Flow in Faults and Fractures

Where present faults and fractures often dominate the flow pattern. Fractures not only increase the permeability of sediments and breach barriers but will also sequester gas from the matrix because of the lower capillary pressure that exists there (Bethke et al. 1991, fig. 17). Fluid can flow passively along fractures held open by asperties, or can actively generate the flow pathway through hydrofracture. If a clay or mud has previously experienced a greater effective stress it will be relatively brittle and can be fractured under tensile stress even though it may appear to be plastic and ductile when handled (Jones 1994). The classical criterion for generating an open tensile fracture in a water-saturated sediment is that the fluid pressure in the fracture Pf must exceed the sum minimum effective stress c3 and the tensile strength or cohesion of the sediment T (Secor 1965):

T is related to grain interlocking and cementation. To keep the fracture open once formed, Pf must be maintained at a value of at least o3. The cohesion in coarse uncemented sediments is often negligible but in muds it is related to electrochemical properties, so it is important to define whether the sediment is fractured open by water or by gas pressure (Harrington and Horseman 1999). If the maximum effective stress is vertical, then the pressure of fluid required to hydrofracture the mud works out at 75-90 % of the lithostatic pressure (Clayton and Hay 1994). As sediment is loaded by continued sedimentation, the lateral stress, Oh in a basin without imposed tectonic stresses is a simple function of the vertical stress av.

where K0 is the coefficient of earth pressure at rest, (Jones 1994)). K0 during monotonic loading takes on very conservative values, typically from 0.6 to 0.75. However, during unloading, or equivalently, overpressuring, the value of K0 increases, even to values between 1 and 3 (Brown 1994). Not only does this inhibit the formation of fractures, but since a3 is now the vertical stress, it suggests that hydraulic fractures formed will be subhorizontal.

Given the existence of elongate flaws with favourable orientation, the criterion for propagation of the fracture in an elastic reduces to one of stress intensity at the tip (Murdoch 1993; Andersen et al. 1994) (Fig. 8).

The preceding analysis suggests that open fractures require an almost isotropic stress field, and almost lithostatic pore fluid pressures. Numerous studies show that fracture flow is important in sedimentary basins even when such extreme conditions are not met. There are several ways around this problem:

1. Fractures may flow under lower pore pressures if propped open by asperities or cements.

2. Secor (1965; cf Chapman 1983) note that the actual failure mode when pressure is increased in a basin with vertical maxium effective stress is usually by shear fracture. Mudrocks will dilate if sheared when overconsolidated, or pathways may open along fault jogs (Ingram and Urai 1999). However shear zones in mudrocks generally seal except at high pore fluid pressure (Dewhurst et al. 1999). Furthermore, Pf required to keep a shear fracture open is even higher than a3 is since it is not oriented normal to the minimum principle stress.

3. A tensile component of the tectonic stress field will reduce the value of ct3. But this must propagate into the upper part of the sedimentary basin (Yassir and Bell 1994).

4. On a slope, the sediments will be under tension in the up-slope region, and compression at the base of the slope (Mandl 1981, Jones 1994, Bjorlykke and Hoeg 1998). Therefore seepage may be localised here at the head wall of slope failures (see Yun et al. 1999).

Tip line: crack

Isostress lines propagation trajectory

Zones of plastic deformation

Figure 8. Fracture propagation (from Murdoch 1993). The mechanism proceeds initially by pressurizing the existing fracture until the elastic stresses in the wall region balance the internal fluid pressure excess. This is time-dependent because some injected water leaks off. Around the tip the sediments will be in tension, leading to a drop in pore fluid pressure. Water is sucked into the tip zone. Gas may exsolve in the tensile zone and pool in the fracture. When the pressure is sufficient, the rupture propagates as a mode I crack.

Fluids sucked into Fracture ti|

Gas exsolutio in lensile zon propagation trajectory

Fluids sucked into Fracture ti|

Gas exsolutio in lensile zon

Fluid flow paths

Pressurized water driving hydraulic fracture

Watts (1987) stated that fractures can form in a seal when hydrocarbons build up beneath it because the buoyancy pressure in the non-wetting phase can exceed the fracture pressure if the hydrocarbon column height, L, becomes large (Fig. 5); i.e., together lead to failure. Watts proposes that the gas pressure is the effective Pf in equation (9). These analyses presuppose that the fracture is completely filled with gas, and the walls are impermeable. When gas invades a fracture in porous sediment gas pressure increases will in part be balanced and dissipated by intrusion of gas into the pores, and the full gas pressure excess available from buoyancy will not be mobilised to do work on the surroundings. Bjorkum et al. (1998) note that gas buoyancy pressure cannot assist fracturing until the gas has already entered the cap rock, and that overpressure in the water will not assist upwards gas drive since the phases are in static capillary equilibrium. Nunn (1996) has modelled the propagation of single, fluid-filled vertical fractures in sediment. The fluid in the open fractures, which may be water, gas or a mixture is less dense than the surrounding saturated sediment, and so generates a buoyant upthrust. Existing flaws, or fractures generated by overpressure in the source bed can propagate upwards when this force is sufficient to overcome the fracture toughness of the sediment.

Was this article helpful?

0 0

Post a comment