In this paper, an efficient modeling method for plate heat exchanger (PHE) is developed. Firstly, the mechanistic model is established according to heat transfer equation and heat balance equation. Due to the existence of multiple unknown parameters in the mechanistic model, sensitivity analysis is employed to study the influence of these parameters on the model output. Then, the parameters which have a significant influence on the model output are estimated using real data, and the values of parameters with low influence on the output are acquired from previous works. With PHE running, the performance of PHE deteriorates due to the fouling deposits on plate surface, the corrosion and erosion damages to plates and other unknown factors. To track the changes of outlet cooling water temperature caused by above factors, Gaussian mixture model (GMM) is used to assess the performance of mechanistic model. According to the assessment results, model output offset updating or model parameter updating is activated. Finally, the feasibility and efficiency of the proposed modeling method are demonstrated by applications to three PHEs in a steel plant. The presented modeling method lays an important foundation for energy-saving optimization of circulating cooling water system. (C) 2018 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.