I'm still not certain as to what is actually required for filtering Magnesol. I was told by the a Magnesol rep that one should *re*circulate the entire mix through a 1 micron filter so as to build up a layer of filter cake which will provide the required particulate removal. He did not mention however whether a nominal or absolute filter should be used but I would guess he was talking about a nominal (?).

Anyhow, this is one uncertainty that is keeping me from trying Magnesol. I'm also still uncertain about the great disparity between the dosage some here are using and what is recommended by Magnesol as well as what was used in the U of Iowa tests.

quote:

Originally posted by NoSmoke:
I'm still not certain as to what is actually required for filtering Magnesol. I was told by the a Magnesol rep that one should *re*circulate the entire mix through a 1 micron filter so as to build up a layer of filter cake which will provide the required particulate removal. He did not mention however whether a nominal or absolute filter should be used but I would guess he was talking about a nominal (?).

I vote for his having meant nominal. I REALLY want to get the patch tests done and get the cleanliness result from several different tries. The Dallas Group states that 10% of the particles are smaller than 20microns. That gives us general idea what we are looking to remove. The 90% that are 20 microns or greater and the easy one. It is the 10% of smaller particles that are of concern. The patch testing will give us consistant repeatable standardized testing to compare results. It also just happens to be able to give ISO cleanliness standards for the fluid. If you can get what cleanliness requirements are required by the actual injectors that you use, you can actually devevope some meaningful numbers. I'm not so worried about early 80's MB. I'm worried about the new injectors out there. Everyone, please contact you injector maker and ask what ISO cleanliness codes are applicable to the diesel that is run through the injectors. Ask if there is another component that is more limiting than the injectors. If there are, ask for their ISO cleanliness codes.
The right way to go about this is to find out how clean the fuel needs to be for the given equipment. Then set up a filtering system to achieve that cleanliness level.
I'm rambling, it's late, and I'll be in classes all day tomorrow.
http://www.emersonexchange.org/home/index.html

quote:

Anyhow, this is one uncertainty that is keeping me from trying Magnesol. I'm also still uncertain about the great disparity between the dosage some here are using and what is recommended by Magnesol as well as what was used in the U of Iowa tests.

Only further testing will get this one figured out. The goal obviously is to get acceptably clean fuel at the lowest cost possible.

-Jim

Good news! I thought that the Dallas Group had sent me all of the documentation available. They though they had sent me all of the information available. It turns out that we missed two documents. One is how to do a soap test. http://www.biodieselcommunity.org/testingforsoap/ covers the same information in a bit clearer format.

The other important piece of information establishes a direct relationship between the ppm soap measured in the raw biodiesel and the correct amount of Magnesol R60 to add.

quote:

Subject: treatment procedures
1. Ensure full transesterification reaction.
2. Separate methyl ester and glycerin phases.
3. Remove any excess methanol via flash evaporation if available (or other method).
4. Add between 0.5 - 2.0% by weight Magnesol R. Treatment temperature should be between 21 - 120+
°C. Engineering Guideline: add 1.0% Magnesol R60 by weight (0.009 kg per 1 Liter Biodiesel) per
800 ppm Soap as determined by AOCS Cc 17-79.) Soap method is attached in separate file.
5. Allow mix/resonance time of 10 - 15 minutes including 5 minute circulation time (below).
6. If using a commercial filter ensure proper filtration by circulating mixture through filter vessel
and back into reaction tank for 5 - 10 minutes or until filtrate is clear.
7. Once filtrate is free and clear of any particulate, fill finished product holding tank.
8. To ensure that your methyl ester is free from particulate filter finished product with a 1 micron
"absolute" filter media.

Now we have a way to determine how much needs to be added. I'm away from home this week and will not be able to do the test any time soon. I've searched the forum here but cannot find much about typical ppm soap levels in raw biodiesel. My initial thoughts on this are that this is just another reason to take the step over into the two step (acid/base) method of processing. In the caustic only method you add extra lye to make up for excess FFA's. These FFA's end up as soap. The more soap, the more Magnesol R60 needed. If you do the acid/base method the FFA's end up as fuel and not as soap, therefore less Magnesol R60 is needed.

I am REALLY interested in what typical ppm soap numbers people are coming up with.

-Jim

I have been testing soap in my BD unwashed, washed, and in the glycerine. I have made several test batches that I am doing bubble wash on and using Magnesol so that I can compare. I am using this procedure to test for soap.

soap test

I have had some luck with the Magnesol. I have been filtering through a 1 mu sock filter that I put diatomateous earth in. The sock absorbes alot of the bd.

On by base only experiments, I start with 2.9% ffa. The soap in the unwashed BD was 2070 ppm and the glycerine was 172000 ppm. After doing a bubble wash, the soap amount was 37 ppm and after using magnesol, it was 0, but I only used 10ml sample to test for soap instead of 100 ml in the magnesol samples because I don't have enough sample to test (I am sending the samples out to a lab to get tested for quality).

