% and we calculated it would take almost another 4 hours to complete, as the conversion progresses in a series of “half lives”. If the half life of a particular reaction is 1 hour, it means that after 1 hour, 50 litres of biodiesel have been produced, after 2 hours we have 75 litres (50 litres plus half of the remaining 50 litres), after 3 hours we have 87.5 litres (75 plus half of the 25 remaining) after 4 hours we have 93.75, after 5 hours we finally have 96.88, but the last 10 litres have taken over 2 hours to produce compared to 50 litres in the first hour.

The results show a 3 startling facts

1. You can make biodiesel at 96.5% in under 7 minutes

2. If it takes 7 minutes to make 25 litres it should also take 7 minutes to make 100 litres, 500 litres or any quantity you choose

3. Our 110 litre pump, which we considered over sized, struggled to process 180 litre, taking 12 times longer than it should, while the 80 lpm pump failed altogether

These facts prompt the question; how on earth does anybody make good biodiesel in a sensible time if they use the Tam 105, or indeed any pump, rated at just 40 lpm, and why are some processors sold on ebay with a laughable central heating pump attached?

For Your Information: most of the above reactions have been videoed, but the actual times recorded are irrelevant and not the important issue. What we can guarantee in every single case, whatever equipment you use, whatever WVO you use, whatever chemical base rate you use, is that reacting a batch volume of just 25 litres in your normal processor will always take considerably less time to complete than reacting a batch of 100 litres or 150 litres. Case proved.

Try it for yourself, please. Get your own figures. It is really very easy and completely free; no need to buy a larger pump to test against yours; no need to make any changes to your processor. Just decrease your batch size.

First, make a 100 litre batch (or whatever size your processor normally makes) in the normal way and time exactly how long it takes to finish. Let's say it takes 1 hour 30 minutes.

Assuming you cannot heat a batch of 20 or 30 litres because of the placement of the heater element, put another full batch into your processor in the normal way. Heat the oil to your normal temperature. Titrate in the normal way. But just before you mix up the methoxide, turn off the heater (and leave it off for the duration) and drain off enough hot oil so that the remaining oil left inside the processor is less than the flow rate of the pump you are using. e.g. if you use a 40 lpm pump, drain off 75 litres of the oil from your reactor into clean drums for processing later, leaving just 25 litres inside. Now mix up enough methoxide for just 25 litres, get it all into the processor as quickly as possible in a single go (even pouring it directly in if that is possible) and then react in a single continuous process and time exactly how long this reaction takes, but be ready for a fast, sometimes startlingly fast reaction.

By the way, if anyone wants to see the most boring 20 minute video of a complete reaction of 100 litres using our 110 lpm pump, start to finish, then we will upload it onto Youtube and give you the link. The fact that we can get the reaction completed, including settling and testing within 20 minutes should say it all. But why watch us doing it? Just try it for yourself, it costs nothing and you get biodiesel at the end of the test.

So why is agitation so important?

The following is pure logic, not chemistry or rocket science.

Heat a batch of WVO in a large steel drum, gently pour in the correct amount of methoxide and then stir it just once gently with a large wooden spoon and sit back and wait. What will happen? We all know what will happen. Nothing. After all, that's why we run the pump/agitator/stirrer during the reaction, isn't it?

Now take another batch, do everything the same, but this time pump/stir it properly throughout the process and we all know what will happen. We will make Biodiesel.

It couldn't be clearer. If we don't agitate enough, the reaction never finishes. If everything else remains identical but we increase the agitation, we make biodiesel.

Put another way, if the energy of the agitation is insufficient, the reaction cannot ever complete. And for those who believe that just leaving it a bit longer will eventually make it work, try heating your house with a single candle by leaving it burning a bit longer. Not enough energy no matter how long you leave it.

Agitation consists of two components; vigour and duration. Stirring gently with a wooden spoon could be considered the least vigorous, while using an industrial strength food whisk at maximum setting could be considered towards the most vigorous. As vigour increases, the duration can be reduced.

Pumps vary in design and vigour. A self priming pump works on the paddle wheel principal. As a vane passes the inlet, it creates a suction and draws in a quantity of oil behind it. As it turns, the oil is moved forward until it reaches the outlet, where it is driven forward and out by the vane behind it. It is considered to be at the very gentle end of the vigour spectrum. A centrifugal pump has an impeller, often containing multiple sharp metal blades or a helical twin plate arrangement. As these spin, oil is drawn into the centre and flung outwards under centrifugal force at high velocity towards the outlet. These are at the very vigorous end of the agitation spectrum.

