Integrated Simulation, Characterization and Enhanced Recovery Strategies For Shale/Tight Formations
Abstract
The recent economic downturn poses a challenge to the development of oil and gas prospects, especially unconventional or shale assets. As a result, it is prudent to improve well design and placement, and to investigate alternative hydrocarbon extraction methods. To help achieve this goal, this dissertation focuses on improving the following engineering approaches: rock and fracture characterization, and huff-n-puff design and model setup. Improved characterization is necessary for reservoir modeling which can then be used to find the optimum well design and placement. Huff-n-puff is an enhanced oil recovery technique that has been successfully applied to conventional reservoirs and may prove equally viable for unconventional assets.
There are several approaches for rock and fracture characterization. One class of methods utilizes readily available production data to inverse model for reservoir parameters. If certain conditions are met, analytical methods can be used to determine reservoir parameters—such methods are referred to as rate transient analysis (RTA). On the other hand, if conditions are not met, sophisticated reservoir simulators are used to history match production data and thereby inverse model reservoir parameters. This dissertation improves existing analytical solutions that results in errors if applied to bilinear flow (a flow regime that is occasionally observed in shale/tight wells) in gas reservoirs. The improvement is incorporated as a correction factor applied to the bilinear flow analysis plot. In addition to this correction, this dissertation presents an innovative framework that utilizes compositional simulation, multi-objective genetic algorithms, and information theory to characterize reservoir parameters. In addition, the framework incorporates flowback data into the analysis to further improve the characterization results. This framework can be applied when conditions for RTA are not satisfied.
The other engineering approach that is improved in this dissertation is huff-n-puff design and model setup. Although huff-n-puff has been successfully applied in conventional reservoirs, its success is yet to be established for unconventional reservoirs. A number of simulation and lab studies have been published to date. While lab studies agree that huff-n-puff is successful for unconventional reservoirs, simulation studies disagree. Consequently, this dissertation uses numerical simulation to investigate possible reasons for this conflict. In addition, this dissertation also endeavours to find key factors that influence huff-n-puff success. Finally, the feasibility of huff-n-puff is evaluated by utilizing genetic algorithms to find the optimum huff-n-puff design that maximizes net present value.
In summary, this dissertation provides new methods for improved fracture and reservoir characterization and huff-n-puff design. A correction factor is derived that can be combined with existing analytical solutions to improve the characterization of fractures. A framework is also provided to characterize unconventional reservoirs; this framework can be used whenever RTA assumptions are violated. Moreover, this dissertation investigates huff-n-puff simulation, and finds that conflicts in the literature are probability caused by model setup, specifically grid refinement, and fracture pseudo-width. Furthermore, it concludes that higher fracture density, higher fracture complexity, presence of natural fractures, and delayed huff-n-puff can improve the ultimate recovery. Finally, the dissertation demonstrates that huff-n-puff can be economical if the huff-n-puff design is optimized. Two major design criteria are short injection, and longer soaking times. These findings can help operators improve reservoir characterization, and choose whether huff-n-puff is suitable for their respective wells. Importantly, these aspects can ultimately help reduce cost, a necessity in today’s economic climate.
Description
Keywords
Engineering--Petroleum
Citation
Kanfar, M. (2017). Integrated Simulation, Characterization and Enhanced Recovery Strategies For Shale/Tight Formations (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/26575