Process-side-corrected monitoring of equivalent coke thermal resistance in an industrial ethylene cracking furnace: a case study

Coke deposition in radiant coils is a persistent thermal-management problem in industrial ethylene cracking furnaces. Tube-metal temperature (TMT) is one of the few continuously available indicators of coke-related heat-transfer deterioration, but inverse estimates depend on operating conditions and thermal-property assumptions. This case study evaluates a process-side-corrected thermal inversion that combines pyrometric TMT with distributed control system data. A normalized process heat-demand factor updates the effective heat-transfer rate, whereas a TMT-independent furnace-temperature factor is examined only as a model-form sensitivity. The method was applied to 2352 hourly samples from six monitoring groups over a full industrial campaign. Using an effective reference coke conductivity of 4.95 W m −1 K −1 , the terminal equivalent coke-layer thickness was 7.50-11.23 mm. The process-side-corrected and fixed-boundary formulations differed by 0.55-2.76 mm (7.9-35.3%) at the terminal period; this comparison is not interpreted as validation. The naphtha-coke conductivity range of 4.4-5.5 W m −1 K −1 changed the group-mean result by −0.41 to +0.35 mm, while a high-temperature scenario of 8.7 W m −1 K −1 increased it by 1.68 mm. A bounded simultaneous-variation ensemble produced group-specific 2.5th-97.5th percentile ranges spanning 6.95-12.81 mm. The reconstructed quantity is therefore an equivalent thermal-resistance indicator referenced to the assumed conductivity, not a verified geometric coke thickness.