BIODIESEL & SVO DISCUSSION FORUMS





Sponsors    Biodiesel and SVO Forums Home    Forums  Hop To Forum Categories  Biodiesel  Hop To Forums  Dry Washing Biodiesel    Simple Amberlite Setup
Page 1 2 3 4 5 

Moderators: The Trouts
Go
New
Find
Notify
Tools
Reply
  
Simple Amberlite Setup
 Login/Join
 
Member
posted
I received my 40kg of Amberlite BD10dry a couple months ago and I started pondering how we could build some columns that meet Rohm & Haas's guidelines and didn't cost us an arm and a leg.

So the Amberlite user guide calls for the following

* 3:1 Column height-width ratio
* A flow rate of 0.36gph/lb dry resin
* Resin raised 2" from bottom with #80 (175 micron) screen

The column you choose should be able to handle pressure up to 25 psi, be easy to clean and be made of a biodiesel compatible material.

For an added bonus the perfect column would be practical for the typical homebrewer, be constructed from readily available materials and scalable for the small scale producer.

My friend and I pondered this for dilema for weeks. We thought of welding gas tanks, propane tanks, 5 gallon plastic buckets, coolers, square tubing, irrigation drain, PVC pipe, 2" hose, and beer kegs. We thought of everything we could, then one by one shot down every idea.

Finally I came up with the super-cheap idea of using 1 gallon natural HDPE containers, putting just 2 lbs of resin in each container, all in a series configuration.

Our target flow rate was 10 gallons / hour.

I worked it out and determined 15 little jugs across a table top would suffice.

Unfortunately I couldn't ever figure out how to get a screen inside the narrow neck of the container. I figured we could cut the top off and melt it back together. Hrmm.

Then we started back down the path of the beer keg. We purchased 88 lbs of Amberlite and we figured (3) 15 gallon beer kegs could hold it all. Plus, it's stainless!

We could setup the kegs in glycerol catcher -> lead -> lag configuration and process 10 gal / hour.

We still had the problem with getting inside the keg. I really didn't like the idea of cutting, welding, etc. Plus the keg is almost square, surely a violation of height to width ratio.

Then it hit me ... Corning Kegs!

Corning kegs are also stainless, handle up to 150 psi, have a large hatch in the top, and even dry connect fittings with a pickup tube routed straight to the bottom! Then came the height to width ratio. I quickly measured one finding the diameter to be 7" and a height of 22" ... 3.14:1. Perfect!

I searched craigslist and quickly found 9 corning kegs for sale for $23 a piece.

Each corning key has a volume of 5 gallons. This would allow for 10 lbs of resin to be added to each column and still have sufficient room for 300% expansion.

Since I just bought 9 corning kegs and I had 88 lbs of resin I decided to add 9.77 lbs of Amberlite to each column (88 lbs / 9 kegs). This worked out to a operating flow of 3.519 gph.

Not exactly speedy, but for the small homebrewer this equates to 75 gallons in 24 hours.

I was still looking for 10 gallons per hour and 3 columns in parallel would get me there!

We proceeded to construct 3 corning columns!

All we had to do is get a screen in the bottom the keg, which is dished .... Hrmm.

(1) Add 2 lbs of glass marbles to the bottom of the keg.

(2) Starting with an 8" square of #80 stainless mesh, cut a 7" circle with scissors.

(3) Use a hole punch to put a 1/4" hole in the middle of the screen.

(4) Install a small rubber grommet in the newly created hole.

(5) Cut a peice of the cheapest flimsy garden hose you can find to the exact inside circumference of the keg.

(6) Curl screen and place inside keg and carefully place at bottom over settled marbles.

(7) Replace stainless pickup tube, oil to easily slide through screen grommet.

(8) Let garden hose act as a bumper around the perimeter of the screen.

(9) Add 1.5 gallons clean biodiesel to keg.

(10) Pour 2 lbs increments of Amberlite in top of keg, rotate as you pour to disperse the Amberlite around the outer perimeter of the screen. This will also help squash the garden hose to the outer edge under the weight of the Amberlite.

(11) After you add 9.77 lbs of Amberlite backflush the column at 2.5 gph. After two hours and all air being purged, the column is ready to use.

That is it ... simple ... and it works.

Initially we had all three columns in series at <3 gph and it was working quite well.

Currently we are cleaning about 170 gal/day, gravity fed via 330 gal stacked tote.

We are using 3 of our columns in parallel, around 7 gal/hour.

