Photovoltaic plants coupled with battery energy storage systems require dispatch strategies that balance market operation, battery ageing, export-profile tracking, and PV forecast uncertainty. This paper presents a mixed-integer linear programming framework for hourly PV-BESS dispatch with grid charging, battery-throughput and curtailment costs, and soft export-profile tracking. Usable capacity is scaled by state of health, charge and discharge efficiencies are derived from round-trip efficiency, and binary constraints prevent simultaneous charge/discharge and grid import/export. The framework is evaluated for the Danfoss Solar Park in Nordborg, Denmark, using a 2.1 MW-peak PV plant, a 4.3 MWh BESS, PVGIS SARAH3 generation data, and DK1 day-ahead prices. Four seasonal days, five battery ageing cases, three PV forecast scenarios, and 6, 12, and 24 h model predictive control horizons yield 200 unique dispatch configurations. Averaged across forecast scenarios, the 6 h MPC yields a 21.28\% lower reported direct operating-cost component than the full-day open-loop optimization benchmark, but exported energy decreases by 457.23 kWh and export-tracking error increases by 502.33 kWh. The results show that short-horizon MPC can reduce the reported operating-cost component under PV uncertainty, but at the expense of weaker export-profile tracking.