High levels of non-dispatchable photo-voltaic (PV) generation integrated with the grid may disrupt its operation. Aligning the PV panels to follow local building demand can partially elevate this. We investigate the economic viability of alignment for buildings owners and grid utility. A bi-level optimization framework is built to determine the design of a building rooftop PV system. An upper-level optimizer selects panel design parameters to lower either cost or grid electricity imports of the lower-level optimization. The lower-level energy-economical model optimization sizes the panels to minimize building electricity cost given building demand, given grid pricing, available roof area, and other factors. A case study of commercial building archetypes in Vancouver BC finds that at the price of (50% reduction of current prices) PV is profitable. For all archetypes profitable alternative grid-friendly designs were found, however, for some buildings large increases in peak power are observed. For some archetypes, a profitable alternative was found that lowers the overall electricity imports (between 4 and 7%) and limits impacts on peak power exports (below 125% of original peak import). Our future work will look into the uncertainty and sensitivity of the framework, technology improvements, and alternative economical models.