Complexity of Minifrac Tests and Implications for in-Situ Horizontal Stresses in Coalbed Methane Reservoirs

Abstract

Abstract Minifrac tests are used to obtain key parameters of reservoirs for geomechanical modeling and hydraulic fracture treatment design. An ideal minifrac test can clearly indicate the leak-off pressure, fracture propagation pressure, shut-in pressure and fracture closure pressure. Repeated minifrac tests in the same interval provide additional value by verifying the fracture parameters with similar results. However, minifrac tests of two wells conducted in a coal seam in a coalbed methane (CBM) reservoir in China showed unconventional results - the fracture profile is less distinguishable for the fracture behaviors, and the subsequent test cycles in each well exhibited higher pressures than the precursors. A close examination of the test results with laboratory experiments and field microseismic monitoring results indicates that the discrepancies in pressure responses are correlated to the constitution of the coal seam. The coal seam contains many cleats, and the minifrac tests in most cases had opened these cleats instead of initiating new fractures. Opening a cleat whose plane is in line with the maximum horizontal stress orientation requires the least pressure but could be inhibited by connecting cleats with any other orientation that needs higher pressure to open or slide, which could lead to opening more cleats from different sites towards different directions. Eventually, multiple zigzag hydraulic fracture paths instead of a single planar fracture would be induced. The zigzag paths and simultaneously opening of multiple fracture paths will cause a stress increase in their vicinity, resulting in higher treating pressures for further advances. This implies that the magnitude of minimum horizontal stress Shmin shall be the least magnitude seen amongst the leak-off and fracture closure pressures of all tests, 0.0143 MPa/m for the case study. A conceptual model has been put forward to reveal the hydraulic fracturing in coalbeds. The mechanisms possibily contributing to the high treating pressure in hydraulically fracturing coalbeds have been explained from the geomechanics point of view, including poroelastic effects, fracture tortuosity effects and multiple fracture opening effects. This paper describes how the physical coal seam model reacts to fluid injection, which assists understanding the nature of minifrac tests and avoids the misinterpretation of the results. The approach can further help to quantify accurately the minimum horizontal stress magnitude from the complex minifrac tests, and hence contributes considerably to accurate geomechanical modelling and fracture treatment design for any CBM reservoirs. Introduction With horizontal drilling and hydraulic fracturing increasingly becoming economical and practical in developing shale gas, tight sand gas and coalbed methane reservoirs, knowledge of the minimum horizontal stress (Shmin) is critical in optimizing well trajectories and fracture treatment design (Ramakrishnan et al., 2009). Although there are a few methods available to assess the Shmin magnitude, the micro-fracture (microfrac) and mini-fracture (minifrac) as well as the extended leak-off test (XLOT) remain the most reliable methods (De Bree and Walters, 1989; White et al., 2002; Howkes et al., 2005; Lin et al., 2008). These methods are similar in principle: to obtain a fracture closure pressure by creating a pair of small fractures through injecting fluid in an isolated wellbore section (Fig. 1), followed by a shut-in process to allow the fracture to close. The fluid pressure varies corresponding to the fracture behavior during a complete test and the variation can indicate the moments of fracture creation, propagation and closure associated with fluid injection and shut-in process. The pressure at the time the fracture closes is termed the fracture closure pressure and is equivalent to the minimum horizontal stress. Repeated minifrac tests provide additional value by verifying the pressure parameters with similar results.