Fracture stimulation production response coupled with the hydrocarbon sales price determines the value of a fracture stimulation treatment. Many factors can significantly effect the production response of a fracture stimulated well. Some examples include stimulation fluid selection, proppant selection, pumping rates, rock properties, reservoir fluid properties, in-situ stresses, stress variations, on-site execution, post-treatment stimulation fluid recovery, and operating practices. The production response in economic terms portrays the net effect of these variables. This paper presents a case study that demonstrates how post-treatment evaluations expressed in economic terms can be used to assess the performance of stimulations and to guide future design choices. Introduction Methods of evaluating fracture stimulation treatments range from comparing offset well performance, comparing pre- and post-treatment well tests and/or production response, to using type curves, analytical models, and numerical simulation. Some techniques even combine aspects of several of the above mentioned methods to increase the chances of arriving at a unique solution. The results of the methods range from relative performance comparisons to detailed analysis of reservoir and fracture flow parameters. While these assessments are useful to varying degrees, none of them attempt to make a meaningful evaluation and comparison in economic terms. The net present value versus infinite conductivity fracture half-length (NPV vs Xfi,) plot is one tool which can help to optimize fracture stimulation treatment designs. Ideally, an economic comparison of fracture stimulation treatments accounts for treatment costs, reservoir fluid and rock properties, operating practices, and the myriad of stimulation treatment variables. Although the treatment design may be based on subjective analysis of the NPV curve, maximizing NPV vs Xfi is a valid tool if good estimates for reservoir parameters can be made prior to the stimulation treatment. The fracture model selected will also influence the design expectations. The impact of this variable on pre-fracture design expectations is presented. The NPV vs Xfi plot can also be used, post-fracture, to assess the success of the stimulation treatment. By using reasonable ranges of reservoir parameters, the actual net present values and effective fracture half-lengths of similar wells can be plotted together for a rapid visual comparison. Including fracture treatment efficiency curves on the plot helps quantify the level of success of the treatment. In this study, the fracture treatment efficiency curves represent the NPV as a function of the ratio of the effective fracture half-length to the design fracture half-length. One minor concern in the post-fracture evaluation is the reliability of estimates for the infinite conductivity fracture half-length. Pressure build-up tests are expensive in terms of deferred production and the tests frequently are not uniquely interpretable. Constant rate draw-down tests can be difficult to conduct and also have a deferred production component. A more cost effective approach is needed. Evaluation of reservoir parameters and well geometry through analysis of production data has been demonstrated by several authors. Recording daily producing rates and pressures enables an analysis to be conducted early in the life of a well. The production data analysis should consider the expected reservoir parameters, the design and execution of the stimulation performed, and the operating practices. This approach is used to estimate the effective fracture half-lengths for this study. The goal is to maximize the value of a well or a field. Since optimizing stimulation design is an iterative process, a decision loop must be developed to evaluate and refine the assumptions. The decision loop starts with an initial design based on a specific economic goal. This is followed by careful execution of the treatment to ensure proppant placement. P. 45

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractThis paper presents an integrated static and dynamic evaluation for selecting an appropriate geostatistical modeling method that best represents the production from a tight, stacked fluvial sandstone reservoir. The critical parameters investigated include sand body geometry, water saturation permeability, and hydraulic fracture properties. This effort involves the construction of several static and dynamic models within a one-square mile section of the field.The study area contains about twenty-four wells with varying productivity or EUR, indicative of the reservoir heterogeneity. Several geological scenarios were developed to characterize facies/sand bodies, and multiple realizations of each scenario were investigated. Reservoir properties including porosity, water saturation and permeability were populated for each realization using sequential Gaussian simulation.The geological scenarios incorporated variations of three widely used facies modeling algorithms: fluvial object, defined object, and sequential indicator simulation. These algorithms have their pros and cons in representing the distribution of reservoir facies. The model constraints, including facies at wellbores, global facies proportion, vertical facies probability, and areal facies probability constructed from well data were held constant between the scenarios, and infill well locations were used to evaluate the static facies prediction.Reservoir dynamic models were used to rank individual scenario performance. The static and dynamic models share the same grid dimensions and no upscaling of reservoir properties was performed. The grid system and the recurrent data (completion, production and pressure histories) for the wells were consolidated in schedule files for input into a reservoir simulator. While there were no attempts to match the historical performance of the wells, boundary conditions and/or constraints derived from field operation were determined and used as production controls for the wells. Average hydraulic fracture properties were assigned to the fracture cells of each well, and gas production under depletion drive was simulated with the flow models thus built. The results were evaluated qualitatively and quantitatively to select the appropriate modeling workflow for constructing the static and dynamic models of the field.

