The increasing penetration of distributed photovoltaic (PV) generation systems due to the expansion of renewable energy has posed challenges to power grid operators, particularly in managing the uncertainty caused by behind-the-meter (BTM) PV systems. These systems are unmeasured, complicating gross load estimation and grid management. To address this issue, a novel disaggregation method utilizing net load measurements and partially aggregated advanced metering infrastructure (AMI) data is proposed. The proposed method uses a weighted least squares (WLS) method by assigning higher weights to nighttime errors to mitigate daytime errors caused by PV generation. Additionally, a first-order time-scaling method is introduced to correct temporal discrepancies caused by partially aggregated AMI data. The proposed method is validated using data provided by the Korea Electric Power Research Institute (KEPRI).