Cost of electricity (COE) is the most widely used metric to quantify the cost-performance trade-off involved in comparative analysis of competing electric power generation technologies. Unfortunately, the currently accepted formulation of COE is only applicable to comparisons of power plant options with the same annual electric generation (kilowatt-hours) and same technology as defined by reliability, availability and operability. Such a formulation does not introduce a big error into the COE analysis when the objective is simply to compare two or more baseloaded power plants of the same technology (e.g., natural gas fired gas turbine simple or combined cycle, coal fired conventional boiler steam turbine, etc.) and the same (or nearly the same) capacity. However, comparing even the same technology class power plants, especially highly flexible advanced gas turbine combined cycle units with cyclic duties, comprising a high number of daily starts and stops in addition to emissions-compliant low-load operation to accommodate the intermittent and uncertain load regimes of renewable power generation (mainly wind and solar) requires a significant overhaul of the basic COE formula.
This paper develops an expanded COE formulation by incorporating crucial power plant operability and maintainability characteristics such as reliability, unrecoverable degradation, and maintenance factors as well as emissions into the mix. The core impact of duty cycle on the plant performance is handled via effective output and efficiency utilizing basic performance correction curves. The impact of plant start and load ramps on the effective performance parameters is included. Differences in reliability and total annual energy generation are handled via energy and capacity replacement terms. The resulting expanded formula, while rigorous in development and content, is still simple enough for most feasibility study type of applications. Sample calculations clearly reveal that inclusion (or omission) of one or more of these factors in the COE evaluation, however, can dramatically swing the answer from one extreme to the other in some cases.