Williams-Otto Reactor Case Study

This case study is same as presented in Miletic and Marlin (1998). The Williams-Otto reactor (Williams and Otto, 1960) is shown in Figure , and has reaction sequence:

  
Figure: Williams-Otto Reactor

The optimization variables are the flow rate of component B () to the reactor and the reactor temperature (), while the flow rate of feed A () to the reactor and the reactor volume () are constant. The profit function for the William-Otto Reactor RTO system is:

Reaction products C and G have no value, and the costs of utilities are neglected. The perfect model structure is used for RTO implementation in this paper:

Material balances:

Reaction rate expressions:

where are fixed model parameters and are adjustable model parameters. For the purpose of this paper, a trust-region constraints for the manipulated variables are introduced by:

and measurement noises are assumed as zero-mean normal distributed with the standard deviation of the flow, temperature and composition measurements are , and of the nominal value, respectively.

• Simulation Results
• Performance Analysis