Artificial neural network approaches on composition-property relationships of jet fuels based on GC-MS

The relationships of composition-properties of 80 jet fuels concerning chemical compositions and several specification properties including density, flashpoint, freezing point, aniline point and net heat of combustion were studied. The chemical compositions of the jet fuels were determined by GC-MS, and grouped into eight classes of hydrocarbon compounds, including n-paraffins, isoparaffins, monocyclopraffins, dicyclopraffins, alkylbenzens, naphthalenes, tetralins, hydroaromatics. Several quantitative composition-property relationships were developed with three artificial neural network (ANN) approaches, including single-layer feedforward neural network (SLFNN), multiple layer feedforward neural network (MLFNN) and general regressed neural network (GRNN). It was found that SLFNNs are adequate to predict density, freezing point and net heat of combustion, while MLFNNs produce better results as far as the flash point and aniline point prediction are concerned. Comparisons with the multiple linear regression (MLR) correlations reported and the standard ASTM methods showed that ANN approaches of composition-property relationships are significant improvement on MLR correlations, and are comparable to the standard ASTM methods.     

Artificial neural network model to predict cold filter plugging point of blended diesel fuels

Diesel fuel blending is an indispensable process in the diesel fuel producing process. It will benefit greatly the refineries to increase their profits if a mathematic model is developed to accurately estimate CFPP instead of substantial experiments. In this article, a back propagation artificial neural network model is established to predict CFPP of the blended diesel fuels, using input parameters of kinematics viscosity, density, refractivity intercept, CFPP and weight percentages of constituent diesel fuels. This model can give satisfactory predicting results for unknown diesel fuel samples either without PPD or with PPD and has been tested by practical industrial applications of produce blended diesel fuels. The mean predicting errors for the unknown samples without PPD are about 1.3 °C and about 2.5 °C for unknown samples with PPD.     

The influence of n-butanol/diesel fuel blends utilization on a small diesel engine performance and emissions

Nitrogen oxides and smoke emissions are the most significant emissions for the diesel engines. Especially, fuels containing high-level oxygen content can have potential to reduce smoke emissions significantly. The aim of the present study is to evaluate the influence of n-butanol/diesel fuel blends (as an oxygenation additive for the diesel fuel) on engine performance and exhaust emissions in a small diesel engine. For this aim five-test fuels, B5 (contains 5% n-butanol and 95% diesel fuel in volume basis), B10, B15, B20 and neat diesel fuel, were prepared to test in a diesel engine. Tests were performed in a single cylinder, four stroke, unmodified, and naturally aspirated DI high speed diesel engine at constant engine speed (2600 rpm) and four different engine loads by using five-test fuels. The experimental test results showed that smoke opacity, nitrogen oxides, and carbon monoxide emissions reduced while hydrocarbon emissions increased with the increasing n-butanol content in the fuel blends. In addition, there is an increase in the brake specific fuel consumption and in the brake thermal efficiency with increasing n-butanol content in fuel blends. Also, exhaust gas temperature decreased with increasing n-butanol content in the fuel blends.     

Biodiesel engine performance and emissions in low temperature combustion

Engine performance and emission comparisons were made between the use of soy, Canola and yellow grease derived B100 biodiesel fuels and an ultra-low sulphur diesel fuel in the high load engine operating conditions. Compared to the diesel fuel engine-out emissions of nitrogen oxides (NO_x), a high-cetane number (CN) biodiesel fuel produced comparable NO_x while the biodiesel with a CN similar to the diesel fuel produced relatively higher NO_x at a fixed start of injection. The soot, carbon monoxide and un-burnt hydrocarbon emissions were generally lower for the biodiesel-fuelled engine. Exhaust gas recirculation (EGR) was then extensively applied to initiate low temperature combustion (LTC) mode at medium and low load conditions. An intake throttling valve was implemented to increase the differential pressure between the intake and exhaust in order to increase and enhance the EGR. Simultaneous reduction of NO_x and soot was achieved when the ignition delay was prolonged by more than 50% from the case with 0% EGR at low load conditions. Furthermore, a preliminary ignition delay correlation under the influence of EGR at steady-state conditions was developed. The correlation considered the fuel CN and oxygen concentrations in the intake air and fuel. The research intends to achieve simultaneous reductions of NO_x and soot emissions in modern production diesel engines when biodiesel is applied.     

