To solve the problem of reserve transfer caused by the wind power，a novel optimization method of reserve capacity allocation is proposed with market fairness being considered.It takes into account of unavoidable wind power and load forecasting error as well as significant economic issues，such as coal saving benefit and power grid loss factor.An integrated term including coal consumption is built as objective function.A set of consistency constraints of reserve capacity are derived and integrated into a nonlinear dynamic optimization model.A series of performance indices such as effective adjustment rate，dispatchable level and warning indicator are established to illustrate the feasibility and efficiency of the algorithm by comparing the pre-defined system scenarios with and without consistency constraints.IEEE-RTS24 system is used as numerical cases to demonstrate that the algorithm has satisfactory performance in reducing reserve uncertainty and dispatch frequency，and improving operation conditions，market fairness and availability of decision support for dispatchers.