Modelling food logistics networks with emission considerations: The case of an international beef supply chain

Highlights

• We present a model for decision makers who are concerned with logistical network problems of perishable products under emissions consideration.

• We consider road structure, vehicle and fuel types, loads, distances and return hauls while integrating emissions into a MOLP model.

• We present the applicability of the model in a beef logistics network based on real data, multiple scenarios, and analyses.

• The Pareto frontier shows how much cost to bear to reduce emissions to different levels.

• Change in fuel efficiency of trucks results in shifts of the Pareto frontier with increase in logistics cost and emissions.

Abstract

Intrinsic characteristics of food products and processes along with growing sustainability concerns lead to the need for decision support tools that can integrate economic considerations with quality preservation and environmental protection in food supply chains. In this study, we develop a multi-objective linear programming (MOLP) model for a generic beef logistics network problem. The objectives of the model are (i) minimizing total logistics cost and (ii) minimizing total amount of greenhouse gas emissions from transportation operations. The model is solved with the $\epsilon$-constraint method. This study breaks away from the literature on logistics network models by simultaneously considering transportation emissions (affected by road structure, vehicle and fuel types, weight loads of vehicles, traveled distances), return hauls and product perishability in a MOLP model. We present computational results and analysis based on an application of the model on a real-life international beef logistics chain operating in Nova Andradina, Mato Grosso do Sul, Brazil and exporting beef to the European Union. Trade-off relationships between multiple objectives are observed by the derived Pareto frontier that presents the cost of being sustainable from the point of reducing transportation emissions. The results from the pie chart analysis indicate the importance of distances between actors in terms of environmental impact. Moreover, sensitivity analysis on practically important parameters shows that export ports' capacities put pressure on the logistics system; decreasing fuel efficiency due to the bad infrastructure has negative effects on cost and emissions; and green tax incentives result in economic and environmental improvement.

Introduction

The progressive increase of food consumption due to growing world population and wealth stimulates higher food production. A recent way for managing the increased production is globalization of food supply chains (FSCs) with the help of improvements in transport technologies, cheaper transportation, reductions in tariffs and other barriers to trade. Globalization has improved the chance of profitability from cross-border operations as well; however it has led to increased distances between partners in supply chains (Elhedhli and Merrick, 2012). The increased distances have enhanced the strategic importance of logistics network decisions such as selection of suppliers, distribution channels and transportation modes, determining production and inventory amounts at each plant and allocation of products (Cordeau et al., 2006, Harris et al., 2011). The need for a well organized logistics network thus has increased in the food sector, which is producing more than ever on a global scale.

Traditional logistics management considers mainly two key logistical aims, cost reduction (efficiency) and improved responsiveness, while dealing with the logistics network problem. However, intrinsic characteristics of food products and processes such as product perishability and food quality, and a growing sustainability trend require extension of the key logistical aims with quality and environmental considerations. This necessity leads to the need for decision support tools that can integrate economic considerations with quality preservation as well as environmental protection in FSCs. Accordingly, literature review shows that there is a need for models that are able to deal with the key challenges in managing quality and sustainability (Akkerman et al., 2010, Soysal et al., 2012). The need in practice and in research forms our main motivation to develop a model that allows to consider perishability of goods and emissions from transportation operations along with cost concerns in food logistics network.

We take the beef sector as a representative of a food supply chain that has both food and environment related challenges. Shelf life for beef that includes several quality factors (e.g. juiceness, tenderness, nutritive value, appearance and palatability) puts an additional pressure on logistics decisions, since the product may become undesirable, even it is not unsafe (Delmore, 2009). Apart from quality concerns, appreciation grows for the idea of a carbon-constrained economy in the livestock sector with the growing awareness towards environment conservation (Robinson et al., 2011). Especially, transportation is one of the main sources of livestock related carbon dioxide (CO₂) emissions (Delgado et al., 1999) and increasing global beef trade results in more fuel consumption for beef related transportation. Therefore, it is wise to address product perishability and emissions from transportation while managing beef logistics chains.

We develop a multi-objective linear programming (MOLP) model for a generic beef logistics network problem. The objectives of the model are (i) minimizing total logistics cost and (ii) minimizing total amount of greenhouse gas (GHG) emissions from transportation operations. Duration of inventory keeping is limited due to the perishability nature of the product. The environmental effect of freight transportation is measured in CO₂ emissions. We provide a case study of the international beef logistics chain operating in Nova Andradina, Mato Grosso do Sul, Brazil and exporting beef to the European Union (EU) to illustrate the applicability of the proposed model for real logistics systems. The rationales for the selected beef chain are the following: (1) Brazil ranks as the largest beef exporter in the world by holding an approximately 21% share of the global beef trade in 2011 (Abiec, 2012d), (2) Brazil has potential to keep its position in the global market, and (3) Beef trade relationship exists (47,693 ton for fresh-chilled beef in 2010) between Brazil and the EU (Abiec, 2012a, Abiec, 2012b). In this case study, we put the main focus on road transportation, which is the only delivery option till the export ports, as rail, inland ship or air need infrastructure that is not available yet. The logistical challenges in the case of Brazil are mainly related to usage of old trucks, inefficient road infrastructure or deficiency of available trucks.

The structure of the remaining paper is as follows. Section 2 presents a literature review on logistics models that take product perishability and/or emissions into account. Section 3 presents a formal definition of the generic problem, the methodology used for emission estimations and the proposed MOLP model for the generic beef logistics network problem. Section 4 presents the case study description, data gathering and the computational analysis of the model. The last section presents conclusions and directions for further research.