Partial torque ride through with model predictive control

Symmetric and asymmetric dips of the grid voltage pose serious problems to gas compression stations powered by drives such as load commutated inverters (LCI). Drive control systems used in industrial practice are not capable to handle reduced grid voltage situations appropriately, and execute a ride-through procedure instead during which no drive torque is provided by the drive. Without drive torque compressors may quickly enter surge conditions, under which the gas flows rapidly back and forth, causing wear and risking damage to the equipment. In this paper we describe a novel control approach developed for load commutated inverters based on model predictive control (MPC). Model predictive control is an optimization-based control method, where a mathematical model of the system is used to determine control inputs which are optimal with respect to some objective function. With the revised control system, the drive is capable to provide partial drive torque during grid disturbances; thus resulting in robustness improvements for electrically-driven gas compression stations. In the case of a voltage dip, the compressor is still supplied with partial drive torque, decreasing the probability of the compressor diverging into surge. The paper includes experimental results executed on two real 41.2 MW LCI-fed synchronous machines each powering a gas compressor.