The cornerstones of quality programs for B100 biodiesel are gas chromatography (GC) test methods for residual glycerides and free glycerin (FG), formalized as ASTM D6584 and EN14105.  These are part of the legally enforceable fuel quality standards, D6751 and EN14214, used in commerce throughout much of the world. This article focuses on some results of ongoing research on impurities in biodiesel—their detection and quantification, discoveries made about their solubility, and implications outside the laboratory, which are truly surprising.

Concerns about residual glycerides and glycerin emerged in the late 1980s during the development of commercial processing in Austria. Excessive levels were found to stimulate exhaust emissions containing the  carcinogen acrolein. Deposits in fuel injector nozzles, on the bottom of storage tanks, and accumulated in blocked fuel filters (Ref. 1) were also discovered. This led to guidance on maximum allowable amounts of total and FG residuals in finished fuel.

The first specification for FG was published in Austria by 1991 and was set at 0.03 percent. By 1997 it was reduced to 0.02 percent by a DIN standard, which was later adopted by both the U.S. and EU in 2002-’03. The total free and bound glycerin were set at 0.24 and 0.25 percent, respectively. In 2012 a new specification for monoglycerides was added, targeting saturated monoglycerides indirectly because of their involvement in fuel filter obstruction incidents. Fortunately, technological solutions to reduce or eliminate these impurities are now known and implemented successfully by many biodiesel manufacturers.

Questions about the reliability of the testing protocols have persisted, however. For example, over many years both ASTM and EN round robin testing have indicated a high level of uncertainty for both total glycerin and FG. And, numerous new analytical challenges have emerged. Early method development focused on relatively simple unused soybean and rapeseed oils as biodiesel feedstocks. Today, a host of new and chemically diverse oils are used that can increase the complexity of the GC fingerprint with added ester components, oxidation products and unknown substances. For FG, there are also no reference materials from the National Institute of Standards and Technology or AGQM to mark the performance of analytical methods or commercially available internal and external standards widely used in the industry. Despite these challenges, the many advantages of the GC approach still prevail.

Earlier this year an article was published highlighting some of the research on data scatter and the solubility of FG by our group (Ref. 2). Any of many possible error sources can ruin data. But among our results was the discovery of heterophases of FG in several B100 samples at initial levels originally measured at substantially below 0.02 percent. When agitated, the FG levels sometimes increased by as much as 217 percent, to well over acceptable limits, clearly demonstrating that these dense heterophases can cause severe underestimates. Moreover, simple agitation was itself found in several samples to increase the data scatter. A promising new and readily implemented procedure was developed, consisting of: 1) controlled agitation, 2) increased sample size, and 3) a pretreatment. This combination greatly reduced both imprecision and underestimates due to the heterophases.

The existence of heterophases at room temperatures and levels below 0.02 percent was previously unrecognized. Solubility defines a threshold for the formation and disappearance of these dense heterophases, which are problematic for the variety of reasons mentioned above. The solubility of glycerin, when measured in a purified FAME, was less than one-fifth of the previously published data, and fell even more with the addition of small amounts of moisture. This brought solubilities down to levels well-below 0.02 percent at 23 degrees Celsius. Additional factors affecting solubility include blend composition, temperature and interacting impurities. The high density of FG heterophases contribute to unwanted accumulations during storage and transport, and the high viscosity of dry FG phases can increase the risk for fuel-flow obstructions.

Contamination of diesel fuels by heterophases is a not a new concern, but, in the case of FG, this has been ignored. For example, the test specification for sediment and water, based upon centrifugation of much larger samples, allows for 0.05 percent by volume of such phases, the equivalent of more than 0.06 weight percent glycerin. Other impurities with limited solubility, such as fatty acid soaps, catalyst residues, filter aids, water and more are likely to have similar influences on test procedures, particularly those that dictate the use of small sample aliquots. Adjustments to test methods will help restore confidence in test data. This renewed confidence will lead to a better appreciation of enforceable limits, and ultimately to an improvement in the consistency of fuel quality.

References:
1. Mittelbach, M., Remschmid, C. (2004) Biodiesel—the comprehensive handbook. Martin Mittelbach, Graz, pp 119–120.
2. Heiden, R.W., Schober,S., Mittelbach, M. (2017) Bias and imprecision in the determination of free glycerin in biodiesel: the unexpected role of limitations in solubility. J Am Oil Chem Soc 94: 285-299.

Author: Richard Heiden
Owner, R.W. Heiden Associates LLC
717-299-6860
rwheidenphd@gmail.com