Modeling and analysis of diagnostic fracture injection tests (DFITs)

Abstract

Diagnostic fracture injection tests (DFIT) are used as an indirect method to determine closure pressure and formation effective permeability in unconventional reservoirs as a first step in formation evaluation. The information obtained from DFIT is particularly useful because it is obtained before any production for a given well is available. In DFIT, a small fracture is created by injecting few barrels of completion fluid until formation breaks down and a fracture is initiated and propagates a short distance into the reservoir. Then, injection is stopped and the pressure decline (or falloff) is monitored. From this pressure decline, the effective permeability of the formation is estimated by Nolte’s G-function, log-log plot, or square root of time analysis. In this research, the viability of the common DFIT analysis techniques was investigated for unconventional reservoirs with and without micro-fractures by using a numerical hydraulic fracturing simulator, CFRAC. The results of numerical simulations were investigated to assess the impact of permeability, residual fracture aperture, and complex fracture networks on conventional DFIT interpretations. For the example considered in this work, the commonly used G-function analysis yielded estimates of permeability over an order of magnitude higher than the simulated matrix permeability. Error in the G-function estimates of permeability were higher for higher matrix permeability and in the existence of a fracture network. On the other hand, straight-line analysis of ∆p versus G-time yielded much closer (in the same order of magnitude) estimates of permeability.