Abstract
Biofuels have always been of a certain interest in the technical-motor field, due to the beneficial effects they can produce when used to fuel internal combustion engines. Liquid biofuels such as biodiesel and bioethanol are being studied in particular applied to Diesel engines, widely used in heavy-duty applications and, therefore, also in agriculture. In this last field, moreover, the interest is even greater, thanks to the possibility to create interesting supply-chain economies and reduce the energy demands of the sector by proposing a partial self-sufficiency.
However, the effects of the many biofuels on the performance and environmental impact of a machine can be different and, sometimes, even antithetical. Hence, the maximisation of biofuels properties requires necessarily the search for a trade-off in their use, e.g. by mixing them in appropriate percentages. A chemical-analytical approach to this problem, even if theoretically possible, could be extremely simplifying at the modelling level and too little generalizable due to the extreme variety of technical solutions (EGR, SCR) adopted in engines and regulations of them. For this reason, the approach to these topic has always been purely-experimental, even if the interpretation of the resulting effects is often difficult and, generally, little predictive. Therefore, we decided to apply the Response Surface Methodology, usually adopted in other areas characterized by very complex phenomena (i.e., in industrial engineering), to process the data collected during some tests at the dyno on a New Holland T4020V tractor. Through this technique we have obtained multi-parameter regression equations that can be useful to describe the tractor outputs (performance curves, pollutants) and identify the optimal fuel-blends composition. Finally, through the same technique, we have found that bioethanol should be added only to blends having a minimum of 8-12% of biodiesel to give advantages in terms of NOx concentration reduction.