Comparing emission estimation models for rail freight transportation

Highlights

• Review of five models to estimate emissions from rail freight transportation.

• Supports the evaluation of the eco-friendliness of rail freight transportation.

• Models: MEET (empirical and steady state), ARTEMIS, EcoTransIT World, and Mesoscopic.

• Experiments demonstrate each model’s results for varying train and trip parameters.

• Python implementation of the models is available in the supplementary data.

Abstract

This study reviews emission estimation models that aim at providing realistic estimates of the emitted greenhouse gases from rail freight transportation. Five models are considered: two models from the MEET project, the ARTEMIS model, the EcoTransIT World model, and the Mesoscopic model. For each of the five models, this paper describes the estimation principles, methodology, and procedure, as well as relevant input parameters. An experimental study demonstrates the impact of train and trip specific parameters on each model’s emission estimate. Results are presented for varying values of a train’s number of wagons, the payload per wagon, the average speed, the trip distance, the number of stops, and the altitude profile along the route. In so doing, given a specific transportation scenario, the paper supports decision makers from industry and researchers to find and apply an appropriate emission estimation model for evaluating the eco-friendliness of rail freight transportation.

Introduction

There is a recent revival of using rail transportation to match the constantly increasing global freight volume (International Energy Agency, 2019). The International Transport Forum (2019) expects global rail freight volumes to grow 2.7% per year between 2015 and 2030, with some countries having significantly higher growth rates (e.g., in 2017, growth rates are 13.3%, 6.4%, 5.5.%, and 5.2% for China, Russia, India, and the USA, respectively). Recent infrastructural developments demonstrate this growth, such as the ‘silk railway’ initiated by the Chinese government to reduce transportation time between China and Europe (The Times, 2020). Further, transportation makes up for about 25% of global energy-related CO₂ emissions (IPPC, 2018, Eurostat, 2020) and the rail mode is often considered as a mean to reduce total emissions from transportation. For example, recent studies show that using intermodal rail/road transportation instead of road-only transportation can reduce the overall transport emissions, e.g., Craig et al., 2013, Kiyota et al., 2015, de Miranda Pinto et al., 2018, Heinold and Meisel, 2018, Heinold and Meisel, 2019. These studies analyze different transport scenarios and apply diverse emission estimation models to measure the environmental impact of the considered modes of transportation. Demir et al. (2011) provide a comparative analysis of several vehicle emission models for road freight transportation. As far as is known, there is no such analysis of emission estimation models for rail freight transportation.

This study fills this gap by reviewing and comparing five emission estimation models for rail freight transportation. The selected models provide an overview of the most common approaches. For this, two models from the MEET project (Hickman et al., 1999), the ARTEMIS model (Lindgreen and Sorenson, 2005a, Lindgreen and Sorenson, 2005b), the EcoTransIT World (ETW) model (EcoTransIT World Initiative, 2019), and the Mesoscopic model (Kirschstein and Meisel, 2015) are considered in this study. In a first step, each model’s estimation principles, methodology, input parameters, and procedures are explained. Then, in a second step, the impact of varying values of a train’s number of wagons, the payload per wagon, the average speed, the trip distance, the number of stops, and the altitude profile along the route is demonstrated in an experimental study. Here, a typical freight train serves as a base scenario and results are presented for each of the considered experimental settings, which results in emission rates for a large amount of potential transport scenarios. Finally, in a third step, this paper discusses scenarios in which it might be appropriate to select one or another of the considered models.

With this, the main contribution of this paper is twofold. Firstly, the models’ detailed description with a consistent notation allows a direct and precise application of the models. All models are implemented in Python and the code is made publicly available as supplementary material. Secondly, the results of the experimental study demonstrate the impact of relevant train and trip specific parameters on each model’s emission estimate. In so doing, given a specific transportation scenario, the paper supports decision makers from industry and researchers to find and apply an appropriate emission estimation model for evaluating the eco-friendliness of rail freight transportation.

The rest of this paper is organized as follows. Section 2 describes general principles of estimating emissions from rail freight transportation. Section 3 describes each of the considered emission estimation models in detail. The experimental study in Section 4 demonstrates the impact of train and trip parameters on each model’s emission estimate. Section 5 discusses the selection of an appropriate model for a given transport scenario. Section 6 concludes this paper.