Flow rate gives an indication of the power of the pump, although it is not, in itself, an important factor. A large flow rate self priming pump, say 60 lpm, will pass more litres through it in any given minute than a small centrifugal pump of say 40 lpm, but will provide far less agitation of the oil/methoxide mixture.

Conclusion: Agitation is more important than flow rate, so centrifugal wins hands down over self priming, and the larger the centrifugal pump, the greater the agitation and the shorter the reaction time. If you want to process batches of 50 litres, a 40 lpm pump might cope. For a batch size of 100 litres you should not consider anything much under 100 lpm. Above 100 litre batches, you should be on a minimum of 100 lpm, preferably 150 or 200 lpm would be even better. * we see that our favourite supplier, Biotecengineering (www.biotecengineering.co.uk), has just launched a new 165 lpm centrifugal pump and we'll be getting one very soon. Thanks Chris!

110lpm pump

Reasons why your agitation energy might be too low:

If your test showed a problem with batch size, there could be many reasons, other than pump size. Here we list all we can think of;

Flow rate too low: If you use a pump to agitate, then be aware that the stated flow rate is measured using water. Oil is a lot more viscous, even when hot, so a 40 lpm pump might achieve less than 20 lpm when pumping oil, and even this will reduce as glycerol forms, as this is not only even more viscous but far denser, too. Get a more powerful pump. 125 litres of oil and methanol might take somewhere between 4 and 8 minutes to go through a complete cycle if connected to a 40 lpm pump, meaning 90% of the oil and methoxide is sitting around doing absolutely nothing all of the time. It should take under a minute to cycle the entire batch size, so if you can't afford a new pump, reduce your batch size.

Wrong type of agitation: Not all pumps operate in the same way. Open up a TAM 105 and you find a multi-bladed brass impeller that is absolutely ideal for thrashing and churning the oil and methoxide into reaction. Other effective pumps have two brass plates fitted together with a gap of a couple of millimetres containing multiple tapering helical chambers. Oil is compressed into the central narrow end of the helical chambers and is flung vigorously outwards under amazing centrifugal force, providing very high mixing and agitation energy.

Open a self priming pump and the operation is totally different. Oil is moved in chunks by a gentle sweeping blade tipped with a rubber or synthetic seal, operating in a smooth chamber. While a particular pump may achieve 60 lpm and sound good, it will do little to mix the oil and methoxide (think of stirring with a wooden spoon as opposed to whisking with a food mixer).

Narrow pipework: Your pump will almost certainly have at least 1” BSP female threaded inlet and outlet, meaning 1” ID or internal diameter, and using Πr² we find the inlet and outlet cross sections are 0.79 sq inches each. Many people install a 1” BSP to 22mm reducer and then use 22mm copper pipes ( I did on my first processor).

But copper pipe is measured OD or outside diameter, so a 22mm pipe having 1.5mm walls will have an ID of 22 – 3 = 19mm and using Πr² you get a cross section of 0.44 sq inches. No prizes for guessing what almost halving the cross section of pipework from 0.79 to 0.44 does to the flow rate? If your pump has a 1” inlet and outlet, then it will produce its maximum flow rate using 1” ID pipes, so use 1” ID pipework and get maximum flow rate. If using copper, switch to 28mm (OD so 25mm or 1” ID).

Restrictions: Many copper cylinders have baffle plates welded inside just in front of the bottom inlets to stop the incoming water from stirring up the contents that have settled to the bottom of the cylinder. These will seriously restrict the flow rate into the cylinder, even if the pipework is of good size. Remove them if possible or find a cylinder with no baffles. 90° bends in copper pipework seriously disrupt the flow rate and create a back pressure, and if you have 2 or 3 in series, you will not be getting anywhere near the pump rating. Reduce the number of 90° bends as much as possible, consider bending 28mm pipes in a gentle continuous curve instead..

Venturi: While the principal of a venturi seems sound, it is usually the exact opposite of what you need and the worst thing you can do. Why?