I could increase the flow rate, but I've decided to play it safe at first.

Our biodiesel settles for quite some time and we always experience soap fall out, we need to filter the biodiesel before introducing it to the ion exchange resin to prevent resin fouling.

We filter the fuel through a 25 micron sock filter when filling the top tote.

On the output of the top tote we placed a 100 micron strainer followed by a 20 micron paper filter to protect the resin beds.

Amberlite is stated to work at ambient temperature or about 65-70F. Since our operation is outdoors and the fuel has most certainly cooled. I needed to hook up a little inline heater. I ended up using a 900 watt inline heater out of a $12 coffee machine and hooked it up to a light dimmer wall switch for heater output control.

We only need to raise the temperature to around 85F so it passes through the resin bed warm. We operate the heater at about 450 watts when flowing at 7 gal/hour.

For now, everything is hooked together with 1/4" clear hose.

So ... long story short.

5 gallon stainless keg, $23.00
2 quick connect fittings + barb nipples, $18.00
4 lbs glass marbles, $4.50
10 ft 1/4" clear hose, $3.00
8"x8" #80 stainless mesh, $6.00
1 rubber grommet, $0.18
21" flimsy yellow garden hose, $0.52

Total column cost = $55.20

Just add $72 of Amberlite and you can wash over 1000 gallons of biodiesel at 3 gal/hour.

Total cost for the homebrewer, $127.20.

So for our purposes we have spent triple the amount to clean 3000 gallons of fuel at 10 gal/hour.

So far we have pushed through 700 gallons and it's working great!

I'll follow up with pictures and hopefully a accurate lifetime once I get through 3000 gallons.



--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com

This message has been edited. Last edited by: sunbreakbiofuels,
 
Registered: March 21, 2006Reply With QuoteReport This Post
member

posted Hide Post
Very informative and detailed rundown, thanks.



** Biodiesel Glycerine Soap - The Guide
- on 5 continents helping people make & sell soap from the Biodiesel Glycerine.


 
Location: :-) Great White North eh ? | Registered: December 10, 2004Reply With QuoteReport This Post
member
2010 Sponsor
arborbiofuelscompany.com
posted Hide Post
Jeff,

Nice job, It sounds like you have found a great answer for homebrewers. Essentially you have achieved the same price point we did for our design.

Question though when you run the 3 units in parallel - do you still maintain a lead lag configuration? So for $127; is this one column or two?

Also are you recommending producers to remove methanol prior to or after the ion exchange beds?

The Dry Wash experts are saying it doesn't really matter much as long as the excess methanol is 2% or less of the total mixture.


Chad
2000 F350 SD 7.3PSD, 1996 F250 SD 7.3 PSD, 2005 Jeep liberty CRD, 2002 TC29D New Holland Tractor, 6K Diesel Generator, Heated Power Washer 'All Burnin Bio'

http://arborbiofuelscompany.com/ Selling Dry Wash Columns.

http://i88.photobucket.com/alb...icationGCresults.jpg
 
Location: S.E Michigan | Registered: May 12, 2007Reply With QuoteReport This Post
member
2010 Sponsor
arborbiofuelscompany.com
posted Hide Post
Oh Jeff, also since your using a stand pipe to pump the purified biodiesel out of the bottom of the corney keg; how are you controlling pump flow rate?


Chad
2000 F350 SD 7.3PSD, 1996 F250 SD 7.3 PSD, 2005 Jeep liberty CRD, 2002 TC29D New Holland Tractor, 6K Diesel Generator, Heated Power Washer 'All Burnin Bio'

http://arborbiofuelscompany.com/ Selling Dry Wash Columns.

http://i88.photobucket.com/alb...icationGCresults.jpg
 
Location: S.E Michigan | Registered: May 12, 2007Reply With QuoteReport This Post
Member
posted Hide Post
Jeff,
that's about how you described it to me when I saw your setup in person. Thank you for the information. Keep us posted here.
P.S. I'm on the hunt for Cornelius keg or two.
Garry Young.


2001 Powerstroke
BH100 (Bio-Homebrew 100%)
 
Location: Portland, Oregon | Registered: March 08, 2008Reply With QuoteReport This Post
Member
posted Hide Post
quote:
Originally posted by CHADS454:
Jeff,

Nice job, It sounds like you have found a great answer for homebrewers. Essentially you have achieved the same price point we did for our design.

