Lower heating value of jet fuel from hydrocarbon class concentration data and thermo-chemical reference data: An uncertainty quantification

• Published in: Fuel (Guildford) 2022-03, Vol.311, p.122542, Article 122542
• Main Authors: Boehm, Randall C., Yang, Zhibin, Bell, David C., Feldhausen, John, Heyne, Joshua S.
• Format: Article
• Language: English
• Subjects: Alkanes; Aviation fuel; Calorific value; Confidence intervals; Fuels; GC × GC; Heat of combustion; Heat of formation; Hydrocarbons; Jet engine fuels; Lower heating value; Mathematical analysis; Sustainable alternative fuel; Thermo-chemical reference data; Uncertainty
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2021.122542

•A detailed LHV determination method with uncertainty quantification is reported.•Determinations leverage a fundamental approach with GC × GC-FID measured class concentrations.•Uncertainty quantifications consider isomeric, GC, and heats of formation variance.•All 17 determinations from this work agree with ASTM D4809 measurements.•The ASTM D3338 estimation method agrees poorly with ASTM D4809 for alternative fuels. A detailed assessment is presented on the calculation and uncertainty of the lower heating value (net heat of combustion) of conventional and sustainable aviation fuels, from hydrocarbon class concentration measurements, reference molecular heats of formation, and the uncertainties of these reference heats of formation. Calculations using this paper’s method and estimations using ASTM D3338 are reported for 17 fuels of diverse compositions and compared against reported ASTM D4809 measurements. All the calculations made by this method and the reported ASTM D4809 measurements agree (i.e., within 95% confidence intervals). The 95% confidence interval of the lower heating value of fuel candidates that are comprised entirely of normal- and iso-alkanes is less than 0.1 MJ/kg by the method described here, while high cyclo-alkane content leads to 95% confidence bands that approach 0.2 MJ/kg. Taking a possible bias into account, the accuracy and precision of the method described in this work could be as high as 0.23 MJ/kg for some samples.