2026, Vol. 13, No. 2. - go to content...

Permanent address of this page - https://resources.today/en/06inor226.html

Метаданные этой статьи так же доступны на русском языке

DOI: 10.15862/06INOR226 (https://doi.org/10.15862/06INOR226)

Full article in PDF format (file size: 527 KB)

For citation:

Komrotov I.S., Chelnokov V.V. Artificial intelligence in monitoring oil biodegradation: modern approaches and prospects (rewiev). Russian journal of resources, conservation and recycling. 2026; 13(2). Available at: https://resources.today/PDF/06INOR226.pdf (in Russian). DOI: 10.15862/06INOR226

Artificial intelligence in monitoring oil biodegradation: modern approaches and prospects (rewiev)

Komrotov Ivan Sergeevich
D. Mendeleev University of Chemical Technology of Russia, Moscow, Russia
E-mail: komxim@yandex.ru
ORCID: https://orcid.org/0009-0005-0893-2907

Chelnokov Vitaly Vyacheslavovich
D. Mendeleev University of Chemical Technology of Russia, Moscow, Russia
E-mail: chelnokov.v.v@muctr.ru
ORCID: https://orcid.org/0000-0002-3065-9776

Abstract. Biodegradation of oil caused by the metabolic activity of microorganisms (bacteria, archaea, fungi) is a serious threat during long-term storage of raw materials in tanks and pipelines. The annual economic damage from microbial corrosion, loss of light hydrocarbon fractions (up to 20 % in six months) and emergency spill response is estimated at $1,5 billion. This entails a decrease in the energy value of oil, accelerated degradation of metal structures and large-scale environmental risks. The article discusses approaches to the use of artificial intelligence technologies to create systems for continuous monitoring, predictive forecasting and proactive control of biodegradation of petroleum products. Three key technological directions are described in detail. The first is the intelligent analysis of streaming data from distributed sensors of the Internet of Things that record hydrogen sulfide concentrations, temperature, and redox potential. The second is the processing of spectroscopic information for rapid detection of chemical composition changes. The third is the metagenomic analysis of microbial communities to assess their destructive potential. Predictive modeling based on hybrid neural network architectures combining physico-chemical laws of kinetics with machine learning algorithms occupies a central place. Examples of successful industrial implementation of artificial intelligence systems at pipeline transport facilities and tank farms are given. The main challenges are outlined: ensuring interpretability of models to justify decisions, seamless integration with legacy automated process control systems and laboratory systems, as well as the critical need for representative sets of labeled data. Overcoming these barriers is necessary for the transition to fully automated preventive biodegradation management.

Keywords: artificial intelligence; petroleum; biodegradation; Internet of things; predictive modeling; metagenomics; microbial corrosion

Download article in PDF format