Question though when you run the 3 units in parallel - do you still maintain a lead lag configuration? So for $127; is this one column or two?



When the 3 units are run in parallel there is no lag column, you are just increasing flow rate. If you wanted a lag column you would need to create another column for each of the 3 parallel columns.

The $127 is what I spent per column; How you configure them is up to you. I purchased 9 Cornelius kegs in all so I could eventually have 3 parallel glycerol -> lead -> lag columns, or 9 columns in total at 10gph.

The "lag" column is really only beneficial for "hot-swapping" columns in a continuous flow setup.

I'm sure some people might like the idea of a "polish" column, but really, if it isn't doing it's job in the first column you haven't sized things correctly or your flowing too fast. The lag column is only there if the first column has become saturated and you need to remove it for resin replacement.

For the typical homebrewer I would imagine you could stop for a little bit Wink

I limit the flow rate with a need valve on the combined output of all three columns, this will obviously change as the top tote drains to a lower level and is made even worse by the head pressure building up in the bottom tote.

I was hoping that a float switch in the top tote could trigger a pump once an hour that ran for 60 seconds to replace the 7-10 gallons that drained in order to maintain a constant "head pressure". I had never taken into account the back "head pressure" from the bottom tote as the fluid level increases. You can compensate for this by moving the drain out of the biodiesel and placing it near the lid. However this removes several feet of head.

All in all the flow "slowing down" can't hurt, it can only clean the fuel more! There was lots of "slowing down" or "slow flow" problems before we started filtering what we pumped into the top tote. We were constantly cleaning strainers and filters every 30-40 gallons. Now we might have to inspect the strainer/filter every 200-300 gallons.

quote:

Also are you recommending producers to remove methanol prior to or after the ion exchange beds?

The Dry Wash experts are saying it doesn't really matter much as long as the excess methanol is 2% or less of the total mixture.


Current Rohm & Hass literature suggests ion-exchange "before" de-methylation. If you understand how this particular resin works it seems obvious. Methanol acts as a solvent to carry the soap through the resin beads. If the soap has precipitated out it will simple clog up the resin bed. When you re-generate the Amberlite for more glycerol removal you wash it with methanol to flush the weak bonded molecules (aka glycerol) out of the resin beads.

So as far as I know, more methanol the better. Although it should work just fine with de-methylated, de-soaped biodiesel produced in the GL1 setup. The Amberlite should last 4 or 5 times the rated capacity using a Graham's methods to initially get the soap down to 100ppm. The biodiesel will need to be "absolutely" dry of methanol as the Amberlite will absorb anything that is left and precipitate "more" soap thus clogging the resin bed. I would think that "initializing" the column with methanol might alleviate the soap precipitate problem.

Of course then your biodiesel will have methanol in it again. Perhaps not such a bad thing if there is no soap?

Once the Amberlite bonds to free catalyst, soap, salts, glycerol, methanol, etc. It has a certain bond strength to the cations inside the styrene matrix. Turns out that most of those bonds are very strong and you would have difficulty washing them out. However, the bond for glycerol is quite weak. It is easy to remove by washing the Amberlite with a more preferential alcohol.

Now I haven't been able to figure out the "displacement order" for Amberlite yet. Although I think it goes something like this:

Water -> Methanol -> Glycerol -> Soap -> Salts -> Catalyst

Basically anything at the beginning of the list you could remove, everything at the end of the list is stuck in there once it bonds. This explains why after the Amberlite becomes "exhausted" you can flush some Methanol through it and reuse it for glycerol removal. Although it will still remove all other impurities, there is less available surface area for the bonds to occur and at the designed flow rate would let impurities through. Since the primary impurity directly after separation is glycerol, the glycerol would attach to the Amberlite first making it available to be cleaned and used again.

Unlike Purolite, Amberlite does not take their spent resin back for regenerating and ultimately selling it back to me for a lower price.

After looking at Rohm & Hass Amberlite BD10dry, Purolite PD206 and LanXess Lewatit GF202, they are all basically the same strong acid cation in a styrene matrix. All the specification listed by the manufacturer seem pretty close as well.

So far I have only used the Amberlite product and after passing over 1000 gallons through a "3000 gal lifetime" column it is working as advertised, outdoors at 65-70F (heated fluid) in unattended gravity feed operation.


--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post



Member
posted Hide Post
quote:
Originally posted by CHADS454:
Oh Jeff, also since your using a stand pipe to pump the purified biodiesel out of the bottom of the corney keg; how are you controlling pump flow rate?


