The grid connectivity requirements imposed today, ask for an accurate controller design able to deal with system constraints and nonlinearities. In order to fulfill grid international standards, this paper details a Finite control set model predictive controller called FCS-MPC as part of controlling an energy conversion system. It uses a discrete time 3L-NPC inverter model (in the α-β frame) to predict the future values of the injected current and DC capacitors voltages. Switching pulses are chosen to ensure the optimum errors between the controlled variables and its desired values. Simulations results for both one-step and two-step horizon have been introduced to verify system robustness and dynamic response against variations of the DC link voltage, AC side current/voltage and reactive power. Low current-distortion levels with an arbitrary power factor prove the system performance when a two-step horizon control algorithm is adopted.