Rapid measurements of heterogeneity in sandstones using low-field nuclear magnetic resonance

Fast spin echo MRI of reservoir core plugs with a variable field magnet

Fast Spin Echo MRI is now widely employed in biomedicine for proton density and T2 contrast imaging. Fast Spin Echo methods provide rapid data acquisition by employing multiple echoes to determine multiple k-space lines with single excitations. Due to the multi-exponential behavior of T2 in typical porous media, and the strong dependence of T2 on the details of the experiment, acquiring a proton density image with Fast Spin Echo methods requires favorable sample and acquisition parameters. In recent years, we have shown the value of pure phase encode Free Induction Decay based methods such as SPRITE. However, in a reservoir rock, a typical T2* is hundreds of µs, whereas a typical T2 is hundreds of ms. Hence, there is merit in considering spin echo-based MRI measurements such as the Fast Spin Echo for rock core plug studies. A variable field superconducting magnet was employed in this study. This is a new class of magnet for MR/MRI. These magnets have the flexibility of operation in the field range of 0.01 Tesla to 3 Tesla. This is advantageous when working with rock core plugs, as it allows one to maximize sample magnetization, by increasing the static field while controlling magnetic susceptibility mismatch effects, and thereby T2 and T2*, through reducing the static field. The magnetic fields employed in the study were 0.79, 1.5, and 3 Tesla. Measurements were undertaken on five brine-saturated reservoir rock core plugs (Bentheimer, Berea, Buff Berea, Nugget, and Wallace). The results show that Fast Spin Echo measurements are more sensitive than SPRITE methods in amenable samples and usually feature higher resolution. Quantification of saturation with Fast Spin Echo methods requires correction for T2 attenuation. The results also show that 3 Tesla is too high a static field in general for rock core MRI studies with either method. While the current study is focused on five representative reservoir rock cores, the conclusions which result are general for MRI of fluids in porous media.

A fluid specific dimension of confinement as a measure of wettability in porous media

The wettability of a reservoir rock is among the most important factors influencing oil recovery. This study presents how the application of a modified CPMG pulse sequence can determine the wettability of porous samples containing oil and water in the same pore network. This method can be applied as an average of the signal over the entire sample as well as with spatial resolution using slice selection. Correlations between dephasing in internal gradients (G₀²D) and the spectral linewidth (Δν) were used to calculate a fluid specific dimension of confinement (DOC) distribution for oil and water separately during forced displacement experiments performed inside the bore of an NMR spectrometer. Two samples, one water-wet and one oil-wet, were investigated. Data averaged over the entire sample show oil and water distribution at endpoint saturations (S_w=1,S_w,i and S_or) while spatially resolved data show which pores are initially occupied by the displacing fluid as well as the displacing efficiency. Both the data averaged over the entire sample and the spatially resolved data can, by itself, indicate wettability. As opposed to relaxation times as an indirect indication of wettability, the DOC distributions provide a direct fluid dimension, which more carefully describes the distribution of oil and water in the pore network.

Controlling susceptibility mismatch effects, signal lifetimes, and SNR through variation of B0 in MRI of rock core plugs

¹H relaxometry measurements of petroleum core plugs are commonly performed on low field magnets (<0.5 Tesla) to reduce the influence of magnetic susceptibility mismatch on measurements of the spin-spin relaxation time, T₂. The Signal to Noise Ratio (SNR) of the MR signal, however, generally decreases with lower magnetic fields. Higher magnetic fields (>3 Tesla) are typically employed in small animal MRI studies to improve SNR and image resolution. For many rock core plug samples, susceptibility mismatch effects can be severe at these higher fields leading to decreased T₂ and T₂*. In this work we seek an answer to the general question of what is the best field for MRI of rock core plugs, anticipating that it will be both sample and measurement method dependent. Free Induction Decay (FID) relaxation time measurements were undertaken to investigate the conditions under which the SNR in Centric Scan SPRITE (Single Point Ramped Imaging with T₁ Enhancement) MRI measurements is maximized. The image SNR benefits from greater signal at higher fields, but is negatively impacted by the correspondingly shorter signal lifetimes. Depending on the noise regime of the sample, the maximum SNR may be predicted for Centric Scan SPRITE MRI with T₂* being B₀ field dependent. In this work we describe a series of simple experimental considerations to determine the optimal B₀ field for SPRITE MRI. Selection of the best field is aided by a new generation of superconducting magnets which allows the experimentalist to readily vary the field strength. Such magnets allow one to experimentally control sample magnetization for high sensitivity MRI measurements of core plug samples, while controlling the effect of susceptibility mismatch on the signal lifetimes.

Investigating pore geometry in heterogeneous porous samples using spatially resolved G0-Δχapp and G0-Δν correlations

This study presents a two-dimensional NMR pulse sequence for obtaining spatially resolved correlations between magnetic susceptibility induced internal gradients (G₀), and both the apparent difference in magnetic susceptibility (Δχ_app) and spectral frequency (Δν). G₀-Δχ_app correlations were utilized to generate spatially resolved pore size distributions, while the G₀-Δν correlations were used to additionally evaluate sample heterogeneity. The spatially resolved measurements were performed on a water saturated heterogeneous porous sample which contains one layer of 5-50 μm glass spheres (top layer) and one layer of 140-165 μm glass spheres (bottom layer). The slice selection was validated by applying the pulse sequence on a liquid mineral oil and water sample as well as on the porous sample. The resulting spatially resolved pore size distributions show very good agreement with results from our previously published non slice selective pulse sequence in the 140-165 μm glass spheres and good agreement in the 5-50 μm glass spheres. The G₀-Δν correlations correctly indicate a slightly higher degree of heterogeneity in the 5-50 μm glass spheres compared to the 140-165 μm glass spheres.

