A simple method to compute the controllable set of chemical reaction systems with enzyme concentrations as the control parameters is presented. The method features the fact that the catalyst enhances the reaction rate while not changing the equilibrium, and it enables the efficient computation of the global controllability of linear- and nonlinear models. The method is applied to the reversible Brusselator and the toy model of cellular metabolism. With the metabolism model, we show that the stoichiometric rays is a powerful tool for quantifying the "life-death boundary" of the cell model.