A Gradient-Based Economic Model Predictive Controller with Zone Control Scheme Applied to Natural Gas Pipeline Networks

This paper addresses a one-layer model predictive control (MPC) strategy that simultaneously deals with safety and economic issues for natural gas networks (NGN). The simulations consider a nonlinear pipeline model based on the non-isothermal flow and non-deal gas behavior. The proposed NGN-oriented MPC strategy uses an adaptive scheme that relies upon the successive linearization of the nonlinear NGN model and the surge prevention constraints of the compression stations, incorporated into the control law to avoid unsafe operating conditions. The controller has the guarantee of feasibility by incorporating a suitable set of slack variables into its formulation, mainly in the surge avoidance constraints. At the same time, the resulting control law is more flexible by adopting output zone tracking cases rather than setpoint tracking. The simulated study, aiming at minimizing the power consumption of the centrifugal compressors, sought to control the pressures in the consumer nodes of NGN into a predefined zone while meeting the process constraints. In all scenarios of zone changes, the controller could lead the controlled outputs in their respective zones, accommodating the operation in steady states with a minimal power consumption of three compression stations considered in NGN. By respecting the surge prevention constraints flexibly, and using the slack variables when necessary, immediately after perturbation, the proposed NGN-oriented adaptive zone MPC controller has proved to be a suitable tool to manage the NGN with control performance, operational safe and economic competitivity.