In zinc hydrometallurgy, an advanced copper removal process purifies zinc sulfate solution through a series of chemical reactions with recycled underflow by using zinc powder in zinc hydrometallurgy. This paper focuses on the kinetic modeling of the competitive-consecutive reaction system in the copper removal process, and proposes an adaptive parameter optimal selection strategy for different industrial conditions. In the system model, copper cementation, one of the removal reactions, is described by a surface controlled pseudo-first-order rate equation; cuprous oxide precipitation, the other removal reaction, is described by a shrinking core model of a noncatalytic fluid-solid reaction. Because there are several kinetic parameters in the system model, parameter estimation plays an essential role. Because of the complexity and variation in the practical removal process, the kinetic parameters are usually sensitive to alterations in the process conditions. This work solves the parameter estimation problem using an optimal selection strategy. In the strategy, the industrial conditions are classified adaptively according to the system model performance, then the kinetic parameters are selected optimally by evolutionary and particle swarm optimization algorithms for different industrial conditions. Three different representative industrial data sets are used to test the effectiveness and flexibility of the proposed modeling and parameter optimal selection approach in various situations. Finally, the kinetic model is applied to the soft measurement of the practical copper removal process with underflow, and the results demonstrate that the model effectively captures the trends of the removal reactions.