Marine Professional – Bringing forensics to biofuels

First featured on Marine Professional (The official IMarEST magazine and newsletter)

Chemical fingerprinting for feedstock authentication

In recent years, the adoption of green marine fuels has seen significant growth, with FAME based biofuels (fatty acid methyl esters) becoming the preferred choice. The preference for FAME stems largely from their drop-in compatibility with existing bunkering infrastructure and shipboard technologies coupled with ready availability in the market.

Although FAME based biofuels are often termed ‘carbon neutral’, concerns persist regarding the greenhouse gas (GHG) emissions associated with upstream activities, particularly the acquisition of feedstocks. These biofuels can be derived from a variety of feedstocks that are cultivated using different methods, collected in various ways, and produced through diverse processes, leading to substantial differences in their upstream GHG emissions.

For marine use, IMO mandates that qualifying biofuels must demonstrate a well-to-wake GHG emissions reduction of at least 65% compared to fossil marine gasoil. This translates to a ceiling of 33gCO2eq/MJ for qualifying biofuels. While certification bodies, like ISCC and RSB, audit biofuels production for sustainability, and standard testing methods (EN 14103, ASTM D7371 and ASTM D7963) are in place for measuring FAME content, there remain significant limitations. Current audits only certify the production process of biofuels rather than the final products themselves. Furthermore, these audits are administered retrospectively, and existing testing methods are unable to identify feedstock origin. These limitations, coupled with a complex, multi-stakeholder marine fuel supply chain, create a substantial opportunity for fraud. Therefore, the industry requires supplemental physical validation to verify biofuels authenticity.

To address this need, the Global Centre for Maritime Decarbonisation (GCMD) has introduced FAME fingerprinting, a method to identify feedstock origin through its unique markers and trace biofuels authenticity along marine fuel supply chains.

Decoding FAME fingerprints
FAME is commonly produced from feedstocks containing triglycerides through a process called transesterification.

Triglycerides typically consist of fatty acid chains (R1, R2 and R3 in Figure 1), and the combination of these chains varies by feedstock source. Upon transesterification, the fatty acid profile of FAME-based products remains true to their source, providing a unique chemical signature – or ‘fingerprint’– specific to their feedstocks.

FAME fingerprints not only function as intrinsic tracers of feedstock origins, they also provide insights to biofuels properties, such as cold flow performance and oxidation stability, all of which are critical to their efficacy as a marine fuel.

Recognising the potential of ‘FAME fingerprints’, and noting the absence of a standardised analytical method, GCMD partnered with VPS, a fuel-testing laboratory, to enhance the existing EN 14103 standard used to determine ester and linolenic acid methyl ester content to determine FAME fingerprints for both neat FAME and their blends in residual marine fuels.

This method relies on chromatographic separation of individual fatty acid components of FAME-based biofuels according to their chain lengths (six to 24 methylene units) and numbers of double bonds, as shown in Table 1.

Since these species have distinct boiling points and polarities, FAME fingerprinting can be performed using a gaschromatography equipped with flame-ionisation detection. VLSFO contains negligible FAME intrinsically; this analysis thus also allows easy distinction of FAME from VLSFO.

Applying FAME fingerprinting

Through fingerprinting, we are able to differentiate between FAME of virgin vegetable oils, including palm oil, soybean oil, rapeseed oil, coconut oil and jatropha oil, as illustrated in Figure 2. Our analyses indicate this approach is also sensitive to FAME that has been exposed to high heat, allowing us to differentiate between virgin and used cooking oil feedstocks.

This fingerprinting method can not only differentiate the types of FAME, but also allows quantification of its concentration, allowing for the verification of FAME-based biofuels in blends with fossil-based VLSFO. The FAME fingerprint of B24 (FAME+VLSFO) retains all the characteristic markers of B100, although their proportions are lower in intensity, matching the expected blending ratio of 24% w/w. This analytical method can be applied across the entire composition range from B0 to B100 biofuels.

Recent cases have highlighted the prevalence and impact of biofuel fraud, particularly with the mislabelling of biofuels. This situation underscores the urgent need for effective tools to verify the origin of FAME-based biofuels.

In February 2025, Malaysia announced a crackdown on fraud in the used cooking oil export industry after investigations showed that some shipments contained virgin oil. In August 2024, the European Public Prosecutor’s Office recovered €3.1m related to biodiesel fraud, highlighting the transnational nature of these schemes and the detrimental impact on the EU’s environmental goals and budgets. Bosnian company Sistem Ecologica was accused of selling US soy biodiesel mislabelled as next-generation biofuels derived from used cooking oil.

To assess the feasibility of using this FAME fingerprinting approach to detect cases of mislabelling, we examined 16 FAME samples from different suppliers. We compared the identities of these based on their FAME fingerprints against the suppliers’ specifications of their feedstock origin.

The investigation revealed:

• FAME 1–3: Their FAME fingerprints confirmed that they are derived from virgin palm, soy and rapeseed oils, respectively, consistent with how they are labelled by their suppliers.
• FAME 4–13: Despite being labelled as used cooking oil-derived biofuels, FAME 4, 5, 6 and 10 exhibit fingerprints resembling virgin palm oil, and not heat-treated palm oil. FAME 7, 8 and 11–13 exhibit elevated levels of linolenic acid, consistent with palm oil that had been exposed to high heat. FAME 9 comprises fatty acids that are associated with palm and soybean oils, aligning with the common consumption of such vegetable oils in China. Quantitative analysis of the relative intensities of the signals attributable to each oil indicates that FAME 9 is derived from a mixture of 50% palm oil and 50% soybean oil.

• FAME 16: Although labelled as food waste-derived biofuel, this sample’s FAME fingerprint appears identical to virgin palm oil. This discrepancy warrants a closer look to determine its actual feedstock.

Within this limited sample size, discrepancies between their FAME fingerprint and labels were identified in about a third of the samples, suggesting mislabelling may be as pervasive as reported.

Advancing FAME fingerprinting

Acquiring FAME fingerprints using modified EN 14103 takes approximately one hour, comparable to the turnaround time for routine marine fuel-quality testing. Furthermore, the estimated incremental cost of fingerprinting is relatively small, around 0.3% of the total cost of biofuels on a 500-ton batch basis.

While these factors indicate the practicality of FAME fingerprinting, further advancements are required to enhance its robustness and wider applicability. This includes developing higher resolution techniques to detect linolenic acid levels to bolster the distinction between virgin and waste-derived FAME, as well as defining quantitative detection of other by-products, such as dihydroxylpropyl elaidate, i-Propyl14-methyl-pentadecanoate and cis-9-hexadecenal, that are uniquely present in biofuels that have been subjected to high heat.

These advancements will strengthen certification practices, justify green premiums of biofuels with genuine environmental benefits, and ultimately ensure the integrity and sustainability of marine fuel supply chains.

Dr Prapisala Thepsithar is director of projects at GCMD. You can download the report from gcformd.org/ourpublications/?report-id=7471