The use of pure vegetable oils in Micro Gas Turbine can damage the injection system or the combustion chamber causing undesired vibrations. An ideal solution would consist in having an available tool able to forecast and/or follow in real time the vibrational state of the combustion device. The present paper describes tests performed on a low emission Micro Gas Turbine for power generation, fueled with different liquid fuels, including commercial diesel oil and its blends with pure rapeseed oil. A particular attention was paid both on the emissions and on the measurements of the micro vibrational distributions and their correlation under the different fueling conditions using a new signal processing based on a nonlinear method and chaos analysis. We observed that the overall behavior of the MGT fueled with the blends was good, and the emission concentrations of CO, NOx and Total Particle Matter were comparable to the pure diesel oil ones. Moreover, the chaos analysis and the proposed methodology came out as a possible tool for the real-time characterization of the combustion process of the MGT and to individuation of the fuel supplied.