Are conventional energy megaprojects competitive? Suboptimal decisions related to cost overruns in Brazil

Cost minimization is arguably the most important criterion governing decisions about energy sector infrastructure construction. Usually, a winning project is picked among similar alternatives based on lowest levelized cost of energy, because, ceteris paribus, economies of scale drive down the unit cost of energy delivered. As such, megaprojects - here defined as costing more than a benchmark US$ 1 billion - are perceived as more competitive than smaller-scale options. However, megaprojects are prone to construction cost overruns and delays that, if included ex ante, may change the optimality of decision for a given project. We hypothesize that optimistic assumptions on techno-economic performance of megaprojects favor their inclusion in the solution of integrated assessment models (IAMs), preventing higher shares of non-hydro renewables, energy efficiency and other low carbon options. To test this hypothesis, we ran the COPPE-MSB energy system cost-optimization model for infrastructure expansion. We estimate a factor (named Z factor, for zillions) to determine cost differences both within Brazil and vis-a-vis international parity and adjust the model's parameters for CAPEX and construction times of projects qualifying as megaprojects. Results show decreased coal and increased wind power generation, and a reduction in the number of new refineries leading to higher imports of diesel and gasoline.