Page 54 - Reservoir Geomechanics
P. 54

38     Reservoir geomechanics


              production from previously drilled wells in what was mapped as the same reservoir.
              Thus, this reservoir appears to be compartmentalized at a smaller scale than that mapped
              seismically, presumably by relatively small, sub-seismic sealing faults that subdivide
              the sand into small compartments.
                Figure 2.10 illustrates compartmentalization in a Miocene sand (the Pelican sand)
              in Southern Louisiana (Chan and Zoback 2007). A structure contour map indicating
              the presence of faults that compartmentalize the reservoir is superimposed on an air
              photo of the region. As shown in the inset, the pressure in the wells penetrating this
              sand in fault blocks I, II and III were initially at a pressure of ∼60 MPa. By 1980
              fault blocks I and III had depleted along parallel, but independent depletion paths to
              ∼5MPa (all pressures are relative to a common datum at 14,600 ft). Wells B and C are
              clearly part of the same fluid compartment despite being separated by a fault. Note that
              in ∼1975, the pressure in the fault blocks I and III differed by about 10 MPa. However,
              the pressure difference between fault blocks I and III and fault block II after five years
              of production is quite dramatic. Even though the first two fault blocks were signicantly
              depleted when production started in fault block II in the early 1980s, the pressure was
              still about 60 MPa. In other words, the pressure in wells E and F was about 55 MPa
              higher than that in wells B and C in the same sand. The fault separating these two
              groups of wells is clearly a sealing fault whereas the fault between wells B and C is
              not.
                An important operational note is that drilling through severely depleted sands (such
              as illustrated in Figure 2.10a) to reach deeper reservoirs, can often be problem-
              atic (Addis, Cauley et al. 2001). Because of the reduction of stress with depletion
              described in Chapter 3,a mud weight sufficient to exceed pore pressure at greater
              depth (and required to prevent flow into the well from the formation) might inadver-
              tently hydraulically fracture the depleted reservoir (Chapter 6) causing lost circulation.
              This is addressed in Chapter 12 both in terms of such drilling problems but also from
              the perspective of the opportunity reservoir depletion offers for refracturing a given
              formation.
                It is worth briefly discussing how pore pressure can appear to increase with depth
              at gradients greater than hydrostatic. In Figure 2.2,at depths greater than ∼11,000 ft,
              pore pressure increases with depth at approximately the same rate as the overburden
              stress increases with depth. This would suggest that a series of compliant, isolated
              reservoirs is being encountered in which the reservoir pressure is supporting the full
              overburdenstress.However,anextremelyhighpressuregradientisseenbetween9000ft
              and 11,000 ft (much greater than the overburden stress gradient). One should keep in
              mind that data sets that appear to show pressure gradients in excess of hydrostatic
              are compilations of data from multiple wells which penetrate different reservoirs at
              different depths, even though a hydrostatic pressure gradient is observed within each
              individual reservoir (assuming that water is in the pores).
   49   50   51   52   53   54   55   56   57   58   59