GL's "Eco-System" biodiesel processor - March 2007

Copyright notice

Introduction

Warning - this page is now obsolete
Please see latest details here.

The Eco-System is a development of the ever-popular Appleseed processor, and of the GL1 day process, with performance, safety and efficiency improvements for both. It is easy to build from standard plumbing parts.

It takes its name from its main benefits, being Economical and Ecologically sound.

What can it do?

• Makes top quality biodiesel without water washing, magnesol, amberlite etc.
• Releases next to no methanol vapour
• Reduces the chance of the heater being accidentally switched on
• Quickly dewaters your oil, right in the processing tank, without giving off oily steam
• Captures most of the expelled water, to help you estimate your oil's % water content
• Effectively mixes methoxide into oil without cavitating your pump, and releases minimal fumes
• Performs a standard transesterification reaction
• Drains off your glycerol without releasing any fumes
• Allows you to extract surplus methanol for re-use, without releasing fumes
• Ensures your biodiesel is methanol-free, without water, so that soaps drop out easily

I've also included notes so you can easily add these features ...

• Cuts off the heating if the pump fails or if the liquid level drops too low
• Cuts off the heating if the pump overheats, and only allows manual reset.
• Sets a well defined low pressure release limit
• Lets you see the pressure / vacuum within your process at any time

Here, I've mixed a sample of the finished biodiesel with water, and allowed the two to separate. You can see that the water is crystal clear and so is the biodiesel. This biodiesel had not been water-washed, because it didn't need to be. This process removes the need for water washing. This photo is not a trick, it is a consistent result I get with this processing method. Honest!

It is based on the Appleseed processor, for which thanks go to Girl Mark of localb100.com, and everyone else who has helped to evolve it over the years.

To make the Appleseed even more ecologically sound, I've tried to reduce the amount of methanol released into the local working environment and to the air / water in general, compared to most current processor designs and processing methods.

This not only protects the environment, it protects YOUR health and the health of those around you.

It also reduces the total amount of methanol needed to process a batch of given size, compared to common methods.

Here's how...-

• The system is loosely vapour sealed, to reduce fume release, even during de-watering, glycerol draining and distillation.
• The process recovers methanol from the raw biodiesel in a closed loop distillation system.
• We DONT need to wash the raw methanol-laden biodiesel with water.

Using water to clean your biodiesel is wasteful. It flushes away valuable methanol. Every litre of water to be processed at a water treatment works costs energy, and water may itself be a scarce commodity, depending on your location.

Plus, in colder climates, you need to keep the wash water warmed, to avoid losing high melting point esters down the drain, and then you need to add heat to dry the biodiesel. You may even occasionally end up with an emulsion, and that takes yet more time and energy to put right.

I really am not a fan of water washing! It simply isn't necessary. This waterless method has passed ASTM quality tests (thanks to 'SpeedRacer' who sent off the samples. See here..)



So, What's NEW??

This processor uses some components you may not have tried before, but that's no problem, because they are really easy to make, and offer some valuable advantages.

The main difference is that this processor uses a Venturi. The venturi is the main reason this processor can do what it does. The venturi acts as a suction pump and as a mixer. It moves liquids and vapours around our processor in ways you don't get on a standard processor. Its a good pump for pumping methanol vapours because it uses no electricity and has no moving parts.

Here's a 5Mb video of my test rig with small home-made venturi at the top, pulling water and air bubbles up from a bucket on the ground. Note that with this method, you do not need to inject methoxide to the inlet of the pump and the venturi acts as a good mixer, and can cope with air and liquid in the tubing, without affecting the pump's performance.

Sorry, the video is on its side. The pic below should clarify the video for you, which is rather cluttered with background material.





Thanks to ebtz on the biodiesel.infopop.cc forum who suggested McMaster Carr do a venturi, for 1" pipe, part number 4563K63 at $54.23 Each. As yet untried, so we're not sure how it will perform, but almost certainly better than my home made version described in the link. I have ordered one on 26 Feb 2007, so will let you know how it works out. (Arrived 6 March - plan to run it on 10 March) OK, I ran it, but with a 60 litre per minute pump I could only lift 4" of water. It needs a MUCH bigger pump.

Thanks especially go to Mike of JTP Inc. in Shillington PA 19607 Tel (610) 777-4890, who made a custom Nylon venturi for me, to my design, which I am trialling now. If you need anything fabricated in plastics, Mike's your man! Lovely job! And he's a serious biodiesel fan.

Make your own venturi

Next, we've got a slim water-jacketed condenser, which lets us recover all that valuable methanol, which would otherwise go out the vent pipe, down the drain, or up your nose. Nothing goes to the ouside world without first going through our condenser, so we can snag the volatile methanol vapours and keep them for our next batch.

Make your own condenser

Next you'll notice our carboys have some kinky looking lids with 2 tubes coming out of them. They help to ensure the methanol vapours stay in the system. They are also easy to make.

