HTW-KI-Werkstatt/FuelRamanSpectraHandheld

Dataset Overview

This dataset contains Raman spectra for the analysis and prediction of key parameters in commercial fuel samples (gasoline). It includes spectra of 179 fuel samples from various refineries.

The dataset is designed for training models, such as Partial Least Squares (PLS) regression, to quickly and easily determine critical fuel characteristics like the Research Octane Number (RON) and the content of oxygenated additives, without the need for time-consuming standard laboratory procedures. The spectra here were recorded with a handheld the handheld Raman spectrometer.

Target Parameters and Concentration Ranges

The dataset contains measured Raman spectra correlated with reference values for the following comprehensive list of key parameters, determined according to respective norms (ASTM/DIN):

• Research Octane Number (RON)
• Motor Octane Number (MON)
• Ethanol Content (%)
• Ethyl Tert-Butyl Ether (ETBE)
• Methyl Tert-Butyl Ether (MTBE)
• Density at 15°C
• Water Content
• Oxygenates Content
• Oxygen Content
• Olefins Content
• Aromatics Content
• Benzene Content

Reference values for RON were available for all 179 samples, while other parameters (including various additive contents) were available for subsets of the samples (e.g., 164 samples for some additives).

Data Acquisition

Raman spectra were recorded using a handheld Raman (h-Ram) spectrometer IDRaman mini 2.0 (Ocean Optics, Dunedin, FL, U.S.A.).

• Instrument: IDRaman mini 2.0
• Laser Wavelength: 785 nm
• Resolution 13 cm^{-1}
• Spectral Range: The raw data ranged from 2300 to 400
• Sample Preparation: fuel was analysed in 2 mL glass vials (Ocean Optics, Dunedin, FL, U.S.A.)

Citation

Users should cite the two original publication when using this dataset.

Full Citation:

• Legner et. al. http://dx.doi.org/10.1021/acs.energyfuels.9b02944
• Voigt et. al. https://doi.org/10.1016/j.fuel.2018.09.006