Autonomous vehicle-based storage and retrieval system (AVS/RS) offers considerable flexibility with respect to throughput capacity in the transfer of unit loads in high density storage areas. AVS/RS relies on autonomous vehicles to provide horizontal movement within a tier and uses lifts to provide vertical movement between tiers. In these systems, vehicle blocking delays in the aisles and cross-aisles could significantly impact system throughput and transaction cycle times. In this research, protocols are developed to address vehicle blocking, and a semi-open queueing network model is proposed to analyze system performance and evaluate design trade-offs. A decomposition-based method is used to solve the queueing network and quantify the effect of blocking. This model is adopted to analyze the effect of varying tier configuration parameters such as number of storage locations, depth/width ratio, number of vehicles, and vehicle utilization on blocking delays. These insights are useful for design conceptualization using AVS/RS.