13th International Colloquium Fuels - September 2021 189 Fast-Ageing method BigOxy for the examination of Fuel Blends Karin Brendel OWI Science for Fuels gGmbH Nina Mebus OWI Science for Fuels gGmbH Summary Though fuel ageing is a well-known phenomenon, there are still open questions concerning chemical reaction pathways and influence of environmental conditions. To examine the process of fuel ageing, especially considering new alternative fuel components, a fast-ageing method has been developed at OWI which has since been used in multiple applications. The method, called BigOxy, is based on the analytical oxidation stability measurement according to DIN 16091. Different fuel blends have been aged and consequently analyzed with various laboratory methods, and thus the course of ageing can be observed. The blends that have been examined include heating oils, diesel fuel, gasoline, heavy fuel oil, various alcohols, and alternative components like hydrogenated vegetable oil, fatty acid methyl esters and synthetic fuels. Results show that the method can successfully compress the natural ageing during storage into a much shorter time scale (down to 72 h). Possible side-effects that could occur during ageing in blends with new components can be found quicker and countermeasures can be enacted. The goal of this test is the prediction of long-term behavior of fuels. 1. Introduction Ageing in fuels is a long-known phenomenon. The first patent for an additive to prevent ageing was filed as early as 1927 [1]. A lot of research has been conducted to assess fuel quality, prevent ageing, and better understand the underlying reactions. This has proven to be difficult due to the great complexity of mineral oils, which can contain hundreds of thousands of different substances [2]. Due to recent improvements in efficiency of both vehicles and heating systems, and development of new technologies like hybrid electric vehicles, storage times for liquid fuels have increased. Additionally, the complexity of fuel is increased by the introduction of new alternative components like hydrogenated vegetable oil (HVO), used cooking oil methyl ester (UCOME) or synthetic fuels (X to liquid, XtL, where X can be various carbon sources). Thus, the need to assess correctly and quickly not only the current fuel quality, but also the future behaviour of fuels and their ageing grows more important. Yet, no quick method exists that allows a prediction of ageing behaviour. The test rig presented in this article is an attempt to develop such a method. 1.1 Ageing processes in fuels Various chemical reactions can happen in fuels during ageing. They range from oxidations to polymerizations and produce a variety of ageing products from aldehydes, ketones and acids to polymers. A good overview over prevalent reactions and previously done research into fuel ageing is presented by Biernat [3] in a review, especially in the subsections by Owczuk [4]. Most of the ageing reactions happen only in the presence of oxygen and are very slow at usual storage conditions. They can be accelerated by influences like temperature, pressure, and catalysts like non-ferrous metal surfaces. Prevention of ageing is usually done using additives, which can stop the peroxide radicals that are essential for most ageing reactions. Fuel ageing changes some of the fuels properties, ranging from viscosity increase due to polymer formation [5], density increase due to evaporation of volatile components [6], over conductivity changes due to oxygenate formation [7], to turbidity due to water increase and deposit formation [6]. All these effects can be measured and quantified with various methods. 1.2 Fast-ageing methods Many methods to accelerate ageing exist. All of them use a combination of the accelerating influences like increasing temperature, pressure, contact with non-ferrous metals, irradiation as well as direct contact with oxygen. Of these methods, a lot are standardized, like various methods that measure oxidation stability in some way Fast-Ageing method BigOxy for the examination of Fuel Blends 190 13th International Colloquium Fuels - September 2021 [8-15]. Most of these methods are designed to measure one comparable parameter that can be used to assess the current stability of a fuel. Some measure the sediments formed under defined conditions, some measure oxygen uptake or detect formed oxygenated products in some way. Only a few methods have been developed expressly to study the ageing process itself, and to conduct measurements over the ageing time, but some of the norm methods have been adapted to this purpose. Järviste et al. have used a similar method to the Rancimat method (DIN 15751) [16], and the method described in this article is derived from the PetroOxy method (DIN 16901). The difference apart from the reactor volume is that the PetroOxy test is conducted with pure oxygen, not with air as in the BigOxy test. 1.3 Sensors The ageing reactions happening in fuels cause changes to physical properties which can be measured with suitable sensors. Table 1 shows an overview of the changes that can happen during ageing, and sensor parameters that correlate to these changes. Table 1: Sensor parameters and their correlation to changes of fuel properties during ageing. Parameter Changes in the sample Water content / relative Humidity Water from condensation, from air, water formation as a reaction product Density Evaporation of volatile components, reactions influencing intermolecular forces Viscosity Evaporation of volatile components, changed interactions of molecules due to polymerization and higher polarity Electrical conductivity Formation of polar components and ions, water Permittivity Increase of polarity through oxidation reactions, changed mobility of molecules due to changes in density Turbidity Formation of sediments and emulsions Infrared Reactions with oxygen, water Due to the conditions in the used ageing method (closed reactor, high temperature), several of these possible sensor parameters cannot be used in the setup. The ones that are left include electrical conductivity and permittivity, as well as viscosity. 