Investigating the effect of biodiesel on diesel engine corrosion and durability with an environmental assessment: Modeling the impact on water resources with machine learning

Introduction

Because of the increasing trend towards the adoption of biodiesel-a veritable renewable alternative to fossil fuel-for purposes of transport, acceleration of research has been initiated over performance and durability of the engine. One major challenge identified can be termed as biodiesel as corrosive while in contact with metallic surfaces of diesel engines. Studies have indicated that biodiesel induces faster corrosion rates in metals like copper, brass, and carbon steel-associated with fuel systems and engine components. Corrosive behavior of biodiesel, being derived factors from its higher porosity, electrical conductivity, and presence of organic acids-forming when subjected to oxidation processes, are some of the factors that induce corrosion of parts in the engine which, in turn, are likely to stimulate higher maintenance costs and reduced life of engines at the end.

Biodiesel corrosion and engine endurance can be assessed using this study as a pioneering attempt to understand the manner in which biodiesel interacts with all metal components in diesel engines and its environmental effects in water pollution using advanced modeling techniques. The thorough investigation of multiple metal types regarding biodiesel combustion presents an exhaustive life cycle assessment-from-their-staged influence of these metals on water pollution. Such a multidisciplinary investigative activity advances materials degradation understanding and pushes the development of enduring engine components that can negotiate the peculiarities of biofuels within their operating environments. Additionally, machine-learning models are among the most significant advances toward predicting corrosion performance from different biodiesel compositions and environmental conditions. Traditional approaches rely on spending a large amount of time conducting experiments, performing static immersion tests, and suffering delays in the delivery of data. Machine learning will help leverage the large data sets for pattern recognition and correlations to help ensure a better selection of materials and fuel formulations.

Materials and Methods

The steps of this research were carried out in four steps. In the first step, biodiesel was produced through the transesterification process of waste cooking oil in the presence of 1.1% by weight of sodium hydroxide (as a catalyst) and methanol with an alcohol to oil ratio of six to one and a mixing intensity of 710 rpm. The production temperature was kept constant in the range of the boiling point of methanol. Then, test fuel samples labeled B0 (pure diesel as a control fuel), B2 (diesel fuel containing two percent biodiesel), B5 (diesel fuel containing five percent biodiesel), and B10 (diesel fuel containing ten percent biodiesel) were prepared with a volume of one liter. Next, the engine test was performed using a single-cylinder diesel engine. Next, engine oil sampling was carried out in volumes of 150 ml at working intervals of 48, 96, and 144 hours. Atomic absorption spectroscopy was used to atomize metal elements in the presence of an oxyacetylene flame. In this process, the digested solution is converted into free atoms containing vapor. Using the calibration and extrapolation method, the obtained absorption value is used to determine the element concentration in the presence of control solutions. In the next step, the life cycle assessment began with the preparation of a life cycle inventory focusing on the inputs related to fuel sample preparation and its effects on the corrosion rate in a diesel engine based on engine operating hours. The IMPACT2002+ method was used to perform the production inventory assessment. The next step included modeling and determining the effective parameter using the support vector machine method (a method of machine learning). In this step, 60% of the data was used as data for the model training stage and 40% as data for the test stage. The root mean square error parameter and the correlation coefficient were used to evaluate the model. The recursive feature removal method was used to perform sensitivity analysis and select the independent variable affecting the dependent variables.Text

Results and Discussion

According to the results, it was concluded that increased biodiesel concentration has a pronounced effect on the indicators, with B10 being the highest level considered for acidification and eutrophication due to increased nitrogen oxides stemming from biodiesel combustion. It thus calls for optimizing combustion processes coupled to nitrogen oxide emission reductions in order to mitigate adverse ecological impacts. Coupled with this, there was also a support vector machine (SVM) modeling approach to predict these water quality parameters, which showed high accuracy and reliability in being modeled. Sensitivity analysis determined relevant independent variables. The analysis was performed on six independent parameters and showed that they affected much of the environment-related indicators to water resources. The critical factor, biodiesel percentage in fuel, emerged as a significant factor affecting all the indicators for environment impact (percentage basis: 0.55 for eutrophication, 0.81 for acidification, and 0.76 for ecotoxicity). Such a trend indicated that changes in biodiesel content would play an important role in incurring-reducing or increasing impacts on the environment. Engine operational hours had a relatively low impact on eutrophication (0.45) and acidification (0.61) with an even smaller impact on ecotoxicity (0.38). This indicates that engine hours do act towards deteriorating the environment but are less efficient than other parameters. Biodiesel production schemes highly impact the water eutrophication.

Conclusion

This operation period extremely matters because, from the analysis of the engine oils, it says that the amount of metal will be highly influenced by operation time. This was further brought about by oxidation or corrosion over time on the aluminum, chromium, copper, and iron content in this trend of attempting to use a higher percentage of biodiesel to achieve a significant increase in welding and corrosion affecting engine parts. Besides these two, acidification and also eutrophication of water resources have also been worsened by the use of biofuels, more especially B10. Excess nitrogen oxides emitted through burning biodiesel act as nutrients and favor algal blooms that ultimately acidify waters and could eventually be destructive to aquatic ecosystems. Hence nitrogen oxide emissions shall be strictly controlled, and one must install after-treatment systems for catalytic mitigation of this kind of fuels on the adverse effects the environment has to face with such fuels.