Dribble the methoxide into the oil too slowly and the pump thrashes mostly pure oil, as there is no methoxide nearby with which to react. Result; unreacted oil ends up back on top of the processor contents, wasted energy and wasted time. Add it too fast, and you thrash spare methoxide but it has no unreacted oil nearby to react with. Result; unreacted methoxide ends up on top of the processor contents, wasted energy and wasted time. If you get the flow rate exactly right, inserting all of the methoxide in the same time that your pump takes to cycle the entire contents of the batch in your processor, it can help, but it might take 10 minutes to get all the methoxide in, and our complete reaction only takes 12 minutes! Get a better pump.

With a decent sized pump, one having good agitation and a flow rate larger than the batch size, the entire contents of the processor are thrashed and churned as much as 2, 3 or 4 times every single minute as they pass through the pump. Oil in our processors never sits still and never gets thrashed without there being plenty of methoxide immediately around it. After the batch has finished the first complete cycle through the pump, usually in less than 60 seconds, the methoxide is already fairly evenly distributed and each subsequent cycle through the pump thrashes it into a reaction with the WVO.

80/20 methoxide: Many people add the methoxide in two batches, 80% at first then, after half an hour or so, you let it settle, drain off the glycerol and add the remaining 20%. The principle is simple. With less volume being processed, the agitation energy is higher. And draining off the glycerol relieves even more strain on the pump. But why are you using a pump that cannot cope with a full batch? And why is it taking half an hour anyway?

This process always leads to ever higher base rates. Glycerol is brilliant at absorbing or dissolving water, methanol, caustic and soap. When the glycerol is drained off after 30 or 60 minutes, it always takes with it unreacted methoxide, caustic and methanol. Adding the next 20% does not replace this. Remember you did the titration and everything that you put in is actually needed. So how can this reaction now complete with 5% or 20% of the methoxide and caustic having being drained off and outside the processor before it has done its job? It cannot unless you overdose with caustic and methanol to compensate. Many people who use 80/20 often also use base rates of 5g, 6g, even 8g NaOH whereas we use 3.5g for good oil (titration value between 1 and 4) increasing to a maximum of 4g for WVO with high titration values (above 5).

Apologies:

Again, if you make consistently good biodiesel, with high percentage completions and high yields in well under an hour, then you might just want to carry on doing what you are doing and ignore this. We won't be offended. You see, we're not trying to sell you anything, just help you. Everything we state, we tell you how to test it for yourself.

Then again, even if you are making good biodiesel in less than an hour, how long does it take to make a 25 litre batch and find out if you could be doing it even quicker? and what harm can it really do? The very worst it can do is prove that your agitation is not good enough, but then you would already know this as it is almost impossible to make 100 litres of 96.5% biodiesel in less than 30 minutes without sufficient agitation energy.

Maybe you're happy doing what you're doing. Fine. But speaking personally, if I need 300 litres of biodiesel, I'd rather make 2 consecutive batches (taking 2 hours in total including heating) of 150 litres each one morning, let it settle for at least one day (three being the optimum), then dry wash it in a single batch (300 litres at 3 lpm takes 100 minutes or less than 2 hours), than struggle making 3 separate batches that each take 2 or 3 hours to complete, taking maybe a whole weekend. Where's the fun in that?

When it all goes wrong:

Sometimes we get it wrong. Sometimes the reaction doesn't complete in 30 minutes, and we know it probably never will. So what do we do?

Simple. As soon as you suspect the reaction has failed, turn the pump off, let the glycerol settle, then drain it off as if it had finished. Don't try to titrate what you have made for another reaction as it is full of caustic and soap and don't, please, do another base reaction using 3.5g per litre and 10% methanol as some people tell you. Try this instead.

Do the 10/100 methanol test, shake it up, let it settle out, suck the drop-out into a syringe and work out how much has reacted and how much hasn't.

Then we take the remaining batch size (what's left after draining off the glycerol) and multiply it by the % unreacted, e.g. if we are left with 100 litres in the processor and the drop-out is 4.2mls, then 58% has turned into methyl esters (and dissolved) and 42% remains as oil. Or to put it in litres, 58 litres of the oil have gone and are now methyl esters and 42 litres is still oil. We also assume that any FFAs have reacted, as they tend to go first, so now we treat what's left as fresh oil.

