Abstract
The process of optimizing the Brazilian hydrothermal system presents an extraordinary challenge, characterized by its continental scale, stochastic flows, and the non-convex nature of hydroelectric generation. Traditional models, based on aggregations of equivalent reservoirs and the linearization of generation functions, have struggled to provide appropriate operational policies in an increasingly complex energy landscape. Furthermore, due to this lack of adherence, the use of the current model as an electricity price setter and a tool for planning the expansion of the Brazilian electricity sector has been questioned by system participants. The Lynx Project introduces a new hierarchical structure designed to overcome these limitations. Its architecture is based on five premises: 1) individual modeling of each hydroelectric reservoir, using nonlinear generation functions to more accurately capture variations in head; 2) the use of fourth-degree polynomial functions to describe the relationships between head, reservoir area, and stored volume; 3) a multi-objective (cost-risk) optimization mechanism that minimizes both the expected operating cost and its respective standard deviation, solved through implicit stochastic optimization supported by the generation of hydrological scenarios; 4) a sophisticated stochastic input module, LYNX-Series, which employs a Contemporary Autoregressive Moving Average (CARMA) model and an original sampling method based on the Mahalanobis distance to select the most hydrologically relevant scenarios in order to reduce computational effort; and 5) an integrated Dynamic Optimal Power Flow (DOPF) model that verifies the feasibility of power dispatch in relation to the physical constraints of the power grid. The main contributions of the Lynx Project include the generation of a set of more appropriate operational policies reflecting different levels of risk aversion, as well as the creation of a powerful analytical tool capable of supporting high-risk strategic decisions in the Brazilian energy sector, including market restructuring and policy analysis.