There is no pump, I have two 330 gallon totes stacked on top of each other. I use a common Fill-Rite pump to fill the top tote. By keeping the fluid level in the top tote relatively constant, the flow rate is basically the same all the time.

If for some reason the flow was to great I have a needle valve on the combined output of all three columns that I can slow things down with.

--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
member
posted Hide Post
Jeff, I applaud your work in this area. I didn't even know what a Corny Keg (aka Corney, Cornelius or Soda Keg, if your searching Ebay or the Web)was until you posted about them!

I would however like to modify or extrapolate some on a few of your comments:

Although a lag tank is not absolutely necessary, it does allow for at least 20% more use (life) out of the lead tanks resin by accounting for the initial impurities that begin getting through the resin bed as it approaches full adsorbtion capacity (this bleeding through will be increased with larger percentages of methanol if your biodiesel isn't demethylated). As a home brewer being that the resin is your largest cost, this 20% could be significant over time however this cost would need to be weighed against the cost of another full column of resin. The additional column cost will not pay for itself initially.

Then: To remove methanol OR Not to remove methanol - that is the question...

As you say the methanol will keep more soap and glycerin in solution therefore allowing the resin in the column to work as designed ("at the level of solubility" - the industry phrase) however it does this at what could be a tremendous cost in the resins life expectancy. The gel-type resins like PD206 or BD10 Dry all quote loading capacities based on total impurity concentrations of 500 ppm (this would include soap, catalyst and glycerin along with other trace postively charged ions like Calcium and magnesium etc.)

With this being the case and that most home brewers utilize used cooking oil with FFA concentrations of typically 1-4%; most test results I've performed indicate that with 1.5 to 2% methanol concentrations (after glycerin removal), the crude methyl ester will contain at a min. 1000 ppm up to 4000 ppm of total impurities). This will rapidly exhaust the resin and significantly increase the cost per gallon.

In my opinion unless resin is operated most optimally then the cost isn't justifiable to the home brewer when compared to water washing or long term settling and not worrying about the methanol. Neutral suggest adding just a small bit of isopropyl to ensure no glycerin or soap falls out if the methanol laden biodiesel has to sit around a while and is apt to loose excess methanol by evaporation. However most home brewers work on a just-in-time basis and even this is probably not a big concern.

Also methanol being passed through your resin column just increases the number of components in a system were flammability is a concern as well as making your resin a flammability hazard when it comes time for disposal or shipment back to purolite for regeneration (the resin can be dried but its just another process hassle for the home brewer which will discourage the use of resin).

Which brings me to a main point - if the parameters/restrictions for resin use (in biodiesel production) are not well understood, then it will not be adopted and from my point of view that would be a sad outcome since I believe it to be the fastest way for home brewers/coops to produce higher quality fuel and have further credibility/larger acceptance...

Lastly, Lewatit's GF202 although similar in function is NOT a gel-type resin but instead a macroporous resin. All this means chemically is the plastic polymer has a much larger number of cross-links when compared to say a gel-type resin. This makes them a more robust/resilent and affords them a life expectancy (5-10 yrs) which far exceeds resins like PD206.

PD206 will also expand more rapidly and more fully with methanol being ran through them as the resins expand with loading, some burst or break their cross-linking bonds. Therefore gel-type resins will typically require ~10% change out to ensure the active resin depth is maintained after elution (flushing out of the glycerin).

Macro resin doesn’t expand any where near as much as the gel-type resins and therefore don’t require the same 3:1 resin column depth nor as much change out after regeneration (if any). Macro does require a deeper bed (36”), may require lower operating temp. and needs a longer residence time. (so, slower flow capacities on a pound for pound basis when compared to PD206 or BD10 Dry - 66%).

That's my bucks worth,

GCG
 
Location: Michigan | Registered: May 08, 2007Reply With QuoteReport This Post
Member
posted Hide Post
Here are the pictures I promised, this was the first day where there was any good light.

Ignore the fact that one of the columns is in a bucket. I had a bad o-ring and was containing the leak before I fixed it. The bucket is no longer needed.








--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
Member
posted Hide Post
quote:
Originally posted by GCG:
Jeff, I applaud your work in this area. I didn't even know what a Corny Keg (aka Corney, Cornelius or Soda Keg, if your searching Ebay or the Web)was until you posted about them!


