For the experiments with
added hydrogen, the measurements grow closer to the predictions
of the GRI and DRM for decreasing temperatures. For the three
models, six of the seven most influential reactions involve methane
and the free radical methyl. As a second step, series of experiments
involving the lean combustion of H2; H2—CO; H2—CO—CO2;
H2—CO—H2O; H2—CO—CH4 and H2—CH4—CO—H2O—CO2 were
simulated for three pressure regimes p ¼ 1:7 atm, p ¼ 12:5 atm
and p ¼ 32 atm. Under these conditions, the three models always
overpredicted the delay for the onset of the ignition. The effect of
the temperature on the goodness of fit was different for the three
pressures. For p ¼ 1:7 atm, the agreement improves with increasing
temperatures, for p ¼ 32 atm it becomes better with decreasing
temperature whereas it is not considerably influenced by the temperature
for p ¼ 12:5 atm. For understanding the cause of this feature,
sensitivity analyses were performed for the bio-syngas blend
H2—CH4—CO—H2O—CO2. This led to the discovery that for low
pressures and temperatures as well as high pressures and
temperatures, the OH-profile is sensitive to five reactions which
are unimportant for low pressures/high temperatures and high
pressures/ low temperatures where the match is satisfying,
thereby leading to the suspicion that their kinetic coefficients
might not be optimally known. Four of the five reactions involve
the HO2 free radical, the other one being CH3 þ H þM ! CH4 þ M