Source: Biodiesel Magazine
U.S. biodiesel fuel quality has been on an ever-increasing trajectory upward since the darkest days of its past in 2006, when a B100 quality survey performed by National Renewable Energy Laboratory showed nearly 60 percent of samples failed specification. Not only has fuel quality improved dramatically since then, but the stringency of ASTM D6751 has also increased significantly.
Teresa Alleman, a senior fuel chemist with NREL, made her first appearance in several years on stage at the National Biodiesel Conference & Expo in Tampa this January to discuss the biodiesel industry’s latest fuel quality report—an expansive compilation of B100 test data from BQ-9000-accredited producers.
BQ-9000 is the U.S. biodiesel industry’s quality assurance program and, according to Scott Fenwick, technical director for the National Biodiesel Board, more than 50 producers in the U.S. and Canada are BQ-9000-certified. “Those producers in the program represent approximately 93 percent of the gallons that are represented by the NBB,” Fenwick said.
Diesel fuel quality is becoming increasingly important as engine technologies rapidly advance to keep up with tougher emissions and fuel economy regulations. “Engine technologies are changing faster and faster, with tighter and tighter tolerances,” Fenwick said. “Fuels need to be cleaner and cleaner. We half-jokingly say that, in the very near future, your fuel will be cleaner than the drinking water that you’ve got in front of you. Understanding where we are [with respect to biodiesel fuel quality] gives OEMs a baseline on where they can begin to adjust their engines for future designs and provide information back to regulators on where we stand.”
What prompted this latest fuel quality assessment, according to Richard Nelson of Enersol Resources, was a fuel quality study conducted in 2017 by Germany’s biodiesel quality management association AGQM, the results of which were released in 2018. The AGQM investigation collected 53 samples from 15 producers and analyzed 1,060 parameters, concentrating on metals. The results were very positive. All samples met or exceeded the DIN EN 14214 biodiesel standard and even stricter AGQM requirements. Only 14 limits were violated, but all violations were within the precision of the method. Since then, AGQM has performed two subsequent surveys, the latest of which involved three sampling campaigns in 2019. The 2019 results, just released this spring, showed significant improvement over 2018. In 59 samples taken, which resulted in 1,062 analysis data, only seven limit violations could be determined.
Nelson said tractor engine manufacturer Deutz wanted to know what the state of biodiesel fuel quality was in order to warranty biodiesel in engines. The OEM warrantied B100 in a specific model for Germany, but Nelson said it only warrantied B20 in the U.S. because the company did not have recent information on where U.S.-produced biodiesel stood in terms of quality. “So, Steve Howell and Scott Fenwick came up with the idea to survey BQ-9000 companies,” Nelson said. “We asked for all 14 parameters from 2017 and 2018, and determined the average, minimum, maximum, and other metrics with that data.”
All BQ-9000 producers were surveyed for their monthly test data on all 14 critical property parameters over two years. This amounted to more than 11,000 property parameters analyzed over the study period: 5,391 in 2017, and 5,766 in 2018 from BQ-9000 producers in both the U.S. and Canada. The results, which are provided below in detail, were very favorable, demonstrating that biodiesel fuel quality—at least high up the chain at the point of production—has never been higher, and the voluntary BQ-9000 quality assurance program works as intended.
Diving into the Data
Once the analysis data from the BQ-9000 producers’ fuel samples were received, Nelson scrubbed it of company and location information, and anonymized and randomized the data. When Alleman at NREL received the data, she had no idea from which producers or location the data came, or how results from any given parameter connected to results from the other parameters. In other words, there was no way to tell, for instance, whether a sample that failed oxidation stability minimums was the same sample that may have had an abnormally high acid value. The data were completely randomized and anonymized.
Metals content is perhaps one of the most important parameters for OEMs, considering future aftertreatment designs may use close-coupled SCR systems whose catalysts are highly sensitive to metals contamination. Unlike today’s systems that feature a diesel oxidation catalyst followed by a diesel particulate filter and then an SCR catalyst, more stringent NOx reductions may require future aftertreatment designs to expose sensitive SCR catalysts directly to exhaust manifolds at high temperatures and putting the catalysts directly in the line of fire of any metals exiting the combustion chamber.
