Selecting the most suitable packaging technology for liquid foods using an ANP model

The aim of this paper is to provide a model which allows identifying, for several different liquid foods the most appropriate packaging technology, that is the industrial process by which the product is hygienically coupled with the container. In order to achieve this aim, several factors were examined, such as cost, safety, quality of product and plant reliability, which may directly or indirectly affect the resulting efficiency of the process. The first step of the work was thus to exhaustively investigate commercial liquid foods and to classify them on the basis of their chemical and physical properties, relevant to identify the most suitable packaging technology. Examples of such properties include CO2 concentration, viscosity, presence and size of solid particles, and foaminess of the product during filling operations. Based on the above properties, cluster analysis was exploited to group the products into 4 sets; as a further step, a mathematical model was developed to predict the group membership for a new product, not previously classified. For each group of products, several different filling and packaging procedures are considered as possible alternatives, and pondered based on criteria that emerged as relevant for the industrial feasibility of the process. Analytic Network Process (ANP) is used as a viable decision tool to this extent. Decision matrixes were completed based both on pieces of information available in literature, and interviews with experts. Results of the study provide a quantitative assessment of the most suitable packaging technology and equipment depending on the liquid foods considered.