Solving a Great Biodiesel Mystery

The best researchers scratch their heads at rare, isolated cases of storage tank filter clogging with biodiesel blends
By Ron Kotrba | September 10, 2012

The introduction of ultra-low sulfur fuel for on-road diesel in 2006-’07 brought unexpected consequences for biodiesel blending. Refiners began hydrotreating diesel fuel to reduce sulfur to 15 parts per million and, in doing so, changed the solubility characteristics of diesel fuel by removing aromatics and other compounds. “Aromatic compounds are good at making relatively polar things soluble,” says Robert McCormick, a principal engineer at National Renewable Energy Labs. “So when you take them out, that could become an issue.”

Sporadic issues with biodiesel blends clogging filters, both vehicle and dispenser, above the cloud point with on-spec B100 arose, recounts Steve Howell, National Biodiesel Board technical director. In general, clogged filters were found to have high levels of sterol gluccosides and saturated monoglycerides. “That prompted the development of the cold soak filtration test,” Howell says, “as a performance test on B100 to pick up any minor biodiesel components that were precipitating out in the field that were not showing up in the cloud point test.” A performance test was desired over measurement of individual components since more than one or two minor components were found on the sampled filters. The CSFT was designed so the cloud point can still be used as a conservative indicator of cold flow operability for blends in the field if the B100 met the CSFT prior to blending.  Two CSFT values were adopted by ASTM in 2008 after being implemented first by industry on a voluntary basis: a year-round 360 seconds maximum filtration time for 300 milliliters of B100; a 200-second maximum CSFT value for B100 destined for use in winter weather (below minus 12 degrees Celsius). After full implementation, reports of vehicle filter clogging with biodiesel blends ceased by all accounts, but rare, isolated instances of storage tank filter clogging at terminals—above the cloud point—continue to haunt the industry.

And this wasn’t just a cold-weather-state problem. While Minnesota has waived its biodiesel mandate for No. 1 ULSD the past few years as the issue is addressed, it’s not clear, according to McCormick, that this is only a problem with biodiesel blends in No. 1 ULSD. “I don’t know you can say one way or another,” he says. “There is a possibility that certain impurities in biodiesel will be less soluble in colder temperatures in No. 1 mainly because No. 1 contains significantly less aromatic compounds than No. 2, in most cases.” Terminals in Washington state, Minnesota and Texas, to name a few, experienced this problem. No fuel can be expected to perform below its cloud point, but seemingly random precipitate formation above cloud point with on-spec fuel using wintertime CSFT values is a mystery, one the industry cannot ignore.

“While the 2008 changes worked for the vast majority of the biodiesel and petrodiesel on the market, it was evident there was some small portion of the market that might need additional controls,” Howell says. “Several refiners and terminals began implementing internal biodiesel purchase specifications controlling monoglycerides at levels in the 0.4 to 0.5 (percent by mass) range, even though there was no official ASTM test method for monoglycerides.”

Howell chairs the ASTM Biodiesel Task Force and formed a cold flow working group to investigate the phenomenon and make recommendations how best to address it. “The efforts furthered the understanding of the phenomenon, with several root causes potentially identified," Howell says, but “with very few problematic samples from the field, it was difficult to pin down specific cause and effect.”

As a result, a useful nonmandatory cold flow appendix was added to the D6751 specification, Howell says, to give users more advice and guidance on when the phenomenon might occur and the potential causes. “With no clear-cut answer, NBB led efforts to involve the leading petrodiesel and biodiesel fuel technical specialists to design a larger set of experiments that could provide a more definitive cause and effect,” Howell says. “That would take some time to design the protocols and secure buy-in, and then secure the samples and funding to execute the work. This work is now in progress, and is estimated to cost over $200,000.”

Before NBB approached NREL to conduct this work late last year, the mono limit implemented by select petroleum companies seemed to be working, Howell says, adding that the cold flow working group had several choices. “After some deliberation,” Howell says, “it was generally agreed … that ASTM should move forward now with the best information available, rather than waiting.” The result was two years of balloting the No. 1-B grade of biodiesel, a voluntary spec developed as an interim measure that features a 0.4 percent by mass monoglyceride limit and a 200-second year-round CSFT. The spec finally passed and at press time awaited publication. Howell says it was developed and passed with the understanding that additional study is ongoing and if more information or better test methods are found, there could be changes to the No. 1-B spec to reflect new information.

 “I think one of the topics of research is to not just understand what these impurities are in biodiesel that come out,” says McCormick, “but try understand exactly what the properties are of diesel fuel—what diesel fuels are exhibiting the same compatibility with biodiesel because it’s a pretty rare problem, there’s a lot of blending in No. 1 ULSD and other low cloud point diesel fuels in the winter months without these problems.” He says NREL looked at biodiesels implicated in this problem, “and they all looked like high-quality material,” he says. “They’re not running up against the top edge of the limits of 6751. That’s the surprise. That is what makes this difficult to solve. We’re looking at the fuel, saying, ‘We don’t see a reason why this stuff should cause problems.’” He says there is not an obvious research path to solving this mystery.

NREL has spent the past 10 months talking to industry and deciding what to do, “rather than going off half-cocked,” McCormick says, “doing a research project and, at the end, not really having solved the problem.” NREL has developed a test matrix that uses a couple of really low cloud point No. 1 diesel fuels as the diesel component and up to eight B100 samples with varying cloud points (feedstocks) and levels of monoglyceride content. “There is some evidence that saturated monoglycerides are the culprit here,” McCormick says, “although they by themselves may not be the culprit. It may be a combination of things, or maybe we will complete this project and find out that’s completely wrong.”