2. Methods 2.1 BigOxy method The BigOxy test method is derived from the PetroOxy method. Figure 1 shows a flow diagram of the test rig. Four reactors are filled with 500 mL each of the samples and are then closed airtight. A pressure of air is applied (4.8 bar for middle distillates and heavy fuel oil, 3 bar for gasoline) and the pressure checked for ten minutes. If no changes occur, the test rig is then heated up to the desired test temperature (standard parameter is 105 °C) and the test runs over a predefined length of time (usually between 8 and 72 h). At the end of the run time, the reactors are cooled down and the samples can then be taken out. Further analysis with laboratory methods is possible, as well as deposit formation tests and many more. Figure 1: Flow diagram of the BigOxy test rig. 2.2 Analytics Various analytical methods have been applied to samples that were aged using the method described in section 2.1. The results discussed in this paper have usually been obtained using norm methods, which will not be described here in detail, but referenced in Table 2 so they can be looked up. Table 2: Short description of laboratory methods for fuel analysis and the corresponding norm numbers. Method name and number Short description PetroOxy DIN 16091 Measures oxidation stability using increased pressure and temperature, determines time scale of oxygen uptake Rancimat DIN EN 15751 Measures oxidation stability using increased temperature and oxygen flow, measures the formation of volatile conductive components Fast-Ageing method BigOxy for the examination of Fuel Blends 13th International Colloquium Fuels - September 2021 191 Acid number DIN EN 14104 Measures the total acid amount in a sample, using a titration Total contamination DIN 12662: 1998 Measures the amount of sediment in a sample by filtration Water content DIN EN ISO 12937 Measures the water content using a titration method developed by Karl Fischer Density DIN EN ISO 12185 Density is measured using the vibration of a u-tube. The frequency relates to the weight, and for a known volume, the density of the sample. 2.3 Long term storage To compare the ageing processes in the test rig to ageing in more realistic storage conditions, long term storage tests have been conducted. For this, fuel samples were filled in bottles with 1 L or 2 L content, and a copper spiral was added as a catalyst for ageing reactions. In addition to that, the bottles were stored open to allow contact with oxygen and humidity in the air. The copper spiral is identical to the ones used in the thermal stability test according to DIN 51371. Every one to three months, depending on the project, bottles were taken out and the fuel analysed thoroughly as described in 2.2. 3. Results 3.1 Correlations between test rig and long term storage In a completed previous research project, a correlation between the analysis data from the BigOxy test rig and long term storage data for conventional and some alternative heating oil was found [17]. Various test parameters were tried with temperatures ranging from 90 °C to 140 °C before it was found that the best results could be gained at 105 °C. Higher temperatures accelerated ageing speed so much that a good differentiation became difficult, and the reproducibility suffered due to the long time it takes for the whole sample to reach the desired temperature. Lower temperatures increased the necessary time the sample had to spend in the test rig, so the 105 °C were a good compromise and showed the best correlation to long term storage. To achieve a good comparison between test rig data and long term data, a stability parameter was calculated from multiple analysis results and the fuels were then ordered according to their stability parameters. This method achieved almost the same order for the long term storage as it did for the test rig, meaning the differentiation according to stability is comparable. Another result from the tests was the calculation of time lapse factors. To do this, criteria that defined the fuels as ‘aged’ were empirically determined or taken from the norm (20 min in PetroOxy, 20 h in Rancimat, 200 mg/ kg water content, 0,25 mg kOH/ g acid number, and 24 mg/ kg total contamination). The analytical data over time was then fitted to a linear function to determine the time at which these limits were reached. Since this was done for both the long term storage data as well as the test rig data, the time lapse factor between the two could be determined. Since the linear function is only a very basic approximation of the analytical data over time, a newer project is currently running to improve on this model. The time lapse factors calculated with this method differed between the various tested fuels and ranged from about 60 for 20 % rape seed methyl ester in heating oil to 250 for pure heating oil. A time lapse factor of 250 means that one hour of ageing in the test rig equals 250 h in real time, so a typical run time of 16 hours equals about five and a half months. 