Data anonymization is crucial to allow the widespread adoption of data-driven technologies, such as smart meters. However, anonymization techniques should be evaluated in the context of a dataset to make meaningful statements about their eligibility for a particular use case. By comparing characteristics of raw streaming data and its respective anonymization with CASTLE, we assessed the suitability of ks-anonymization for data streams generated by smart meters. Additionally, we investigated the influence of specific parameters on the publication process and on utility metrics like, e.g., the information loss.