In this project I explore different methods for anomaly detection for energy consumption data. The data I use is taken from PG&E (cited in the notebook) and contains the average hourly load profile for a single residential building between 2000 and 2024. Anomaly detection is important because it can help people identify data points to investigate for various malfunctions, which can help prevent similar problems in the future. In the notebook, I implement isolation forest, one-class SVM, LOF, and an RNN model to detect anomalies. I also visualize and discuss the results of each model.