This study employs Large Language Models (LLMs) to analyze 8,560 public comments submitted to the Federal Energy Regulatory Commission (FERC) regarding hydropower projects, providing insights into the evolving nature of public opposition to renewable energy infrastructure. Using automated text classification, we estimate a 20% increase in expressions of antagonism toward new hydropower facilities over a ten-year period, with concerns about insufficient community input emerging as the dominant factor driving opposition. Our findings reveal that traditional economic concerns such as property values and transmission infrastructure play surprisingly minor roles in stated opposition, while procedural considerations are prominent. However, results for some LLM annotation tasks vary greatly depending on variation in wording and structure. This highlights credibility concerns for the methodology that has yet to be resolved in the literature.