Computer simulation of hydrogen-diesel dual fuel exhaust gas emissions with experimental verification

The drawback of lean operation with hydrocarbon fuels is a reduced power output. Lean operation of hydrocarbon engines has additional drawbacks. Lean mixtures are hard to ignite, despite the mixture being above the low fire (point) limit of the fuel. This results in misfire, which increases un-burned hydrocarbon emissions, reduces performance and wastes fuel. Hydrogen can be used in conjunction with compact liquid fuels such as gasoline; alcohol or diesel provided each is stored separately.Mixing hydrogen with other hydrocarbon fuels reduces all of these drawbacks. Hydrogen’s low ignition energy limit and high burning speed makes the hydrogen/hydrocarbon mixture easier to ignite, reducing misfire and thereby improving emissions, performance and fuel economy. Regarding power output, hydrogen augments the mixture’s energy density at lean mixtures by increasing the hydrogen-to-carbon ratio, and thereby improves torque at wide-open throttle conditions.This paper involves the simulation program for determining the mole fraction of each of the exhaust species when the hydrogen is burnt along with diesel and the results are presented. The proportion of hydrogen in the hydrogen-diesel blend affecting the mole fraction of the exhaust species is also simulated. Experimental investigations were carried out, in hydrogen-diesel dual fuel mode, which showed a good agreement between the predicted and experimental results. The program code developed is valid for any combination of dual fuels.     

Cetane Number Prediction of Biodiesel from the Composition of the Fatty Acid Methyl Esters

In the present work, the measured cetane numbers (CN) of pure fatty acid methyl esters (FAME), as well as the FAME compositions and the reported CN of 59 kinds of biodiesels collected from literature were used to develop a simple model involving as more FAME component as possible for predicting CN of biodiesel from its FAME composition. Two different regression equations correlating the CN of pure FAME with the carbon number of fatty acid chain were obtained by regression analysis, which shows that the dependence of the CN on the carbon number varies with the unsaturated degree of fatty acid chain. The 59 biodiesels were divided into two categories and used, respectively to develop and test a multiple linear regression model (MLRM) correlating the CN of biodiesel with its FAME composition. A simple and convenient regression equation with a high accuracy and a good reproducibility (average absolute error of 0.49 CN for testing set and 1.52 CN for all data) were developed, showing excellent correlation (R ²: 0.9904 for testing set). The model developed in the present work can be used conveniently to give a satisfactory predicted CN of biodiesel from the FAME composition.     

Molecular reconstruction of complex hydrocarbon mixtures: An application of principal component analysis

Three methods for reconstruction of the detailed molecular composition of complex hydrocarbon mixtures, based on their global properties, are compared: a method based on the Shannon entropy criterion, an artificial neural network and a multiple linear regression model. In spite of the broad range of naphthas included in the training set, the application range of the last two methods proved to be limited. Principal component analysis allowed to identify their three‐dimensional ellipsoidal application range. In this subspace, the artificial neural network is more accurate than the multiple linear regression model and the Shannon entropy method. However, outside its application range, the performance of the neural network, as well as the regression model, decreases drastically. In contrast, the performance of the Shannon entropy method is not influenced by the characteristics of the considered naphtha, but rather depends on the number of available commercial indices. The Shannon entropy method yields comparable results to the artificial neural network, provided that a sufficient amount of distillation data is available to supply information on the carbon number distribution. Combining the reconstruction methods with a fundamental simulation model illustrates the necessity of having accurate feedstock reconstruction methods since they allow to capture the full power of fundamental simulation models for the simulation of industrial processes. © 2010 American Institute of Chemical Engineers AIChE J, 2010     

Effects of air/fuel combustion ratio on the polycyclic aromatic hydrocarbon content of carbonaceous soots from selected fuels

Patterns of polycyclic aromatic hydrocarbon (PAH) content were observed from GC/MS analysis of the extracts of soots at various air/fuel combustion ratios of three commonly used fuels: n-hexane, JP-8 (Jet fuel), and diesel. With increasing air/fuel ratio, from a simple diffusion flame up to an air/fuel ratio of 3.94, there is a significant loss of high molecular weight PAHs and an increasing abundance of oxidized lower molecular weight aromatics. The formation of high molecular weight PAHs is favored for JP-8 and diesel fuels at higher air/fuel combustion ratios than is the case with n-hexane, probably due to the aromatic content in JP-8 and diesel fuels acting as centers for large aromatic and soot nucleation. The efficiency and reproducibility of two techniques, Soxhlet and supercritical fluid extraction (SFE), used for extraction of PAHs from soot were compared. Electron paramagnetic resonance (EPR) measurements were performed on the soot both before and after supercritical fluid and Soxhlet extraction, and a substantial decrease in the spin density of soot following extraction indicates that extractable molecules are associated with 40–50% of the unpaired electrons in soot. This analysis generally supports trends observed in our earlier work for surface oxidation, surface area, unpaired electron spin density, hydration, and ozone oxidation.     

