You are here

A general dynamic model of a complete milk pasteurizer unit with fouling and CIP models integrated

12 September 2019
5:00 pm
San Francesco Complex - Classroom 2


Milk fouling is a common problem in pasteurization process. Periodic cleanings needed to restore the performance of the pasteurizer unit and safeguard the process operation, generate downtime and the need to treat large amounts of waste water. System performance of the unit is dictated by the overall pasteurizer configuration, including not just a plate heat exchanger (PHE) milk heater (which have been studied), but also a holding tube, cooler and energy recuperation via a “regeneration” preheater, which are all tightly integrated. Currently, none of the models available considers fouling-cleaning cycles for the overall pasteurizer unit.

The objective of this research is to develop a general dynamic model of a complete milk pasteurizer unit with fouling and Cleaning-in-Place (CIP) models integrated. First a 2D dynamic and distributed thermal model coupled with a fouling and a CIP model is developed for a full milk pasteurizer unit. This unit consists of three PHEs (heating, regeneration, and cooling), a non-isothermal holding tube, and two non-isothermal tubular connections. The thermal, fouling and CIP models of individual PHEs extend and modify the 2D moving boundary model developed previously1,2. For the tubes, the thermal model of Coletti3, Diaz-Bejarano and Macchietto4 was adapted to account for milk deposition in the tubes. The fouling and CIP models used for PHEs was modified and applied to both holding tube and tubular connections. Finally, the entire unit was modelled by joining the component models via suitable boundary conditions. Additionally, the model is coupled with on-line controllers to ensure the required pasteurization temperature as well as the required storage temperature are met.

The generality and flexibility of the model are demonstrated in two common pasteurization processes: high temperature short time (HTST) and ultra-high temperature (UHT) treatments. The thermal model of the whole pasteurizer unit was first validated against experimental data for a HTST process, with excellent agreement. The evolution, extent and location of fouling, and its impact on the process was then assessed for both processes. For the HTST process, it was observed that deposition mass was not significant for the first 2.1 days. For the UHT process, no experimental data on a configuration and measured performance of the unit was available, so a realistic case was simulated. Results indicate a much more severe fouling in main PHE heater, but also in the regeneration section. The evolution of protein concentration profiles and deposition thickness in various locations were obtained. It is noted that experimental validation is still needed to confirm the prediction of deposition mass. The two applications demonstrate that the model can capture the main features of different pasteurization processes. This modelling approach gives a wholistic view of the operation of the unit, which provide valuable insights into the process and help identify cleaning strategies and schedules. It represents an advance over current systems.

Keywords: food processing, pasteurization, plate heat exchanger, fouling, dynamic model, HTST, UHT


1.        Sharma, A. & Macchietto, S. Fouling and cleaning of Plate Heat Exchangers for Milk Pasteurisation: a moving boundary model. in 29th European Symposium on Computer Aided Process Engineering (2018).

2.        Guan, S. & Macchietto, S. A novel dynamic model of Plate Heat Exchangers subject to fouling. in 28th European Symposium on Computer Aided Process Engineering (2018).

3.        Coletti, F. & Macchietto, S. A Dynamic, Distributed Model of Shell-and-Tube Heat Exchangers Undergoing Crude Oil Fouling. Ind. Eng. Chem. Res. 50, 4515–4533 (2011).

4.        Diaz-Bejarano, E., Coletti, F. & Macchietto, S. A new dynamic model of crude oil fouling deposits and its application to the simulation of fouling-cleaning cycles. AIChE J. 62, 90–107 (2016).

Note: a modified version of this Abstract has been submitted to ESCAPE-30


Mengjia Zhu - Imperial College London