Abstract

Purpose: Among the Sustainable Development Goals (SDG), the global warming and malnutrition relationship relies on food consumption patterns. It calls for a decisive change in food consumption patterns by offering nutritionally adequate, acceptable, affordable, and environmentally friendly menus in massive food services. Therefore, the research question can be defined as: Are there menus considering several servings, nutritionally adequate, acceptable, affordable, and environmentally friendly? Methods: To address this problem, we focus on the dilemma between minimizing the equivalent carbon dioxide (CO2 eq) emitted in the entire production process of a menu, the monthly acquisition cost, and the selection of the most nutritionally adequate menu possible. Specifically, we formulated three multi-objective quadratic mixed integer programming (QMIP) models: (i) minimization of CO2 eq emissions and the selection of the most nutritionally adequate menu possible, (ii) minimization of the monthly acquisition cost of the system and the selection of the most nutritionally adequate menu possible, and (iii) the weighting of CO2 eq emissions, the monthly acquisition cost, and the selection of the most nutritionally adequate menu possible. All are subject to the same nutritional constraints, operational requirements, and cultural acceptability. Chile is the case study to illustrate the proposal’s usefulness, which is one of the wealthiest countries in the Global South with the highest prevalence of overweight. In addition, Chile has addressed various international initiatives, where greenhouse gas reduction is one of the most important commitments. Consequently, alternatives to change food consumption patterns are required. We selected a prestigious Chilean public university in Santiago, which houses all the communities on a single campus. The university offers a daily lunch menu consisting of a starter, main dish, and dessert, considering a set of nutritional aspects: proteins, lipids, carbohydrates, saturated fatty acids, fiber, cholesterol, sodium, and calories. Results: As a result of the 6-month optimization, minimizing acquisition costs proposes cheaper menus with higher CO2 eq emissions than those proposed by the emission minimization. However, when both objective functions are weighted by 50%, results reach saved costs equivalent to 53,030 USD and emission reduction by around 94,900 kg CO2 eq in the 6 months, compared to the data the menus delivered by the university. Conclusions: It shows that, despite assessing conflicting objectives, it is possible to design nutritionally adequate, acceptable, affordable, and environmentally friendly menus for the case study presented. Future research considers solution approaches for multi-objective optimization as genetic algorithms and includes maximizing the number of servings to deliver.