A mechanistic and probabilistic model estimating micronutrient losses in industrial food processing : Vitamin C and canned green beans, a case-study

Clémence Rigaux (1) , Stéphane Georgé (2) , Isabelle Albert (1) , Catherine Renard (3) , Frederic Carlin (3)

1 MIA-Paris - Mathématiques et Informatique Appliquées
2 Centre Technique de la Conservation des Produits Agricoles
3 SQPOV - Sécurité et Qualité des Produits d'Origine Végétale

Changes in vitamin C concentration along the processing chain of canned green beans were predicted with a probabilistic and modular process model using vitamin C concentrations in raw green beans, constants of chemical reactions (activation energy, reaction rates and diffusivity) and process descriptors (time and temperature, dissolved oxygen concentration). The model accounts for the statistical uncertainty and/or variability of these quantities. The initial vitamin C concentration in green beans, partly established with n ¼ 65 proper vitamin C assays, was (mean [variability interval at 95%]) 17 mg vitamin. 100 g1 [4.1, 30.3] and markedly decreased to 7.5 mg 100 g1 [0.6, 14.9] after blanching, to 6.2 mg 100 g1 [0.3, 12.5] after sterilization, and to 2.3 mg 100 g1 [0, 5.6] after storage. The model predictions were globally on agreement with independent observations that include specific vitamins C assays (n ¼ 26) at different steps of processing.

Mots clés: process risk model; model validation; scenario analysis; sensitivity analysis; uncertainty; ascorbic acid

Domaines: Ingénierie des aliments

Clémence Rigaux, Stéphane Georgé, Isabelle Albert, Catherine Renard, Frederic Carlin. A mechanistic and probabilistic model estimating micronutrient losses in industrial food processing : Vitamin C and canned green beans, a case-study. LWT - Food Science and Technology, 2016, 69, pp.236-243. ⟨10.1016/j.lwt.2016.01.051⟩. ⟨hal-01275142⟩