Production from the extremely low permeability reservoirs found in shale gas plays results in very large pressure gradients at and around the sand face. Finite-element modeling technology, in which a large number of closely-spaced nodes are placed near hydraulic fractures, is the best approach for replicating the reservoir and completion complexity required to model horizontal shale gas wells. The primary objective of this work is to demonstrate how the performance of three Haynesville Shale gas wells were modeled using this technology. Each of these three wells exhibits the large early production decline typically observed in the Haynesville. Case histories for these wells show how the finite-element software can be used to simulate the mechanisms thought to contribute to the observed production decline. The workflow used to model performance demonstrates how petrophysical, stimulation, and flow modeling are utilized to understand the key reservoir and completion mechanisms that affect production behavior of shale gas wells. IntroductionUnconventional reservoirs have gained prominence because they hold the supply potential necessary to achieve energy independence for many countries. It has always been a challenge to model the reservoir and completion aspects of unconventional reservoirs accurately. In his recent (2010) review, King pointed to the variability in shales, even within the same lateral, and to the need to collect sufficient data to optimize the design of shale completions and stimulations. Thompson (2010) analyzed well performance data from Haynesville Shale wells and determined that the phenomenon of pressuredependent permeability is responsible for performance behavior. Taylor (2010) used reservoir and fracture stimulation data in a shale oil field to determine the optimum fracture spacing.

The use of multi-stage fractured horizontal wells has made production from shale gas plays feasible. The production history from these wells is characterized by long transient linear transient flow. The key reservoir (nature) and completion (nurture) properties that impact flow behavior of the well are permeability, Original Hydrocarbon in Place (OHIP), number of hydraulic fractures, fracture area and fracture spacing. Understanding the interrelationship between these parameters is critical for optimal development of shale gas resource plays. The proposed shale gas workflow uses a hybrid analytical model. An analytics based diagnostic process analyzes well performance history and a numerical model validates the feasible history match models for representative forecasts. The model results allow the user to capture the range of uncertainty in estimating individual reservoir or completion properties, such as permeability, fracture area, fracture spacing, etc. The diagnostic process provides the relationship between key factors dictating well performance such as OHIP, effective fracture area, effective reservoir permeability, effective fracture spacing and well spacing. In this paper, wells from the Marcellus play, in Pennsylvania, USA are evaluated using the proposed shale gas workflow. Surveillance is an integral part of this workflow. The paper shows how surveillance can be used for resource management and exception management. Along with data mining, the workflow accelerates the learning curve to evaluate the effectiveness of current field practices. The results help with understanding the effectiveness of proppant pumped, the potential number of contributing clusters and production issues. Continued surveillance reduces the uncertainty surrounding all parameters. This knowledge base can then be used to optimize the asset development strategy, maximizing the return on investment (ROI).

A key characteristic in unconventional reservoirs development is that they need massive hydraulic fracturing to create high permeability conduits to connect the reservoir to the wellbore and assure appropriate flow rates to make the development economical. In general, production performance for this type of reservoirs show an early high inflow followed by a steep decline. Refracture jobs have been a common practice to mitigate the flow rate decline and revitalize wells productivity.In this paper, a wide range of possible refrac configurations are evaluated using synthetic models. Parameters such as fracture spacing, matrix permeability, fracture conductivity, fracture orientation and refrac locations were varied to study their impact on the success of the refrac job. Results showed significant impact on both the overall productivity, recovery and the economic value. The results show the impact of refracturing in higher permeability cases (greater than micro Darcy) is not as economical as in the low permeability scenarios (less than micro Darcy). Results also show a possible threshold in fracture spacing (60 -80 ft between clusters) below which refracturing may not an economical completion strategy. In scenarios where the initial fracture conductivity is severely degraded, refrac can be an alternative to improve the net present value (NPV) and estimated ultimate recovery (EUR). Under the conditions of this study, fracture re-orientation cases show improvement in recovery factor (RF) but does not have a significant impact on the overall NPV. The timing of refrac has a large impact on both the NPV and recovery for the wells. In all cases, NPV was used to define optimal condition for each scenario evaluated.All cases presented demonstrate potential impact of refracturing for a typical range of properties in shale reservoirs. The results shown in the paper could be used as a catalogue for better decision support to estimate the impact of refracturing in field cases under similar configurations to determine whether refrac will beneficial.

Over the last few decades, shale gas has become an increasingly important global source of natural gas, especially in the United States. According to Polczer (2009) and Krauss (2009), shale gas will greatly expand worldwide and is expected to supply as much as half the natural gas production in North America by 2020. Due to extremely low shale matrix permeability, shale is considered an unconventional source of gas and requires fractures to provide a flow path to the wellbore. Due to uncertainties in quantifying the gas-in-place and identifying flow behavior; estimating the ultimate recoveries in shale gas reservoirs requires new techniques. In this paper, we used four approaches to estimate the ultimate recovery in shale gas wells: two empirical methods (conventional and modified decline curve analysis), analytical modeling and numerical modeling. All four approaches were applied on wells from four different shale plays (Barnett, Haynesville, Marcellus and Woodford).

In shale plays, as with all reservoirs, it is desirable to achieve the optimal development strategies, particularly well spacing, as early as possible, to minimize loss of capital or reserves. The understanding of parameters influencing well spacing is a vital for the economic development of unconventional reservoirs. Previous papers on this subject have concentrated on unique history match solution, or a parametric study to evaluate optimal well spacing to maximize returns on investment. The optimal well spacing decision is a tradeoff between maximizing the ultimate recovery from an asset and the cost associated for that recovery. Development of shale gas resource will require drilling a large number of wells. Most of the shale gas reservoirs are early in their development cycle with very few wells having long term production data for error free forecasting. Shale gas wells have a long production life but most of the economic value of the well is recovered in the first few years of its life. During the field development it is critical for the operators to obtain a good understanding of the Stimulated Reservoir Volume (SRV) initially, and the contribution of the External Reservoir Volume (XRV) to the SRV in the long run. This paper presents a stochastic forward modeling workflow capturing uncertainties both in the reservoir and completion properties. The workflow was applied to evaluate an optimal number of wells required in a section in various shale gas resources in North America. The forecasted rates for all models are evaluated with an economical model to determine the optimal well placement in the section. Unlike the deterministic approach, advantage of the stochastic approach is in capturing the uncertainty in Net Present Value (NPV) by providing reasonable bounds for NPV that reflects the uncertainties associated with reservoir and completion parameters. Examples of application of this workflow in Marcellus, Woodford, Fayetteville and Haynesville shale gas resources are presented. The workflow discussed in this paper can be used by the operators in unconventional reservoirs to determine optimal well spacing and completion strategies earlier in the lives of these reservoirs, which could accelerate production and improve economic value of shale gas assets. Introduction The exploitation of unconventional gas reservoirs has become essential for the continuous gas supply of North America. Shale gas reservoirs have been explored and produced for over a decade and will continue to be a major energy supplier in the future. The boom in the development of the shale gas reservoirs, especially in North America (i.e. Haynesville, Marcellus, Woodford, Barnett), is being supported by the continuous development and improvement of the critical completion and stimulation technologies. The use of the horizontal drilling and advances in transverse fracturing of shale gas reservoirs are two revolutionary technologies that have improved the gas recovery and consequently the production economics. Understanding the morphology and growth of the hydraulic fractures is essential in the development of the shale gas resources. Based on the completion design and the generated stimulated reservoir volume, shale gas operators must decide on the optimal number of wells that need to be placed in a given area to maximize the ultimate gas recovery and economic return.