Thank you very much, we were really trying hard to come up with a solution that could fit almost any budget.

quote:

I would however like to modify or extrapolate some on a few of your comments:

Although a lag tank is not absolutely necessary, it does allow for at least 20% more use (life) out of the lead tanks resin by accounting for the initial impurities that begin getting through the resin bed as it approaches full adsorption capacity (this bleeding through will be increased with larger percentages of methanol if your biodiesel isn't demethylated). As a home brewer being that the resin is your largest cost, this 20% could be significant over time however this cost would need to be weighed against the cost of another full column of resin. The additional column cost will not pay for itself initially.



So once the column has reached 80% saturation and slight impurities start making it through the resin bed couldn't you just run the slightly impure biodiesel through the column a second time?

For my setup I'll be using a lag column so I'm not entirely concerned. Personally I would like to see homebrewers move away from water washing. So I would like to imagine that a single column could be workable.

quote:


Then: To remove methanol OR Not to remove methanol - that is the question...

As you say the methanol will keep more soap and glycerin in solution therefore allowing the resin in the column to work as designed ("at the level of solubility" - the industry phrase) however it does this at what could be a tremendous cost in the resins life expectancy. The gel-type resins like PD206 or BD10 Dry all quote loading capacities based on total impurity concentrations of 500 ppm (this would include soap, catalyst and glycerin along with other trace positively charged ions like Calcium and magnesium etc.)

With this being the case and that most home brewers utilize used cooking oil with FFA concentrations of typically 1-4%; most test results I've performed indicate that with 1.5 to 2% methanol concentrations (after glycerin removal), the crude methyl ester will contain at a min. 1000 ppm up to 4000 ppm of total impurities). This will rapidly exhaust the resin and significantly increase the cost per gallon.



I must admit that the first barrels of fuel that went through the column had settled several weeks. The biodiesel we are feeding through the columns now has been static settled for 2 months.

The Rohm & Hass representative said that the 900-1600 lbs / dry lb of resin figure was obtained from real world commercial operations using either a centrifuge or 8-12 hours of settling with standard transesterfication using either KOH or NaOH with a FFA level of 1% or less.

I was given no details as to the amount of NaOH or KOH used or if the catalyst was in fact sodium or potassium methylate.

We process using the 2 stage acid base method. Our oil will typically titrate at 1.4 or less using NaOH after the acid stage. Our glycerol is seperated using Graham's electro-static 1000Vdc method in about 30 minutes and then is put into a 14 tote settling rotation that allows the fuel to settle for at least a month outdoors at 40 ish F.

From my understanding a titration of 1.4 should represent a FFA of 1.07%.

So all of my reactions should be within Rohm & Hass's guidelines.

Perhaps my extreme settling time is making the resin's job easier.

I also understand what you're saying, and I completely agree that the resin can only absorb so many grams of impurities before it's saturated.

I'm just interested if Amberlite will do what it says if you follow the companies published requirements.

As far as I can tell, I am following the requirements and even exceeding some. So I would hope it works.

quote:


In my opinion unless resin is operated most optimally then the cost isn't justifiable to the home brewer when compared to water washing or long term settling and not worrying about the methanol. Neutral suggest adding just a small bit of isopropyl to ensure no glycerin or soap falls out if the methanol laden biodiesel has to sit around a while and is apt to loose excess methanol by evaporation. However most home brewers work on a just-in-time basis and even this is probably not a big concern.



I figured that the most it would cost is double the manufactures stated cost of 3.5-7.0 cents / gallon. Personally I dislike water washing and I think many other people dislike it as well. I think for some people even 14 cents/gal is quite reasonable. Perhaps I'm wrong.

quote:


Also methanol being passed through your resin column just increases the number of components in a system were flammability is a concern as well as making your resin a flammability hazard when it comes time for disposal or shipment back to purolite for regeneration (the resin can be dried but its just another process hassle for the home brewer which will discourage the use of resin).



An interesting point, I should try igniting some of my biodiesel to see how flammable the methanol makes it.

quote:


Which brings me to a main point - if the parameters/restrictions for resin use (in biodiesel production) are not well understood, then it will not be adopted and from my point of view that would be a sad outcome since I believe it to be the fastest way for home brewers/coops to produce higher quality fuel and have further credibility/larger acceptance...



I couldn't agree more, I had to dive through tons of incomplete documentation and several conversations with Rohm & Hass before I felt I had something that worked.

