Page 481 - Enhanced Oil Recovery in Shale and Tight Reservoirs
P. 481

Air injection                                                445


              Table 13.10 Average (base) parameters used in the Frank-Kamenestskii method.
              d c          A      r       MW      l         Q       E       Ta
              Dimensionless  1/s  kg/m 3  g/mole  W/(m$K)   J/mole  J/mole    C
              0.878        50,000  850    200     2.6       200,000  33,000  80


                 In Abu-Khamsin et al.’s (2001) experiments, the maximum temperature
              increase was about 10 C, indicating spontaneous ignition could not occur.

              The Frank-Kamenestskii’s method is used to check whether spontaneous
              ignition could occur. In their experiments, radial packed reactor was used.
              So d c is 2 according to Table 13.9. Other required data were not reported
              in their paper. Using the base values of other parameters above, the critical
              length L is estimated to be 0.001 m, indicating spontaneous ignition could
              occur. This mismatch between the theory and the experiment could be
              caused by incorrect parameter values used. They reported that the failure
              of spontaneous ignition could be caused by excessive heat loss by sweeping
              gas or running out of LTO fuel. To simulate an excess heat loss, the thermal
              conductivity l is increased by 10,000 times, then the estimated critical length
              is 0.1 m which is several times the diameter of the sand pack. This analysis
              shows that the failure was not caused only by heat loss.
                 Similarly, Jia et al.’s (2012a) experiments are also analyzed using the
              Frank-Kamenestskii method. In their experiments, a cylinder-shaped core
              was used. d c was 2. The activation energy of 26 kJ/mol was reported by
              Jia et al. (2012a). For the rest of the other parameter values, the base values
              above are used. When the thermal conductivity l is increased by 1000 times,
              the critical length is 0.1 m. The above two examples and the base case all
              show that it should be easy to achieve spontaneous ignition. However,
              the reality did not show that. It seems that the Frank-Kamenestkii method
              may not predict the reality. Now we conduct a sensitivity study of each
              parameter in the Frank-Kamenestskii method.
                 The kinetic parameters E and A, thermal conductivity l, and the
              enthalpy Q are studied about their sensitivity on the critical characteristic
              length L c at which spontaneous ignition occurs. The reservoir is assumed
              to be an infinite slab (d c is 0.878) and its temperature is 80 C. The results

              are presented in Table 13.11. The minimum and maximum values of E of
              20,000 and 70,000 J/mole, respectively, are tested. Even when the
              maximum E is used, L c is 0.28 m, indicating that spontaneous ignition can
              readily occur. For the rest of three parameters, their base values are increased
              or decreased by 10 times. The results show that L c hardly changes and it is
   476   477   478   479   480   481   482   483   484   485   486