In this study, pyrolysis
experiments were conducted with a saturate-rich
Tertiary source rock-derived oil from the South China Sea basin
using a fixed-volume pressure vessel at temperatures from 350 to 425
°C for 24 h (0.92–1.85% Easy R
0) to investigate pressure effects up to 900 bar on the generation
and stable carbon isotopic fractionation of light hydrocarbons in
the C6–C7 range. The results demonstrate
that the pressure retards oil cracking to light hydrocarbons, but
the retardation depends on the thermal evolution. In the peak oil
to early wet gas stage (350 and 373 °C, 0.92–1.15% Easy R
0), the light hydrocarbon generation is low
but it is still suppressed by increasing pressure. In the late stages
of the wet gas window (390, 405, and 425 °C, 1.35–1.85%
Easy R
0), the light hydrocarbon generation
is suppressed significantly from 200 to 470 bar, followed by promotion
and promotion-suppression as pressure is increased up to 900 bar.
Meanwhile, the distributions of branched alkanes, cycloalkanes, and
aromatic hydrocarbons are pressure-dependent. The medium to high pressures
result in increasing Mango K
1 values and
toluene/n-C7 ratios and decreasing n-C7/methylcyclohexane ratios, suggesting that
pressure benefits the occurrence of cyclization and aromatization
during oil cracking, probably involving bimolecular reaction pathways.
Preferential aromatization and isomerization with increasing pressure
lead to significant carbon isotopic fractionations of aromatic hydrocarbons
and branched alkanes as up to 4‰ and 2‰, respectively.
However, stable carbon isotopic compositions of cycloalkanes show
almost no fractionation under pressurized cracking. Therefore, caution
must be taken with respect to the application of light hydrocarbon-derived
parameters in deep petroleum reservoirs usually at high temperatures
and pressures. The carbon isotopes of branched alkanes and aromatic
hydrocarbons could be potential measures to identify the pressure
effects, while carbon isotopes of cycloalkanes could be an effective
index for oil–oil/oil–source correlations.