While solar PV generation is well-established on single-family houses, there is still a lack of installations on apartment buildings. To understand the effect of sharing distributed generation, we developed two energy sharing models: 1) a welfare optimization, and 2) a game theoretical (bi-level) model. We introduced two type players: 1) the owner of distributed generation (e.g., solar PV and energy storage), and 2) the consumers. Furthermore, we included consumer preferences by multiple objectives such as emissions reduction and distributed generation in addition to cost in the model. We applied both models to a numerical example using data from the electricity market in Texas, USA. The results showed that welfare is maximized in both models, but shared differently between the owner of the generation and the consumers. One exception is the bi-level model with uniform price auctions which results in a reduction in system welfare to maximize owner revenues.