Transient flow modeling of diagnostic fracture injection tests in fractured formations

Abstract

The Minifrac, also referred to as Diagnostic Fracture Injection Test (DFIT) is generally considered as the most important test on-site before the main hydraulic fracture injection treatment. Critical parameters gathered from a minifrac test are required for the tuning of the main fracture stimulation treatment. The minifrac is a pump-in/shut-in test that employs the injection of a small volume of KCl water. The test is intended to break down the formation and create a short fracture, shut-in and then record pressure falloff. The test can provide critical reservoir and rock parameters for the design of fracture stimulation operations. Information gathered includes initial reservoir pressure (pi), reservoir transmissibility (kh/μ), closure pressure, and leakoff types. In this research, the viability of the common DFIT analysis techniques was investigated using a coupled geomechanics and reservoir flow model. The model simulates the minifrac response in dual-porosity systems and is used to study the effect of natural fracture density, permeability, injection volume and stress variations on the pressure falloff. The results of numerical simulations were investigated to assess the impact of permeability, natural fracture density, and stress variations on conventional DFIT interpretations as well as the effect of permeability and injection volume on time required to observe late-time flow regimes. DFIT analysis begins after the ISIP, which is taken as the incipient fracture extension pressure. In most cases, the ISIP is not clear, and it is not instantaneous due to the near-well pressure drop, fluid expansion, and after-flow effects. ISIP is certainly a subjective measurement that could be argued in some instances. This thesis presents a novel iterative method to determine the ISIP. The results show that, regardless of the first pick of fracture closure pressure, the ISIP converges toward a fixed value within four iterations. The improvement of the ISIP estimate improves the semi-log derivative, which is used to qualitatively describe the subject reservoir and leakoff type, and determine the fracture closure pressure. Often, DFITs fail or are interrupted before fracture closure pressure is obtained. This thesis also presents an integrated workflow to predict fracture closure pressure, reservoir permeability, and time to reach late-time flow regimes from interrupted or non-ideal DFIT signatures, with the application of supervised machine learning algorithms.Conventional pressure transient techniques have been adopted to analyze the pressure falloff from DFITs. Nolte and Smith (1979) pioneered the interpretation of pressure response during hydraulic fracturing. Interpretation of fracturing pressure can yield valuable information about fluid efficiency, fluid loss coefficient, fluid leakoff behavior, fracture closure pressure, fracture closure time, and fracture half-length. Fracture extension, conductivity, and proppant transport can be inferred qualitatively from the analysis of fracturing pressure. The integrated methodology introduced in this thesis is based on the transient analysis of post-stage-fracture falloff pressure. The Nolte-Smith log-log pressure vs. time plot is used to identify flow regimes. The underlying principles of the post-stage-treatment falloff analysis proposed in this study are similar to that of DFIT falloff behavior. However, the falloff period of a minifrac test or DFIT may need to be several weeks to reach the late-time flow regimes in tight formations (Belyadi et al. 2019). The pressure falloff after the stage treatment only needs to last for a fraction of an hour since the main interest is in the early flow regimes such as bilinear flow. In this thesis, post-stage-treatment pressure decay analysis was completed for two horizontal wells, and the findings from the treatment were then compared with the traditional well testing methods, namely PTA and RTA.Finally, the results prove that the use of the conventional PTA techniques to analyze the post-stage fracture pressure decay is useful to characterize fractures, such as the fracture geometry (half-length), conductivity, and proppant selection, as well as the near-field characterizations of skin. Monitoring and analysis of pressure data during stage fracturing enable the operators to review and revise fracture design and implementation in near real-time. The stimulated reservoir area obtained from buildup analysis provides a reasonably good match with the SSRA calculated from post-stage fracture pressure decay. This suggests that the post-stage fracture pressure decay analysis approach is applicable.