Recently it has been demonstrated that rate-transient analysis (RTA) performed on flowback data frommulti-fractured horizontal wells (MFHWs) can provide timely estimates of hydraulic fracture properties. This information can be used to inform stimulation treatment design on upcoming wells as well as other important operational and development decisions. However, RTA of flowback data may be complicated by rapidly changing operating conditions, dynamic hydraulic fracture properties and multi-phase flow in the fractures, complex fracture geometry, and variable fracture and reservoir properties along the MFHW, among other factors. While some constraints on RTA model assumptions may be applied through a carefully-designed surveillance and testing program in the field (e.g. to constrain fracture geometry), still others require laboratory measurements.
In this work, an integrated flowback RTA workflow, designed to reduce uncertainty in derived hydraulic fracture properties, is demonstrated using flowback data from MFHWs producing black oil from low-permeability reservoirs in the Montney and Duvernay formations. The workflow includes rigorous flow-regime identification used for RTA model selection, straight-line analysis (SLA) to provide initial estimates of hydraulic fracture properties, and model history matching of flowback data to refine hydraulic fracture property estimates. The model history matching is performed using a recently-introduced semi-analytical, dual-porosity, dynamic drainage area (DP-DDA) model that incorporates primary (propped) hydraulic fractures (PHF) as well as a dual-porosity enhanced fracture region (EFR) with an unpropped (secondary) fracture network. Inclusion of both the PHF and EFR components addresses the need to incorporate both propped and unpropped fractures and fracture complexity in the modeling. The DP-DDA model is constrained using estimates of propped fracture conductivity and unpropped fracture permeability (measured as a function of stress), and unpropped fracture compressibility values, obtained in the laboratory for Montney and Duvernay core samples. Use of these critical laboratory data serves to improve the confidencein the modeling results.
The case studies provided herein demonstrate a rigorous workflow for obtaining more confident hydraulic fracture property estimates from flowback data through the application of RTA techniques constrained by both field and laboratory data.