NEWS
The project creates the best factors to produce anodes and reduces costs A student team at UOB develops a model that identifies optimal conditions for aluminum manufacturing
Sakhir – University of Bahrain (Ali Al Sabbaq)
29 January 2024
A student team at the University of Bahrain has developed a model that identifies optimal operating factors for aluminum manufacturing conditions, and reduces faulty anodes, which helps reduce production costs in aluminum smelting plants.
The project was completed by female students in the Department of Chemical Engineering at the College of Engineering at the University: Nariman Najeh Faris, Suad Mohammed Khalifa, and Hadeel Rashid Al-Muqahwi, who trained at Aluminum Bahrain Company “Alba”. Their project won second place in the competition of the graduation projects exhibition at the College of Engineering in the category of chemical engineering, after competing with 13 projects of the same category.
The project was supervised by Dr. Mohamed Ali Bin Shams, Head of the Chemical Engineering Department at UOB, along with Eng. Ahmed Abdel Ghaffar, Head of Planning in Alba’s Carbon Circles Group.
Regarding the student project, student Nariman Fares explained that carbon anodes are of great importance in the production chain of any aluminum smelter, and for the Bahrain Aluminum “Alba” smelter, the quality of these anodes is of great importance in terms of components and manufacturing process.
She added: “Our project aims to find the best operating factors, such as temperature and pressure, to produce these anodes in a way that reduces overhead costs, while maintaining the same quality standards adopted during manufacturing.”
She stated that the student team used machine learning algorithms within the Python programming language, taking advantage of the available data provided by the company, and then using predictions to reach optimal operating factors, which will reduce production costs and increase productivity.
Regarding the project results, student Suad Khalifa stated that the project provides the engineer responsible for the production process with an expectation of the carbon anodes’ quality, which are produced based on the operating factors determined by the engineer himself, noting that the student team reached the optimal values for the variables involved in the formation of carbon anodes using the machine learning model.
Khalifa said: “The model showed that it will produce acceptable anodes, which means that the losses resulting from the production of excludable carbon anodes will be significantly reduced after avoiding them if this model is applied,” noting at the same time to the study of another solution to reduce the number of excluded carbon anodes, through modifying the devices and equipment’s maintenance plan, the amount that would be saved if the proposed plan were approved, was mathematically estimated, and this solution resulted in saving approximately one hundred thousand Bahraini Dinars annually.
Commenting on her assessment of the student team’s experience, Hadeel Al-Muqahwi said: “It was a rich experience during which we learned Python programming language and machine learning algorithms as tools to achieve the concept of artificial intelligence in the aluminum industry, and we were able to develop strong skills in collaboration, communication and problem solving, as well as data analysis skills – which are essential in exploring the capabilities of Python – and building a machine learning model.” She added: “We trained for two months at the Alba plant and learned a lot of skills from the engineers and workers there, which increased our knowledge of operating systems and the environment surrounding the production process.”
Dr. Abdulla Habib Ahmed, Chief Operatins Officer at Aluminium Bahrain (Alba), praised the project of the female students at the University of Bahrain, which contributes to reducing the rate of exclusion of defective anodes after heat treatment in furnaces, which is included in the aluminum production process, noting that the project helps reduce costs.