You can't really half ass the resin bed setup or it would simply cost you money and waste your time.

quote:


Lastly, Lewatit's GF202 although similar in function is NOT a gel-type resin but instead a macroporous resin. All this means chemically is the plastic polymer has a much larger number of cross-links when compared to say a gel-type resin. This makes them a more robust/resilent and affords them a life expectancy (5-10 yrs) which far exceeds resins like PD206.

PD206 will also expand more rapidly and more fully with methanol being ran through them as the resins expand with loading, some burst or break their cross-linking bonds. Therefore gel-type resins will typically require ~10% change out to ensure the active resin depth is maintained after elution (flushing out of the glycerin).

Macro resin doesn’t expand any where near as much as the gel-type resins and therefore don’t require the same 3:1 resin column depth nor as much change out after regeneration (if any). Macro does require a deeper bed (36”), may require lower operating temp. and needs a longer residence time. (so, slower flow capacities on a pound for pound basis when compared to PD206 or BD10 Dry - 66%).

That's my bucks worth,

GCG


I was unaware of the differences with the GF202 resin. I'll have to dig through my documentation again. One of the things I didn't understand with the GF202 was if you could regenerate it for soaps/catalyst/etc or only glycerol. Their literature implies it can be regenerated for everything. Although I never saw it stated anywhere other than for glycerol.

Thanks for the terrific reply and fantastic input.

Where have you been learning about different resin beds? I thought I had read just about every message on this forum and you have clearly pointed out a few tidbits that aren't out there?

--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
member
posted Hide Post
Jeff,

If you search GCG post you will find numerous post related to this topic. I work in Nuclear Power and we resins for various purposes; everything from condensate purification to GC columns.

I have never really liked the idea of water washing for obvious reasons and have been researching elegant industrial processes for biodiesel production for over 2 years. This has led to many late night conversation with leading Phd's in Germany (Lewatit) on resin use for every thing from vegetable oil purification, FFA esterification, to direct transesterification.

There is a Taiwanese company [URL=http://www.sunhobiodiesel.com/] Phd who has figured out how to immobilize lipase enzymes on the Lewatit's macro resins and can do both esterification and transesterification of oils directly. However I can't get them to let me license the technology and a small economy plant takes a year to deliver and about 1.8 million dollars FOB... Most interesting is the glycerin by-product comes out nearly pharmaceutical grade and all methanol is recovered. Cool stuff!

Anyway we're a budding biodiesel producer in SE Michigan just about ready to do larger scale production 50,000 gallons and up. Presently awaiting our ASTM pedigree.

Thanks for taking my comments positively - it's very easy to take the written word harshly and I take some time to compose my responses in order to foster the highest degree of open source material.

GCG
 
Location: Michigan | Registered: May 08, 2007Reply With QuoteReport This Post
member
posted Hide Post
Oh by-the-way in terms of elegant and cost effective production of biodiesel for the home brewer or small coop, Arbor Biofuels Company is testing a small scale coalescing unit and will be able to post some numbers soon on its initial effectiveness.

This unit would ensure the simplest and fastest way to go directly from separation to ion exchange resin columns - from Oil to ASTM fuel for the typical 30 gallons of biodiesel from an Appleseed processor in 7-8 hours (mostly depending on demethylation time of 90 minutes)

Reaction: at 50-60 deg. C (90 minutes)
Separation: (60-90 minutes - 95%)
Demethylation: using venturi circulation at 65-75 deg. C (90 -120 minutes)
Coalescion Big Grin: (15 min.)(poetic license)
Purification: Ion exchange (180 min.)

Add a little time for transfers and the like and high quality fuel in 8 hours (preheating and filtering the raw used oil could be included in this).

Wish us luck,
GCG
 
Location: Michigan | Registered: May 08, 2007Reply With QuoteReport This Post



Member
posted Hide Post
quote:
Originally posted by GCG:
If you search GCG post you will find numerous post related to this topic. I work in Nuclear Power and we resins for various purposes; everything from condensate purification to GC columns.


Great, good reading material for this evening.

quote:


There is a Taiwanese company Phd who has figured out how to immobilize lipase enzymes on the Lewatit's macro resins and can do both esterification and transesterification of oils directly. However I can't get them to let me license the technology and a small economy plant takes a year to deliver and about 1.8 million dollars FOB... Most interesting is the glycerin by-product comes out nearly pharmaceutical grade and all methanol is recovered. Cool stuff!



