Unconventional natural gas systems include fractured shale gas (FSG), tight gas sands (TGS), basin center gas (BCG), shallow basin methane (SBM), and coalbed methane (CBM). Recently, more operators are focusing attention on shale reservoirs. The most notable shale play being developed in the north Texas region is the Barnett shale. This success has encouraged operators to investigate producing potential of the Woodford and Caney shales in Oklahoma. Shale plays are unique in that they often are both the source rock and producing rock contained in the same package. This duality leads to difficulty in log and reservoir interpretation. To date, conventional log interpretation has proved inadequate in identifying producing potential. Simply perforating areas of high porosity and pumping massive hydraulic fracture (MHF) treatments do not always yield commercial production results. Identifying areas of high producing potential using gamma ray, density, resistivity, and sonic transit time to locate high total organic carbon (TOC) has also yielded mixed production results. We believe the addition of mechanical properties of the rock can help identify shale intervals with a high propensity to contain natural fractures and a high probability to create a fracture network during hydraulic fracturing. We propose a linkage between the mechanical properties of the rock and the hydraulic fracture network created during a MHF treatment and the resulting production outcome. Desirable combinations of mechanical properties are selected to help locate areas in the shale that have a propensity to fracture as a network with sufficient aerial extent to impact production results. We use these mechanical properties in addition to TOC and porosity to select fracture initiation sites and give a qualitative assessment of producing capacity. In this paper, we describe calculated log parameters that illustrate this technique. Introduction Finding areas of productive shale is difficult because most logging techniques are based on conventional formations and are calibrated for standard rock types such as limestone, dolomite, or sandstone. The mineral composition of shale is very complex and variation in density, resistivity, and radioactive material content can cause serious errors in porosity and saturation calculations. Specialized logs are used to identify areas of possible hydrocarbon production by locating high total organic carbon (TOC), mature kerogen type III organic material, mineral composition, and areas prone to fracture network development. Identification of these areas is key to successful completion of a shale play. Obtaining a reasonable estimate of shale reservoir quality through logging has evolved by modifying standard log interpretation to fit the complex producing units we call shale. Our initial conceptual model of a shale play as a homogeneous clay-rich, organic source rock is not a valid model. Instead we encounter a multi-layered litho package, comprised of quartz, dolomite, lime, chert, and feldspar, sandwiched in a black shale bulk matrix. These are the shale productive units that most typify commercial Oklahoma shale. Productive shale plays are unique in that they are source rock, reservoir rock, and trap. The primary storage and producing mechanism of shale is a topic of heated debate. In general, most agree that there is a "free" and "adsorbed" component of hydrocarbon present in most shale plays. Exactly where the hydrocarbon is and how it flows out of the rock is still a topic of research. A conceptual model that seems to have some degree of acceptance is one where the "free gas" is stored and produced from micro-porosity in lamina and natural fractures and the "adsorbed gas" is stored and produced from the bulk shale matrix (Fig. 1). The importance of high-resolution log interpretation calibrated to laboratory measurement of mineralogy, organic content, maturation, gas saturation, and mechanical properties is apparent when dealing with multiple storage/flow mechanisms encountered in shale plays. The following will discuss log parameters used to help determine the quality of a producing shale unit. Production potential will be viewed as a mix of geology, geochemistry, mineralogy, hydrocarbon storage and flow, as well as, rock mechanics and its effect on fracture network creation. A commercial shale play will have the necessary mix of all these attributes, and each will have its own impact on the ultimate production.

Unconventional reservoirs require an effective system of completion. A major difference between unconventional and conventional reservoirs is the plastic nature of the rock itself. Within a very short period of time, certain portions of the reservoir will flow, sealing off the productive mechanism and capacity of the formation. Complicating matters even more is the near universality of horizontal drilling in these reservoirs.Dipole sonic logs have been used for many years to establish rock mechanical properties for conventional formations. The use of Young's Modulus and Poisson's ratio in frac design is well established. This same relationship exists in the unconventional world, but the complication of plasticity in the formation makes the relationship more difficult to understand. Traditional twodimensional models have not been able to adequately identify intervals of the horizontal reservoir that could maintain the fracture treatment.The calculation of three-dimensional rock mechanical properties has been advanced as a way to overcome this difficulty. This paper presents the application of this advancement in unconventional reservoirs. This paper will also discuss the differences in two-dimensional and three-dimensional data and the effects of ignoring this information.

Unconventional reservoirs require unconventional solutions. Horizontal drilling into these reservoirs followed by intensive perforating and fracture treatment techniques have been the accepted standard for producing these wells. The general thought process was that the closer these perforations could be placed and treated, the greater the production rates.Recent papers that show the results of production logs in the horizontal sections of these wells present a dilemma. Even though many different intervals have been attempted in the wells, only one or two of the intervals in each well are actually producing-either the completions are ineffective, or the production logs are suspect.In addition, data published recently also show the plastic nature of these unconventional reservoirs. In one study, the rock was placed in three-dimensional stress conditions and then perforated. Within 48 hours, the perforation tunnel had closed off, eliminating the ability of fluids to migrate to the well, which could also be a source of decreased production. A technique was required to find intervals of the well that could support open perforations to maintain contact of the formation to the well.Magnetic resonance imaging logs were added to the logging program to establish additional parameters that could be used for reservoir description. The Bray-Smith equation, which has been used for several years to describe permeability, was used as a base for adjustments. The observation for the relationship of the bin-size definition to permeability was established through laboratory NMR work. The algorithm was adjusted to apply specifically to unconventional reservoirs. We have already used this relationship to select additional intervals for perforation and treatment and have eliminated other sections that would have been completed under the previous completion programs.In this paper, we present the base case for this relationship and establish the general relationship between bin size and unconventional permeability. We expect to publish conclusive relationships of unconventional reservoir permeability to production rates as we acquire a history of data.

These two fields in Washita County, Oklahoma present extensive traditional log interpretation problems. The wells contain a variety of sandstone reservoirs, including Britt, Boatwright, Cunningham and Schneberger. Each provides unique log interpretation issues. A clear definition of log derived porosity is extremely difficult for these reservoirs in the best of circumstances. Permeability is also an issue. Well to well correlations are unreliable due to the variability within the formations. In many cases, the expectation for the performance of the well was subject to little science and much conjecture. In order to arrive at effective porosity and an estimate of permeability, NMR logs were added to the traditional logging program. At the time of logging, the technique employed for analysis of this log relied heavily on work done in the area. Parameters for analysis of a specific reservoir in a specific area were held constant with raw measurements of the tool considered to make changes to the base model. The Bray-Smith permeability model was applied to present estimates of permeability utilizing only the data generated by the NMR device. This paper attempts to compare the results of this new permeability measurement technique with the production actually observed from this field. We will compare and contrast the results of three methods of developing permeability from NMR measurements. These will all be compared to the actual results achieved in the wells in the Rocky and Ammunition fields in Western Oklahoma.