Multi-criteria analysis applied to the hydro-chemical classification of bottled waters marketed in Algeria

Introduction/purpose: Mineral and spring waters play a crucial role in daily nutrition due to their taste and health benefits. In Algeria, the bottled water market is rapidly growing, yet there remains a lack of precise, objective systems to classify these waters based on organoleptic qualities and safety. This study aims to develop a reliable numerical model for the hydrochemical classification of bottled waters marketed in Algeria. Methods: A hybrid multi-criteria decision-making approach was employed, integrating the DIBR II method to determine the weights of criteria based on expert judgments and the Bonferroni operator to aggregate these opinions. To effectively manage uncertainty and imprecision in the data, Triangular Neutrosophic Numbers (TNN) were utilized. Subsequently, the CODAS (COmbinative Distance-based ASsessment) method was applied to rank the bottled water brands. The evaluation focused on four key quality criteria: nitrate concentration, hardness, total mineralization, and hydrochemical nature. After constructing and aggregating the neutrosophic decision matrices and performing de-neutrosophication, CODAS computed a proximity coefficient for each brand, resulting in the final ranking. Results: The study identified the top five brands of mineral waters (Daouia, Mansourah, Youkous, Alma, and N'Gaous) and spring waters (Taya, Ariaf, Arwa, Sidi Rached, and Hirouche). Sensitivity analysis confirmed the robustness of the ranking, validating the model's reliability and showed that nitrate content is the most influential criterion, highlighting its key role in assessing water quality. Comparative analysis also demonstrated that CODAS produces consistent and reliable rankings compared to other multi-criteria methods, with good discrimination among closely ranked alternatives. Conclusion: This study developed an innovative and robust numerical model combining DIBR II and CODAS within a neutrosophic framework, providing a reliable hydrochemical classification of bottled mineral and spring waters in Algeria. The model incorporates expert weighting and accounts for data uncertainty to generate a clear and coherent brand ranking. Sensitivity analysis confirmed the stability of the rankings, particularly emphasizing the dominance of the nitrate criterion, demonstrating the model's value as a decision-support tool for various stakeholders in the sector.