Adaptive predictive control of the sulfur recovery process at Pemex Cadereyta refinery

SUMMARY This paper describes the design and application of an optimized control system, based on adaptive predictive expert control methodology, to the sulfur recovery units (SRUs) of the Pemex Refinery in Cadereyta, and presents comparative results obtained in relation to those of the conventional control system operating the plant. The sulfur recovery process represents the final stage in the refinery production chain, and is responsible for removing the sulfur content, usually in the form of hydrogen sulfide (H2S) waste gas generated by other refinery processes. The main goal is minimizing the sulfur content in the gasses released into the atmosphere to comply with emission level requirements. This goal can be obtained by a precise control of the ratio between hydrogen sulfide and sulfur dioxide (SO2) in the tail gas, which determines, through a Claus reaction, the maximum sulfur recovery. The SRUs operating in Cadereyta were connected to a common amine acid gas collector and a common ammonia acid gas collector. Optimized control strategies were designed to control the pressure on both acid gas collectors and the flow rate of air entering each SRU. The results show that the H2S:SO2 ratio in the tail gas is considerably more stable under optimized control, particularly in the presence of abrupt changes in the acid gas load. The peaks observed in this ratio under conventional control disappear under optimized control showing a significant improvement in the process operation. Copyright © 2012 John Wiley & Sons, Ltd.