Online Program

Return to main conference page
Friday, October 4
Fri, Oct 4, 11:45 AM - 1:15 PM
Evergreen I
Speed Session 3

Chemical Reactive Stability Modeling Using Bayesian Analysis of Differential Equations (306570)

*Fangyi Luo, Procter & Gamble 

Keywords: Chemical Reactive Model, Product Stability Prediction, Differential Equations, Bayesian Analysis

Formulated Consumer Products may experience stability problems over time caused by chemical reactions (e.g. degradation of active ingredients, color change, etc.). Chemical reactions can be typically described by systems of differential equations. We have developed a chemical reactive stability model using Bayesian analysis of differential equations on a simple chemical reactive stability problem where data were collected over time from accelerated stability testing. Using this approach, we were able to predict long-term (3 year) product stability degradation and failure risk with uncertainty estimate. This method can be applied to many other chemical reactive stability problems in formulated consumer products.