Blockchain adopts a chain data structure, and the characteristics of blocks that can only be added and cannot be deleted make the total number of blocks accumulate over time, forcing resource-constrained nodes to become degraded nodes in order to alleviate increasingly severe storage pressure. Degraded nodes only store partial blocks, although improving the scalability of blockchain storage and reducing data redundancy will lead to a decrease in data availability. To address the problem of storage scalability, quantitative research is needed on data availability. Based on a summary of the existing definitions of data availability, we propose a definition of data availability for blockchain. By analyzing the data synchronization process and the transaction lifecycle, key factors affecting data availability were extracted, and a data availability measurement model was constructed based on node types. On this basis, a relationship model linking data availability and storage scalability was constructed to find the range of data redundancy that meets the target data availability. The experimental results indicate that the data availability measurement model for blockchain can measure the data availability levels of different scalable storage schemes. The model of the relationship between data availability and storage scalability can guide the setting of data redundancy in scalable storage schemes.