George V. Chilingar

Acoustic and Vibrational Enhanced Oil Recovery


Скачать книгу

two independent polarization types are available in a solid isotropic medium, the total displacement vector u(r,t) should be presented in the form of in-plane ul and lateral u r displacements:

      (2.20)image

      As an indicator of efficiency at a certain frequency may serve encompassment radius within which are maintained certain interrelations between the threshold values of vibration parameters—vibratory displacements ξ and vibratory accelerations image. These parameters are determined from the density of vibratory energy flow E at a given point of the medium and through vibration frequency f as follows:

image

      where ρC is the wave resistance or acoustic impedance of the medium.

      Numerical modeling was conducted by Sherifulling et al. [26] in order to determine the vibrations’ space-energy distribution. This enabled the computation of the energy picture of the wave distribution field for the assigned vibration frequency and of the vibratory accelerations and offsets field in the reservoir accounting for petro-physical properties of the top and base of the reservoir. Modeling was conducted using a method of the wave spreading statistical testing in the plane dissecting the reservoir with the plane-parallel boundaries [20]. The source of harmonic waves was distributed along the circumference of a well with the radius Rc with the center in the origin. The top and base of a reservoir having thickness H was positioned parallel to the X axis. At calculating by the statistical testing method, the spreading of the low-frequency harmonic waves is modeled using acoustic “quanta” flying out from the source in a random direction. Acoustic energy is assigned to each “quantum”, the energy equal to surficial density of a cylindric pulsator with the pressure amplitude P0:

image

      For the evaluation of acoustic properties of the saturated porous medium and description of two-dimensional processes of wave transfiguration on the boundaries, one can use the approach proposed by Sharifullin based on Bio’s linear theory. The offsets of solid and liquid phases (appropriately, u and V) may be determined from Bio’s equations:

image

      Here, λ1, λ2, Q, and R are Bio’s elastic moduli for the medium; b = m2μ/k is Bio’s resistance coefficient; μ is the fluid’s dynamic viscosity; k the permeability of porous medium; m is porosity; ρ11, ρ12, and ρ22 are Bio’s density parameters expressed by rock matrix density ρ1 and fluid density ρ2 as follows:

image

      where ρ12 is the so-called “reduced density”.

      The Bio’s moduli may be expressed directly by the measured parameters: rock matrix triaxial compression modulus K, shear modulus G for a “dry” porous medium, and compressibility factors β1 and β2.

      On assuming that in plane OXZ on the separation boundary of two saturated porous media (the boundary coinciding with the OX axis), first or second kind compressional wave or shear wave incidence angle to OZ axis is α. The OZ axis is directed into the second medium. In the general case, six types of waves are generated on the boundary, three are reflected and three are refracted.

      This approach was implemented for computing phase velocities, free space attenuation factors, angular wave transformation factors, and energy reflection parameters at reservoir’s boundaries. These parameters are needed for a “raffle” of the acoustic “quanta” trajectories.

image

      where A is amplitude of wave potential in the medium and Cm is the phase velocity of m-wave.

      The exit angle of each “quantum” trajectory ray of 0° to 180° was “raffled” using pseudorandom number sensor for the numbers uniformly distributed in the zero to one interval. The field near every of 5,000 observation points was averaged on the surface of a cell with the side H/50, where H is the reservoir thickness. Thus, phase displacement at the quanta superposition in a given point was taken into account by that separate cells have been provided for accumulating two orthogonal components of the wave vector at every point of the reservoir. Modeling was being performed in the mode of continuous information accumulation with releasing results over a certain number of histories— “acoustic quanta exits” from the source. The result at each release was normalized per a number of hits in every reservoir cell. For calculating the hit numbers, a special two-dimensional massif was included in the memory.