Analysis of temperature distribution and
heat penetration data for in-container
sterilisation processes

Historically, heat processes have been designed to assure the safety of the end-product. Heat processes based around a 'botulinum cook' or 'worst case scenario' are undoubtedly safe. However, studies suggest that there is significant scope for reducing the degree of processing through alternative approaches to data analysis - thereby improving process efficiency with concomitant benefits to product quality and nutritional content - without compromising product safety. Despite this, the established methods still appear to be used universally.

This review looks afresh at this issue and specifically sets out to:

• Identify the alternative data analysis methods that have been proposed
• Identify why these approaches have not been adopted
• Propose the best means to initiate appropriate changes

It considers how systematic assessment of the sources of variability in measurements of heat penetration and distribution can be conducted - including variability in the process requirement, process application and the product - and how the resulting data should be analysed to ensure that process modifications have a firm scientific base. It highlights the need for a reduction in the variability in heat penetration and heat distribution data as a first step towards using statistical methods for evaluation of thermal processes.

Contents:

• Introduction
• The nature of the data being analysed
• Allowance for instrumental error
• Alternative approaches to basic data analysis
• Statistical interpretation of process evaluation data
• Statistical modelling of processes
• Conclusions
  - Designing models - simplification of the problem
  - The problem of verification
  - Can industry expect a return on research investment?
• References