Page 231 - Fundamentals of Enhanced Oil and Gas Recovery
P. 231

Waterflooding                                                                       219



                      This correlation was applied in our analysis and assuming @P c =@r is negligible,
                   various capillary pressure was estimated for increasing water saturation. Note that the
                   free water level (FWL) was assumed at a height of 20 ft; hence we could predict our
                   threshold pressure to be
                                                    P d 5 Δρgh                           (7.55)
                                                       3                          3
                      Assuming water density of 1005 kg/m and oil density of 900 kg/m with g taken
                             2
                   as 9.81 m/s . P d was calculated to be 0.91 psi. Brooks and Corey related the rock
                   parameter λ to the distribution of pore sizes. For narrow distribution λ is .2 and for
                   wide distribution λ is ,2. For this analysis, a normal distribution was assumed and
                   λ 5 2. A plot for both systems were generated as shown in Fig. 7.6.
                      The plot above shows that the effect of capillary pressure drastically reduces the
                   water cuts with respect to water saturation. Buckley Leverett equation for linear sys-
                   tems showed a deviation of about 23% for a water saturation of 50%. However, it was
                   observed for the absence of capillary pressure; the water cuts were the same for both
                   systems.
                      Similarly, the waterfront correlations for both systems were also analyzed. For a
                   duration of 100 days, the location of the waterfront populated with respect to water-
                   flooding time enabled the development of the plot in Fig. 7.7 which showed that lin-
                   ear system had higher distances invaded as compared to radial system.
                      From the chart, one can see the discrepancies of waterfront for both systems. If the
                   linear displacing mechanism is applied for waterflooding scenarios that follow a radial
                   displacement, then error would be introduced into the analysis. For example, at 70 days
                   of flooding, linear system estimated the waterfront to be at 17 ft, whereas the radial sys-
                   tem estimated 11.5 ft; thus, it has been overestimated with a deviation of nearly 48%.
                             5


                             4
                                                                                 vs S
                                                                               F w  w
                           Fractional flow 3 2                                 dF /dS  vs S w
                                                                                  w
                                                                                w
                                                                               F'  vs S  (radial)
                                                                                w
                                                                                   w
                             1
                                                                               (dF /dS )' vs S
                                                                                w  w  w
                                                                               (radial)
                             0
                            –1
                              0   0.1  0.2  0.3  0.4  0.5  0.6  0.7  0.8  0.9  1
                                             Water saturation (S w )
                   Figure 7.6 Plot of fractional flow and water saturation.
   226   227   228   229   230   231   232   233   234   235   236