Editor Rene Gonzalez editor@petroleumtechnology.com tel: +1 713 449 5817 Managing Editor Rachel Storry rachel.storry@emap.com Editorial Assistant Lisa Harrison lisa.harrison@emap.com Graphics Peter Harper Business Development Director Paul Mason Paul.Mason@petroleumtechnology.com tel: +44 7841 699431 Managing Director Richard Watts richard.watts@emap.com Circulation Fran Havard circulation@petroleumtechnology.com EMAP, 10th Floor, Southern House, Wellesley Grove, Croydon CR0 1XG tel +44 208 253 8695 Register to receive your regular copy of PTQ ptq PETROLEUM TECHNOLOGY QUARTERLY Vol 31 No 3 Q2 (Apr, May, Jun) 2026
The challenges of AI deployment W ith generative AI (gen AI), refinery planners can rapidly go from planning to production at scale by layering gen AI on top of data sources. The industry has been generating massive amounts of data going back to the pre-internet era. Fast forward to 2026, and refiners have deployed gen AI for only a few years, with varying levels of success. Accenture recently reported that only 13% of these companies have created significant enterprise-level value. Many operate on decades-old systems that lack native AI compatibility. Nevertheless, a 24/7 continuous chemical processing facility requires carefully planned revamps that can benefit from the ability to detect anomalous equipment and operational behaviour. AI providers give the opportunity to buy time and extend revamp intervals through early detection of incipient stages of equipment ‘chatter’. ExxonMobil is already said to applying AI and advanced analytics across its refin - ing and chemical operations, having built a massive data infrastructure (‘data lake’) that aggregates plant data and collects trillions of operational data points from sites worldwide. Technology such as model-based predictive monitoring and diagnostics for rotating equipment is not new, but AI enhances existing capabilities, gaining new insights from vast quantities of data at scale to monetise previously unachievable efficiency. To accelerate their digital transformation journey, refiners are team - ing up with data and AI sources of expertise. For example, Cognite and Koch Ag & Energy Solutions, LLC recently announced a collaboration that will result in agile, data-driven execution in maintenance strategy, turnaround execution, and general efficiency gains across all plants, leveraging AI to automate analysis and increase velocity for decision-making. With this type of collaboration, the onus is on creating a holistic view (digital twin) powered by contextualised data and AI to run predictive analytics and effectively monitor the health of critical equipment. Access to all relevant data and insights in one place also enables operators to avoid siloed workflows and improve collaboration and decision-making processes. Maximising plant efficiency reduces time to value by identifying and contextualis - ing the data from numerous disjointed IT, operational technology (OT), and engi - neering data sources, and leveraging hybrid AI to perform real-time optimisation via visualisations, simulators, and optimisers. Honeywell and TotalEnergies have recently announced a collaboration at the Port Arthur Refinery in Texas, aiming to support and empower operators in making timely and informed decisions while enhancing operational autonomy. TotalEnergies has already implemented an AI-assisted solution at the Port Arthur site’s delayed coking unit. Preliminary results show the AI-assisted solution has successfully fore- casted five potential events, helping minimise downtime and reduce emissions from flaring. The predictions were made an average of 12 minutes before an alarm inci - dent, enabling operators to quickly implement corrective actions before an event. Even as AI proves its value across refinery offsite operations, deploying it in real-world brownfield environments presents its own set of hurdles. Legacy infra - structure, fragmented data, and cultural resistance often stand in the way of real- ising AI’s full potential. Among these, data quality and integration remain the most foundational and frequently underestimated barriers. As digital transformation accelerates, AI is not just a trend; it is a strategic asset in building the refinery of the future. Refinery conferences and seminars worldwide are increasingly dedicating more sessions to generative AI, digital twins, and related technologies, as will be discussed throughout 2026 in PTQ and Digital Refining. Rene Gonzalez
PTQ (Petroleum Technology Quarterly) (ISSN No: 1632-363X, USPS No: 014-781) is published quarterly plus annual Catalysis edition by EMAP and is distributed in the US by SP/Asendia, 17B South Middlesex Avenue, Monroe NJ 08831. Periodicals postage paid at New Brunswick, NJ. Postmaster: send address changes to PTQ (Petroleum Technology Quarterly), 17B South Middlesex Avenue, Monroe NJ 08831. Back numbers available from the Publisher at $30 per copy inc postage.
5
PTQ Q2 2026
Powered by FlippingBook