Brent's Appleseed Eductor:

quote:

The Whirlpool would just spin the whole batch in circles

Which increases the path length the falling glycerin takes to the bottom and keeps it up longer. It works well for me in a short squat 400lt tank.

Ant,

Were talking about eductors here. A whirlpool would rob energy from the eductor's mixing capability to create a whirlpool, thus reducing it's efficiency. Remember it takes energy to keep the oil plastered to the sides of the tank with a hole going down the middle. The main purpose of an eductor is to improve the efficiency of pump mixing allowing a smaller more energy efficient pump than would normally be needed.

quote:

A whirlpool would rob energy from the eductor's mixing capability to create a whirlpool, thus reducing it's efficiency. Remember it takes energy to keep the oil plastered to the sides of the tank with a hole going down the middle

Well I don't go for the hole down the middle. Just an anticlockwise swirl. It seems to work. The eductors energy is going to direct the flow somwhere why not around the rim? I thought the eductor increased mixing by sucking in more than was pumped through if you follow me. So why would it not still do that just as well if you keep it all swirling? I'm not saying you are wrong, just asking for clarification as the swirl works well for me.

Ant,

My guess is that when you changed things to get the swirl, you made an improvement other than the swirl itself. Yes one of the things happening is setting up a rotation on the tank, down in the middle and up on the sides. But even this is mixing, turning the tank in on itself. A swirl is spinning the contents of the tank as a whole.

Like the dead genius said, "everything is relative." If you imagine yourself inside the spinning tank, spinning with the tank, then it would appear to you that you were in a motionless tank with the world spinning around you. With the flow pattern set up by the eductor, it would appear to be more jacuzzi like with lots of moving liquids.

Ant,

This may sound like backtracking, but I'm not sure a little spin matters either way. Several of the beta testers installed their eductors pointing off centerline. I'm sure that created a swirl. They had good success. Like you've said before, just a little bit of "in the tank mixing" goes a long way. That's why I'm betting that even these home made eductors will see impressive gains, even though they don't do nearly as good a job of mixing as the ones designed by engineers and sold commercially.

Ant,

One of the reasons that a swirl is avoided is it induces a centrifuge effect on the fluid. Put a spoon full of sugar in a glass of water and stir...notice the sugar swirls into a pile in the center? Not really centrifugal but an eddy effect.

A straight down blast into the center of the reactor will produce a toroidal flow carrying the lighter liquid at the top deep into the reactor producing a very efficient mixing action.

Reflecting on what you both have said and thinking about the way I do it I think I know what is happening.

I use a venturi to mix the meth and oil with a long tube to provide a mixing zone. The swirl just keeps everything gently mixed after the venturi has done the main job.

I think it was Neutral or possibly Tilly who established a long time ago that if you have vigourous initial mixing you only need gentle motion to keep it all going for the duration of the reaction,

Ant,

What is the flow rate on your pump? With 100L in a short tank, anything over 30 Liters per min with a venturi would give good mixing. You would probably still get good mixing down to 20 liters/min.

It is a 400lt plus tank. Holds 350lt or oil plus 70lt meth and a few kilos of caustic.

My slow one with great pressure is 40lt min.
That worked well to mix and make bio but sucked slowly via the venturi.

I had another that did 80lt min.
That sucked better.

I have just got one that should do around 160lt min but haven't tried it yet.

They all did pressure. Two are multistage centrifugal and one was a turbine thingy but I melted the plastic rotor with 90DegC oil. The surviving two are all stainless throughout the pump.

Update on the eductor testing. I did a single stage batch at 20%. Not much sucess. I tested at 30 minutes and every 15 minutes till 1:45. None passed the 3/27 test. I even tested my last batch to make sure there was not some other problem in my testing, but it passed. So I pulled out the glycerol out and mixed 2% more methanol with 1gm/L (125 gm) KOH. Mixed and passed in 10 minutes. My yield is also down, Lost about 1 1/2 gallons where I usually only loss about 1/2 gallon with a 2-stage process

There was a reduction in the amount of stuff settling out in the 3/27 tests. So perhaps I was getting close. But at 1:45 it was much longer than It took for a 2-stage process and a lower yield.

