Bidirectional electric vehicles can provide fast active-power flexibility to renewable-rich microgrids, but their repeated use for frequency support depends on both technical feasibility and the economic treatment of vehicle owners. Most EV-based frequency-control studies coordinatethe fleet as an aggregate controllable resource, whereas most local energy-market studies focus on energy trading, tariff design, or bill reduction without explicitly representing short-term frequency-service quality. This paper develops a market-aware predictive dispatch framework in which prosumer-owned EVs are remunerated for local active-power frequency support. A supervisory model predictive controller computes the aggregate regulation request from forecasted PV–load imbalance, a real-time correction layer adapts the request using frequencyand fleet-state information, and a projection stage converts the corrected request into feasible EV-level commands under charger, state-of-charge, availability, and departure constraints. The proposed economic layer introduces an internal regulation-service settlement signal that values upward and downward active-power support according to local service need and EV flexibility availability. Payments are calculated from the feasible EV-level action actually delivered, rather than from the unconstrained aggregate request. The framework is assessed using measured irradiance, temperature, and load profiles as operating inputs, while the frequency trajectory is obtained from the reduced-order microgrid model. The analysis quantifies the trade-off among frequency quality, battery throughput, prosumer remuneration, payment distribution, and settlement cost per delivered frequency-support service.