Victoria Salmerón Marazuela1, Miguel Ortega-Mier1, Eduardo Caro Huertas1 and Mig
Victoria Salmerón Marazuela1, Miguel Ortega-Mier1, Eduardo Caro Huertas1 and Miguel Gutiérrez1
1 Universidad Politécnica de Madrid, ETSII, c\José Gutiérrez Abascal 2, 28006 Madrid, Spain.
Keywords: Air pollution, Automobiles, Data analysis, Environmental labelling.
1. Motivation and objectives
Concern for the environment, and in particular air pollution, is very present today and is reflected in new social behaviours and policies. Following the development of the 2030 Agenda and the Sustainable Development Goals [1] various competent authorities have taken part in the endeavour by imposing new traffic restrictions specific to certain areas or pollution episodes. To this end, the Spanish Directorate General of Traffic (DGT) made a classification of vehicles divided into five categories: four under environmental badges and a fifth one covering the remaining vehicles. The DGT’s initial idea was to base the classification on the identification of the least polluting 50% of the fleet [2], approximately 16 million vehicles. This half would be divided into four categories depending on their motorization. Therefore, cars with alternative engine systems would belong to two categories that would benefit the most from traffic policies.
This led car companies to take advantage of this distribution to produce hybrid or micro-hybrid vehicles that pollute as much as traditionally powered vehicles but enjoy the advantages of an ECO or Zero badge. Moreover, it was expected that, after the recovery of the economic crisis a renewal of the fleet would take place which has not happened. Hence, the growth of the car fleet and the criteria for the choice of the various categories have meant that the effectiveness of this classification has not been optimal.
2. Current Vehicle Fleet and Proposed Classification
Within the statistical analysis of the current fleet, two very important phenomena were found: progressive ageing and dieselization. Purchased vehicles before 2000 represent only the 19% of the fleet, but still produce 54% of the particulate matter emitted and 34% of NOx. A constraint for this study has been the provision of data by the Spanish government on the emissions of vehicles for sale in Spain. For this reason, the analysis of the fleet’s polluting power was carried out using general national data, but for the specific study of the badges, data from vehicles sold in the UK in 2019 was used [3]. Although it is true that the fleets are different, the models of the vehicles put on sale that year are practically identical, so this is the best approximation to the Spanish data.
The ineffectiveness of the classification can be seen from the analysis of the vehicles that went on sale in 2019 in the UK [4]. The ECO models are intermingled with those of category C, reaching very high NOx pollutant emission values, basically produced by diesel micro-hybrids.
The reclassification offered by this study was carried out by means of unsupervised statistical learning mode