I want to increase the mixing of my processor and have been researching static mixes and eductors.

In his post on March 2007 Johno gives a great explanation of a static mixer with directional elements jamesrl has one that works differently.

http://biodiesel.infopop.cc/ev...551/m/5261047821/p/3

Jacoby-Tarbox has a whole link on intank and inline eductors.

1877eductors.com/index.htm

Rickdatec and many others have offered home made eductors.

Can I please have your opinions? What do you use? Pros and cons of each.

Thanks, Tim

I think there's no substitute for a good pump.

If I had to choose, I'd go with an eductor. I think static mixers are not correct for our application. I know some swear by them so I'll just leave it at that.. To each his/her own.

Murphy,

At first I was looking at one of your pumps (remember the hemi in a Volkswagen analogy you quoted.)

Anyway I have read here that just pump flow may not mix the whole tank thoroughly I am getting good bio I just thought I would try a mixer to see if I can get quicker conversion.

Tim

When they say "the pump may not be enough", they are refering to the little blue pump because it doesnt have much volume.

Think for a second about what a pump does and how it does it. An impellor spinning at 3600 RPM's is beating the living daylights out of the liquid. So long as you pass all the liquid through the pump a few times, it will get a near perfect mixing.

There could be problems. If you turn your pump off and let your tank sit at any point and time during your process, there is always one thing that will happen and that is that the liquid will begin to stratify and split into different layers. Some of molecules will want to rise up, some will want to sink down. But they will go one direction or another.

The stuff that sinks isnt much of a problem because it will sink right into the pump's suction line and get sucked right up. The stuff that wants to rise is another story. And this is where pumping volume becomes a big issue. If the pump you are using doesnt have enough volume to turn the entire contents of the tank over faster than the tank wants to stratify, you can end up with unreacted fuel. This is where eductors come in and the game changes to velocity. By providing enough velocity, it is possible to push the stratifying top down towards the bottom so the pump can get it.

The best option is to turn the tank over fast enough so none of the above becomes an issue. Hence, my comment "there is no substitute for a good pump"

Hope that helps,

One would use a static mixer when one uses a low shear pump. The little blue pumps are high shear regenerative turbine pump that pass the liquid through the blades hundreds or thousands of times resulting in a really good mix. If you choose a gear pump or a diaphragm pump with low or no shear mixing, then a static mixer would improve mixing.

Pump mixing calls for higher power consumption than propeller mixing. For example to mix the same 40 gallon batch would take less than .1 hp using a propeller mixer and almost a 1/2 hp to achieve the same mixing with pump mixing. A tank eductor lets you have the same mixing capability with less motor, say 1/3 hp instead of 1/2 hp.

Each has their place.

Murphy / Rick

Very good points, thank you. This forum always amazes me at the information you can get if you just ask. My current pump has been tested @ 16GPM through the venturi with thick/cold WVO. So with a batch size of 20 gal that is mixing the whole batch every 1.25 min. That being said, and reading your explanations I think I should (don't fix if it ain't broke) leave it alone for now.

Thanks again, Tim

I made a something that is not quite an eductor, but does increase velocity. It decreased my mixing time from 5-6 hours to 2 in my 50 gallon appleseed. I just used one of the pipe adapters that b100supply sells, screwed a 36"x1/2" pipe to it, and used a brass fitting (female 1/2" pipe thread to flare fitting". This probably necked down to about 3/8" or maybe a little less) on the end of that. I had to grind the hex head of the brass fitting so I could lower the whole thing through the top.

I was really surprised at how much it helped.
irokcj5

What about adding a second pump that would draw a few inches from the top. Then have the bio pumped in at the bottom, but forcing the bio to be pushed to the side. Causing the contents to swirl, maybe add a small baffle for a better mix depending on the force of the swirl.

I have done a lot of work on inline mixers and and eductors, single and 2 stage.
You cannot compare the difference, an eductor out performs a mixer by a long way.

On my experimental reactor, the one that all my ideas are tested on, I've gone from a 90ltr/min pump to a 90ltr + inline mixer to an HFpump single stage eductor to an HF 2 stage eductor.
The results are:-

90ltr pump, 2 - 2.5hr conversion

90ltr + inline mixer, 1.5 - 2hr

HF pump + eductor, 45 - 55mins

HF pump + 2 stage eductor, 35 - 40mins

All reactions were done at 62c/144f using NaOH and 22%methanol, all passed a 3/27 with flying colours.

To get the best from an eductor it needs to be vertical and 3" under the WVO + methoxide level. If you're using an HF the nozzle needs to be no more than 1/4" ID.
The extra load on the HF to work the eductor is 50watts with a nozzle velocity of 40tf/min and a mix ratio of 1:5, adding a second stage to the eductor brings the ratio to 1:30.