Completion Optimization Through Advanced Stimulation Technology and Reservoir Analysis: A Case Study in the Red Fork Formation, Okeene Field, Major County, Oklahoma J.D. Harkrider, SPE, M.L. Middlebrook, SPE, C.H. Huffman, SPE, W.W. Aud, SPE, Integrated Petroleum Technologies, Inc.; G.A. Teer, SPE, Lomak Petroleum, Inc.; and J.T. Hansen, SPE, Gas Research Institute Abstract This paper illustrates the use of advanced stimulation technologies coupled with reservoir analysis to improve gas production from a low permeability formation. Modern stimulation techniques used include real-time treatment data analysis, stress profiling, three dimensional fracture modeling and fluid quality control procedures. Implementation of these technologies was based on an evaluation of previous and current completion and stimulation approaches in the study area. A statistical review was performed to characterize the reservoir and establish a baseline from which to compare results and quantify benefits of the completion optimization process. Part of the project was performed under the Gas Research Institute Advanced Stimulation Technology Deployment Program. Through the use of modern completion and stimulation practices, the operator was able to nearly double the average initial production rate in the Red Fork formation from 300 Mscf/d to over 600 Mscf/d. Ten year reserve estimates have increased about 38% from 390 MMscf to over 540 MMscf. Acceleration of reserves has allowed the operator to produce in less than 5 years the same amount of gas that was previously recovered in 13 years. The combination of improved reserve recovery and accelerated production has increased the discounted cashflow about 43%. Introduction This project, from the beginning to the end, attempted to integrate the complete package of engineering practices to optimize costs and results. A multi-phase program was outlined and included an initial phase of evaluating previous completion and stimulation approaches in the area. The following technologies and techniques were implemented in baselining previous results:–Integration of practical and theoretical considerations to evaluate prior completions.–Advanced 3-D fracture modeling of breakdown and fracture treatment pressure responses.–Reservoir simulation of production and pressure responses.–Iteration between fracture treatment and production response on all wells to achieve consistency of overall interpretation.–Establishment of a production response baseline from offset well history. Once the baseline analysis was completed, field deployment was implemented and included a continued evaluation and evolution of approaches. This phase employed the following technologies and techniques:–Intense surface and in-situ fluid and equipment quality control before and during each fracture treatment.–Advanced real-time evaluation of the treating pressure response on all treatments.–On-site, real-time integration of fluid and equipment quality control with pre-treatment diagnostics and main fracture treatment execution.–Pre-treatment diagnostics to identify closure pressure of the Red Fork and adjacent layers, observe the leakoff response of various fluids and determine the quality and complexity of the near-wellbore and far-field fracture geometry.–Real-time execution of fracture treatments to optimize near-wellbore and far-field proppant placement/conductivity.–A coupled approach to acquire both post-treatment pressure decline data, which yields a better understanding of the fracture treatment, and rapid flowback to enhance fracture conductivity and minimize formation damage. The final phase of the project was a cost benefit analysis. This comparative analysis of wells using modern completion practices to the offset production baseline quantified the benefits of optimization. The following were used in this phase of the project:–Comparison of long-term production response on new wells to previous wells. P. 357^

The ability to predict well inflow performance for varying well and reservoir conditions is important when optimizing production. Many methods exist to estimate a well’s current productive capacity (IPR curve) and extensions to the methods are available for predicting future well performance. The extensions to predict future inflow performance behavior account for changes in relative permeability and assume an average reservoir pressure. The applicability and accuracy of the methods depends on knowledge of reservoir parameters which may be difficult to obtain in low permeability reservoirs. Several authors have presented methods of analyzing and history matching well performance. These methods typically yield reservoir parameters which may be used in the well inflow performance methods in order to investigate the results of varying well production parameters. These methods are particularly useful in low permeability settings where interpretable welltest data may be difficult to obtain or prohibitively expensive. Currently, the analytical history matching approach is most accurate when applied to single-phase systems. Predictions of black oil reservoir performance below the bubble point can exhibit large error since depletion of the total reservoir energy is not accounted for using the constant gas-oil ratio approach typical for these methods. This paper presents a method to analyze well performance of black oil systems below the bubble point. The method incorporates a material balance approach to account for changing gas-oil ratios as the reservoir is depleted. Prediction of future well performance is also presented. Along with reservoir characterization, another benefit of the method is the ability to construct IPR curves at any point in order to optimize production. The proposed method uses a pseudo pressure transform to account for changes in fluid properties as the reservoir pressure is depleted. Relative permeability changes can be incorporated in the pseudo pressure transform. Comparisons to finite difference simulation results and actual productio data are presented. Comparisons of future IPR curves generated by other methods are also presented.

No abstract