Make your own carboy vents

Finally, there's an optional valve V1 and large U-tube made out of pvc drain pipe, part filled with water. This helps us to keep the methanol vapour in our system while we are distilling. It also acts as a well defined pressure relief valve and as a pressure/suction indicator - I added this just so I could see what happens during the process, how much vapour leaves the system, how much enters.

It isn't essential, just makes you more aware about the effects of dosing methoxide rapidly into very hot oil, so tends to help you to regulate your methoxide feed to give minimal out-gassing, so is useful if you are 'learning the ropes'.

Make your own U-Tube vapour trap

Safety Notes

This system does NOT guarantee there will be no methanol vapour in your workspace. It is intended to help you achieve lower levels than you might expect with other current methods, but it does depend on your system being properly made and maintained.

Ensure all joints are tight, regularly check carboy lids are tight during the process.

Flexible hose should be straight and un-kinked.

To attach a carboy, rotate the carboy, not the lid.

The dosing carboy should be low enough for you to safely handle, no higher than waist height. With a more efficient venturi, you should be able to have the dosing carboy at ground level. A safe, stable way of mounting the dosing carboy is needed. I propose a wooden stand with 3 castor wheels, to allow it to be easily rotated, without risk of it tipping over.

The carboys I use are the 25 litre ones my methanol comes in.

The green tubing is clear PVC tubing, held in place with jubilee clips.

The condenser is fed with cooled water, to aid condensation of methanol. I use a barrel of rainwater as my water store and recirculate it through the condenser with a home heating water pump. In the summer you could add a fan-cooled radiator in the water loop to get rid of excess heat, or you could use that heat from the condenser to help warm up the next batch of oil for processing. There are plenty of opportunities to further increase the efficiency of the system.

You can use a simple 45 gallon drum for the processor, tilted 10 degress to help draining glycerol. The lower outlet pipe would then need to be near the edge of the drum at the lowest pont. The drum must be sealable, because we don't want methanol vapour leaking out of the drum, or air leaking in.







This is the processor with the optional vapour trap fitted. This allows you to see immediately if you are dosing methoxide too fast, distilling with too much heat power for the condenser, etc.

The aim is to control the energy input to your processor in such a way that it doesn't burp out large amounts of methanol vapour. The water level in the sight tube gives you instant visual feedback about the pressure balance and vapour loss in the system. It's like balancing on a tightrope with your eyes open, instead of with your eyes closed





Electrical system ideas ...

The procedure, and what's going on at each stage:-

(see also the state diagram)

Attach the condensate collection and glycerol collection carboys.
Refer to the State diagram to see which valves to operate and what to switch on or off at each stage, and look at the processor diagram, for referencing to this description.

Load filtered oil into the reactor and heat to 50C if dry, or 90C if wet.

Attach a water collection carboy to the end of the condenser if you want to measure condensate.

If the oil is wet, then you can open valve 2, which will cause the venturi to draw air into the reactor, through the circulating oil. The air will be intimately mixed with the hot oil in the venturi, so will saturate with any water present in the oil. The air will then seek an exit, as it becomes pressurised in the reactor, and can only escape through the condensor. Cold water passing through the condenser jacket will cause the water vapour to condense and fall into the condensate collector vessel. The now dryer air will be returned to the venturi for the process to continue.

When you find that the condensate collection rate starts to slow, switch the heater to the 50C setting. The water removal process will cool the oil, through evaporation, so you can use the stored heat in the oil to continue the drying process, while the evaporative cooling will bring the oil temperature down towards the desired processing temperature.

When the processing temperature is reached, you may take a sample of the oil, as by now it should be sufficiently free of water.

Titrate the oil and make up the methoxide mixture to suit. Attach the methoxide dosing carboy to its vented lid.

Check the valve and power settings to suit methoxide dosing and gradually open valve 3. The heater must be OFF. Methoxide should be drawn up, into the venturi, where it will become mixed with the oil, and then passes into the reactor, to be recirculated thru the pump.

You'll see that there is no coughing or pump stalling if you draw air in from the methoxide tank, unlike traditional systems with methoxide intake on the pump inlet.

Aim for a valve opening which will allow the dosing to take around 20 minutes. The venturi does a good job of mixing the methoxide and oil. I suspect the methoxide boils as soon as it hits the 50C oil, due to the low pressure in the venturi throat. It then collapses rapidly as the mixture enters the processing tank. A drawn off sample, taken immediately after the venturi, just before entering the tank, shows reasonably good mixing, there is no apparent separation of methanol from the oil.

When all the methoxide has been dosed, you can close valves 3 and 4 and allow the pump to recirculate for around an hour to 90 minutes, depending on your oil quality.

You can take periodic samples from valve 6 to do a 23/7 test, to judge completeness of reaction, as there is no point in continuing the process if the reaction has reached a satisfactory level of conversion.