On samples using the acid/base procedure, I start with 10.3 ffa%. After the acid treatment, the %ffa fell to 0.68. For the unwashed BD, the soap content was 2700 ppm and the glycerine was 120000 ppm. The bubble washed sample has 330 ppm and the magnesol washed has 0 ppm soap.

I did do some brown grease that started with 32 ffa%. I got the ffa% to fall to 0.053. The soap content on the unwashed bd is 6100 ppm and the glycerine is 170000 ppm. For the bubble wash, the soap content was 138 ppm and the magnesol was 301 ppm.

There have been a few samples that the bubble wash was lower in soap than the magnesol samples. An exaple of this is where I started out with ffa% 5.2 and ended with 2.0 after an acid treatment. The unwashed sample has 1300 ppm and the glycerine has 160000 ppm soap. The bubble wash sample has 6.8 ppm while the magnesol has 90 ppm.

I have a feeling if I change the filtering system, I will get better results. The filter sock is quick but absorbs a lot of the bd, so I have been squeezing the sock a little to get more product. I know, bad idea. I am being more patient and letting is filter over night.

The instructions say to us 1% by weight based on a soap level of 800ppm. If you are getting 2.7ppm that works out to:
2.7/800 * 1% = .003375% by weight.

Can that be right?!! Check my math please!

Heck, even if you triple that it is only .01% by weight. That means that one pound of Magnesol R60 used to treat raw biodiesel with a soap level of 2.7 ppm will treat 10,000 pounds. Biodiesel is about 7 pounds per gallon (correct me if I am wrong) so that works out to 1428.6 gallons per pound!

Is 2.7ppm soap typical?

I have got to be missing something or doing this wrong. Please check my math and logic.

That should dramatically reduce the cost of treating with Magnesol!!!

PLEASE test and treat according to what the Dallas group has recommended and let's see how it works. I'll start on some this weekend when I get back, but since I have a stake in this my opinion really doesn't matter. We may find that there is a base level that we really don't want to go below, no matter how low the soap level in the raw biodiesel.

-Jim

I'm thinking that they mean to use 0.5-2% by weight and add an additional 1% for every 800ppm of soap.

I think I might have screwed up my soap results. According to the soap procedure I have, the soap result would be g soap/g sample and if expressing as ppm, we multiply our result by one million. I thought by not dividing by 1000 would give me ppm. So, the soap value should actually be 0.0027x10^6=2700ppm. WOW! I fixed the soap values on my previous post.

I was looking at their web page, and they have soybean bd with 551 mg/kg soap and use 1%. They also have yellow grease bd with 2458 mg/kg soap and only use 2%.

I can't figure it out from here. I'm going to have to figure it out when I get home. I need to mess with it myself until I can get my head around it.
-Jim

I have been informed that the Dallas Group recommends using 1% by weight Magnesol per 1000 ppm of soap. I am still waiting for a reply from the rep, but I believe that if you measured 2070 ppm soap, use 2% by weight and if you measure 500 ppm, then use 0.5% by weight.

So, where can one get Magnesium Silicate in Austrlia?

You can get plain old talc (hydrated magnesium silicate) quite cheaply by the 25Kg bag from Unimin in Sydney or Melbourne. But when I tried some on a 1 litre test batch I didn't notice any cleaner subsequent water wash. I'm assuming there's something special about the Magnesol product, judging by the price. Anyway, I'll give the no frills variety another go to check my results.

I have confirmation of the Dallas Group's recommended appication rate of 1% by weight Magnesol per 1000 ppm of soap.

quote:

if you measured 2070 ppm soap, use 2% by weight and if you measure 500 ppm, then use 0.5% by weight.

I had a batch that had 7000 ppm soap (this was an experiment on an acid/base/base method, normally I get 2000-3000 ppm soap on either base or acid/base method). Mixed 7% by weight for about 30 min at 77 C, filtered through DE with a vaccum filter twice and ended with clear BD with no soap. When I put some in water and shook, I did get some emulsion that did not go away, but the BD and the water is clear. I am sending some samples off to get tested.

Maybe I missed it but has anyone derived where we can get magnesium silicate? And also this earth stuff needed for the top of the filter?

I would love to be able to avoild washing but it requires special or barely available products then it becomes a moot point.

Thank for that very interesting article.