A friend of mine and I are actually trying to grow some metarhizium anisopliae fungus in hopes to reproduce some of the results we have read about.

I have a brochure on Amberlyst BD20 for pre-treating high FFA feedstock. The BD20 doesn't really offer much compared to sulfuric acid for FFA levels below 20%, but beyond 20% and it really starts to shine. The downside is that their catalyst resin is just as susceptible to water as sulfuric acid. So you must remove all methanol and water from oil, re-add more methanol and then pass the solution through a second stage column. Repeating the process 2 or 3 times allows the use of 100% FFA. Pretty impressive but also energy intensive. It all sounded so good when it was "room temperature" but then followed by distillation. uGh.

I wonder if you could use a molecular sieve to remove the water? I think I read somewhere that Zeolite 3A could be used but that the similarity between water and methanol was to close in terms of size.

When I first saw this I was hopeful of just converting all the oil to FFA through hydrolysis and just processing 100% FFA at room temperature. Of course then you would have water in your FFA which would screw up the Amberlyst.

Super critical methanol has been shown possible at lower pressures and temperatures (1500psi, 620F) if you use high FFA feedstock. Plus super critical methanol actually prefers some water in solution to help with simultaneous hydrolysis of remaining mono,di,triglycerides.

quote:


Anyway we're a budding biodiesel producer in SE Michigan just about ready to do larger scale production 50,000 gallons and up. Presently awaiting our ASTM pedigree.



We are in the same boat, we process 3 batches a week; yielding just over 250 gallons per batch. Last year we processed about 10,000 gallons.

This year I would like to see that number triple, currently we are having a hard time keeping up with all the oil that comes to us.

quote:


Thanks for taking my comments positively - it's very easy to take the written word harshly and I take some time to compose my responses in order to foster the highest degree of open source material.

GCG


I'm an open source advocate, so I completely understand. No advancements were ever made in the middle of a flame war. Wink

quote:
Originally posted by GCG:
brewer or small coop, Arbor Biofuels Company is testing a small scale coalescing unit and will be able to post some numbers soon on its initial effectiveness.

This unit would ensure the simplest and fastest way to go directly from separation to ion exchange resin columns - from Oil to ASTM fuel for the typical 30 gallons of biodiesel from an Appleseed processor in 7-8 hours (mostly depending on demethylation time of 90 minutes)

Reaction: at 50-60 deg. C (90 minutes)
Separation: (60-90 minutes - 95%)
Demethylation: using venturi circulation at 65-75 deg. C (90 -120 minutes)
Coalescion Big Grin: (15 min.)(poetic license)
Purification: Ion exchange (180 min.)


We use Graham's Lamings electro-static glycerol coalescing technique.

Most people on the forums seem to have only tested the method with small volumes (< 200L).

We adopted the idea right before winter. Just in time too, as our glycerol was settling up in our hoses within 10 minutes at 23F. No time for long term settling at those temps.

I have successfully employed the technique on our 330 gallon (total volume) processor without any problems on over 17 batches.

We achieve good seperation in about 30-40 minutes using 1000Vdc current limited to 600uA with nothing more than 2" of exposed wire for probes, lowered to the middle of the tank. I have even used the tank itself for the negative lead. This also was effective, although this justifies more testing.

This is a bit more current than Graham had specified (I think he specified 4uA).

Although the seperation seems to work well, I also settle the seperated fuel for about 1 month before actually running it through the resin bed.

Once I get through the 3,500 gallons of biodiesel that I need to clean. I'll attempt to run a freshly reacted, electro-static coalesced batch directly into the ion-exchange column as Rohm & Haas claims can be done.

--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
member
posted Hide Post
When it comes to electrostatic separation I talked over a year ago with the leading Phd (Director of R&D) of NATCO about building me a small 10 gpm unit.

He laughed at the idea of it (they talk in terms of barrels per minute crude oil) but was willing to begin the process in their down time - which was going to be 3 months away. He said their lab size unit was a minimum of 4 feet in diameter and given their process it was unlikely that they could get it down to my desired flow rates but for 10 thousand he'd begin testing and specking out used cooking oil.

He openly discussed what I would have thought was proprietary info: exact dimensions, ratios, geometry and voltages (incl. amps) DC and AC currents are both used and they use a minimum of 8000 volts. And do indeed apply voltage/potential to the vessel as part of the process.

Maybe I should dig up those numbers...
 
