Generation capacity expansion planning has been evolving in rencent decades. First, the long-term planning procedure is taking more detailed considerations of short-term operation impacts. Second, as more renewable resources being integrated into the grid, a new strategy of dealing with the non-dispatchable renewable energy should be developed, with more ancillary services needing to be procured from thermal units. These trends are expected to continue. This thesis describes a methodology in generation capacity expansion planning. The screening curves method can be used to estimate optimal generation mix for a target year. This thesis first introduces three screening curves methods, which are classified based on their ability to deal with detailed shortterm operational issues. It then includes ancillary service and wind integration impacts. Finally, it presents a case study of a projected ERCOT 2030 system. / text