The energy storage systems of grid-connected electric buses (eBuses) equipped with Vehicle-To-Grid technology are a valuable source of flexibility that the power system operator can manage to mitigate the impacts of grid contingencies and the randomness of renewable power generators. In this context, many market schemes have been recently deployed to enable these flexibility sources to offer ancillary grid services in the electricity markets, and a large number of optimization methods have been proposed for flexibility resource allocation, and for defining strategic bidding strategies. However, after the ancillary service market settlement, the eBuses owner is compelled to adapt the charging/discharging profiles of all the grid-connected eBuses according to the system operator's request. In this paper, this scheduling problem is formalized by a multi-temporal mixed-integer programming problem, which aims at maximizing the welfare of all the grid-connected eBuses. The insight is to define proper utility functions describing the cost/benefit of the charge/discharge power profiles in function of the eBus parameters and its state of charge and to propose proper linearization techniques enabling the prompt solution of the welfare maximization problem by conventional mixed integer linear programming solvers. Detailed simulation results obtained on realistic case studies are presented and discussed to prove the effectiveness of the proposed methodology.