So I think I will try the 2 stage on the next batch at 18%. But that will be in 2 weeks away, as all my fuel tanks are full.

Brent

Thanks for the update Brent!

I have just completed a 2 stage batch at 18%. 1 Hour on the first stage and <30 minutes on the second stage. It passed at 15 minutes but I ended up stopping and retesting at 30 anyways. Few changes occured this time. I slowed down the mix in rate of my methoxide. Before it took 5 minutes for the first stage and about 2 for the second. This time it took 13 minutes and 5 minutes repesctivly. I immediatly pulled about 3L (2.2%) of methanol out of the BD putting my net consumption at <16%. Plus I'm sure I'll get some more back them I distill the methanol out of the glycerol waste. The glycerol was noticably thicker than in the past.

If I get repeatable results (several batches) I may remove the eductor and run an identical batch to get a true feel of the effect it has. But without the eductor I was never able to pass with out 22% methanol so it must be helping some.

Brent

2 quick questions...will these eductors work with acid, and why wouldnt a tube that runs to the bottom of the processor,then u turns and shoots a stream straight back up work just as well. If made out of 1/2 black pipe, it probably could be squirrelled down into an appleseed and the screwed tight. Im going to try it on my drum processor, so I know its going to be a little different than the appleseed. Im actualy going to run one shooting down on one side of the tank, and one shooting up on the other. Just a quick idea as I was reading. It would keep everything to the top until the pump was shut off.

The next processor I build will be an appleturnover. I intend to mount the nozzle almost flush with the bottom pointing straight up. That will let me easily hardpipe the connection from pump to nozzle. I don't know how well it will work, but it's my next processor project. I just have to figure out what I want to use for a nozzle.

Mechanically this will work great with acids. Chemically, you want to pay attention to materials compatibility.

Thanks for the update Brent.

Rick,

I've been perusing McMasters alot these days to find a solution for the flow issue I'd have by implementing an eductor and a venturi off the same feed pipe coming from the pump's outlet. That aside, I found some parts that might be useful for the DIY eductor concept you've shown on this page. Here is a cobbled together page of what I found. The 2nd quote in red is something you wrote here that I wanted to keep as reference:



I may purchase and build one with these parts as I have to order some parts from here soon.

Also, can you comment here? I welded in a 1" fittingup at the top. I wanted to weld it offset from center and pointing vertically downward, but I messed up the angle as it is pointing to the wall of the processor instead of straight down. Do you think this will be an issue?



Thanks,
Bob

I think it will be okay. The angle might induce a little swirl but you should still get a good mixing effect and churning of the in-tank eductor.

If you have concerns about using the same pump/plumbing on the venturi and eductor if you can plumb one pump to the venturi and one to the eductor.

Paul,

I found a Y-fitting (technically called a "wye") on McMaster, so I'm going to try and see if that works. I need to order that soon. If that doesn't work, I'll consider your idea.

But for now, I'm only going to route the venturi and try that out first before adding on the piping to the eductor.

Bob

Those reducers look like they would do a good job.

A slight angle shouldn't make that much of a difference. What were looking for is for the main flow to reach the bottom instead of bouncing off the side of the tank.

On eductors, rather than trying to engineer it, we may be better of with experimentation. I'm thinking an eductor on water with whatever pump we use, should give a good pencil like stream that can reach to tree top level without a lot of spreading. I've been able to get that with a variety of different styles of nozzles. I tried to photograph experiments showing this, but I couldn't get far enough back to see the full stream and still make out the water stream clearly.

Rick,

I'll go back to the processor tonight to guesstimate at what point the nozzle hits the side of the tank so I can get a better idea if it'll hit near the bottom or if it's hitting the side. Since I'm welding a section of 1/2" pipe to a 3/4" bushing, I can make that a little higher on one side so that I can wrench the bushing+pipe off center as well.

Good thought about testing outside. and see how good of a pencil stream I can get. I have a few busy body neighbors I would love to douse!