It has been known for over a century that carbon steel corrodes in syngas at moderate temperatures due to reaction with carbon monoxide to form gaseous iron pentacarbonyl, Fe(CO)5. However, commercially produced syngas also contains various minor contaminant species, and their impact on corrosion remained unclear. In this work, experiments have been carried out for carbon steel in syngas flowing through a reaction chamber, using Proton Transfer Reaction Mass Spectroscopy (PTR-MS) to measure the carbonyl concentration in the effluent gas. We have investigated the effects of temperature and contaminants (H2S, H2O, CO2 and O2). For comparison, we have also used online ultrasonic wall thickness monitoring of piping in an operating industrial plant. Results show that in clean syngas containing 21 bar CO and flowing at up to ~ 5 m s-1, the corrosion rate peaks at ~ 200C, but remains 0.1 mm/yr. The corrosion process is very strongly catalyzed by ppb levels of H2S, much less strongly activated by H2O and CO2, and can be inhibited by ppm levels of oxygen. Both in the laboratory and field work, corrosion rates up to ~ 0.2 mm/yr have been measured. The peak in the temperature dependence of corrosion rate is caused by a transition from kinetic control at lower temperatures to thermodynamic control at higher temperatures, but the position of the peak is dependent on the flow rate and other details of the experimental setup. The carbonyl corrosion reaction has been modeled using the Langmuir-Hinshelwood (LH) formalism for interactions at the gas-solid interface.
Remans et al. (Fri,) studied this question.