Open Access
Yang, Yun
Graduate Program:
Energy and Mineral Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
August 28, 2017
Committee Members:
  • Luis F Ayala H, Thesis Advisor
  • Hamid Emami-Meybodi, Thesis Advisor
  • Turgay Ertekin, Committee Member
  • Flowback Analysis
  • Rate Transient Analysis (RTA)
  • Production Data Analysis (PDA)
  • Fracture Characterization
  • Two-phase Flowback Model
  • Single-phase Flowback Model
  • Shale Gas
  • Tight Sand Gas
Unconventional gas reservoirs, such as tight gas and shale gas, appear to have great potential to supply future demand for hydrocarbon. Economics of these reservoirs are tied closely to the performance of multi-fractured horizontal wells (MFHWs), which is the most direct indicator of stimulation effectiveness. Thus, greater understanding and analysis of the factors affecting performance of MFHWs are critical for the efficient exploitation of such reservoirs. Hydrocarbon production data analysis (PDA) techniques have been commonly used to characterize hydraulic fracture (HF) and, ultimately, to evaluate hydraulic-fracturing jobs. Recent studies have shown that rate transient analysis of flowback data can also provide early insight into HF attributes. While PDA methods seek long-time production data, flowback analysis can be conducted using early water and gas production data obtained immediately after the completion of stimulation jobs. However, in comparison with the long-term hydrocarbon production period, the physics of the process is more difficult to capture during flowback production because of its short duration, at which one or more flow regimes may occur. In addition, the flowback flow system could be single- or two-phase, depending on reservoir type. According to reported field data, single-phase flowback can be observed in tight sands, but two-phase flow is expected in the case of shale gas. Although various mathematical models have been proposed to analyze single-phase (water) and two-phase (gas and water) flowback data, analytical models for interpretation of data are still at an early stage of development. The objectives of this study are first to reproduce the relevant analytical models available in literature and understand their advantages and limitations; then, to develop single-phase and two-phase analytical models capable of predicting HF attributes such as fracture half-length and fracture permeability using early water and gas production data. In this study, a set of numerical simulations was conducted using CMG (IMEX) to examine the capacity of available mathematical models. It was found that most of the single-phase flowback models in the literature are accurate only under pseudo steady-state conditions, where a boundary-dominated flow regime with a constant production rate has been established. Another limitation of current models is that they can only estimate one fracture attributes: kf or xf. Knowing the shortcomings of current models, I developed a set of analytical models for both single- and two-phase systems, which were validated against numerical simulations. The single-phase model can closely estimate HF attributes, such as permeability and half-length under constant pressure as well as constant flowrate condition, for both transient and boundary dominated flow periods. Furthermore, I extended the developed single-phase model to variable bottomhole conditions by employing superposition principle. In the case of two-phase flow system, I developed an analytical model under fracture depletion mechanism for both early gas production (EGP) and late gas production (LGP) periods. In the case of EGP, gas flux from matrix to HF is assumed to be negligible. Comparisons of numerical results with those obtained from the analytical model show that the developed two-phase model for EGP can accurately predict fracture attributes. In the case of LGP, a coupled model is developed to include the effect of gas influx from matrix to HF on flowback data, where a uniform pressure decline rate is assumed in fracture-matrix system. The two-phase model has the advantage of linear behavior of water properties and avoids the computational complexity. With typical Barnett shale properties input in the numerical simulation, the analytical model can accurately estimate fracture attributes within a 10% error margin. Sensitivity analyses of fracture conductivity and initial water saturation in fracture have been conducted to illustrate the validity of two-phase flowback model applied in LGP. The results reveal that, within the physical range of fracture conductivity and initial water saturation, the two-phase flowback model can accurately evaluate fracture attributes. However, the model is more accurate for cases with smaller fracture conductivity and higher initial water saturation in fracture.