A pressing question facing policy makers today in developing a long-term strategy to manage carbon emissions from the electric power sector is how to appropriately balance investment in R&D for driving innovation in emerging low- and zero-carbon technologies with investment in commercially available technologies for meeting existing energy needs. Likewise, policy makers need to determine how to allocate limited funding across multiple technologies. Unfortunately, existing modeling tools to study these questions lack a realistic representation of electric power system operations, the innovation process, or both. In this paper, we present a new modeling framework for long-term R&D and electricity generation capacity planning that combines an economic representation of endogenous non-linear technical change with a detailed representation of the power system. Die model captures the complementary nature of technologies in the power sector; physical integration constraints of the system; and the opportunity to build new knowledge capital as a non-linear function of R&D and accumulated knowledge, reflective of the diminishing marginal returns to research inherent in the energy innovation process. Through a series of numerical experiments and sensitivity analyses-with and without carbon policy-we show' how using frameworks that do not incorporate these features can over- or under-estimate the value of different emerging technologies, and potentially misrepresent the cost-effectiveness of R&D opportunities.