The trade-off between costs and carbon emissions from lot-sizing decisions

Logistics decisions can have a significant impact on carbon emissions, a driver of global warming. One possible way to reduce emissions is by adapting a lower delivery frequency, which enables better vehicle utilization or the usage of relatively effcient large vehicles. We study the situation in which a decision maker decides on the amount to be shipped in each period, where he/she can order items in each period and keep items on inventory. If the shipped quantity is large, vehicle capacity is well utilized, but many products have to be stored. Existing studies in this field of research, called lot-sizing, have introduced models for incorporating carbon emissions in the decision making, but do not focus on realistic values of the emission parameters. Therefore, we conduct a survey of empirical studies in order to establish the possible marginal emissions from holding inventory and performing a shipment with a truck. We consider a case study based on real-life considerations and on the findings of the survey study, and introduce a novel bi-objective lot-sizing model to find the Pareto optimal solutions with respect to costs and emissions. In our initial experiments, we consider various demand scenarios and other relevant factors, such as product properties and driven distances. We find that it is often costly to reduce carbon emissions from the cost optimal solution, compared to carbon prices in the market. The cases in which carbon emissions can be reduced most cost-effciently are those in which carbon emissions are large relative to costs, typically because costs are the results of past investments and can be considered sunk.