In this paper, we present a nonlinear model based dynamic optimization approach for optimal control for thermo-mechanical pulping (TMP) processes. The method is based on nonlinear model predictive control (NMPC) technique with an economic objective function. In our previous work [1], an economically oriented NMPC (econNMPC) was presented for a two-stage TMP process with high consistency (HC) refiners. The two-stage TMP process was simulated up to the secondary refiner. In this study, we extend the previous study to multiple stages that include low consistency (LC) refining. Here we can exploit the complexity of interactions of multiple stage TMP for both performance and optimal operations. The first process is a conventional refining process with two stages of HC refining with a primary and a secondary refiner, followed by a latency chest and a third-stage LC refiner. The second TMP process has only a HC primary refiner, and the secondary stage HC refiner is replaced by multiple stages of LC refiners after the latency chest. Both TMP schemes are dynamically optimized against disturbances to produce the same target final pulp qualities while minimizing specific energy consumption of the processes. The simulation results show the potential economic benefits of the proposed method, through a reduction in total specific energy, about 24%, for the second TMP process. (c) 2013 Elsevier Ltd. All rights reserved.