Effect of temperature and pressure on the speed of sound and isentropic bulk modulus of mixtures of biodiesel and diesel fuel

The density and speed of sound of blends of biodiesel with No. 2 and No. 1 diesel fuels were measured from atmospheric pressure to 32.46 MPa at temperatures of 20 and 40°C. The isentropic bulk modulus was calculated from these quantities. The results show that the density and isentropic bulk modulus can be accurately modeled as having a linear variation with blend percentage. Speed of sound is better correlated by a second-order equation. Correlation equations are given and a blending rule is developed that allows the density, speed of sound, and isentropic bulk modulus of blends to be calculated from the properties of the biodiesel and diesel fuel.     

Optimization of Fischer‐Tropsch Synthesis Using Neural Networks

Fischer‐Tropsch synthesis is an important chemical process for the production of liquid fuels and olefins. Optimization of hydrocarbon products such as diesel and gasoline produced by Fischer‐Tropsch synthesis usually requires the knowledge of the complex polymerization mechanism and the kinetic parameters associated with it in order to optimize production. The Fischer‐Tropsch reaction mechanism is still not fully understood, making optimization a hard task. In this work, a neural network was used in substitution to the reaction mechanism to optimize diesel and gasoline production based on few experimental data for the reaction. The neural network has yielded satisfactory predictions of the product distribution (with prediction errors lower than 5 %) and the optimum operating conditions for gasoline and diesel production were found for a commercial iron based catalyst.     

Effect of the alcohol type used in the production of waste cooking oil biodiesel on diesel performance and emissions

Experimental results were obtained by testing two different alcohol-derived biodiesel fuels: methyl ester and ethyl ester, both obtained from waste cooking oil. These biodiesel fuels were tested pure and blended (30% and 70% biodiesel content, volume basis) with a diesel reference fuel, which was tested too, in a 2.2 l, common-rail injection diesel engine. The operation modes were selected to simulate the European Driving Cycle. Pure biodiesel fuels, compared to the reference fuel, resulted in a slight increase in fuel consumption, in very slight differences in NOx emissions, and in sharp reductions in total hydrocarbon emissions, smoke opacity and particle emissions (both in mass and number), despite the increasing volatile organic fraction of the particulate matter. The type of alcohol used in the production process was found to have a significant effect on the total hydrocarbon emissions and on the particulate matter composition. As the alcohol used was more volatile, both the hydrocarbon emissions and volatile organic fraction of the particulate matter were observed to increase.     

Emission characteristics using methyl soyate–ethanol–diesel fuel blends on a diesel engine

A blend of 20% (v/v) ethanol/methyl soyate was prepared and added to diesel fuel as an oxygenated additive at volume percent levels of 15 and 20% (denoted as BE15 and BE20). We also prepared a blend containing 20% methyl soyate in diesel fuel (denoted as B20). The fuel blends that did not have any other additive were stable for up to 3 months. Engine performance and emission characteristics of the three different fuels in a diesel engine were investigated and compared with the base diesel fuel. Observations showed that particulate matter (PM) emission decreased with increasing oxygenate content in the fuels but nitrogen oxides (NO_x) emissions increased. The diesel engine fueled by BE20 emitted significantly less PM and a lower Bosch smoke number but the highest NO_x among the fuel blends tested. All the oxygenate fuels produced moderately lower CO emissions relative to diesel fuel. The B20 blend emitted less total hydrocarbon (THC) emissions compared with base diesel fuel. This was opposite to the fuel blends containing ethanol (BE15, BE20), which produced much higher THC emission.     

Experimental investigation of fuel properties, engine performance, combustion and emissions of blends containing croton oil, butanol, and diesel on a CI engine

Emission problems associated with the use of fossil fuels have led to numerous research projects on the use of renewable fuels. The aim of this study is to evaluate the effects of blends containing croton mogalocarpus oil (CRO)-Butanol (BU) alcohol-diesel (D2) on engine performance, combustion, and emission characteristics. Samples investigated were 15%CRO-5%BU-80%D2, 10%CRO-10%BU-80%D2, and diesel fuel (D2) as a baseline. The density, viscosity, cetane number CN, and contents of carbon, hydrogen, and oxygen were measured according to ASTM standards. A four cylinder turbocharged direct injection (TDI) diesel engine was used for the tests. It was observed that brake specific energy consumption (BSEC) of blends was found to be high when compared with that of D2 fuel. Butanol containing blends show peak cylinder pressure and heat release rate comparable to that of D2 on higher engine loads. Carbon dioxide (CO₂) and smoke emissions of the BU blends were lower in comparison to D2 fuel.     

