Page 117 - Formation Damage during Improved Oil Recovery Fundamentals and Applications
P. 117

Formation Damage by Fines Migration: Mathematical and Laboratory Modeling, Field Cases  99


              Table 3.2 Rock properties
              Initial permeability  Porosity  Diameter  Length   Pore volume
              (mD)              (%)       (mm)         (mm)      (mL)
              235               38.54     38.04        46.00     20.14


              suspended concentration is zero, and the strained concentration is con-
              stant with time. As the core outlet is defined at point X 5 1, the time of
              stabilization, where the core contains no suspended particles, will be at
              T 5 1/α. At this time, the distribution of strained particles, and conse-
              quently the impedance, will be constant with time. By Eqs. (3.43 and
              3.45), both the strained concentration and the impedance grow mono-
              tonically with time until stabilization.

              3.3.4 Analysis of laboratory data

              A laboratory coreflood on an artificially prepared sand-kaolinite core was
              performed to simulate permeability decline due to fines migration at high
              velocities. The core was comprised of 10% kaolinite and 90% sand by
              weight. The properties of the artificial core are presented in Table 3.2.
                 The core was compacted in a core holder to create a reproducible sample
              with stable permeability. Pressure drop across the core and outlet suspended
              concentration were measured during the injection period. A full description
              of the experimental design has been presented by Russell et al. (2017).
                 The core was flooded with a sodium chloride solution with ionic
              strength of 0.01 M. Experimental results with the fitted model for a single
              injection cycle are shown in Fig. 3.12. The injection rate prior to this
              injection cycle was 40 mL/min (Superficial velocity, U 5 5.869 3 10 24
              m/s). Increases to velocity prior to this point resulted in negligible per-
              meability decline. The data shown demonstrates the response of the core
                                                                      24
              to increasing the injection rate to 50 mL/min (U 5 7.336 3 10  m/s).
              The result of changing the velocity was a reduction in permeability from
              235 mD to 219 mD. The results are presented in the form of the dimen-
              sionless pressure drop, or the impedance, and the accumulated outlet con-
              centration, which is defined as:
                                          1
                                    8
                                    >  2   ð exp 2αΛTÞ 2 1Þ; T , 1=α
                                               ð
                                    >
                       ð T          >     Λ
                                    <
                C acc 5 α  C 1; yÞdy 5                                : (3.56)
                           ð
                        0           >    2  1  exp 2ΛÞ 2 1Þ;  T . 1=α
                                    >        ð   ð
                                    >      Λ
                                    :
   112   113   114   115   116   117   118   119   120   121   122