Author Topic: Degumming WVO  (Read 747 times)

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Degumming WVO
« on: February 13, 2017, 03:53:11 PM »
Recent discoveries this year is there are more problems with WVO than removing the high melting point triglycerides, and thinning, and filtering it.  It turns out there is gum in WVO, which I believe it acquires during the frying of plant and animal matter.  Here is a link that I found to an article on the topic.

Small scale degumming techniques for oil entering the biodiesel market.
Quote
In the case studied in this paper, a small scale (1 million gallon per year capacity) soybean, sunflower, and canola crushing facility wanted to improve the quality of their finished product through degumming.  Their product market was mainly to biodiesel producers which helped define an acceptable level of gums and metals removal, and their desired capital investment was relatively small.  After testing a variety of methods with crude soybean oil which could be applied to their scale, the most cost effective method for achieving oils for the biodiesel market was determined to be simple water degumming.  Crude Canola, however, may require more testing to ensure that water degumming achieves low enough levels of metals for the biodiesel industry.

Background
Gums in vegetable oils refer to a variety of compounds which are generally removed during the refining process, including hydratable and non-hydratable phosphatides, lecithin, and other impurities. Gums vary significantly between oil types, and some oils will naturally drop gums out of solution after pressing.  However, significant amounts of gums can remain, specifically the non-hydratable type which are soluble in the oily layer

1.  This material is removed during the chemical refining process which usually includes an acid addition step (phosphoric, oxalic, or citric acids) and an alkali addition step (typically NaOH).

2. For biodiesel producers, gums are a concern for two reasons: metals contamination in the final product (P and Ca specifically) and yield loss.  Previous work by Van Gerpen and Dvorak indicates 50ppm phosphorous is the upper limit before yield loss is noticeable in soybean oil

3.  This work also showed that phosphorous contamination in the finished fuel was not an issue, probably because alkali was used during the transesterification step.  However, little previous work has evaluated degumming methods which achieve less than 50ppm phosphorous for various oilseeds using mechanical pressing. While a multitude of information exists on the chemical refining process for RBD quality oils (edible oils), very little exists describing the minimum degumming required to create a high quality biodiesel feedstock.  Identification of this minimum level of refining required and further defining of “biodiesel quality” crude oil will help biodiesel plants understand the product, and give refiners the knowledge of what the industry specifications are for crude oils. 

Materials and Methods
Three 2.5 gallon samples of soy, Canola, and sunflower oil were taken from the crushing facility from side ports out of the finished product storage tanks.  The sample did settle out visible gums during storage in the lab, and were shaken vigorously before use during testing.  There were 3 sections of the testing: degumming trials, transesterification trials, and final yield trials.  Degumming and transesterification trials effectively screened the methods under various conditions using crude soy only, while the final yield trial was used to obtain more accurate yield data on the best performing method for crude soy, crude Canola and crude sunflower.

Degumming Trials
All glassware (100ml graduated cylinder, 100ml mixing container, and centrifuge tubes) was heated to 85C to help maintain the oil temperature once removed from the water bath.  The oil to be tested was heated in a hot water bath.  Then, the oil was measured to 98 ml in a 100ml graduated cylinder and transferred to the capped 100ml containers.  The appropriate amount of acid or water was added to the container and shaken vigorously for 1 minute.  After this stage, the method for each samples varies according to the descriptions in Table 2.  Crude oil metals levels for phosphorous and calcium are stated in Table 1.

Yield Trials
Once the preferred method was determined to be simple water degumming, a larger sample of crude soy was prepared and run through the entire process to acquire more precise yield data (degumming, transesterification, purification).  Oil was degummed 100ml at a time (the 100ml centrifuge tube limited the size of each degumming batch) until 500ml of degumming soy was obtained.  That 500ml was then transesterified with 2 step addition of methoxide (60% in the first stage for 30 minutes, 40% in the second stage for 1 hour).  24% methanol by volume oil was used, along with .79% KOH by weight oil.  The reaction took place at 130F.  Once reacted, the biodiesel was drained after settling the sample overnight.  The entire sample was then washed 6 times with 200ml of water and dried, then treated with .5% magnesol (magnesium silicate) by weight biodiesel.

