Small data and Gaussian fields: how mathematics helps to produce oil

The University mathematicians are developing new algorithms for spectral geological modelling of oil-bearing formations designed for arbitrary shaped wells. The scientists have created a mathematical tool that makes it possible to reproduce the most probable formation structure based on rather scarce well data. The mathematicians are constantly optimising and adapting the used algorithms for specific fields with the help of the programme tailored for these purposes.

Mathematically, the problem of modelling the formation using scarce well data is akin to reconstructing a photograph using 50 pixels. Everyone is talking about data mining these days — big data is a buzz word today. The project looks further: how to extract information from small amounts of data is still an emerging trend in small data.

Sergey Tikhomirov, Professor at St Petersburg University, project manager, coordinator for relations with industry at the Chebyshev Laboratory

‘The algorithms and software modules developed in cooperation with St Petersburg University have enabled Gazprom Neft to create a new digital product. As a result of more precise modelling of target formations, carried out in real time, the programme makes it possible to boost the drilling efficiency of high-tech wells. The effective application of the new product at the company’s fields enables us to appreciate the prospects for its commercialisation on foreign markets,’ says Alexei Vashkevich, Director of Gazprom Neft for Technology Development.

At the moment, about 10 associates from the Chebyshev Laboratory and the Faculty of Mathematics and Computer Science at St Petersburg University are involved in the project. Postgraduate and undergraduate students of the faculty are involved in the work together with professors and research associates. The remarkable thing is that this project is not their only collaborative effort with the Gazprom Neft Science and Technology Centre. In total, more than 30 research associates of the laboratory and students of mathematics are involved in various joint projects with the company.

Geological modelling problems are addressed in a variety of ways throughout the world. The Chebyshev Laboratory has vast scientific experience in Gaussian fields and the spectral method. We are one of the world leaders in this topic, which makes it possible for us to develop a unique product.

Sergey Tikhomirov, Professor at St Petersburg University

‘Scientific cooperation between the Chebyshev Laboratory and the Gazprom Neft Science and Technology Centre goes beyond the project related to spectral modelling. For example, we have a joint project on the selection of optimal methods for enhancing oil recovery. In this project we are working both on the analysis and creation of physical and mathematical models, and on the mathematical support and interpretation of field and laboratory experiments. Recently, the project on spectral modelling has begun to develop into the field of quantum computing. It seems fair to say that this is the hype of the future,’ said Sergey Tikhomirov.

For reference: St Petersburg University and the Gazprom Neft Science and Technology Centre have been cooperating for much more than one year. This includes the areas related to petroleum engineering, mathematics and IT, economics, management and human resources. Since 2017, scientists and students of the University have participated in 30 joint research projects with the Gazprom Neft Science and Technology Centre. Additionally, St Petersburg University, together with the company, is implementing the master’s programme ‘Geological Support for Hydrocarbon Reservoirs Development’ in the ‘Petroleum Engineering’ field of study. St Petersburg University students regularly do practical training and internships at the company with the possibility of further employment.