Electric power systems are vulnerable to extensive damage due to hurricanes with most of the damage concentrated on overhead distribution systems. There is evidence that climate change will affect future hurricane patterns. Additionally, wood poles, which are most commonly used in distribution systems, are susceptible to decay. The scarcity of resources and increasing demand for higher reliability warrant the use of optimization techniques for wood pole maintenance planning. This paper presents a framework for optimal maintenance of wood poles subjected to non-stationary hurricane hazard and decay. Maintenance cost, service life, and system performance are considered separately and simultaneously in the optimization. Periodic chemical treatment and repair of decayed poles using fiber-reinforced polymer are considered. The distribution system of a virtual city assumed to be in Florida is used to demonstrate the framework. The results of the single-objective optimization indicate that the objective that maximizes service life resulted in higher optimal maintenance time. However, delaying maintenance will lead to a larger probability of pole failure, higher corrective maintenance cost, and lower system performance. The result of the multi-objective optimization is closer to the result of the cost-based optimization because the cost function is more sensitive to the variation of maintenance time.