Required accuracy of thermochemical information

Opinions about the accuracy of thermochemical data needed for combustion research are educated guesses rather than facts. Unfortunately, no systematic investigation has been carried out yet on this question. 

The well-determined species, such as those appearing in JANAF tables or API Project 44 tables or those published in the Journal of Physical and Chemical Reference Data, are calculated to the third or fourth digit after the decimal point. In the majority of cases the third digit after the decimal point can be well reproduced but the fourth is sometimes doubtful. However, for most calculations even the second digit after the decimal point is more than is ever necessary. Thus, Cp and S may be known to two digits after the decimal 

One property we would want to know to at least two digits after the decimal point is log Kp. The reason for this is that Kp is the property used to calculate back reaction rate constants when forward reaction rate constants are known or assumed. In chemical kinetics calculations this is of primary importance and a difference of one unit in the second digit of log Kp causes a 2% difference in the value of the calculated reverse reaction rate constant. On the other hand, in the calculation of shock tube experimental properties where reactive species are highly diluted in a noble gas or nitrogen, there is hardly any importance to knowledge of the exact properties of any species other than the diluent itself. 

Even the thermochemical values for well-determined species, such as those appearing in JANAF tables, are changed from time to time as physical constants or atomic weights change or mistakes are found. This should concern combustion science only if the changes affect the second digit after the decimal point. On the other hand, it is important to use reliable sources as much as possible, and if there is a choice not to mix different thermodynamic sources since they tend to differ at temperatures higher than 3000 K, where anharmonicity effects may be important. 

One must also realize that except for elements, stable compounds which burn cleanly in oxygen, and small molecules or radicals whose electronic spectra in dissociative regions have been thoroughly analyzed, standard enthalpies of formation must be evaluated through difficult experiments subject to interpretive difficulties. This means that the largest errors—by far—in thermochemical calculations arise from uncertain enthalpies of formation. Fortunately, it is a fairly straightforward matter to modify the polynomial representations to change enthalpies of formation when new experimental results become available or sensitivity checks on AH 29s e to be made (cf. comments in Appendix C). 

---