FOWTs pose several control challenges. This work addresses reduced order modelling of FOWT to develop an MPC. They key objective is tackle the platform pitch instability while improving the performance in other KPIs. An existing reduced order model, which considers platform pitch and surge, and rotor speed DOFs is adapted and extended, with tower flexibility. The models are validated with open and closed-loop simulations in HAWC2, using the IEA-15MW RWT and the WindCrete spar-buoy floater. Based on these models, an ADKF is designed to estimate several unmeasured states and the wind speed. The MPC aims at reducing the platform pitch motion and tracking the reference rotor speed utilising collective blade pitch angle. The proposed holistic weight tuning procedure conveys the use of Pareto fronts and metrics from several relevant KPIs to guarantee an effective trade-off. Results show a major effects of the simulated wind conditions on the optimal tuning, therefore the process should account for realistic met-ocean conditions. The final tuning is compared to the baseline PI controller with torque compensation. Significant improvements are observed in terms of platform motion reduction in surge, pitch and roll, improved power quality, and greatly reduced fatigue and extreme loads on the tower base and shaft, under all simulated wind conditions. Limitations on rotor speed tracking near rated wind speed show that further development could lead to increase performance.