Revamps 2026 Issue

The coking-and-carburisation propensity index from the I-PAMS soft sensor declined by 71% over the same period. Sensor data uptime over the 18 months that have followed the deployment was 99.2%, with the small remaining down- time attributable to two short DCS bus events unrelated to the sensor system itself. Deployment The deployment reframed the operator’s relationship with the SMR. Before, operating decisions about burner adjust- ments and tube management were guided by 12 weld-pad TC values, eight pass-outlet TCs, and a weekly pyrometer snapshot.6 Afterwards, the operator could see a 36-tube heat-map updated continuously, with the worst-case tube highlighted and the trend of every measurement point retained in the historian. The conversation in the control room shifted from “is the reformer within alarm?” to “where is the reformer drifting and which burners need trimming?”⁷ Three classes of limitation are worth noting. First, even at 36 instrumented tubes out of 240, the majority remain dependent on periodic pyrometer surveys for individual tube data.6 The instrumented selection was sufficient to support combustion engineering decisions and to trigger inspections of specific zones, but generalisation from instru - mented tubes to neighbours requires careful interpretation. The next phase plans to extend instrumentation to a fur- ther 48 tubes, prioritised by the historical exposure pattern revealed by the first phase. Second, the technology relies on a robust mechanical thermal contact between the waveguide and the tube outer wall. In the case study, the stainless-steel banding scheme survived 18 months of operation without re-tensioning. On tubes with significant ovality or thick scale, however, more aggressive surface preparation or alternative attachment may be required. Third, the soft-sensor models on the I-PAMS platform require a representative training dataset that spans the operating envelope. Thirty days of correlation data acquired during commissioning were sufficient for a first model, but rolling retraining is required as the catalyst ages, the feed composition drifts, and the burner population evolves.8 Scalability is straightforward. The same approach is appli- cable to fluid catalytic cracking (FCC) unit air heaters and CO boilers, hydroprocessing feed heaters, and the thermal reactor and waste heat boiler of a sulphur recovery unit (SRU). The same waveguide technology in its process-side variant has applications in SMR transfer lines, secondary reformers where present, and downstream heat exchangers during fouling studies. Conclusions The deployment of multi-point ultrasonic skin tempera- ture sensors on 36 of the 240 catalyst tubes of a refinery hydrogen plant SMR demonstrated that a measurement density of approximately 290 points across a representa - tive tube selection is sufficient to fundamentally change the operability of the asset. The hidden hotspots that the gold- cup pyrometer and weld-pad TC array had not detected became visible within 72 hours of commissioning, and the

structured intervention that followed delivered three quan- tifiable outcomes: a reduction of tube-to-tube temperature standard deviation from 28°C to 12°C, a 1.4% increase in hydrogen production at unchanged firing, and a projected tube-life extension of 22 months. The carbon-cost equiv- alent of the energy reduction was approximately 9,000 tonnes of CO₂ per year. The same approach is applicable across the principal high-temperature equipment in a typical refinery, including FCC regenerator-side equipment, hydroprocessing fired heaters, and the SRU thermal section. The technology is ret - rofit-friendly during planned shutdowns, requires no weld - ing to pressure-containing components, and integrates with existing DCS and historian infrastructure through standard industrial protocols. The broader implication for refining is that the long-stand - ing assumption – that SMR tubes must be managed with sparse and periodic temperature data because dense data is impractical – is no longer accurate. Multi-point wave- guide-based sensing closes the measurement gap and brings combustion engineering into the same data-driven regime that has already transformed rotating equipment monitoring and process control.

μTMaps, μSTMaps, and I-PAMS are marks of XYMA Analytics Pvt Ltd, Chennai.

References 1 Rostrup-Nielsen J R, Catalytic Steam Reforming , Springer-Verlag, Berlin, 1984. 2 American Petroleum Institute, API Standard 560 – Fired Heaters for General Refinery Service, 5th Edition, 2016. 3 Larson F R, Miller J, A Time-Temperature Relationship for Rupture and Creep Stresses, Trans. ASME, 1952. 4 Periyannan S, Rajagopal P, Balasubramaniam K, Ultrasonic Waveguide-Based Multi-Point Temperature Sensing, Insight , 2014. 5 API Technical Data Book , Heater Tube Life Assessment and Creep Evaluation Methods. 6 Latham D L, McCann B P, Pyrometry in Steam Methane Reformer Furnaces, Hydrocarbon Processing , 2017. 7 Zecevic N, Bolf N, Application of Advanced Multivariable Predictive Control on a Hydrogen Plant Steam Reformer, Petroleum Science , 2020. 8 Lee J, Bagheri B, Jin C, Introduction to Cyber Manufacturing and Predictive Maintenance, Computers in Industry, 2016. Balaji Nambi is Global Growth Lead at XYMA Analytics. As a down - stream oil and gas professional, Nambi specialises in asset integrity and high-temperature inspection technologies. He works closely with refinery operators across the region to address the limitations of con - ventional inspection methods and deliver sensor-driven solutions that improve safety, yield, and operational reliability. With a deep under- standing of critical refinery units including crackers, SRUs, reformer tubes, and DCUs, Balaji brings both technical depth and a practical operator mindset to every engagement. Dr Nishanth Raja is the CEO at Xyma Analytics, with a strong back - ground in ultrasonic guided waves, high-temperature sensors, non-de- structive evaluation, and data analysis and interpretation, acquired through his PhD in engineering from Anna University and post-doctoral research at IIT Madras. Prior to founding XYMA Analytics, Dr Raja was engaged in pioneering work with DRDO (Government of India).

Revamps 2026