Transient testing of soy methyl ester fuels in an indirect injection,compression ignition engine

An evaluation of the exhaust emissions from a compression ignition engine for fuels composed of 100 and 30% methyl esters of soy oil (SME) is described. These fuels were compared with a low-sulfur, petroleum #2 diesel fuel in a Caterpillar 3304, prechamber, 75 kW diesel engine, operated over heavy- and light-duty transient test cycles developed by the United States Bureau of Mines. More than 60 h of testing was performed on each fuel. The objective was to determine the influence of the fuels upon diesel particulate matter (DPM) and gaseous emissions. The effect of a modern diesel oxidation catalyst (DOC) also was determined in an effort to minimize emissions. Neat SME produced a higher volatile fraction of the DPM, but much less carbon soot fraction, leading to overall DPM reductions of 23 to 30% for the light- and heavy-duty transients. The DOC further reduced the volatile fraction and the total DPM. The SME fuel reduced gaseous emissions of CO by 23% and hydrocarbons by over 30% without increasing NO_x. The DOC further reduced CO and hydrocarbon levels. Mutagenicity of the SME exhaust was low. Results indicate that SME fuel, used with a proper DOC, may be a feasible emission reduction technology for underground mines. References to specific products do not imply endorsement by the U.S. Bureau of Mines, a now defunct agency.     

Engine-out characterisation using speed-load mapping and reduced test cycle for a light-duty diesel engine fuelled with biodiesel blends

In this two-phase experimental programme, key effects of different biodiesel fuels and their blends on engine-out responses of a light-duty diesel engine were investigated. Here, coconut methyl ester (CME), palm methyl ester (PME) and soybean methyl ester (SME) were tested to represent the wide spectrum of degree of saturations in the fatty acid composition. Fossil diesel which served as the blending component was used as the baseline fuel for benchmarking purposes. Phase I examined how engine speed and load affect patterns of variation in tailpipe emissions and engine performance parameters for the test fuels. Here, the trends in engine-out responses across the operational speed-load map for all the tested biodiesel fuels were similar and consistent throughout. However, there were marked differences in the levels of equivalence ratio and specific fuel consumption, as well as exhaust concentrations of CO, UHC and smoke opacity. This is mainly due to differences in fuel properties, especially fuel-bound oxygen content, density and impurity level. Phase II appraised the performance of 31 different fuel blend combinations of fossil diesel blended with CME, PME or SME at 10 vol.% interval under a steady-state test cycle. The use of biodiesel fuels with low to moderate degree of unsaturation was found to conclusively reduce regulated emission species of UHC, NO and smoke opacity levels by up to 41.7%, 5.4% and 61.3%, respectively. This is in contrast to the performance of the highly unsaturated SME, where CO, UHC, NO and smoke opacity levels are higher in relation to that of fossil diesel. Simultaneous NO-smoke reduction can be achieved through the introduction of at least 1 vol.% of PME or 50 vol.% of CME into diesel fuel, although minor trade-off in the higher specific fuel consumption is observed.     

Characterization of the key fuel properties of methyl ester-diesel fuel blends

Characterizing of the fuel properties of diesel fuels, alternative fuels and their blends can assist the researchers who work on alternative fuels for diesel engines. Therefore, in this study, methyl esters were produced from five edible vegetable oils (sunflower, soybean, canola, corn and cottonseed) and blended with two different diesel fuels at 2%, 5%, 10%, 20%, 50% and 75% on a volume basis to characterize the key fuel properties of the blends such as density, viscosity, pour point, distillation temperatures and flash point. The results showed that the fuel properties of the blends were very close to those of diesel fuels at low concentrations upto 20% of methyl esters.     

The evaluation of viscosity and density of blends of Cyn-diesel pyrolysis fuel with conventional diesel fuel in relation to compliance with fuel specifications EN 590:2009

The production of synthetic fuels from alternative sources has increased in recent years as a cleaner, more sustainable source of transport fuel is now required. In response to European renewable energy targets, Ireland has committed, through the Biofuels Obligation Scheme of 2008, to producing 4% of transport fuels from biofuels by 2010 and 10% by 2020. In order to be suitable for sale in Europe, diesel fuels and biodiesels must meet certain European fuel specifications outlined in the EN 590:2004 and EN 14214:2009 standards. The aim of this project is to prepare blends of varying proportions of synthetic diesel fuel (Cyn-diesel), produced from the pyrolysis of plastic, versus regular fossil diesel. The viscosity (mm²/s) and density (kg/m³) of these blends as well as of the regular diesel fuel were analysed in relation to compliance with the European fuel standard EN 590.