Elsevier

Fuel

Volume 96, June 2012, Pages 284-290
Fuel

Use of radiocarbon analyses for determining levels of biodiesel in fuel blends -- Comparison with ASTM Method D7371 for FAME

https://doi.org/10.1016/j.fuel.2012.01.026Get rights and content

Abstract

A simplified radiocarbon (14C) analytical method was investigated for measuring the biodiesel (FAME) content of diesel fuel blends. That method involves mixing a fuel sample with a fluor and then analyzing it in a liquid scintillation counter (LSC) without further processing. The method is referred to as the direct LSC method because it analyzes the samples directly rather than employing the sample preparation procedures (including sample combustions) typically used in radiocarbon analyses. Biodiesel produced from four different feedstock materials was mixed with petrodiesel to produce fuel blends containing 2–20% FAME, which were subsequently analyzed using the direct LSC method. The fuel blends were also analyzed by several laboratories using FTIR in accordance with ASTM Method D7371, as well as by alternate IR-based procedures being considered for inclusion in that method. Accurate results could be obtained using the direct LSC method if color quench within the samples was sufficiently alleviated. When using counting times of 360 min, results obtained on canola, soy, and white grease biodiesel blends were accurate to within 0.3% (absolute) or better, while results on the coconut biodiesel blends were accurate to within 1.0% (absolute) or better. Results from the direct LSC analyses were comparable to those obtained by Method D7371 for the canola, soy, and white grease biodiesel blends, and were much more accurate than Method D7371 for coconut biodiesel blends. The results reported by different laboratories using Method D7371 varied by 1% (absolute) or more for all the samples, and by 2% (absolute) or more for samples containing 15–20% FAME. Therefore, the practice of reporting results to two decimal places does not appear to be justified in view of the data scatter observed between laboratories.

Highlights

► Radiocarbon analysis of biodiesel fuel blends to determine biofuel percentage. ► Analysis of fuel blends containing biodiesel from four different feedstocks. ► Alternative analytical approach involving direct analysis of fuel in a LSC. ► Direct LSC radiocarbon method works at least as well as IR methods. ► Results are generally accurate to within 0.3% (absolute) or better.

Introduction

Biodiesel is produced by reacting triglycerides with an alcohol in the presence of a catalyst. Methanol is the alcohol that is almost always used because of its relatively low cost. This produces biodiesel consisting of fatty acid methyl esters (FAMEs). Analytical methods are needed to verify the FAME content of diesel fuel blends, and numerous approaches for performing those analyses have been explored. These include fluorescence spectroscopy [1], [2], X-ray spectrometry [3], gas chromatography [4], [5], liquid chromatography [6], [7], nuclear magnetic resonance [8], [9], [10], [11], [12], infrared spectroscopy [12], [13], [14], and radiocarbon (14C) analysis using accelerator mass spectrometry [15]. Some of those methods have been summarized in a paper by Knothe [16]. ASTM Method D7371 [17] and European Standard DIN EN 14078 [18] use infrared (IR) spectroscopy to determine the concentration of FAME in diesel fuel blends by measuring the ester content (specifically the carbonyl bonds) of the fuel, and those are the only methods used commercially for verifying the FAME content of diesel fuel blends. In fact, the ASTM fuel quality specifications require the use of IR-based methods for such determinations [19].

Radiocarbon measurements are commonly used to distinguish between modern carbon and fossil carbon in highly diverse sample types using the benzene synthesis and accelerator mass spectrometry (AMS) procedures described in ASTM Method D6866 [20]. Radiocarbon is unstable and has a half-life of about 5700 years. Therefore, fossil carbon (e.g., from petroleum) has no 14C remaining since it is millions of years old, while modern carbon still has all of its 14C present. Similarly, petrodiesel has no 14C since it contains only fossil carbon, while biodiesel has a relatively high 14C content since it is derived from modern-day feedstock materials. As the biodiesel content of a diesel fuel blend increases, the 14C content of the fuel should increase proportionally.

