MOSCOW, May 18. A model for evaluating the suitability of objects for burying environmentally dangerous excess carbon dioxide in them was proposed by scientists from the Tomsk Polytechnic University. According to them, the development takes into account many different parameters and has a high forecast accuracy. The results are published in Expert Systems with Applications.
In recent decades, with the development of industry, the combustion of hydrocarbon fuels has increased, which leads to an increase in emissions of anthropogenic carbon dioxide (CO2) into the atmosphere. The intensive release of this particular gas, according to experts, is one of the main causes of global warming. //ria.ru/20220926/tyumgu-1819010529.html» data->
One of the most promising methods for eliminating carbon dioxide is geological disposal, the specialists of the university noted. According to them, there are two main approaches to this process: «spent» carbon is either used to increase the efficiency of hydrocarbon production, or pumped into deep-lying aquifers.
Scientists of the Tomsk Polytechnic University (TPU) proposed a model for evaluating the efficiency of CO2 binding in a particular rock, taking into account a large number of parameters. According to them, the development predicts with high accuracy the behavior of carbon dioxide under given conditions, which makes it possible to select the optimal objects for its disposal.
«
“Not every geological formation can become a reliable repository of CO2. Deep-lying aquifers, as well as depleted oil and gas fields, are primarily considered as storage facilities,” said the author of the work, research engineer at the Heriot-Watt Center of TPU Shadfar Davoudi.
New model based on the application of machine learning methods. According to them, the scientists managed to achieve high accuracy of the forecast due to the large volume of the training sample and detailed planning of experiments.
«5450 data series were used as initial data for training. variables and result, and then learns to predict similar dependencies on new data,” explained Shadfar Davoudi.
In the future, scientists plan to improve the quality of the model prediction by optimizing the algorithm and applying a new technique for preprocessing the initial data .
TPU research is supported by the program of the Ministry of Education and Science «Priority-2030».