Revamps 2026 Issue

Multi-point temperature mapping of steam methane reformer tubes

A retrofit field deployment of ultrasonic waveguide skin temperature sensors in a refinery hydrogen plant

Balaji Nambi and Nishanth Raja XYMA Analytics

I n refinery hydrogen plants, the steam methane reformer (SMR) operates with tube wall temperatures within tens of degrees of the catalyst tube alloy metallurgical limit. A 20°C average overshoot can halve creep life; a 20°C under - shoot can trim approximately 2% from hydrogen output. Conventional measurement – periodic gold-cup pyrometer surveys and a small number of weld-pad skin thermocou - ples – captures neither the per-tube spread nor the temporal drift that drives both losses. This article reports on a retrofit field deployment of an ultrasonic multi-point waveguide skin temperature sensor on a 100,000 Nm³/h refinery hydrogen plant SMR with 240 catalyst tubes. The deployment delivered continuous, per-tube skin temperature data on 36 representative tubes, integrated into the existing distributed control system. A structured campaign of burner trim, feed rebalancing, and excess-air optimisation reduced tube-to-tube temperature standard deviation from 28°C to 12°C, lifted hydrogen pro - duction by 1.4% at unchanged firing, and projects a tube- life extension of approximately 22 months. The technology, deployment, results, and operational lessons are discussed in detail. Dominant constraints Refinery hydrogen plants are the quiet workhorses of the modern refinery.1 Whenever a hydrocracker, hydrotreater, or continuous catalyst regeneration (CCR) unit needs a step change in capacity, the hydrogen plant has to deliver, usu - ally through one or two SMRs operating close to their ther - mal limits. The SMR radiant box typically contains 200 to 500 vertically suspended catalyst-packed tubes,2 each fired externally by row-mounted or wall-mounted burners that supply 70% to 80% of the heat duty for the endothermic methane-steam reaction (see Figure 1 ). The dominant constraint on SMR operation is tube wall temperature. A typical HP-modified or 35Cr-45Ni-Nb micro-alloyed catalyst tube is designed for an average wall temperature of around 880°C, a hotspot allowance near 920°C, and a creep-rupture life of 100,000 hours. The rela - tionship between operating temperature and creep life is exponential through the Larson-Miller framework,3 and field experience yields a widely quoted rule of thumb: a sustained 20°C overshoot at the tube wall can halve creep life, while a

sustained 20°C undershoot below the design optimum can cost approximately 2% of hydrogen output through reduced methane conversion (see Figure 2 ). Despite this sensitivity, the dominant practice for SMR tube wall temperature monitoring remains a combination of weekly hand-held gold-cup pyrometer surveys and a small number of weld-pad skin thermocouples (TCs). Pyrometer surveys are operator-intensive, depend on the line of sight available through observation doors, and capture only one snapshot in time per tube, typically once per week, and rarely once per shift. Weld-pad thermocouples on a tube operating at 900°C have a typical service life of three to nine months before drift, sheath oxidation, or weld-pad detach - ment forces a removal that has to wait for the next planned shutdown. The combined measurement chain therefore leaves the operator with an incomplete picture of one of the most thermally aggressive assets in the refinery. Hot tubes drift undetected; cool tubes carry less than their share of duty; burner imbalance is corrected only at the next planned tuning. The deployment described in this article addresses

Figure 1 SMR with the radiant box exposed to show ver - tically suspended catalyst-packed tubes operating at radi - ant heat. The asset depicted is representative of a refinery hydrogen plant primary reformer

Revamps 2026