The use of AMS for analyzing the modern carbon content of liquid fuels has recently been reported in the literature [15], [21], [22], [23], [24], and to a much lesser extent, the use of benzene synthesis for such applications has been reported as well [24]. However, both the benzene synthesis and AMS procedures require considerable skill and training, and the AMS method requires a multi-million dollar accelerator. In this study, the merits of using a greatly simplified 14C analytical method was explored as an alternate approach for determining biodiesel concentrations in diesel fuel blends. Unlike the procedures used in ASTM Method D6866, the alternate approach involves simply mixing a fuel sample with a suitable fluor. By using that approach, all of the typical sample preparation steps are avoided, including sample combustions, collection and cleaning of the CO2 evolved during those combustions, and multi-step catalytic conversions of the CO2 to either benzene (for the benzene synthesis method) or graphite (for AMS). This reduces sample preparation time and cost by at least 90%. The alternate radiocarbon method is referred to as the “direct LSC method” since it analyzes the samples directly (no sample conversions) in a liquid scintillation counter (LSC). That approach has shown promise for determining the bioethanol content of gasoline [21], [22]. However, since the fuels are analyzed directly rather than first converting all of the sample carbon to CO2, it cannot be assumed that the direct LSC works well for all types of liquid fuels in view of the very diverse nature of such fuels. The direct LSC analysis of biodiesel fuel blends has not been previously reported in the literature. Therefore, in this study, the applicability of using the direct LSC method to measure biodiesel concentrations in diesel fuel blends was explored, and results from the direct LSC analysis of a variety of biodiesel blends were compared to results obtained by FTIR using ASTM Method D7371. In addition, results were compared to those obtained using an alternate IR-based procedure being considered for inclusion in Method D7371.

Section snippets

Instrumentation

Radiocarbon analyses were performed using a Packard Tri-Carb 3170 Liquid Scintillation Counter (Perkin Elmer Life Sciences, Boston, MA). The radioactive decay of 14C produces counts in the analyzer, which are reported as counts per minute (CPM) over a specified period of time. However, color and chemical quenching within any given sample decreases the CPM values. Since the degree of quenching can be highly variable, the CPM values must be corrected for sample quench [25] in order to obtain the

B24 reference samples

Counting efficiencies for the B24 reference samples ranged from 87% to 90%. The background-subtracted DPM values ranged from 38.8 to 40.4 for the canola, soy, and white grease B24 samples. The values varied by 4% from the mean for the 3-h counting times and by 1% from the mean for the 6-h counting times. For the coconut B24 blend, the background-subtracted DPM values were 6–8% lower than for the other B24 blends.

When using counting times of 6 h, the calibration factors (“k” values) obtained for

Conclusions

The basic principles behind the direct LSC radiocarbon method for analyzing biodiesel blends appear to be sound, and accurate results can be obtained if color quench is adequately addressed. Although good results were obtained using counting times of 180 min, counting times of 360 min may be desired for improved accuracy. Results from the direct LSC analyses compared well with results obtained from the IR-based methods for the fuel blends made with soy, canola, and white grease biodiesel. For the

Acknowledgements

This work was supported by a Grant from ConocoPhillips Company. We thank Marge Rover with the Center for Sustainable Environmental Technologies at Iowa State University for performing the total carbon analyses on the B100 samples.

References (33)

  • R.E. Pauls

    A review of chromatographic characterization techniques for biodiesel and biodiesel blends

    J Chromatogr Sci

    (2011)
  • T.A. Foglia et al.

    Determination of biodiesel and triacylglycerols in diesel fuel by LC

    Chromatographia

    (2005)
  • V.M. Mello et al.

    Determination of the content of fatty acid methyl esters (FAME) in biodiesel samples obtained by esterification using 1H NMR spectroscopy

    Magn Reson Chem.

    (2008)
  • B. Diehl et al.

    Analysis of biodiesel, diesel and gasoline by NMR spectropscopy – a quick and robust alternative to NIR and GC

    Lipid Technol

    (2007)
  • G. Knothe

    Determining the blend level of mixtures of biodiesel with conventional diesel fuel by fiber-optic near infrared spectroscopy and 1H nuclear magnetic resonance spectroscopy

    J Am Oil Chem Soc

    (2001)
  • C.M. Reddy et al.

    Determination of biodiesel blending percentages using natural abundance radiocarbon analysis: testing the accuracy of retail biodiesel blends

    Environ Sci Technol

    (2008)
  • Cited by (22)

    • Liquid scintillation analysis: Principles and practice

      2020, Handbook of Radioactivity Analysis: Volume 1: Radiation Physics and Detectors
    View all citing articles on Scopus
    View full text