Results and Discussion

Degumming Trials
The best performing samples from this trial have minimal loss of triglycerides as approximated by the volume of the gums layer, while also having an acceptable reduction in phosphorous and other metals for biodiesel producers (note – many metals were tested along with calcium and phosphorous. However, these two metals were chosen because they showed greatest variation, and the most difficulty to remove.  The other metals did not present a significant removal problem).  Finally, a simple and inexpensive method is preferred over a more complicated and expensive one.  As evidenced in Table 1, there is variation in gums volume and metals among the samples.  Sample 13 had the smallest amount of gums but was also the most complicated process, requiring phosphoric acid addition and water wash, 20 minute slow stir after the reaction, and cooling and reheating.  Other good performers in terms of gums volume were the water only samples, numbers 1 and 2.  In terms of metals removal, most samples performed adequately for the biodiesel industry.  The KOH which is used in the biodiesel industry as a catalyst also helps remove metals, so some amount of phosphorous and calcium remaining in the oil is acceptable.  All samples removed sufficient amounts of phosphorous after transesterification to pass the ASTM specification.  Calcium was also mostly removed, though some samples, like number 11, fared more poorly in this category. 

The method used for sample 1 was selected as the best performer according to the previously stated criterion of minimal loss, acceptable metals removal, minimal capital cost, and ease of use.  The 500ml minibatch results are described in Table 3, along with 100ml degumming trial results using Canola and sunflower.  Overall yields from the water degumming was not excellent, at only 87.8%.  This may be due to loss of temperature during centrifugation, as the unit was unheated.  However, transesterification yields were quite good, at 96.81% from the degummed oil.  The final product passed the above critical specifications except cold soak filtration.  Cold filtration of the oil may improve this, or further magnesol treatment. The results of the crude Canola and sunflower trials using the best performing method resulted in somewhat surprising results.  The Canola, which typically has lower starting levels of phosphorous, resulted in higher metals levels after degumming at 47ppm phosphorous and 46ppm calcium.  This is probably due to a higher number of non-hydratable phosphatides which cannot be removed by water only.  Sunflower performed on par with the soy sample, at 13ppm phosphorous and 10ppm calcium. Both had a similar sized gums layer, indicating a similar yield to the crude soy.

Conclusion and Future Work
After screening a variety of degumming methods, it was determined that a simple water degumming of soybean oil was sufficient to reduce phosphorous, calcium, and other metals low enough for biodiesel producers to use the degummed oil as a feedstock. In addition, biodiesel producers should see high yields at 96.81% using this water degummed oil.  While the degumming yield was relatively low at 87%, this could have been as a result of non-heated centrifugation. For Canola, though a full transesterification was not performed, the remaining phosphorous levels after water degumming were sufficiently high to require further testing to determine if metals levels are acceptable for biodiesel producers. Future work should use a heated centrifuge for the degumming step to reevaluate the water degumming yields.  In addition, work should be done to determine if the remaining levels of calcium and phosphorous in the degummed Canola are acceptable for ASTM quality biodiesel.  Finally, trying a wider range of water addition levels during the degumming step will help optimize the process.
« Last Edit: February 14, 2017, 10:37:37 AM by Jhanananda »
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Re: Degumming WVO
« Reply #1 on: February 25, 2017, 07:11:29 PM »
I have been pondering this new information to me that vegetable oil has gum in it.  If all vegetable oil is properly degummed before it is sold as a commercial product, then where is the gum that I am observing in my fuel coming from?  It must be that as vegetable oil is used to cook with, it acquires gum from what it is cooking.  Thus, WVO might just have a lot more gum in it, than the original vegetable oil.  Thus, the need to degum WVO in the process of turning it into fuel. 

In fact, the presence of gum in WVO might just explain why heated two tank SVO systems, as well as engines that run blends, fail prematurely due to, at least in part, the presence of gum in the WVO, which is not being removed in either method of fueling a diesel engine. Whereas, the reason why people who burn biodiesel on their diesel engines do not have their engines fail prematurely, due to gum accumulation.

Examining the de-gumming process, as described above, it suggests that the biodiesel process removes gum in WVO, and does not need a degumming process before hand.

