Page 450 - Enhanced Oil Recovery in Shale and Tight Reservoirs

Air injection 417

are measured. By comparing these two profiles, the temperature intervals of
the exothermic and the endothermic process can be identified. Relevant
kinetic data can also be estimated through analyzing SBR data.

13.3.5 A simulation approach to estimate kinetic parameters
and heating values
Based on the Arrhenius equation, TG and DSC experimental data should
present a straight line with the slope proportional to the activation energy
and the intercept representing the frequency factor. However, actual
experimental data show straight lines in several temperature ranges, as shown
in Fig. 13.12 for one oil sample, because different reactions occur in different
temperature ranges. In practice, a straight line for the whole tested temper-
ature range is generated. Obviously, it is an approximation and it is assumed
a single reaction occurs. Sakthikumar et al. (1995) observed that the activa-
tion energy determined from ARC for LTO could not be used to forecast
core flood kinetics. To improve this method, a simulation approach was
proposed (Guitirrez et al., 2012) and extensively used by Huang and Sheng
(2017a,b,c). In a simulation model, it is possible to define several different
reactions in same temperature range, and/or define one reaction covering
a large temperature range depending on the values of kinetic parameters
(activation energy, frequency factor, and reaction enthalpy), as these kinetic
parameters control the reaction rates in different temperature ranges. Such
simulation model describes the reactions that actually happen in a reservoir.
The description of the reactions is achieved by using several sets of the
keywords of STOREAC to input stoichiometric coefficients of reacting
components, STOPROD to input stoichiometric coefficients of produced
components, EACT (activation energy), FREQFAC (frequency factor),
and RENTH (Reaction enthalpy) in CMG’s (2016) STARS, with one
set being for one reaction scheme. The activation energy gives the
dependence of reaction rate on grid block temperature. For chemical reac-
tions (e.g., combustion), it is positive, that is, the reaction rate increases with
increasing temperature; the reaction rate is independent of temperature
when EACT is zero. To avoid too low or too high reaction rates, two
keywords “RTEMLOWR” and “RTEMUPR” in CMG STARS are
used to set the two temperature limits. If the reaction temperature is lower
(higher) than the RTEMLOWR (RTEMUPR), the reaction rate is set to
the one calculated at the temperature RTEMLOWR (RTEMUPR).
Sometimes a modeled reaction has different rate responses at different
temperatures, in which case a plot of natural logarithmic of reaction rate