Fault-tolerant control design for safe production of hydrogen from bio-ethanol

This work presents a novel framework for fault-tolerant control design to achieve a safe and reliable production of hydrogen from bio-ethanol, which includes: (i) a nominal controller (NC) for fault-free operation, (ii) a reconfigurable controller (RC) for accommodate the faults anticipated at the design stage. The methodology produces alternative NC and RC solutions allowing to achieve a trade-off between performance and dimension of the structures, subject to the availability of healthy components. It is based on the evaluation of individual steady-state squared deviations in the context of Internal Model Control theory, and efficiently solved with genetic algorithms. The obtained structures are decentralized and can be implemented with conventional controllers. A reconfiguration mechanism which smoothly commissions the RC is also presented. The approach is tested in a hydrogen production plant to demonstrate its effectiveness against critical faults such as actuator blockade and loss of sensor measurement.