Yesterday a number of our US team were in Houston giving a seminar on the application of advanced data analysis methods to get a competitive edge.

Entitled From pain to profit: Find out how world-leading organizations have turned major business and operational challenges into competitive advantages, the seminar gave a background into multivariate data analysis and design of experiments, and how these tools are used in R&D, production and quality control.

Our experts gave detailed examples from industries such as renewable energy, upstream oil and gas, petroleum refining and chemical manufacturing. Examples covered at the seminar included:

Developing models for predicting failure in rotating machinery and turbines: Using data from sensors and other factors (e.g. age of equipment), we were able to predict in real-time when a client’s machinery was likely to break down or require maintenance. This helps significantly reduce maintenance costs, ensure more reliable machinery performance, improve Overall Equipment Efficiency (OEE) and improve the return on investment in plant and machinery.

Blend optimization: Through deeper analysis of raw material properties, a client was able to optimize their blending operations. By combining Near Infrared (NIR) spectroscopy with Multivariate Statistical Process Control, they improved raw material management, process performance, reduced unintended ‘give away’ of expensive components and were able to blend cheaper components into the product while still meeting specification.

Real-time reaction monitoring in a chemical process: A client implemented in situ FTIR monitoring to replace costly and time consuming off-line HPLC analysis and for real-time reaction monitoring. Using powerful chemometrics methods such as PLS regression and Principal Component Analysis (PCA)they were able to develop a model for reaction end-point.

There were many astute and detailed questions from the audience, highlighting the interest and growing awareness of how multivariate data analysis can help resolve many operational challenges.

A big thank you to everyone who attended the event. It was great to meet you and we hope you found the information valuable.

For more information on any of these cases or other examples please contact us.

You can see a selection of slides from the seminar below:

Example Applications of Multivariate Data Analysis

Case1 Overview

Case3 Blending Process Control

Case3 Blending Process Control

Case4 PLS Regression for percentage reagent in reaction

Case4 Important variables for modelling quality

Case4 Project samples from batch6 on PCA model