Fracture Diagnostics Using a Low-Frequency Electromagnetic Induction Method
- P. Zhang (The University of Texas at Austin) | J. Shiriyev (The University of Texas at Austin) | C. Torres-Verdín (The University of Texas at Austin) | M. M. Sharma (The University of Texas at Austin) | Y. Brick (The University of Texas at Austin) | J. Massey (The University of Texas at Austin) | A. E. Yilmaz (The University of Texas at Austin)
- Document ID
- American Rock Mechanics Association
- 50th U.S. Rock Mechanics/Geomechanics Symposium, 26-29 June, Houston, Texas
- Publication Date
- Document Type
- Conference Paper
- 2016. American Rock Mechanics Association
- 8 in the last 30 days
- 99 since 2007
- Show more detail
|ARMA Member Price:||USD 10.00|
|ARMA Non-Member Price:||USD 20.00|
The implementation of a low-frequency electromagnetic induction tool for propped-fracture detection and diagnostics requires in-depth examination of the reliability and accuracy of the method across fractures with realistic geomechanical features. Likewise, the method relies on the effective placement of electrically conductive proppant within the generated fractures for measurable sensitivity. We invoke a fast Fourier-transform-based volume integral equation simulation method for forward modeling to simulate the electromagnetic response of realistic fracture geometries generated in hydro-fracturing operations and under practical rock conditions. Improved performance is achieved by removing model features which are of little significance to the results and yet cause a significant computational overhead. To properly account for the effective proppant conductivity, we introduce a new technique for the measurement of proppant conductivity using a resistivity core holder. The conductivity of petroleum coke—a potential candidate to be used in the field—is tested in the laboratory. Resulting measurements are used to forward simulate the tool's response to multiple fracture geometries and sizes. Distinct results are obtained for fractures with different spatial distributions of proppant, indicating the possibility of distinguishing between them using a future inverse solver.
Hydraulic fracturing continues to play a crucial role in unlocking oil and gas production in unconventional reservoirs. To better design hydraulic fracturing treatments, it is important to predict how fluid-driven cracks will propagate under various conditions; several models, including fully coupled porous flows and geomechanical models [1-4], have been proposed for this purpose. Validation of these models’ ability to predict complex fracture networks, however, requires the capability to measure fracture dimensions with diagnostic methods. From another perspective, prior knowledge of the Stimulated Reservoir Volume (SRV) provided by fracture diagnostics can greatly improve the reliability of reservoir simulations.
The need for and efforts on fracture diagnostic methods have been growing with the improvement of hydraulic fracturing techniques. One common technique is to use proppant tracers, which can be detected by radiation sensors, for near-wellbore proppant detection [5, 6]. One shortcoming of such methods, aside from the obvious radioactive material safety concerns, is that the radiation from proppant located far from the wellbore is harder to detect, thus reducing the sensitivity to the size of the propped fracture. Near-wellbore temperature and water hammer measurements can record the signature of fracture spacing, length and formation inhomogeneity. However, these measurements suffer from non-uniqueness, which makes the extraction of fracture properties very challenging [7, 8].
|File Size||793 KB||Number of Pages||10|