Process integration in sugarcane biorefineries allows reducing steam consumption. As a consequence, the bagasse surplus can be diverted to second generation ethanol production. Furthermore, sugarcane plants can vary the production of ethanol and electricity, depending on the demand. For those reasons, equipment present in the plant might be required to operate under different conditions. This study presents the energy integration of a sugarcane biorefinery. A Mixed Integer Nonlinear Programming (MINLP) optimization model is proposed to solve the problem of synthesizing a Heat Exchanger Network (HEN) able to periodically operate under the distinct conditions required in the biorefinery, i.e., a multiperiod HEN. For solving the MINLP problem, a hybrid metaheuristic approach was used, which combines Simulated Annealing and Rocket Fireworks Optimization. The proposed strategy achieved lower HEN total annualized cost (TAC) when compared with the project energy integration that is commonly found in Brazilian plants. This reduction in TAC, in particular in utilities demand, allows the surplus bagasse to be available for the most suitable application: produce 2G ethanol or more electricity.