A new method for the analysis of pharmaceutical plant cleaning strategies

Abstract

Plant cleaning in the pharmaceutical industry is an undervalued but critical stage of processing. Cleaning with solvents or other cleaning agents is often the only method capable of removing residual particles. In a typical multipurpose pharmaceutical plant, cleaning challenges can cost companies millions of pounds’ as they must clean plant equipment effectively to satisfy regulatory constraints. Failure to do this right first time can result in missed processing schedules often with financial consequences. Furthermore, cleaning is often only considered once the process chemistry has been optimised. The research presented in this thesis describes a new approach to understanding the science behind cleaning using multivariate data analysis, principal component analysis (PCA). This approach utilises the fact that cleaning agent selection can be determined based on the identification of chemical functional groups and physiochemical properties of pharmaceutical products. Using PCA, a set of products were identified which could potentially be cleaned utilising the same approach. This means that the selection of a cleaning agent can be determined for other products with the same chemical functional groups and physiochemical properties. Adopting this methodology helps decide if the cleaning agent used is appropriate to the process chemistry and therefore cleaning can be carried out right first time and as cost effectively as possible. The findings from this research were developed into a tool and used to support the design of manufacturing processes taking cleaning into account from early stages of development thereby saving time and money during the processing stages. Ultimately, the tool will be incorporated into a suite of original and adapted Britest Ltd tools, entitled Fundamental Understanding of Science and Engineering (FUSE) used to identify, understand and provide solutions for cleaning challenges. The tool was applied to industrial case studies to assess its potential.