You can now either stop the pump and allow the glycerol to settle, so that it may be drained off, or you can do a 5% prewash. The prewash seems to improve conversion and reduces the amount of soap left in the biodiesel. The only minor downside to the prewash is that your recovered methanol will have some water content, but this is not a big issue if you add your recovered methanol to fresh methanol. I've not had any noticeable change in processing ease or final quality from this.

If you choose the 5% prewash, add the correct amount of water to the dosing carboy, re-fit it and start the pump. Gradually open valve 2 to draw water into the reactor, where it will become mixed with the raw biodiesel and will tend to be more attracted to the soaps and glycerol than the biodiesel. After 15 minutes of circulating, stop the pump and allow th glycerol/soap/water byproduct to settle. Settling time can be as little as an hour, but if you have time, leave it longer, and you will have less glycerol in your biodiesel. I should mention that your reactor must be able to drain well, you want as little glycerol left behind as possible, because when we do the distillation, any glycerol remaining in the reactor will have a tendency to reduce your biodiesel conversion. You may find a small percentage of the methyl esters you've just made get converted back to glycerides. This is unlikely to cause big problems unless a lot of glycerol remains in the tank, but you probably want the best conversion you can get.

When your glycerol has settled, you can drain it into the glycerol collecting carboy. It will be hot and laden with methanol vapours, but the balancing tube will ensure the vapours remain in your system.

You can now begin the distillation process.

We want to extract as much methanol from the biodiesel as possible, so we can re-use it in our next batch. We also want as little methanol in our biodiesel as possible, because methanol dissolves soap, and methanol rich biodiesel can hide soap. Biodiesel with no methanol in it will not be able to support the soap, which will drop out under gravity. So, let's do a good job of ridding the biodiesel of every last scrap of methanol!

We add heat and circulate the biodiesel through the reactor, constantly exposing fresh, methanol-rich biodiesel to the top of the reactor space. The methanol vapour will expand as it is heated, and will force its way out of the reactor, via the condenser. It will collect as methanol liquid in the collecting carboy.

As the system pressurises, you will notice that the u-tube sight glass level will go up, and you will hear the air glooping out of the system as it is displaced by methanol vapour. This will soon stop as we reach an equilibrium of methanol vapour being produced by the heat energy of the heater, and methanol vapour being condensed by the cooling effect of the condenser. If you heater is too powerful, you will find the u-tube continues to gloop, this is a sign to recuce heating power, as you may be losing methanol vapour through the vent. I use a 3kW heater and a 1 metre long plumbers delight condenser, for 80 litre batches, and the heater is not too powerful for this combination. It seems about right. With practise, you will get a good feel for the way your system performs from the beginning to the end of the process.

Eventually, you will have condensed all the methanol you will be able to with the power of your heater, at 90C.

You may now want to switch off the pump and heater, and put aside the methanol you have recovered. How much did you recover? Typically you might expect to recover from 1.5% toi 2.5% of your biodiesel volume as methanol. This depends mainly on how much methanol you used in your reaction. I use 22% and get just under 2% back on average. It will not be pure, but can most likely be safely mixed with a proportion of new methanol for your next batch. I mix 1 part recovered methanol to 14 parts new methanol. Make sure you mark up your collected methanol as "RECOVERED METHANOL - TO BE BLENDED".

We can now drive off the last remnants of methanol and water from the batch. Put an empty collecting carboy back in place and set the valves and electrical controls for the Purge Water step. Now we will vent to the outside vent, as we will be circulating air through the hot biodiesel and passing it through the condenser. What you now recover will be unusable methanol, as it will be heavily contaminated with water. The air picks up the methanol and water readily from the biodiesel, and combined, they will collect in your carboy. As the concentration of methanol and water drops, to low levels, the condenser will no longer be able to liquify them, so when you see the recovery rate drops to a drip or two, you can stop this part of the process.

After this process, assuming the 23/7 test results or any other quality tests you may perform are acceptable to you, the biodiesel may be left for a few days, to allow the soap to settle, or may be passed thru a soap filtration stage, using bag filters or magnesol or Amberlite or any combination of these you may prefer.

I simply leave the biodiesel to sit in a 45 gallon drum for a few days, which allows the soap to settle. I test the upper layers for soap using bromophenol blue and find it can be used directly, without filtration. You may prefer to filter it. You'll probably find 3 bag filters, rated 100 micron, 10 micron, 5 micron in series will do the job for you.

If you choose to settle the soap out, it will settle at the bottom of your container, so do not completely drain it. I leave the last 1/3 in the tank, and simply add the next batch on top of it.

After every 10 batches or so, you can scoop out most of the soap from the bottom of the tank. I let it sit in a large sieve, so the biodiesel can drip out, leaving mostly soap. You can then use this as a workshop soap, or you can split it into FFA using acid, to turn it into more biodiesel.

Work in progress, article not complete.