(268a) Fast Computations of Pressure Distributions in Complex Plants

Pressure is commonly assumed to propagete with the speed of sound, which in most cases is, and can be considered fast compared to the dynamics of mass and energy communcations in the plant.

Based on a generic network description we argue analytically for a split of the model into two sections, a event-dynamic momentum dissipation description and a capacity-driven dynamics of the mass and energy distribution in the same plant. The network representation being a comprehensive systematic desription of the plant provides the means of a compact representation of the pressure distribution model. This includes any type of multi-phase fluid transport systems, multi-phase fluid storage, jets of one phase in another and multiple inflow/outflow systems. Thus we cover the whole range of equipment configurations.

The pressure distribution model is constructed both using our visual modelling approach supported by a compact matrix/vector representation of the mechanical energy behaviour of the plant. The result is based on a bi-partite graph representation of the plant model, providing both the capacity event-dynamic behaviour as well as the gradients over the fluid transport systems.

Applications are in generic model-based computational engineering activities and specifically in the analysis and detection of process faults, thus fault detection and fault isolation. The same approach can also facilitate a core component to Hazop analysis in the design phase.