Revamps 2026 Issue

The reforming train comprises a primary SMR (R-101) and a fired heater for the steam superheater (R-101), which has 240 vertically suspended catalyst tubes (HP-modified, 14 m heated length, 110 mm outside diameter) arranged in six rows of 40 tubes. The radiant section is fired by 144 wall- mounted ceramic-burner tiles arranged in three vertical lev - els along each side wall (see Figure 5 ). Prior to the deployment, the unit was instrumented with 12 weld-pad type-N skin TCs, eight pass-outlet TCs in the transition section, and a programme of weekly gold-cup pyrometer surveys covering all 240 tubes. The pyrometer survey provided a snapshot of the average tube wall tem - perature distribution but was inherently periodic and opera - tor-dependent. The skin TCs, biased toward the upper third of the tubes, had been replaced twice in the preceding 18 months following weld-pad detachment. Three operational issues motivated the deployment. First, the post-shutdown inspection in 2023 had identified accel - erated creep on a small population of tubes of which the temperature history had not stood out in the pyrometer record. Second, hydrogen production had been reduced by 3% on three separate occasions in the previous 12 months when an isolated skin TC reading approached the alarm threshold. Subsequent inspection suggested that the local - ised excursion had been real but representative of only a few tubes, not the heater as a whole. Third, the burner tiles on the south wall were known to be ageing unevenly. However, the existing measurement chain provided no means to local - ise the resulting heat-flux distortion. The objectives set for the deployment were therefore threefold: to obtain a per-tube tube-skin temperature distri - bution at sufficient density to identify hotspots and burner- side imbalance, to detect tube-skin drift earlier than the existing alarm scheme allowed, and to support a measured Figure 4 Multi-point waveguides clamped to the external surface of reformer tubes. Stainless-steel banding at reg - ular intervals secures the waveguide against the tube wall; no welding to pressure-containing components is required

campaign of burner trim and feed alignment without com - promising tube life. Implementation and methodology Installation was carried out during an 11-day planned shut - down. Thirty-six tubes were selected for instrumentation: six tubes per row, distributed across the radiant box, with two tubes per row on each side wall and two in the centre of the row. Each instrumented tube was fitted with a single μSTMaps waveguide running 8m along the upper radiant section of the tube, with eight reflector stations per wave - guide. Total measurement count: 288 new skin temperature points, against the 12 weld-pad TCs that had previously been the only continuous source of skin data. Mechanical preparation followed the standard pattern for ultrasonic waveguide retrofit. The tube outer wall was cleaned at each contact zone to remove oxide scale, a high-temperature thermal-coupling material was applied, Figure 5 Radiant section of the SMR showing wall-mounted burners (right), waveguide-instrumented catalyst tubes (left), and the field-mounted electronics unit (foreground). The waveguides leave the radiant box through floor pene - trations and terminate at the acquisition unit

Revamps 2026