Location: Michigan | Registered: May 08, 2007Reply With QuoteReport This Post
Member
posted Hide Post
Does anybody know how effective a single column would be using multiple passes? I like the idea here, but I wouldn't have the space needed for the 3 columns, but would have time for multiple passes.
 
Registered: October 02, 2006Reply With QuoteReport This Post
member
posted Hide Post
Initially this technic of multiple passes on the same column will work - especially since initially one column with proper residence time is enough.

The second column (lag, if you will) is primarily in place to ensure the "lead" column is fully used, since as the unit begins to load say initially 20%; a small amount of carrier over of glycerin or pass through will begin to occur.

The second bed provides a "polish" on the biodiesel at these initial stages of glycerin loading on the "lead" column.

Later as the first column reaches say 50% loading capacity, the second column now starts to provide additional purification.

I haven't yet determined or inquired to the exact point of loading when the lag tank is necessary to ensure ASTM compliant biodiesel...(when the percentage of glycerin removed from crude biodiesel fails to be enough to get a sample to <0.24%).

GCG
 
Location: Michigan | Registered: May 08, 2007Reply With QuoteReport This Post
Member
posted Hide Post
I have no experience with Amberlite.

Does the set up have to create a pressure vessel? Or could it work with just gravity allowing the bio to pass through the Amberlite on its own?
 
Registered: April 09, 2008Reply With QuoteReport This Post
Member
posted Hide Post
I have 2 cornelious kegs i bought some time ago(not sure what i was gonna use em for).Not sure I followed how to set em up for amberlite but would be interested in finding out more about setting them up. Any help would be appreciated.
 
Location: Athens Al | Registered: July 22, 2006Reply With QuoteReport This Post



Member
posted Hide Post
quote:
Originally posted by Tiger744:
I have 2 cornelious kegs i bought some time ago(not sure what i was gonna use em for).Not sure I followed how to set em up for amberlite but would be interested in finding out more about setting them up. Any help would be appreciated.


Which part of the instructions did you have difficultly understanding. Let me know and I can edit the original post and hopefully make it easier for everyone.

--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
Member
posted Hide Post
quote:
Originally posted by Chad613:
I have no experience with Amberlite.

Does the set up have to create a pressure vessel? Or could it work with just gravity allowing the bio to pass through the Amberlite on its own?


Large pressures are not required, it was just a nice benefit of using a Cornelius keg. They are rated for pressure and would have no problems sealing liquid tight.

The setup in the pictures above is gravity only. One problem with a gravity feed is that the build up of clean biodiesel in the bottom tote combined with the drop in crude biodiesel from the top tote will reduce the flow rate.

So we are constantly emptying the bottom tote and filling the top tote to keep the head pressure pretty much the same.

So far this has been the only draw back to our setup so far. We have also had some problems with cold weather performance as we filter cold and heat right before the column, the flow gets restricted so much by the filter that the columns aren't passing enough heated biodiesel to keep the beds warm. You can feel the column and it is quite obvious where the heat starts to dissipate and the column goes cold. When everything is working correctly you can feel slightly warm fluid exiting the column.

This last weekend I changed the columns out for the first time after 2300 gallons through our first 3 columns. The fuel coming out still passed my soap test but they weren't flowing like they used to. Our output the day before we changed out had dropped from 170 gal/day to about 50 gal/day.

I have set these columns aside. I plan to tear them apart and take a look at the resin and see how much I can clean it up with methanol and hopefully rebuild it as a glycerol catcher. I'm also very curious as to how much "crud" is stuck in all of the resin. I'm almost certain that we fouled the beds over that 2000+ gallons.

We have since switched to a 5 micron filter instead of 20 micron. Although it clogs more often it also is catching more of the very fine precipitated soap bits.

With the new columns in place, our flow went right back up to 170 gal / day as long as we keep the bottom tote pumped out. If we just let it go without emptying the bottom we will typically get about 120 gal / day with a gradually decreasing flow throughout the day.


--
Jeff Brandt
Sun Break Biofuels, LLC
http://www.sunbreakbiofuels.com
mailto:jeffb@sunbreakbiofuels.com
 
Registered: March 21, 2006Reply With QuoteReport This Post
  Powered by Social Strata Page 1 2 3 4 5  
 

Sponsors    Biodiesel and SVO Forums Home    Forums  Hop To Forum Categories  Biodiesel  Hop To Forums  Dry Washing Biodiesel    Simple Amberlite Setup

© Maui Green Energy 2000 - 2014