Therefore, if the above is true, and I believe it is, then both blenders, and heated two tank SVO people, should not burn unmodified triglycerides.  Instead we both should be making biodiesel out of their WVO.  I plan to do that from now on.
« Last Edit: February 27, 2017, 10:00:34 AM by Jhanananda »
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Re: Degumming WVO
« Reply #2 on: March 28, 2017, 09:40:34 AM »
This last weekend I tried several methods of water de-gumming WVO.  I was not really happy with the results.  The water came out light yellow, and there was a dark, amber, sludge floating on top of the water.  It might be gum.  If so, then it was very successful.  But, it looks like just emulsified oil.  So, I left it to settle out for a few weeks before I look at my samples again.
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)

philbri

  • vetted member
  • Newbie
  • *
  • Posts: 19
Re: Degumming WVO
« Reply #3 on: July 06, 2017, 04:46:49 AM »
Hi Jeffrey,

Hope you're doing well :)

I'm still doing the blend of 80% WVO 20% Gasoline but had some repairable engine fuel system damage in the past years. So if I understood well you are now doing biodiesel , no more blending ? Thanks

Smithy

  • vetted member
  • Newbie
  • *
  • Posts: 26
Re: Degumming WVO
« Reply #4 on: July 06, 2017, 11:29:13 AM »
jhanananda, Hi -so do I hear you correctly that you have finished blending and are going to take the biodiesel route to avoid the gum problem.

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Re: Degumming WVO
« Reply #5 on: July 06, 2017, 11:42:06 AM »
Hello, philbri and Smithy.  I have not yet started making biodiesel, and I am still experimenting with blending to reduce gum.  I have had success in doing so, and plan to post my results soon.  At present I can no longer recommend the 80/20 blend for a naturally aspirated diesel engine.  I have been experimenting with a 60-20-20 blend WVO/diesel/gasoline blend.  It seems to have solved the gum and coking issue.
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Re: Degumming WVO
« Reply #6 on: September 13, 2017, 09:17:52 AM »
This last weekend I tried several methods of water de-gumming WVO.  I was not really happy with the results.  The water came out light yellow, and there was a dark, amber, sludge floating on top of the water.  It might be gum.  If so, then it was very successful.  But, it looks like just emulsified oil.  So, I left it to settle out for a few weeks before I look at my samples again.

I have done a great number of de-gumming experiments in the last 3 years.  It seems like it will take a number of methods to reduce gum.

1] I now put about 4 gallons (16L) of WVO into a 5 gallon (20L) steel bucket, plus about 1 quart (1L) of tap water; then I put a band heater on the bucket, and the bucket on top of a hot plate, and set both heaters to 250F (120c).  I leave it over night.  The next morning I find the bucket is simmering nicely, so I turn off the heat, and pour out the translucent WVO into another steel bucket.  At the bottom I find about 1qt (1L) of yellowish water, and above that, and below the WVO is a dark red liquid.  I believe it is gum.

2]  It appears that gum precipitates out of solution with WVO proportional to how much petroleum distillate is added.  I find when the added petroleum distillate reaches 50%, then most of the gum has been removed.  However, it takes about 2 weeks for the gum to fully settle out at the bottom.

3] With the above practice now in place I have returned to blending WMO with an equal part of WVO plus diesel fuel at 20%.  Then I leave this for about 2 weeks to a month, and find black sludge at the bottom, and translucent fuel blend floating on top.  When I am ready to make fuel, then I pour this blend into gasoline at 5%, then I leave it to settle in a pressurized fuel settling tank for at least 2 days to 1 week.  At the end of the settling period I drain off any liquid that is dark.  At this time I find about 1 qt (1L) of dark liquid to be drained off.
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)

Jhanananda

  • Administrator
  • Hero Member
  • *****
  • Posts: 2083
  • I blend waste oils with solvents
    • Beyond Biodiesel
Re: Degumming WVO
« Reply #7 on: September 14, 2017, 09:56:39 AM »
Another thing that I have implemented is a heat sink on the oil filter.  I simply rapped a piece of aluminum flashing around the filter, which was twice as long as the filter.  I changed it after a month and found the filter heavy with sludge, so I call it a success, and plan to keep changing the filter as needed.

Another experiment for de-gumming and removing high melting point triglycerides was to settle my blends for 2 weeks in a galvanized bucket.  It has turned out to also be very successful at removing high melting point triglycerides and gum.
« Last Edit: September 14, 2017, 09:59:05 AM by Jhanananda »
I have run various blends of waste oils and unleaded gasoline since Feb, 2007 in a 1983 Chevy G-20 van with a 6.2L diesel V-8 engine, with a Stanadyne Rotary DB2 IP. I have started the engine with no difficulty on an 80/20 (WVO/gas) blend down to 3F (-16c).  I now run 60/20/20 (WVO/diesel/gas)