Robinson, Judy (Rutgers University Newark) | Slater, Lee (Rutgers University Newark) | Keating, Kristina (Rutgers University Newark) | Robinson, Tonian (Rutgers University Newark) | Parker, Beth (University of Guelph) | Rose, Carla (University of Guelph) | Prasad, Manika (Colorado School of Mines)
Summary Geophysical length scales derived from complex resistivity (CR) and nuclear magnetic resonance (NMR) measurements offer promise for the noninvasive estimation of permeability. Petrophysical relationships linking NMR length-scales to permeability have been used in both sandstone and mudstone formations. In contrast, petrophysical relations linking CR length-scales to permeability have been limited to sandstones and unconsolidated sediments. The purpose of our study was to evaluate the predictive capability of CRbased permeability relations when applied to mudstones and to compare the predictions against those determined from NMR models. Introduction The robust estimation of permeability from geophysical measurements remains one of the grand challenges in hydrogeophysics.
In this laboratory study, permeability models using spectral induced polarization (SIP) and nuclear magnetic resonance (NMR) data are tested on a large set of sandstone cores in order to compare the effectiveness of each method. Models from the literature are first tested to assess and compare their applicability to the sandstone cores. In order to correct for the effect of variations in mineralogy on the models, the samples are split into two distinct groups: cores without large amounts of iron oxides and cores with visible concentrations of iron oxides. Established petrophysical models are used to determine fitting parameters that for each group of cores, suggesting that mineralogy is a controlling factor in SIP and NMR data. This simple categorization according to the mineralogy is enough to reduce scatter in the predicted permeability values. In general, the NMR models outperform the SIP models, although accounting for mineralogy improves the SIP models to a greater extent.
In hydrogeophysics, in situ permeability estimation remains a formidable challenge. Recently, significant progress has been made in developing spectral induced polarization (SIP) models of permeability in the laboratory (e.g. Revil and Florsch, 2010). However, these SIP models often provide inadequate predictions over a wide range of permeabilities and across different mineralogies, including clay and iron content (Revil et al., 2013). The result is that these models often are not able to predict the permeability of geologic material within an order of magnitude. Nuclear magnetic resonance (NMR) has been used by the oil industry for permeability prediction for almost fifty years (Seevers et al., 1966) and has, more recently, been adapted for hydrogeophysics applications (Behroozmand et al., 2014). While NMR estimates of permeability have been found to be reliable for reservoir characterization, this is largely due to the practice of calibrating NMR models to specific lithologies. The result is a model that may not be applicable for a wide range of lithologies, a situation common in hydrogeophysical applications.
In recent publications, the joint use of SIP and NMR to estimate hydraulic properties of rocks, such as permeability has been suggested (Müller et al., 2005; Revil 2013). However, before such a joint approach can be attempted, a better understanding of the limitations of each of these methods is necessary. Weller et al., (2010a) compared the ability of SIP and NMR data to estimate permeability for cretaceous sandstones and concluded that, because the SIP data were sensitive to anisotropy in some samples, the method produced better models. In their study, the authors used purely empirical models, limiting the potential applicability of their results to mechanistic and quasi-mechanistic models frequently used in the literature.
The volumetric water content (θ) of peat soils below the water table is largely controlled by the production of biogenic methane-rich gas bubbles that are subsequently released to the atmosphere, thereby having significant implications for carbon cycling. Geophysical methods have recently shown promise for improving studies of gas storage and release in peatlands. We investigated the relationship between dielectric permittivity and volumetric water content in organic peat soil using ground-penetrating radar. We developed a novel approach for controlling water content using a pressurized test chamber to reduce the volume of bubbles under high pressure as described by the ideal gas law. This method simulates the bubble-rich natural conditions much more closely than previous studies that utilized drying to vary water content. Our results cover a range of highly saturated peat that is commonly observed in poorly decomposed near-surface peat and we demonstrated that a linear model can be used to estimate water content in peat for a range of water contents (i.e. θ>90%). The data collected from samples taken from different peatlands suggests that it is possible to use our resulting model to convert dielectric permittivity extracted from ground-penetrating radar data into free-phase gas concentration via the water content.
Che-Alota, Vukenkeng (Oklahoma State University) | Atekwana, Estella (Oklahoma State University) | Atekwana, Eliot (Oklahoma State University) | Sauck, William A. (Western Michigan University) | Rossbach, Silvia (Western Michigan University) | Werkema, Dale (U.S. EPA, ORD, NERL, ESD, CMB) | Davis, Caroline (University of Missouri Rolla) | Nolan, Jay (Rutgers University) | Slater, Lee (Rutgers University)
Britt, Larry Kevin (NSI Technologies, Inc.) | Dunn-Norman, Shari (U. of Missouri Rolla) | Smith, Michael Berry (NSI Technologies, Inc.) | Atekwana, Estella (University of Missouri-Rolla) | Slater, Lee (Rutgers University) | Gupta, Anuj (Lousiana State University) | Numbere, Daopu Thompson (U. of Missouri Rolla) | Fontana, J.V. (Direct Geochemical) | Viellenave, J.H. (Direct Geochemical) | Pelger, J. (J-Environmental, Inc.)
Normally, the objective of hydraulic fracturing is to design and execute a fracture treatment that achieves the desired fracture dimensions (length & conductivity) to maximize a wells production rate and reserve recovery. Treatments are most typically applied in reservoirs with in-situ stresses that yield vertical fractures. Case studies of horizontal fractures and identification of the key parameters unique to horizontal fractures are uncommon, in part due to the fact that oil and gas reservoirs occur in relatively deep subsurface formations.
A research project, sponsored by the U.S. Department of Energy, was undertaken to demonstrate a development method for the significant heavy oil reserves that exist at ultra-shallow depth in the Pennsylvanian sands in Southwestern Missouri and Southeastern Kansas. The principal objective is to demonstrate an economically viable and sustainable method of producing the shallow heavy oil using a combination of microbial enhanced oil recovery (MEOR) treatments and horizontal fracturing in vertical wells. In this application, the purpose of hydraulic fracturing is to expose additional reservoir surface area for subsequent microbial injection. Once the hydrocarbons are contacted by the microbes, the horizontal fractures are to provide a conductive pathway back to the wellbore for the lower viscosity hydrocarbons to flow.
Two wells in the study area, Fauvergue 1 and 5, were cased, perforated, and hydraulically fracture stimulated. Analysis of the treatments with surface tiltmeters confirmed that horizontal fractures were created. Geomechanical studies conducted prior to fracturing confirmed that the Pennsylvanian sands are more competent and have a greater Young's Modulus than previously reported in the literature.
This paper provides documentation of the extensive laboratory study conducted to develop an understanding of the elastic properties of the Warner and Blue Jacket sandstones and shale, fracture fluid sensitivity, and embedment testing.
The paper also reports the horizontal fracture stimulation execution and pressure analysis. Surface tiltmeter data are reviewed with the hydraulic fracturing performance, to determine whether the treatment objectives were achieved.
Heavy oil exists in the shallow Pennsylvanian sands in Southwestern Missouri and Southeastern Kansas. These sands occur over an area of about 8,000 sq. miles that extends for 250 miles along the Kansas-Missouri border as shown in Figure 1. The area reaches a width of roughly 80 miles and covers portions of the Northeast Oklahoma Platform, the Cherokee Basin and the Forest City Basin.1 Heavy oil (ranging from 8o to 25o API) is found in deposits distributed throughout this trend most frequently in rocks of the Cherokee Group of the Middle Pennsylvanian age.2