Diagnostic fracture injection tests (DFITs) are often used to estimate formation properties such as closure stress, pore pressure, and matrix permeability. These estimations are typically based on analysis of pressure data assuming the closure of simple planar fractures in homogeneous reservoirs. These interpretations are incorrect when dealing with complex reservoir environments such as layered reservoirs with different properties and stresses. This paper investigates the impact of such complex environments on DFIT interpretation and presents a systematic method to analyze the data.

A 3-D implicitly integrated poroelastic fracture-reservoir-wellbore model is used to simulate DFITs. DFIT fracture propagation and well shut-in are simulated with implicitly computed fluid leak-off and fracture closure. The model is validated by simulating a DFIT for a homogeneous reservoir and the implicitly calculated surface pressure is interpreted to obtain the simulation inputs (stress, pore pressure, permeability, etc.). A multi-layer reservoir model is then built in the numerical simulation domain and a DFIT is simulated in the target layer. The properties and thickness of the layers are varied to analyze their impact on the observed DFIT signature.

We analyze the impact of layer thicknesses, layer stresses, pressure and permeability of each layer, stress contrast between the layers, fracture interaction with bedding planes and the rock roughness and hardness of each layer on the DFIT pressure signature. We show that the layer property variations can cause different but characteristic DFIT pressure responses. Fracture propagation into layers with different stresses induces multiple closure events in the observed pressure signature, which provides a quantitative representation of the fracture height growth. The emergence of these closure events in the pressure signature are found to be dependent on the hardness and modulus of the rock layers and the fluid communication between the closing parts of the fracture. The DFIT signature patterns are also found to correlate with the interaction of the fracture with bedding planes (cross/arrest/divert) and provide valuable insights into fracture containment.

In this work we present best practices for performing DFIT analysis in layered reservoirs. Results from simulated DFITs in layered reservoirs clearly show the effect of key heterogeneity parameters on DFIT responses. The results from this work can be used to more accurately determine reservoir closure stress, pore pressure, reservoir permeability, fracture compliance, fracture conductivity, and fracture containment in heterogeneous reservoirs.