The increasingly widespread utilization enables plug-in electric vehicles (PEVs) to provide ancillary service and support economic and secure operation of power systems. This paper proposes a novel model to optimize day-ahead spinning reserve requirement (SRR) considering PEVs' contribution in providing operating reserve. Based on the probabilistic criteria, the cost of expected energy supplied by PEV (EESEV) is formulated. The costs of reserves from thermal generators and PEVs as well as expected energy not supplied (EENS) are considered together to determine optimal SRR. The post-contingency PEV reaction time is taken into account in the formation of EENS and EESEV. The capacities of PEV interruptible charging demand and vehicle to grid service are calculated, respectively, under the conditions of both immediate charging and smart charging strategies. The effects of PEVs on system SRR quantification, unit commitment are comprehensively analyzed using IEEE reliability test system. Numerical results systematically demonstrate the effectiveness of PEVs' participation on the reduction of operation costs and the improvement of power system reliability. Sensitivity analysis of PEV penetration level and compensation cost to PEV owners is compressively conducted to provide meaningful reference for future implementation.