Traditionally, the big and easily available oil fields offshore producing from a topside processing plant have already been developed. The oil and gas industry now focuses on increased oil and gas recovery from brownfields and on development of smaller oil and gas greenfields. Subsea processing will be a key factor both for increasing oil and gas recovery on existing fields and for making new, smaller oil and gas fields economically feasible.

Subsea processing plants are small compared to topside processing plants and have fast dynamics and strong interaction between units. This fast dynamics poses challenges with transients and modelling which are important for analysis and design of online multivariable control and optimization. The control oriented models need to be fast and sufficiently accurate to describe the physical properties of the subsea processing plant.

Thus, the objective of this sub-project is to create standardised models for a complete subsea processing plant, analyze and improve interacting control loops and design a multivariable control structure to utilize the strong interactions and fast dynamics to handle transients and optimize production, e.g. linear/ nonlinear Model Predictive Control.

An important research objective will be to find the minimum buffer volume necessary to sufficiently control the process and handle transients with multivariable control. The models to be developed will start from recent separation technology, (e.g. the CompactSep separation system developed by Statoil in collaboration with Chevron, Petrobras, Total and FMC) and extend with boosting and cooling, including actuator dynamics, processing time delay and transmission of signals and power. An intension is that the modelling work should result in a general framework for standardized modelling of subsea processing plants.

A laboratory of several microcontrollers is to be constructed and run with various parts of the complete control model of compact subsea process to study the effect of increased time delay subsea. During the modelling work, focus will be given to numerical solutions of the models for the simulation to be fast enough.