Quantitative Analysis of Multi-Phase Flowback From Multi-Fractured Horizontal Wells

摘要

Due to a decline in conventional reserves, particularly in North America, recent development has focused on low to ultra-low permeability unconventional reservoirs. Due to the low permeability of these plays, extensive hydraulic fracturing is required for commercial production. As a result of this growing trend, operators are looking for new methods to characterize hydraulic fractures, especially early in the well life. Several authors have identified high frequency flowback data as an early-time method for fracture characterization.;This dissertation starts with, and builds on, two publications by Clarkson and Williams-Kovacs (2013a) and Clarkson and Williams-Kovacs (2013b), which set the ground work for quantitatively analyzing flowback from multi-fractured horizontal wells (MFHWs) completed in shale gas and light tight oil (LTO) reservoirs respectively.;First a new shale gas model was developed to better capture the physics of the flowback problem. The tool was built using a similar conceptual model to that assumed by Clarkson and Williams-Kovacs (2013b) for LTO applications, although significant modifications were required to account for the complexities of shale gas reservoirs. A focus was also placed on stress-dependant porosity/permeability as a result of fracture closure during flowback. Although this new model yielded comparable results to the model developed by Clarkson and Williams-Kovacs (2013a) in the case study presented herein, by better capturing the physics of the problem the new model is applicable to more cases and creates a much improved platform for further development.;Secondly, the LTO model developed by Clarkson and Williams-Kovacs (2013b) was expanded to account more complex problems, such as stage-by-stage flowback, multi-well analysis and multi-layer flowback for wells contacting multiple productive intervals. These advances of the base model greatly broaden the applicability of the developed methods to many of the complexities faced by operators in unconventional formations, particularly with advancements in completion technology and development strategies currently being employed in these formations.;Thirdly, stochastic simulation and several assisted history-matching techniques were applied to an LTO data set to quantify the uncertainty in key fracture parameters and to optimize the history-match for the most accurate estimation of key fracture parameters. Application of stochastic simulation allows the operator to determine realistic bounds for future potential well performance while assisted history-matching leads to significantly improved results.;Fourthly, several LTO case studies were conducted to address topics such as assessment of the potential economic value of conducting flowback analysis, and development of a modified model for analyzing flowback from LTO wells completed with an oil-based fracture fluid. Numerical simulation was also conducted to confirm the sequence of flow-regimes interpreted from field data. These case studies validate the simple analytical models developed, quantify the value to operators for applying such methods as well as demonstrate extensions which are required to analyze flowback from many modern completions.;Lastly, a salinity model was developed to compliment the flow models primarily to confirm fracture surface area and volume. It is possible that this model could be applied in place of detailed flow modeling if enough of the key transport parameters can be accurately estimated. The flow simulation was successfully validated using the developed salinity model.