3.2 Sensor data in the test rig One critical point of the previously described evaluation was the low number of data points over time. While the number was seven for the long term storage, there were only four data points in case of the test rig data. To solve this problem, a sensor has been integrated in the test rig to measure continuously during the ageing process, and in shorter intervals during the long term storage. The inclusion in the test rig necessitated several constructional changes, the reactor top had to be First tests with the sensor in the BigOxy have yielded continuous data, but due to the very temperature sensitive nature, the full evaluation is still pending. Overall results were that a slight increase of viscosity could be observed during the 64 h runtime in the reactor, and a slight decrease of the conductivity was measured over time. The data from the long term storage shows no certain trends yet. 3.3 Pressure curves of different fuels and alternative components The BigOxy method can be used to age different alternative components, both gasoline and middle distillates as well as heavy fuel oils. The pressure curves differ slightly, depending on the oxygen uptake of the sample. For hydrogenated vegetable fuel (HVO), fatty acid methyl esters (FAME), paraffinic synthetic fuels (Fischer Tropsch products, FT) and conventional fuels, this works very well. It was found however that the use of oxygenated methyl ethers (OME) poses problems, because under the applied conditions, it can decompose into methanol and formaldehyde. In addition to these components being toxic, the resulting pressure increase is also undesirable since the reactors are only tested up to 10 bar. All other components show typical pressure curves, which increase in the beginning due to the temperature increase, and later start decreasing due to oxygen uptake. The maximum pressure achieved during the test run and the time of the oxygen uptake depend on the sample properties. Fast-Ageing method BigOxy for the examination of Fuel Blends 192 13th International Colloquium Fuels - September 2021 In Figure 2, several different pressure curves are shown. The differences between the samples originate both from the composition and the fuel stability. It must be noted that the pressure can only fall by about 20 % since that is the oxygen content of air. Any additional pressure drop stems either from dissolved other gases, or slight leaking of the reactor. The presented fuels include heating oil and diesel, octanol as a pure substance and as a mixture with diesel, and ethanol and butanol in a synthetic methanol-to-gasoline fuel (MtG). Figure 2: Pressure curves over a run time of 64 h for different fuels and blends. 3.4 Conclusion The developed method allows a good approximation of real ageing behaviour, and thus can be used to investigate effects of ageing in various fuel components. Additive reactions in aged fuels can be examined, and deposit formation observed. This is especially interesting regarding the many new candidates for alternative fuels, as they can be tested in a short time, and especially mixing and ageing interactions can be observed. The method will be used in further research projects to examine new fuel components and investigate ageing behaviour of mixtures. Additive testing is also a possible application, especially the effectiveness testing for stability additives can be done using the test rig. References [1] W. W. Evans, Oil composition, etc.: Patentschrift (1927), US1752946A. https: / / patentimages.storage.googleapis.com/ 57/ 08/ 94/ bd0a47b4e05835/ US1752946.pdf. [2] S. K. Panda, J. T. Andersson, and W. Schrader, Angewandte Chemie (International ed. in English) 48, 10 (2009). [3] Storage Stability of Fuels, Ed. by K. Biernat (InTechOpen, 2015). [4] M. Owczuk and K. Kołodziejczyk, in Storage Stability of Fuels, Ed. by K. Biernat (InTechOpen, 2015). [5] A. Agoston, C. Ötsch, and B. Jakoby, Sensors and Actuators A: Physical 121, 2 (2005). [6] K. Brendel and A. Duchowny, DGMK Forschungsbericht, Vol. 778: Untersuchung zur Vermeidung von höhermolekularen Alterungsprodukten in Mitteldestillaten mit alternativen Komponenten unter anwendungstechnischen Randbedingungen (DGMK, Hamburg, 2020). [7] S.-I. Moon, K.-K. Paek, Y.-H. Lee, J.-K. Kim, S.- W. Kim, and B.-K. Ju, Electrochemical and Solid- State Letters 9, 8 (2006). [8] ASTM, Standard Test Method for Oxidation Stability of Distillate Fuel Oil (Accelerated Method) (2014), ASTM D2274. [9] ASTM, Assessing Middle Distillate Fuel Storage Stability by Oxygen Overpressure (2015), ASTM 5304. [10] ASTM, Standard Test Method for Middle Distillate Fuel Storage Stability at 43 °C (2016), ASTM 4625 - 16. [11] ASTM, Oxidation Stability of Aviation Fuels (Potential Residue Method) (2018), ASTM D873 - 12. [12] DIN EN 16091, Flüssige Mineralölerzeugnisse - Mitteldestillat- und Fettsäuremethylesterkraftstoffe und Mischungen - Bestimmung der Oxidationsstabilität mit beschleunigtem Verfahren und kleiner Probenmenge ICS 75.160.20. [13] DIN 15751, Kraftstoffe für Kraftfahrzeuge - Kraftstoff Fettsäuremethylester (FAME) und Mischungen mit Dieselkraftstoff - Bestimmung der Oxidationsstabilität (beschleunigtes Oxidationsverfahren) ICS 75.160.20. [14] DIN 51471, Flüssige Mineralölerzeugnisse - Bestimmung der Lagerstabilität von Heizöl EL ICS 75.160.20. [15] DIN 51371, Flüssige Brennstoffe - Bestimmung der thermischen Stabilität von Heizöl EL ICS 75.160.20. [16] R. T. Järviste, R. T. Muoni, J. H. Soone, H. J. Riisalu, and A. L. Zaidentsal’, Soil Fuel Chem. 42, 2 (2008). [17] W. Koch, S. Eiden, S. Feldhoff, and D. Diarra, DGMK-Forschungsbericht, Vol. 763: Entwicklung einer neuen Prüfmethode zur Bewertung der Stabilität von Heizölen mit biogenen Anteilen: Development of a new stability test method for bio heating oils (DGMK Deutsche Wissenschaftliche Gesellschaft für Erdöl Erdgas und Kohle e.V, Hamburg, 2017) [ger].