This study investigates an alternative approach to rooftop photovoltaic (PV) deployment aimed at mitigating the "duck curve"—the mismatch between peak solar energy generation and typical electricity consumption patterns. Instead of maximizing overall yield on individual rooftops, the analysis focuses on the temporal distributionof solar production across an urban area. By segmenting rooftops using normal vector clustering and simulating hourly irradiance, each segment's contribution to the daily production curve was evaluated. A curve fitting algorithm was then used to identify and weight segments based on how well they helped flatten the collective production profile. Results from simulations in Gävle, Sweden, show that while significant reductions in total output occur, the energy distribution becomes more balanced, particularly during off-peak hours. This geometry based method requires no additional hardware and can complement more conventional solutions like battery storage, offering a low complexity strategy for improving solar grid integration.