Dispersion coefficients for turbulent flow are commonly available, but the common models are not easily implemented for fast computation of dispersion in time-dependent large flow systems. Criteria were developed to aid in the choice of parameters for a model combining plug flow with the N-continuously stirred-tanks model to obtain the best agreement with the pure axial-dispersed plug-flow model, while simultaneously being computationally efficient. Solution accuracy and computational savings were demonstrated for a realistic food industry example. The proposed model includes a method of structuring the discretisation to handle the simulation of pressure drop and momentum balance simultaneously with simulation of dispersion. The criteria were derived from the analysis of Laplace transforms. It was also shown that the proposed model predicts the response to a step change in concentration in agreement with the exact solution of the axial-dispersed plug-flow model. The model was written in the object-oriented language Modelica as an object in a library structure which is being developed to simulate complex liquid food process lines and their control systems. (c) 2007 Elsevier Ltd. All rights reserved.