In this paper, the focus will be on approximating original model of process systems using block-structured models. The context of model reduction is to improve the computational efficiency and simulation time (by reduction in model complexity). Hammerstein structures have been used to approximate the process's mathematical non-linear model. Initially Input-Output Hammerstein structure had been used but the technique is extended to Input-State Hammerstein structure. It is shown that Input-State Hammerstein structure can be derived from Taylor series. Approximation accuracy has been improved by including second order terms. The approximated Input-state Hammerstein block structure model gives good approximation of the original non-linear system. The Input-State Hammerstein structure provides opportunities for model reduction in context of reducing the computational load by order reduction of states and Jacobians. The methodology has been applied to a high purity distillation benchmark and satisfactory results are obtained as far as approximation is concerned. Reduction in states and Jacobian size by 70% is attained.