The metals limit in ASTM D6751, the biodiesel fuel quality standard, for combined sodium and potassium is 5 parts per million (ppm). Alleman said results from the recent analysis of BQ-9000 producer data show averages well below the spec limit. For 2017 and 2018, averages were 0.68 ppm and 0.69 ppm, respectively. For 2017, 416 samples were tested and showed a minimum value of zero, a maximum value of 6 ppm—the only, single sample out of spec for metals—a median of 0.30 ppm, with 95 percent of all samples coming in at or below 2.23 ppm. For 2018, 431 samples were tested and showed a minimum of zero, a maximum of 4.5 ppm, a median of 0.42 ppm and the 95th percentile at 2.20 ppm or below.
The metals limit in D6751 for combined calcium and magnesium is also 5 ppm. For 2017 and 2018, averages were 0.30 and 0.31, respectively. None of the 408 samples from 2017 nor the 426 samples from 2018 were off spec for calcium and magnesium. The minimum value for both years was zero. The maximum for 2017 was 4 ppm, and the max for 2018 was 4.4 ppm. The median for 2017 was 0.06 ppm and 0.08 ppm for 2018. The 95th percentile for both years was 2 ppm or below.
Phosphorous levels were all very close to zero. “Only a handful of data points were any larger than zero,” Alleman said. The average for 2017 was 0.41 ppm and 0.37 ppm for 2018. The 95th percentile for 2017 and 2018 was 1 ppm and 0.98 ppm, respectively.
On oxidation stability, the latest quality results show the biodiesel industry is doing very well. This is another critical parameter for OEMs. Only one sample failed the three-hour minimum in 2017 out of 421 samples. The failed sample time was 2.7 hours, the minimum value for that year. The max time in the 2017 lot was 40.6 hours, and the average for the 2017 samples was 9.8 hours with a median of 9 hours. In 2018, all 452 samples passed oxidation stability, with a minimum time of 4.4 hours, a maximum of 27.5 hours, an average of 9.5 hours and a median of 8.7 hours. The 95th percentile for oxidation stability, since it relies on a minimum time vs. a maximum limit, is not applicable for this parameter.
“Acid number—this is another key parameter that helps talk about the health of the fuel,” Alleman said. None of the 421 samples in 2017 was off spec for acid number, the max value of which is 0.50. The 2017 average was 0.26 and median was 0.27 with a minimum of 0.05, a max of 0.47 and the 95th percentile at 0.41. For 2018, one sample of 453 was off spec at 0.51. The average as well as the median was 0.27 with a minimum number of 0.02, a max, again, of 0.51, with the 95th percentile coming in at 0.43.
There is no specification for cloud point in D6751, but Alleman shared results anyway. In 2017, the minimum was minus 4.1 degrees Celsius and maximum was 15 C, with an average of 0.8 C and the 95th percentile at 9 C. For 2018, the minimum cloud point was 5.6 C, the maximum was 17 C, with an average of 1 C and the 95th percentile at 8.8 C.
On the cold soak filtration test (CSFT), Alleman said, “I’ve spent a lot of time thinking about this test, running this test, and teaching people how to run this test. Some of our early work showed bimodal distribution—either they passed or failed, there was no in between. I’m standing up here telling you today, everything is over here on the passing side. The industry reacted to those previous quality hiccups. Believe me, I remember standing up here saying, ‘We’re having some issues.’ It is very pleasant to be able to say something else today.”
In 2017 and 2018, all 409 and 441 samples, respectively, passed the CSFT. In 2017, the minimum time was 54 seconds with a maximum of 324 seconds, an average of 108 seconds, a median of 105 seconds, and the 95th percentile at 144 seconds or below. For 2018, the minimum was 66 seconds, the max at 338 seconds, an average of 104 seconds, a median of 100 seconds, and the 95th percentile at 144 seconds. The CSFT requires filtration in 360 seconds or less, or 200 seconds or less for No.1-B grade biodiesel.
Total glycerin is “another key quality parameter and something we found early on, back in the day, producers were having a hard time meeting,” Alleman said. “That’s not a problem anymore.”