Biodiesels from different process technologies are another variable. For example, Teresa Alleman, a senior researcher with NREL, says different ways to deal with free fatty acids in feedstock may play a role and, therefore, this information is of value to the investigation. “It’s very difficult for me to say what the production process technology has to do with this problem,” Alleman tells Biodiesel Magazine, “but we’re trying to cover our bases.” There are also different techniques for removing the monoglycerides and impurities from crude biodiesel coming out of the reactor. Some of these may have unintended consequences.

The first phase of testing (B5 blends) is slated for completion this autumn. Testing all the samples is one thing, but interpreting massive amounts of data is another. “It takes a while to really absorb that and really understand what’s there,” McCormick says. “What we’d like to do is identify some properties of biodiesel and say, ‘If it’s above or below some level, there’s a problem.’ Or identify a test we can do with a blend that would reveal that the cloud point is not going to predict the operability temperature, so we would need to pick some other temperature.”

All D6751 requirements, plus a number of additional analytical chemistry tests on B100, will be run. For blends, NREL will look at cloud points, cold filter plugging points and the low temperature flow test (LTFT). “There were some indications from our study that with cold soak filtration, samples with high times would also give you an LTFT result in a blend that was significantly above cloud point,” McCormick says, “so I think we have some preliminary indications that even samples that pass cold soak can sometimes give you, in a blend, a high LTFT result relative to the cloud point. Maybe that’s a test that could be used to identify samples with a problem, even if you don’t know what’s causing the problem.” They are also holding samples for days at temperatures slightly above cloud point to measure precipitate formation. “Not that anyone would want to do a four-day hold on their biodiesel to see if it will work,” McCormick says, “but maybe that long of a hold will identify that this type of B100 or this particular B100 is going to have a lot of precipitates, and then we can go back and look at all the other data and see if there’s some more practical test that predicts it.” 

Another consideration is that the error bars on the bound glycerin part of the free and total glycerin test method, D6584, “are kind of big,” McCormick says. “We are working on that,” says Alleman, who chairs the task force working to improve the 6584 method. “It is something that the industry has needed to address, and it is something that we are well aware of and are trying to improve.” Jeff Fetkenhour, owner of Gorge Analytical, says he thinks 6584 is a fine analytical method. “Where we need to see improvement is in how people correctly interpret the chromatography,” Fetkenhour says, adding that rather than using the reference material as an appendix recommendation, ASTM should make it a part of how the method is applied. “Just relying on retention time indices, there can be a number of reasons you have variability from the tabled values and could lead to selection of the wrong peak. This could be overcome by proper application of the method.” 

As for the new No. 1-B grade voluntary biodiesel spec, “It’s not a slam dunk,” McCormick says. “We’re hopeful it will work, but we’re not 100 percent sure. The problem is not that widespread, and it’s kind of hard to get your arms around, so that’s why I think ASTM members including myself are hopeful, but we’re also worried that it’s not going to do it.”

Author: Ron Kotrba
Editor, Biodiesel Magazine
(701) 738-4942
[email protected]

Array ( [REDIRECT_REDIRECT_HTTPS] => on [REDIRECT_REDIRECT_SSL_TLS_SNI] => [REDIRECT_REDIRECT_STATUS] => 200 [REDIRECT_HTTPS] => on [REDIRECT_SSL_TLS_SNI] => [REDIRECT_STATUS] => 200 [HTTPS] => on [SSL_TLS_SNI] => [HTTP_HOST] => [HTTP_CONNECTION] => Keep-Alive [HTTP_ACCEPT_ENCODING] => gzip [HTTP_X_FORWARDED_FOR] => [HTTP_CF_RAY] => 772596cbfa318023-IAD [HTTP_X_FORWARDED_PROTO] => https [HTTP_CF_VISITOR] => {"scheme":"https"} [HTTP_USER_AGENT] => CCBot/2.0 ( [HTTP_ACCEPT] => text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 [HTTP_ACCEPT_LANGUAGE] => en-US,en;q=0.5 [HTTP_CF_CONNECTING_IP] => [HTTP_CF_IPCOUNTRY] => US [HTTP_CDN_LOOP] => cloudflare [PATH] => /sbin:/usr/sbin:/bin:/usr/bin [SERVER_SIGNATURE] =>
Apache/2.2.15 (CentOS) Server at Port 443
[SERVER_SOFTWARE] => Apache/2.2.15 (CentOS) [SERVER_NAME] => [SERVER_ADDR] => [SERVER_PORT] => 443 [REMOTE_ADDR] => [DOCUMENT_ROOT] => /datadrive/websites/ [SERVER_ADMIN] => [email protected] [SCRIPT_FILENAME] => /datadrive/websites/ [REMOTE_PORT] => 23104 [REDIRECT_QUERY_STRING] => url=articles/8691/solving-a-great-biodiesel-mystery [REDIRECT_URL] => /app/webroot/articles/8691/solving-a-great-biodiesel-mystery [GATEWAY_INTERFACE] => CGI/1.1 [SERVER_PROTOCOL] => HTTP/1.1 [REQUEST_METHOD] => GET [QUERY_STRING] => url=articles/8691/solving-a-great-biodiesel-mystery [REQUEST_URI] => /articles/8691/solving-a-great-biodiesel-mystery [SCRIPT_NAME] => /app/webroot/index.php [PHP_SELF] => /app/webroot/index.php [REQUEST_TIME_FLOAT] => 1669832342.554 [REQUEST_TIME] => 1669832342 )