Jon Van Gerpen: looking at the biodiesel process in a new way
“There are a lot of things we don’t know about the biodiesel process.” This is not a comment expected from one of the most knowledgeable researchers in the biodiesel industry, Prof. Jon Van Gerpen of the University of Idaho. But Van Gerpen, who gave the closing presentation at the Collective Biofuels Conference, said biodiesel production is a “deceptively simple process.”
He said the industry can learn from stepping back and looking at the production process in a new, microscopic way, and brought up questions that perhaps should have been asked back when biodiesel was in its infancy. These questions include, “What affects the reaction, and why?” and “Can we guide the reaction in some way?”
There are limitations to methanol’s solubility in oil. It becomes more soluble at higher temperatures. Most processes run at 60 to 65 degrees Celsius, and methanol solubility is about 10 percent, but excess methanol is used (double the solubility or 20 percent) in order to facilitate the reaction. “When you’re talking solubility, you’re talking polarity,” Van Gerpen said, adding that if you take oil, methanol and catalyst, mix for a minute and then separate it, all the catalyst will be with the methanol.
He said when you look at the methanol droplets from the excess methanol, there’s only about 1 percent oil in those methanol droplets. “The reaction only occurs on the surface of those droplets,” Van Gerpen explained. “The diffusion process moves the oil to reaction zone, and then moves the methyl esters away.” He said the biodiesel transesterification process is conventionally referred to as a mass transfer limited reaction, but really it could be chemical reaction limited. “We usually don’t know which,” he said. “There’s a balance between the relative significance of diffusion and reaction.”
Van Gerpen said he is working to develop mathematical equations to address this. “Diffusion we can do,” he said, “but the piece we’re missing is we don’t know what the chemical reaction rate is in the droplets.”
He said some calculations have been done by industry to address this, but so far only for specific reactors, and those calculations depend on specific sets of mixing parameters. There are also scale-up issues with chemical reaction rates.
Van Gerpen spoke of David Boocock from the University of Toronto, who developed the cosolvent process used by Biox Corp. Using a cosolvent changes a two-phase reaction into one phase. With use of a cosolvent, there is no longer a mass transfer limitation so, as Van Gerpen said, “The measured rates should be the true chemical reaction rate.”
“An important conclusion is that we cannot call the biodiesel process a mass transfer limited reaction anymore,” he said.
Also, Van Gerpen said there is no good data for the solubility of monoglycerides and diglycerides in methanol and oil.
Plus, there is a very powerful force working at microscopic levels known as surface tension, or interfacial tension for liquid/liquid materials, during transesterification. This is why droplets are round. He said this is why to achieve better reactions some use high-shear processing equipment to skew those droplets to facilitate the reaction.
And then there are surfactant effects. Surfactants are molecules that have one polar and one nonpolar end. Soap, for instance, is a surfactant, which is why when mixed with water it helps remove greasy materials during hand washing.
Van Gerpen also said the effects of stoichiometry on monoglycerides are of interest. Modeling can tell us where monos and diglycerides reside in the droplets. Also, it can help understand how methanol droplets get smaller as the reaction proceeds, and how glycerin droplets behave in the methyl esters. “Can we optimize the timing of agitation and the settling process for the fastest overall process?” he asked.
Ultimately, Van Gerpen said the purpose of this presentation was to get people to think more fundamentally about the biodiesel production process. “This knowledge has to help design a better process,” he said, “but I’m not sure how just yet.”





5 Responses
John (Jake) Ferris
2012-08-22
1I wonder if Jon has any suggestions on recent research on costs of conversion of soybean oil, other vegetable oil and feedstock with higher FFA. I have been using his analysis with M. Canakci published in a paper entitled, "A pilot plant to produce biodiesel from high free fatty acid feedstocks." In my monthly article for Jacobsen Publishing, I have adjusted his estimates from the 2001 paper by changes in the prices of the inputs. However, with new technology, I need to check these cost figures.
carlos muñoz
2012-08-23
2As Jon is suggesting,the best reaction can be reached with high temperature and "high-share processing...to skew those droplets" Our BD JET protocol uses sodium methoxide, works at 95°C, and has a very strong mixing by forcing the liquids through an on-line mixer. *Thus, with crude (not degummed) soyabean oil, we get over 99% conversion in one single step
Larry Sullivan
2012-09-12
3The Kreido technology uses extremely high shear in a laminar flow to regime to eliminate surface tensions and complete the reaction in less than one second. See US Patent 7,678,340. The general field is covered in this RSC article: http://pubs.rsc.org/en/Content/ArticleLanding/2000/CP/b000218f
Kirk Cobb
2012-09-12
4Comparing the kinetics of the typical transesterification reaction with methoxide catalyst, to other types of oleochemical reactions, the biodiesel TE reaction is really quite fast; if the reaction is allowed to operate at atm. pressure with methanol hot enough to reflux, the equilibrium temperature is ~ 170 F (~ 77 C), well above the normal boiling point (~64 C) of methanol, because in a steady state continuous system the presence of methanol is at a reduced mole fraction in the liquid, so it does not express its full vapor pressure; that extra 10 to 15 deg C nearly doubles the rate of reaction; also the continuous system will have steady state concentrations of mono and di-glycerides which act as emulsifying agents; I am surprized that people are still focusing on the biodiesel TE kinetics; the larger plant design issues really should focus on the physical separations - decanting, methanol stripping, and methyl ester distillation. I am amazed that Biox is bothering to use a nasty co-solvent like tetrahydrofuran; their recent announcement to build a new plant in the New York harbor, a 100-million Liter/year plant (~ 26-mgpy) at a cost of $60-million is an astronomical cost for a modest sized plant. The co-solvent seems totally unnecessary and its routine recovery results in a very expensive plant.
Hippie
2012-09-21
5Yours is a clever way of tnhiknig about it.
Leave a Reply
Biodiesel Magazine encourages civil conversation and debate. However, comments containing personal attacks, profanity, business solicitations or other advertising will be deleted.
Comments are closed