Gas shales have become an important resource for the production of hydrocarbons in North America, and are being explored as a resource on other continents as well, based on their rapidly increased importance to the North American market and promise to boost domestic production elsewhere. There are numerous pricing and geopolitical reasons for this active exploration, but regardless of where in the world they are being explored, gas shales share some fundamental properties that make them virtually impossible to analyze with conventional core analysis methods or conventional petrophysical models based on log data. These properties are basically that gas shales are tight, with permeabilities in the 10s to 100s of nD, have low (effective) porosity, typically less than 10%, and have high kerogen and clay content. While there are some variations of these themes (e.g. shales with higher detrital input, making them siltier or silty-laminated), in general the tightness of the rock and abundance of clay minerals and kerogen pervades, and that causes a number of challenges to analysis. We have developed analytical methods for evaluating these reservoirs on core by using crushed material to enable better access to the pore space, retort analysis to measure separately free, bound and structural water saturations and also distinguish water from oil, and pressure transient analyses for the determination of permeability. Conventional core analyses (e.g. Dean Stark), applied to kerogen- and clay-rich rocks fails in separating free from bound waters and water from light oils, thereby missing critical inputs into calculating effective saturations, effective porosities and clay-bound water volume. In addition, the amount of oil recovered from retort, as an independent quantity, can be a useful proxy for kerogen maturity. From a permeability standpoint, unconventional reservoirs are usually too tight to allow for steady-state permeability measurements, and microfracturing is often too pervasive to allow for reliable permeability measurements on whole plug samples. As a result, we have developed a crushed-sample pressure decay system to measure the nD permeabilities typical of shales, and a stepped confinement pulse-decay method for measuring micro-Darcy (and higher) permeabilities in more texturally complex, siltier or sandier unconventional reservoirs that typically have these higher permeabilities.

Due to the complexity of tight shale reservoirs, core analysis has become an increasingly important source of data for evaluating these systems. However, as there are no generally agreed upon testing protocols, there are competing methods for obtaining such primary data as fluid saturations and porosity. The two most commonly employed commercial methods are Dean Stark toluene combined thermal and solvent extraction and thermal extraction by retort. However, the impact of these protocols on the rock and its fluid phases is different, primarily due to the abundance of clays. While the Dean Stark extraction produces a total porosity and total water saturation, data suggest significantly elevated values of these parameters relative to what is measured through the retort process, resulting from significantly higher amounts of water recovered. This distinction is fundamentally important for using core analyses in shale for calibrating logs and/or determining reserves, as both methods claim to report the same parameters. To understand this effect, we have assembled a data set of compatible core analyses from various laboratories from several wells for a tight-gas shale reservoir in the United States. In addition, we conducted thermogravimetric analysis and Karl Fischer Titration with methanol extraction on splits of the same samples. The retort, thermogravimetric, and Karl Fischer data generally agree in the amount of water eluted from the samples (per gram of rock), while the Dean Stark data show significantly more water. We suggest this excess water could be a portion of the structural water in the clays, which should not contribute to porosity and saturation. Additionally, there is a relationship between this excess water and the total clay content from XRD. This correlation to XRD analyses suggests that a correction can be determined, leading to more accurate porosity and saturation values necessary for proper reserves estimations.

Proper characterization of reservoir quality of tight shales is critical for evaluating reservoir potential. These reservoir quality properties typically include hydrocarbon filled porosity, permeability, organic content and maturation, and pore pressure. Of these, permeability measurements are among the most complex to obtain, and have been subject to much discussion. Key concerns are the lack of analytic modeling, poor reliability, poor consistency, and ignoring stress sensitivity in the measurements. This paper reviews the pressure decay permeability method using crushed rock, and includes laboratory test results to validate the findings. Part of the review is a numerical model for the pressure decay method. This model includes significant processes of pipe flow in the equipment, thermal effects, diffusion into rock fragments, Klinkenberg effects, and size, shape and anisotropic permeability of the fragments.The paper shows that the measured permeability stress dependence, in tight shales, arises from coring induced microcracks. These result from failure of weak contact planes that are naturally occurring within tight shales and fail during coring and core retrieval. Permeability stress dependence in-situ is slight, as is show by compression test measurements to greater than thirty thousand psi. Crushing the rock to create small fragments for permeability measurements effectively removes these microcracks, and allows evaluation of real in-situ properties. Alternatively, closing microcracks by applying confining stress on plug samples, as routinely done for steady state and pulse decay measurements, is possible, but problematic because the critical stress required for microcrack closure changes from rock to rock facies.The pressure decay permeability method on crushed rock is shown to provide very consistent results that agree with other measurement techniques. The numerical model relates specific ranges of fragment sizes and testing conditions, to measured ranges of permeability. This allows permeability measurements and numerical model analysis for a broad range of variability in permeability that can be measured in heterogeneous tight shales. With some exceptions (e.g., Cui et.al. 2009), the fundamental understanding of the petrophysical properties of tight shales have not previously included rigorous confirmation of experimental measurements by analytical methods.

The M4 field is located in the North of the Central Luconia Province in the Sarawak Basin, East Malaysia. The reservoir is approximately 2000 m below sea-level where the water depth is approximately 120m. A study for CO2 geological storage has been carried out to address the feasibility of injecting and storing CO2 in the M4 depleted carbonate gas reservoir using 3D coupled geomechanical modeling. The water level in the reservoir has risen close to the cap-rock which implies a strong aquifer. Laboratory tests were carried out on core samples before and after injection of a CO2 saturated brine solution, and the results were used in the determination of material property strength and elastic property degradation due to acid-carbonate interaction. Triaxial compression tests on carbonate samples from three different depths at the peak loading stage of the tests for different confining pressures were performed. The effects of CO2 saturation on UCS, Young"s modulus and Poisson"s ratio were determined. The permeability measurements from the pore volume compressibility tests and permeability measurements obtained during the measurements of petrophysical properties were used to evaluate the effects of acidcarbonate interactions. The 3D geomechanical model coupled the reservoir pressures derived from a dynamic model with a stress simulator in order to calculate changes in effective stress and volumetric strain within the 3D model. The derived changes in volumetric strain were related to a change in porosity and permeability which were then passed back to the dynamic model in a staggered solution scheme. At each "stress step" in the solution process, a further modification was made to geomechanical material properties due to the increased CO2 saturation from injection. The material parameter modifications were based on the results of the laboratory tests above. In this way, the material property degradation due to CO2 injection was accounted for during the coupled reservoir geomechanical simulations. The paper discusses the results of these tests and the derived variation of material permeability, elasticity and strength parameters with CO2 saturation for subsequent input to the coupled geomechanical solution scheme. In this way, the potential risk due to the chemical interaction from CO2 injection could be evaluated quantitatively.