Optimized Operation of Plant Using Process Analytic Techniques

Lead Research Organisation: University of Manchester
Department Name: Chem Eng and Analytical Science


This project aims to embed a new generation of digital tools into Unilever's research & development and then into factories. Batch production of personal care products, such as shampoo, currently requires correction stages to account for intangible variations in raw materials, formulation and processing. These correction stages are time consuming and result in lost production time. Raw material quality, compositional and process factors affecting product quality and the need for quality adjustments as a final step of the process will be investigated. State-of-the-art process analytic techniques, supplied by Perceptive Engineering, and material characterisation approaches will be used to understand how product quality evolves during the process. Opportunities for early intervention will be identified to ensure good quality is achieved at the end of the process without the need for a final adjustment step. Predictive batch control will be developed at the lab scale to eradicate the correction stage. Right first time production will be demonstrated at pilot scale.


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509565/1 01/10/2016 30/09/2021
1788740 Studentship EP/N509565/1 06/09/2016 31/03/2020 Stephanie Elizabeth Cunliffe
Description The use of spectroscopic instruments as in-line techniques for the control of shampoo processes has been investigated and determined to be an appropriate and successful technique. In-line measurements of a concentrated surfactant raw material have been shown to be a successful reference for the prediction of the product viscosity.
Off-line analytical methods such as mass spectrometry gas chromatography and NMR spectroscopy have highlighted the main sources of compositional variation between batches of surfactant when different suppliers are used to meet the demand for volume. The compositional variation in the surfactant directly responsible for the manufacturing errors in the shampoo products was identified to be the mass percentage of unreacted alcohol feedstock in the surfactant. This amount varies between suppliers and batches provided by the same supplier.
Near infrared, mid infrared and Raman techniques have been extensively studied and combined with chemometric data analytics to evaluate the mass percentage of unreacted alcohols in a commercial surfactant. The use of spectroscopic techniques to determine surfactant composition within a concentrated surfactant has not previously been reported in the literature. The main finding of this piece of work is that the NIR method combined with data analytics provides better compositional estimations than the MIR and Raman techniques.
The relationship between the alcohol content of the surfactant and the product viscosity has been extensively studied and empiracally modelled to generate a linear relationship between product formulation / composition and viscosity.
The NIR model has been placed into an in-line manufacturing environment, and combined with the viscosity prediction model, has been proven robust for the application of in-line feedforward control of shampoo manufacturing, increasing product success rate from 57% to 90%.
Exploitation Route The main findings of this project can be directly used by Unilever to implement control measures into manufacturing environments. The measurement techniques can be easily implemented into current manufacturing equipment and the chemometric analysis built into existing control systems. This could result in a immediate reduction in manufacturing time. The chemometric models could be taken forward by others to expand the surfactants that can be modelled, increasing the product portfolio that the technique could be used for.
Sectors Chemicals,Manufacturing, including Industrial Biotechology

Description The NIR measurement method, after having been proven a successful technique for quantifying unsulfated alcohol in sodium lauryl ether sulfate, has been installed for a trial period directly into a main plant manufacturing line to collect data on how the composition of raw materials entering the factory are varying. Currently this has only been a data collection exercise and not used for process control, however the outcomes from the installation are expected to be used by the company for decisions on future installation of this technique into other manufacturing facilities.
First Year Of Impact 2018
Sector Manufacturing, including Industrial Biotechology
Impact Types Economic