Reports: AC8
47463-AC8 Primary Migration of Oil Through Self-Propagating Fractures
In the 2008 – 2009 period, we have mainly focused on the effects of interactions between multiple fractures on the primary migration behavior of oil, including migration velocity, oil flux, and migration period. Two typical problems have been considered. The first one consists of an array of oil-filled, parallel fractures propagating vertically due to buoyancy. The second one consists of an array of oil-filled, collinear microcracks subcritically propagating in the horizontal direction due to volume expansion of kerogen-oil transformation and material anisotropy. Next we describe the approach for solving the problem and the results obtained.
1. Oil migration through buoyancy-driven vertical propagation of an array of oil-filled, parallel fractures
We adopt the assumptions that the sediment is linearly elastic, the oil-filled fractures are parallel in the vertical direction and have equal length, and the fractures propagate at a constant velocity. We further assume that the fractures are periodically distributed so that a two-dimensional, semi-analytical approach becomes applicable. We use an integral equation method to derive the expressions for fracture opening displacement, fracture propagation velocity, and mass flux of oil migration. In the numerical calculations, we use the following properties for the source rock and oil: Young’s modulus of 2.215 GPa, Poisson’s ratio of 0.42, rock fracture toughness of 0.1 MPa-m1/2, rock density 2150 kg/m3, oil density 840 kg/m3, and oil viscosity of 0.01 Pa-s.
Our numerical results show that for a given fracture spacing (or density), there is a critical fracture length below which the fractures do not propagate critically. The fractures will propagate when they become longer than the critical value, and the longer the fractures, the higher the velocity. The propagation velocity increases dramatically for fracture lengths just larger than the critical size and levels off for longer multiple fractures. The fracture density affects significantly the propagation velocity. For a given fracture length, the velocity becomes significantly lower for the multiple fractures with decreasing fracture spacing (increasing fracture density). For example, the velocity is about 31,544 m/year for the single fracture 20 meters long. The velocity decreases to 2768 m/year and 1028 m/year for multiple fractures with densities of N = 0.5 and 1 (N is the number of fractures in a horizontal meter), respectively, which represents an order of magnitude reduction.
We found that whereas the opening of the single fracture increases dramatically with increasing fracture length, the opening for the multiple fractures is relatively insensitive to the fracture size. The opening for the single fracture is significantly higher than those for the multiple fractures. For example, the opening is 0.104 mm for a single fracture 20 meters long. The corresponding openings for the multiple fractures with fracture densities of N = 0.5 and 1 are only about 0.037 mm and 0.026 mm, respectively.
The reduced propagation velocity and fracture opening for multiple fractures lead to lower mass flux of oil migration. For example, the mass fluxes for fractures 20 meters long with densities of N = 0.5 and 1 are calculated as 27,611 and 55,222 kg/m2/year, respectively, without considering the fracture interactions. The fluxes decrease by two orders of magnitude to 856 and 452 kg/m2/year, respectively, when the effects of fracture interactions are included. Hence, ignoring multiple fracture interactions will result in overestimated mass flux of oil migration.
2. Oil migration through subcritical growth of collinear microcracks due to volume expansion of kerogen-oil transformation and material property anisotropy
Microcrack propagation around kerogen particles in finely laminated black shales has been observed by a number of researchers to occur parallel to horizontal bedding planes. Horizontal microcracking is favored because of the orientation of flat kerogen particles and the strength anisotropy of the source rock. The microcracks may connect with pre-existing vertical fractures leading to vertical oil migration.
Critical propagation of a horizontal microcrack occurs when the stress intensity factor equals the rock fracture toughness. The stress intensity factor, however, may not reach the fracture toughness as the crack surface pressure induced from kerogen-oil conversion may remain at relatively low levels. In this case the microcrack may propagate subcritically if the stress intensity factor reaches the threshold value which is usually a fraction of the toughness.
We have investigated subcritical propagation of single microcrack as well as multiple, collinear microcracks driven by the oil pressure. For the case of multiple crack propagation, we assume that the cracks have equal length and are periodically spaced. Natural microcracks in shale source regions are neither exactly collinear nor periodically spaced. The collinear cracks model is employed for the purpose of setting a possible lower limit for the time period required for the formation of large horizontal cracks that may connect to preexisting vertical fractures. We have derived the fracture mechanics formulations for subcritical propagation of oil-filled single and collinear cracks during the process of kerogen-oil transformation.
Our results for a source rock 150 0C hot with properties similar to those in the above part 1 show that it takes more than 6 million years for a single crack 100 micrometer long to grow to about 1300 micrometer long when the kerogen is completely converted to oil. This time duration is controlled by the kinetics of kerogen-oil transformation. For collinear cracks with the spacing the same as the crack length, however, it takes less than 60,000 years for the collinear cracks 100 micrometer long to coalesce. If the crack spacing is twice the crack length, it will take about 100,000 years for the cracks to get interconnected. Those numbers, of course, also depend on the temperature and material properties of the rock, for example, Young’s modulus, fracture toughness, and subcritical crack growth parameters.
3. Impact
The grant allows Scott Johnson, an expert in structural geology and tectonics, to expand his research into petroleum geology, and allows Zhihe Jin, a junior faculty member in mechanical engineering, to develop a new research program in geophysics. The grant is crucial for the PIs to develop a strong, successful multidisciplinary program in geology and engineering mechanics, and is supporting the PhD dissertation of Zhiqiang Fan.
