New technologies such as artificial intelligence are quickly being adopted into various oil and gas applications to optimise production and maintenance processes, with AI making substantial advancements in the area of asset integrity management.
As the upstream oil and gas sector matures, there is a growing focus on preserving ageing production assets while also achieving higher productivity at a lower cost. Effective asset management by oil and gas operators is important for the safety of workers, the efficiency of production, regulatory compliance, and enhancing company reputation.
The most common failures oil and gas assets are prone to include corrosion, wear and tear, structural integrity, and outdated control systems and instrumentation. Incorporating AI into asset management enables operators to implement predictive maintenance systems that can forecast probable equipment failures by analysing real-time data collected from sensors, machinery, and previous records.
The predictive nature of this maintenance strategy lowers operating expenses, minimises downtime, and increases the lifespan of important assets. McKinsey has estimated that AI-powered predictive maintenance solutions can reduce maintenance costs by up to 40 per cent by preventing unplanned downtime and reducing the need for emergency repairs.
A paper published earlier this year, titled Predictive maintenance in oil and gas facilities, leveraging AI for asset integrity management, noted that through the integration of AI-driven analytics and real-time data monitoring, oil and gas companies could enhance their asset integrity management practices and ultimately drive cost savings and operational excellence.
The authors said: “The integration of AI in predictive maintenance marks a paradigm shift, offering a proactive approach to asset management.
“Leveraging AI-driven analytics and real-time data monitoring, oil and gas facilities can fortify their asset integrity management practices.
“Through predictive algorithms and machine learning models, these technologies empower companies to forecast equipment malfunctions with unprecedented accuracy, allowing for timely interventions and mitigating potential risks.”
They noted that traditional maintenance practices have typically followed reactive or preventive approaches, with reactive maintenance often resulting in unplanned downtime, costly repairs and safety hazards, while preventive maintenance may lead to the unnecessary servicing of fully operational equipment.
However, the paper also detailed several challenges around implementing AI-driven predictive maintenance at oil and gas facilities, including data quality and availability, algorithm complexity, and integration with existing systems.
The authors added: “Moreover, the deployment of AI models in industrial environments requires careful consideration of safety, regulatory compliance, and cybersecurity concerns.
“Addressing these challenges requires close collaboration between data scientists, domain experts, and operational personnel to develop robust and reliable predictive maintenance solutions.”
The market for AI in oil and gas was estimated to be worth US$3.14 billion in 2024, according to Mordor Intelligence, and will have a compound annual growth rate of 12.61 per cent to reach US$5.7 billion by 2029.
Mordor said: “The increasing application of AI in reservoir analysis, drilling optimisation, anomaly detection in pipelines, safety monitoring, [and] emissions reduction is expected to fuel the growth of the market.
“The emergence of predictive maintenance powered by artificial intelligence in the oil and gas market is transforming companies in the sector’s asset management. “This ensures better reliability and reduces operational risks, which is expected to drive the growth of the market in the future.”
A key driver of growth in the oil and gas asset integrity market is the higher prevalence of digital technologies incorporated into asset integrity solutions, including predictive analytics/maintenance, machine learning, Internet of Things (IoT), and robotics.
Machine learning has been adapted and implemented for pipeline inspections, with the technology used to predict material loss, enabling more accurate scheduling of inspection and maintenance.
Moreover, digital twins are becoming an integral tool in oil and gas operations as companies strive to further optimise asset performance and minimise unplanned outages. Brazilian offshore solution provider Ocyan last year completed a three-year partnership with Vidya Technology on an AI-driven solution that combined digital twin technology, reality capture and AI computer vision to autonomously locate, map and identify visual corrosion anomalies.
The collaborative project boasted up to 90 per cent AI accuracy and was able to identify 256 per cent more anomalies than the usual approach, transforming data, images and reports into safer and more efficient maintenance actions. Using an existing Vidya Technology corrosion management solution enhanced with AI, 3D models within a digital twin and data contextualisation, the project employed the application for the riser balcony area of the FPSO Cidade de Itajai, which Ocyan operates.
The approach started with reality capture execution, from which the photos were synchronised to a 3D model and processed by AI deep neural networks to autonomously identify and classify visual anomalies. Results of the pilot project showed the application was able to identify affected areas by corrosion in metres squared, drastically reduce the necessary people on board for inspection, feed corrosion predictive models, integrate AI into the operation, and transform images into maintenance actions.
It also made it possible to build a continuous flow of data between the FPSO and its virtual representation and promote operational awareness of the integrity process, as well as achieving a scalable return on investment.
Vidya said: “The offshore operational context has thousands of square metres of painted area to control corrosion on the topside – thousands of square metres of thermal insulation, and hundreds of temporary repairs.
“Furthermore, conventionally the approach managing this scope relies on inspections with a high number of people on board exposed to risks for several days.
“Additionally, it relied on manual mapping, several systems, and spreadsheets, resulting in a complex environment of diffuse information.”