Wayne

It has been a while since I posted. Too much traveling for work, then fall break for my family. I finally got a chance to do some processing this weekend.
Yesterday I filtered some WVO from a mom and pop diner in town down to 150 (nominal) microns. I heated it in a 55 gallon barrel by burning 2/3 biodiesel 1/3 kero in a torpedo heater. While heating I put two aquarium pump tubes down into the oil to blow dry air in and hopefully vent wet air out. I pulled a sample and tested it for water using the scale/microwave method. (I think it came from the diner dry.) Then I put 34 liters of it into a HPDE-2 15 gallon barrel. I heated the barrel in a tiny bath tub we have downstairs. (It is too small for me, we use it for bath time for the dogs.) The hot tap water runs about 128 degrees there. I titrated the oil. It was dark, and I was worried, but it came out at 3. I used 5g/l NaOH + 3g/l NaOH for a total of 272grams of NaOH. I added the 272 grams of NaOH to 7.8 liters of warm Methanol (22.9%). After an hour of so it all finally dissolved. For safety's sake I took a shot of Jack Daniel's, purely for medicinal value. I batch added the Methoxide, screwed the cap back on and shook the heck out of the 42 liters or so in the barrel. I did this off an on for several hours and kept the temperature of the tub nice and hot. Then I called it good enough and went to bed. This afternoon I took the barrel outside and propped it up at an angle to make it easier to pump out the glycerin. I got two gallons or so of glycerin out and about 34 liters of nice biodiesel. Beautiful separation! Still smells like fries though.
This evening I made up my Brophenol Blue from powder. One little piece of advice. Do this over the sink and wear gloves. I have three blue fingertips on my left hand, but no stains in my wife's kitchen!
I was doing good on the soap test right up to the point where you titrate with the HCL solution. No HCL. Oh, well. I will get some tomorrow.
This looks like it will be a good representative batch of what you should expect from an average process.
Average oil (titration of 3)
Average amount of Methanol used (22.9%)
Average amount of NaOH used (5g/l base + 3g/l for titration)
Heating should be about right. Mixing could have been better.
The soap level in the demethanoled, unwashed biodiesel should be a good indicator of what people should expect to see in their processes.
Now if I can just finish and see what that level is...
By the way, will there be a problem using the 100ml of isoheet (Red Bottle), 5ml of Brop. Blue, and 100 grams of biodiesel that I have in a sealed mason jar? That was how far I got when I realized that I did not have the HCL.

-Jim

Iso-Heet is fine. Girl Mark has posted how to make up the HCl soln here.

http://www.biodieselcommunity.org/testingforsoap/

You won't need to use 100 ml of unwashed bd to test for soap. I use 5 g to test soap on unwashed and 30 g for washed bd. Also, I only use 1 ml of the indicator, not 5. On Girl Mark's web page, she states she uses 0.5 ml, but in the picture it looks more like 5 ml.


I had oil that I used the same amount of lye (5g + 3g NaOH) and ended up with 1580ppm soap in the bd and 188000ppm of soap in the glycerine layer. I am curious as to what you get.

Man, I must have been tired! 5ml! Well, I should not miss the color change! Girl Mark's page on testing
http://www.biodieselcommunity.org/testingforsoap/
is what I am working off of. I missed the 5/.5 part, but it says 100 grams of biodiesel. It would seem that varying that would really change your results. How do you adjust for this when using 5g of biodiesel?
Nevermind, I just answered my own question, the equation to convert the results into a final number includes the weight of the sample.
soap in parts per million=
(ml of .o1N HCL solution)/weight of sample*30.44

The part that I am worried about is mixing all of them together and then leaving them in a sealed container for 24 hours. My worry is that it will change my results.

-Jim

It should be ok, but you might want to do another titration just for verification. I have noticed that after the sample has sat for awhile after titration the color doesn't change like it does when using phenolphthalein.

Jim D, this has nothing to do with the subject "Waterless Soap Removal", but your 34L batch in a 15 gallon barrel is just what I have been wanting to try. My concern is - how do you "shake the heck out of the 42 liters"? My concern is that it won't be mixed well enough.

I'm 6FT 1IN tall and weigh 285 pounds. The barrel weighs around 80 pounds loaded. I picked up the barrel and shook the heck out of it. I did that every 5 minutes for about 20 minutes. Then it went down to every 15 minutes for a while. Then every half hour for a couple of times. Then one good shake before I went to bed.

-Jim

This is what I use to filter Magnesol. I've been adding 2% by wt to ~300L batches and circulate through the filter and back to the tank until I can see pressure on the filter inlet gage. I then continue circulating until all the magnesol is held up in the filter plates.

You guys have done a lot of work since I tuned in a couple of months ago. The Dallas Group info is great. I was not a aware that the Magnesol Conc. was based on soap Conc. My bromophenyl blue should be here this week and I'll give it a try.

I'm still trying to use my stacked disk centrifuge to water wash the BD and making whipped cream... My WVO starts around 15% FFA and I go Acid/Base/Base. I'll try to post results next week.

Plate_and_Frame_Press.jpg (256 KB, 123 downloads) Plate & Frame Press

foxcreek

Nice filter press. I am curious as to your results.

How do you like the results of the acid/base/base method? Are you doing the FATTA method for your acid stage? Did you get glycerine on the second base stage? I have not, so I am curious. Also, after removing the methanol, the bd became a gel, but after washing or using Magnesol, the bd stayed liquid. I have noticed that doing the acid/base/base causes the soap content to be a little higher on the first base stage than doing an acid/base method, and then the second base stage causes the soap content to at least double. On the first base stage I found the soap to be 3000 ppm, where usually it is 2500-2700 ppm for acid/base, and the soap content after the second base stage is 7000 ppm. I am sending some samples from an acid/base/base method to see how the quality is, but if it isn't significantly better than an acid/base method, I will stick with the acid/base method.