A generation and storage expansion co-optimization model is formulated as an MILP. This expansion model is characterized by the inclusion of unit commitment modeling, such as the minimum run requirement, the ramp-rate limit, minimum up/down time and start-up/shut-down costs. Storage operation considering energy and power capacity limits and energy losses are also modeled. Storage and generation expansion is modeled as continuous variables, and an operating reserve requirement is included. The new formulation is tested in the IEEE RTS-24 test system, and the numerical results show the importance of considering unit commitment in generation and storage operation expansion modeling. Furthermore, the complementary and substitution relationships between intermittent resources and storage are quantified using the model; sometimes storage enhances the values of wind, but in other circumstances the reverse is true.