Stochastic Modeling of Two-Phase Flowback of Multi-Fractured Horizontal Wells to Estimate Hydraulic Fracture Properties and Forecast Production

Abstract

Although high-frequency fluid production and flowing pressures (hourly or greater) are commonly gathered in multi-fractured horizontal wells (MFHW), this data has rarely been used by industry in a quantitative manner to characterize hydraulic fracture or reservoir parameters and there has been ongoing debate about the usefulness of this data. It is likely that the multi-phase flow nature and the possibility of early data being dominated by wellbore storage have deterred many analysts.This work will expand on the flowback analysis work presented by Clarkson (2012b). Consistent with that work, our interpretation is that the early flowback data corresponds to wellbore + fracture volume depletion (storage) and it is assumed that fracture storage volume is much greater than wellbore storage. From this flow-regime, bulk permeability (dominated by fracture permeability) and effective fracture half-length can be estimated. However, as pointed out by Clarkson (2012b) there is a large degree of uncertainty in this type of analysis as a result of the number of unknowns which are being adjusted to provide an adequate history match. To better understand the uncertainty and the impact of each parameter, stochastic simulation was used to provide a range of parameter values, which provide an adequate fit of the data, and to determine which parameters have the greatest impact on the match. Stochastic simulation is also used to derive a long-term forecast using parameters derived from flowback analysis. Additional improvements over previous work include the consideration of different fracture geometries, the use of the matchstick model to estimate fracture permeability and additional constraints on relative permeability curve selection. The field cases presented by Clarkson (2012b) for shale gas reservoirs are reanalyzed for proof of concept and demonstration of the developed techniques.

Stochastic Modeling and Assisted History-Matching Using Multiple Techniques of Multi-Phase Flowback from Multi-Fractured Horizontal Tight Oil Wells

“…The same limits were used for the application of the assisted-history matching techniques, which will be discussed in the following section. This analysis is comparable to that conducted by [16], although the number of uncertain parameters was reduced from 10 to 5, placing the focus on the most important parameters. Using this approach allows reasonable coverage of the sample space with a smaller number of iterations.…”

“…In this work, several algorithms were tested for the purpose of uncertainty analysis and assisted history-matching of flowback data. In previous work, [16] [17] applied MC simulation to the same data set, although less iterations were conducted with significantly more input parameters and with wider bounds, bringing the results into question. The GAPS algorithm has also been tested by [4] for history-matching three-phase flowback with a numerical simulator and [3] attempted to use a combination of algorithms to try to decouple parameters in one of their analysis tools, although more rigorous methods could have been applied.…”

“…Alternate criteria, such as those applied by [16] [17], could also be used if a reasonable baseline deterministic match is not available. In these two papers, the authors used an R 2 value of each phase (in terms of rate) greater than 0.9 (as suggested by [18]) and a total cumulative production for each phase within 10% (as suggested by [19]) for each phase to determine a successful match.…”

“…In these two papers, the authors used an R 2 value of each phase (in terms of rate) greater than 0.9 (as suggested by [18]) and a total cumulative production for each phase within 10% (as suggested by [19]) for each phase to determine a successful match. This approach was shown to be relatively successful in the studies by [16] [17]. However, some successful solutions in both cases were found where a high R 2 value existed but the match was poor, due to the inherent nature of R 2 as a match fit indicator, especially when dealing with highly nonlinear problems with a large number of data points.…”

A semi‐analytical model for capturing dynamic behavior of hydraulic fractures during flowback period in tight oil reservoir

“…(b) The most accurate method for quantitatively analyzing the fracturing fluid flowback data is the numerical simulation method, 4,40 which can rigorously account for multiphase flow, gravity segregation, and pressure-dependent permeability. Many numerical flowback models 5,6 have been established to comprehensively describe the flow regimes and rigorously characterize the reservoir flowback. (c) The semi-analytical model describing the multiphase flow during flowback has received more attention because of its convenience and efficiency.…”