This paper presents a new method for the coordination of decentralised model predictive controllers (MPC) for application to plantwide chemical process control. The key enabling mechanism for this strategy is the concept of dissipativity and dissipative trajectories. The latter is a novel technique which is promising for application to the analysis and
design of model predictive control systems. Results pertaining to both stability and performance criteria of plantwide systems are presented in this framework. A key advantage of this approach is that, unlike existing approaches, it effectively handles interactions between unit processes. Such interactions are often responsible for performance degradation and stability problems in plantwide process control. As the proposed approach is based on the decomposition of plantwide systems to interconnected subsystems, it is scalable as much of the complexity of plantwide systems lies in interactions. Thus this approach is both intuitively and practically appealing for application to plantwide chemical processes. Possible extensions of the methodology presented in this paper are also discussed.