Even though the batch size is still 100 litres, there are only 42 litres of fresh oil in there. 42 litres needs 8 litres of methanol at 20% and 147g NaOH (3.5g base rate x 42 litres). Mix a fresh batch of methoxide and chuck it in for a second reaction. This one will finish.

Closing thought:

I know it's fun making your own processors, inventing bits, overcoming challenges, but ask yourself this question and give yourself an honest answer.

If you wake up one morning and decide you want to become a successful Formula Renault motor racing driver, would you:-

a) visit a motor racing dealer who is a specialist in Formula Renault, ask for and take his advice and then buy the best complete car (new or used) that you can afford, or

b) head down the scrap yard and local motor accessory shop, buy loads of bits, download plans from the internet and then spend weeks and weeks building your own?

And if you chose b) what are your chances of winning anything at all? Ever?

So why, if we want to become a successful biodiesel producer, do we always choose option b)? I know I did and the only reason I did it, all those years ago, was to try to save money.

But if I now look in my garage and count up the cost of copper cylinders, pipes and fittings, the drills and thread cutting taps, the bench drills, angle grinders, blow torches, and so on, and so on, and so on, I know that I have spent far more, in total, than if I had bought a basic complete processor. I also know the number of batches I have disposed of, the number where the yield was pitiful, the vast quantities of “chicken soup” emulsions, the huge amount I've spent on wasted chemicals, the huge electricity bills and so on, and so on. Then add in the sheer frustration, the anger as yet another batch fails, the money spent on fossil diesel because mine wasn't ready when I needed it.

I have not saved any money. In the end, I bought professional equipment anyway. And what a dream it is to use.

I can fill the processor with 150 litres of oil in minutes, heat it in well under an hour, titrate, add the methoxide, set it going, then go and have a cup of tea and a biscuit, go back and turn it off. It has always has finished, I barely need to test it any more. After settling, it washes out at 3 litres per minute using minimal quantities of Eco2Pure, and the result is always 97% or more methyl esters, and if I shake a sample in a bottle with fresh water, it settles out within seconds, leaving clear water below and clear biodiesel on top, no thin white band, and if I PH test the water, it is nearly always between 7 and 7.5. Any reading outside this means that the Eco2Pure needs changing, which is a messy but simple job.

The only failures I ever get now are when I say “ I wonder what will happen if I …....” at the start of a batch.

If you know what you are doing and love DIY, go for it and build your own. It will give you hours of fun and frustration. If you are a newcomer and/or just want good quality biodiesel with the least hassle, you are unlikely to save money by doing it yourself. If budget allows, get a ready made processor from somebody who knows what they are doing and proves it to you with a personal demonstration BEFORE you part with any money. No reputable company would object to that. Although the cash outlay is more up front, it will be easier to operate and you can be up and running the next day, producing usable biodiesel for your car at low cost per litre. You choose, but in the long run I know that I would have saved a lot of money by following my own advice learned the hard way.

And buying a ready made processor does not spoil our “discovery” fun. There's loads still out there to experiment with and discover. After all, some of you, despite having made biodiesel for years, despite having built and designed your own processor, and despite having successfully run your cars on your own fuel for ages, still have a TAM 105 40 lpm pump attached to a processor making 100 litre batches, and wonder why it takes so long to react. Be honest, I bet you never knew before now that your agitation energy was too low, did you?

P.S. I have nothing against a Tam 105, in fact I have one and it's brilliant, but I use mine on the wash cycle. I pump the raw biodiesel from my processor into a 205 litre settling tank, where it stands for as long as possible, but often only a day or so. From there it is pumped into a 22mm copper pipe leading to a brass 'T' piece. One leg goes through a 22mm ball valve back into the settling tank, where it sprays onto the surface, releasing any last traces of methanol. The other leg goes through a 22mm ball valve into the dry wash tower. Adjusting the ball valves allows me to control the flow rate through the dry wash tower. I've had the pump for years and it's still plodding on, long after central heating pumps have packed in. But put a Tam 105 on a biodiesel processor handling more than 20 litres of WVO? Never, ever, ever. If my 110 lpm pump slows down each time when increasing batch size from 25 litres through 100, 150 then to 180 litres, I dread to think what happens using a TAM 105 on a 100 litre batch, and I'm pleased that I will never, ever find out the answer.