Allowing for the drop in flow rate caused by the 1/4" nozzle you'll be mixing 600ltr/min, 159galls US/min.
Not bad for a little blue pump.

How the Single stage eductor is fitted.



Heres my prototype 2 stage eductor.




Conclusion, Don't mess with inline mixers go for an eductor, single or 2stage.

jamesrl

-very interesting post thanks Tom

jamesrl i'm amazed at the numbers, thank you for your time and help in this matter , i'll be looking into the eductor more so now than ever.
jonathan

quote:

Conclusion, Don't mess with inline mixers go for an eductor, single or 2stage.

Your conclusion follows from your test. However, if you think about the different effects of mixers vs eductors, you may see the need for more changes to the design to make mixers work correctly.

In this reactor , I would pass 3/27 in as little as 45 min using 19% methanol on a 145 L batch.

The difference is that this design addresses the weakness of inline mixers. They are strong on micro mixing. But they slow flow rate, reducing macro mixing. So, by taking fluid from the upper layer and the lower layer at the same time and mixing them in the pump, the weakness of inline mixers is apparently over come. The upper pump input is near the oil fill level.

Nothing against eductors. I am working on a new processor where I plan to see if I am happy with only an eductor. If not, I will use a mixer as well.

James it would have been good to see what the HF pump time was on it's own. Rick says they are a high shear regenerative turbine pump that pass the liquid through the blades hundreds or thousands of times resulting in a really good mix. Maybe there is really no gain in time when using an inline mixer.

The HF pump produces a 3/27 pass in the same time as the 90ltr/min centrifugle pump proving the better mixing of the turbine periferal hi-shear pump.

The results of the single and 2 stage eductors are now being reproduced by a few brewers over here, one brewer has 2 x 2 stage eductors and is singing their praises.
The idea of the two inputs, higher and lower, to the pump have been used here for 2years now but never really took off in a big way but it did work well for those who tried it.

We tend to stick to the 22% methanol, some are using 25%, as we are now doing a Whole Batch Demeth and recover most of the excess methanol at 98% pure. Using the stoichiometeric measure we work out how much methanol we should recover and distill that off.
For example:-
100ltr batch
22ltr methanol
Stoich. measure 12.5% of oil batch size = 12.5ltr
22 - 12.5 = 9.5ltr to distill.
No pre-wash will return the 9.5ltr at 98% pure. If your conversion is good, a perfect 3/27, you'll get no reverse action. I've been using this method for a couple of months now without any problems as have many others.
NO acid is used to neutralize the lye.

All I know is that our little HF pump worked fine on the appleseed, then we went with a 110gal tank and as our batch size increased we had problems with emulsions. It was just too much volume to be adequately mixed so we got ahold of an HVI-5 and problems vanished. We put a T on the end of the pipe that drops vertically into the reactor which is typically below the WVO/methoxide level. Plus we get a nice effect when drying the finished fuel that sprays out the side part of the T

Jim, you mean distilling the BD and glycerine by-product together?

quote:

Originally posted by Paulus:
Jim, you mean distilling the BD and glycerine by-product together?

Hi Paulus, long time no speak.

Yes, that's exactly what I mean, there's loads of us doing it with great success. Chug is getting 5ltrs/hr from his Plumber's delight condensor, I'm running a different type of plant also smaller volume but get 2 1/2 - 3ltrs/hr back. We've reduced our methanol cost by over 33% at a stroke.

quote:

If your conversion is good, a perfect 3/27, you'll get no reverse action.

How do you know?

The reason I ask is genuine interest.

Also, did you get a chance to look at my post about my old processor using static mixers which worked very well? They can be very helpful if used correctly. Perhaps better, or at least just as good as eductors, at least from what I have seen.

Andrew, yes I did have a look and I agree that in the right situation with the right static mixer they work well but the simplicity of an eductor becomes a 1 step solution to the same old problem, intank aggitation.

The greater the aggitation the faster the conversion, we have an experiment going on that converts 40% more wvo with a given amount of methoxide and very little aggitation in a suprisingly short time.

A good 3/27:-

I no longer use the 3/27 it's too small to show minute drop out, my standard test is now 5/45. If the methanol is not crystal clear from the first to the last drop of bio being tested with zero dropout its a fail. I always see how much bio I can get in to the measuring cylinder without fall out, my record to date is 50% of the methanol, in my case thats 22.5ml of bio in 45ml of methanol and still crystal clear albeit golden in colour.
The original test was 25/225 now that will show the slightest drop out. The methanol use in the test is not waisted it goes into the methoxide vessel for the next batch.

So Jim no acid means no need to add phenolphthalein and therefore no need to check for colour change? I guess if you have a build-up of MeOH-rich glycerine from previous batches waiting for distillation you could add a little into each latest batch and process that way?