Only one sample of 421 in 2017 surpassed the limit of 0.24 weight percent. In 2017, the minimum was 0.003 weight percent and the maximum was the off-spec sample of 0.256 weight percent. The average for the 2017 samples was 0.093 weight percent with a median of 0.103 weight percent. The 95th percentile was 0.151 weight percent. In 2018, out of 453 samples, the minimum was 0.005 weight percent, the max was 0.225 weight percent with an average of 0.093 weight percent, a median of 0.099 weight percent and the 95th percentile at 0.164 weight percent. “We’re seeing the product is readily meeting the specification,” Alleman concluded.
“For monoglycerides, this is not a mandatory parameter in the U.S.,” Alleman said, “and I can’t tell you if any of these were a No.1-B grade or not.” Out of 421 samples in 2017, the minimum monoglyceride content was 0.001 weight percent, the max was 0.8 weight percent with an average of 0.271 weight percent, a median of 0.299 percent and a 95th percentile of 0.436. For 2018, of the 453 samples, the minimum monoglyceride content was zero, the maximum was 0.642 weight percent with an average of 0.269 weight percent, a median of 0.288 weight percent and the 95th percentile of 0.474 weight percent.
On free glycerin, out of 421 samples in 2017 and 453 in 2018, only less than a handful of samples were off spec, with one 2017 sample well above the 0.02 weight percent max at 0.083 weight percent. The minimum for both years was zero. The max for 2017 was 0.083 percent with an average of 0.007 weight percent, the median at 0.007 weight percent and the 95th percentile at 0.014 weight percent. For 2018, the max free glycerin content was over the limit at 0.030 weight percent with an average of 0.006 weight percent, a median of 0.005 weight percent and the 95th percentile at 0.014 weight percent.
Moving on to sulfur, Alleman said, “EPA cares a whole bunch about sulfur.” In 2017, of 420 samples, the minimum sulfur content was zero and the max was 17.3 ppm—over the spec limit of 15 ppm. The average was 3.4 ppm with a median of 1.8 ppm and the 95th percentile at 10.8 ppm. For 2018, of 448 samples, the minimum was zero and the maximum was below the 15-ppm limit at 14 ppm. The average sulfur content in 2018 was 3.4 ppm, the median was 2 ppm and the 95th percentile was 10.9 ppm.
“Flashpoint—fire marshals care a lot about this data,” Alleman said. “[What we see is this is] trending very high, well in excess of the spec for both years.” The minimum flashpoint for biodiesel under D6751 is 93 C. In the 2017 data, out of 385 samples, all passed and met the minimum threshold. The minimum was 95 C, the max 211 C with an average of 155 C and a median of 160 C. One of 413 samples in 2018 came in at 89 C, which is below the minimum threshold, with the max at 193 C, an average of 153 C and a median of 161 C.
“So, we really have collected a huge amount of data to see where the U.S. and Canadian biodiesel industry is,” Alleman said. “The [metals data are] easily orders of magnitude under the spec limit. We are so far below these limits on the metals—that’s a very important story. I hope my OEM friends in the audience are taking good notes.”
“This data is a snapshot in time, at the point of production,” Fenwick said. “Quality does not get better as it goes along. There is a myriad of ways in the distribution system to [pick up contaminants].”
One might argue that, given all the data used was from BQ-9000 producers, this analysis tested the best of the best, which is true. Not all biodiesel in the U.S. market is produced by NBB member-producers, and not all NBB member-producers are BQ-9000-certified. However, NBB member-producers and BQ-9000-certified producers represent a vast majority of the biodiesel fuel sold in the U.S. market.
“Never before has NREL been able to provide this level of fuel quality information,” Alleman said. “By partnering with NBB’s National Biodiesel Accreditation Commission, BQ-9000, we now have a simpler, more efficient way to collect, analyze and determine the quality of biodiesel. These reports will act as a reoccurring insight into biodiesel fuel quality each year.” As Alleman indicated, this will be a yearly assessment moving forward.
“These reports show what we’ve been saying for years,” Fenwick said. “Biodiesel fuel is of the absolute highest quality, and these reports prove it. Having this data every year will go a long way in demonstrating to OEMs and regulatory agencies that biodiesel should be the fuel of choice for any diesel engine.”
Author: Ron Kotrba
Editor in Chief, Biodiesel Magazine
PHOTO: JENNA HIGGINS ROSE, IOWA BIODIESEL BOARD