Quantitative Analysis of Micron-Scale and Nano-Scale Pore Throat Characteristics of Tight Sandstone Using Matlab

Based on micro-scale casting thin sections, nano-scale SEM images, and the pore distribution map identified through a binary image in Matlab, the pore size distribution and pore throat coordination number of the strata of Upper Paleozoic He8 section tight sandstone in the southeastern Ordos Basin were quantitatively analyzed with the above experimental data. In combination with a high-pressure mercury injection experiment, the pore throat distribution, the pore throat ratio, and the relationships between the characteristics, parameters, and pore permeability were investigated clearly. The results show that the tight sandstone pore space in the study area is dominated by micron-sized intergranular pores, dissolved pores, and intragranular pores. The nano-scale pore throat consisted of clay minerals, intercrystalline pores, and the flake intergranular pores of overgrowth quartz grains. Kaolinite and illite intercrystalline pores occupy the pore space below 600 nm, while the ones above 800 nm are mainly dominated by the intergranular pores of overgrowth quartz grains, and the 600–800 nm ones are transitional zones. The permeability of tight sandstone increases with the average pore throat radius, sorting coefficient, median pore throat radius, and average pore throat number. The porosity is positively correlated with the average pore radius and the average pore throat coordination number, and negatively correlated with the median pore throat radius.

An MR/MRI compatible core holder with the RF probe immersed in the confining fluid

An open frame RF probe for high pressure and high temperature MR/MRI measurements was designed, fabricated, and tested. The open frame RF probe was installed inside an MR/MRI compatible metallic core holder, withstanding a maximum pressure and temperature of 5000 psi and 80 °C. The open frame RF probe was tunable for both ¹H and ¹⁹F resonance frequencies with a 0.2 T static magnetic field. The open frame structure was based on simple pillars of PEEK polymer upon which the RF probe was wound. The RF probe was immersed in the high pressure confining fluid during operation. The open frame structure simplified fabrication of the RF probe and significantly reduced the amount of polymeric materials in the core holder. This minimized the MR background signal detected. Phase encoding MRI methods were employed to map the spin density of a sulfur hexafluoride gas saturating a Berea core plug in the core holder. The SF₆ was imaged as a high pressure gas and as a supercritical fluid.

Local diffusion and diffusion-T2 distribution measurements in porous media

Slice-selective pulsed field gradient (PFG) and PFG-T₂ measurements are developed to measure spatially-resolved molecular diffusion and diffusion-T₂ distributions. A spatially selective adiabatic inversion pulse was employed for slice-selection. The slice-selective pulse is able to select a coarse slice, on the order of 1cm, at an arbitrary position in the sample. The new method can be employed to characterize oil-water mixtures in porous media. The new technique has an inherent sensitivity advantage over phase encoding imaging based methods due to signal being localized from a thick slice. The method will be advantageous for magnetic resonance of porous media at low field where sensitivity is problematic. Experimental CPMG data, following PFG diffusion measurement, were compromised by a transient ΔB₀(t) field offset. The off resonance effects of ΔB₀(t) were examined by simulation. The ΔB₀ offset artifact in D-T₂ distribution measurements may be avoided by employing real data, instead of magnitude data.

Characterizing dispersivity and stagnation in porous media using NMR flow propagators

Low-field nuclear magnetic resonance (NMR) displacement probability distributions (flow propagators) are presented for water flowing through heterogeneous porous materials. Four sedimentary rocks have been chosen as example systems: Dolostone, Bentheimer sandstone, Berea sandstone, and Indiana limestone (in order of decreasing permeability). The fluid displacement is characterized by pre-asymptotic Stokes' flow and so the probability distributions are bimodal, with peaks corresponding to stagnant fluid in dead-end pores and flowing fluid in the connected porosity. Cut-off Gaussian functions are used to fit the flowing and stagnant peaks independently. An effective dispersivity length scale Lv (also known as the mixing length scale) is estimated by fitting the portion of the probability distribution corresponding to the flowing fluid. For the relatively homogeneous Bentheimer sandstone, the ratio of effective dispersivity length scale to effective transport diameter dt is Lv/dt≈16, which is an order of magnitude larger than for randomly packed glass beads where Lv/dt≈1.8. We compare these dispersivity parameters to similar values extracted from a cumulant analysis of the entire propagator. Fitting a cut-off Gaussian avoids the usual complications of analyzing dispersion in the presence of the ubiquitous stagnant fluid, and results in a clear demonstration of the influence of long-range heterogeneities on the dispersivity for flow in real sedimentary rocks.

Direct correlation of internal gradients and pore size distributions with low field NMR

Internal magnetic field gradients Gint, which arise from the magnetic susceptibility difference Δχ between solid matrix and fluid in porous media relate to the pore geometry. However, this relationship is complex and not well understood. Here we correlate internal-gradient distributions to pore-size distributions directly to examine internal gradients in detail at low field NMR. The pore-size distributions were obtained by the method of Decay due to Diffusion in the Internal Field (DDIF), and the internal-gradient distributions were measured with the Carr-Purcell-Meiboom-Gill (CPMG) method. The internal-gradient-pore-size distributions correlation maps were obtained for water in packs of glass beads with different diameter and in a sandstone sample. The relationship between internal gradients and pore structure is analyzed in detail by considering the restricted diffusion of fluids in porous samples. For each case diffusion regimes are assigned by plotting normalized CPMG data and comparing the diffusion lengths, the dephasing lengths and pore diameters. In the free-diffusion limit, the correlation maps reveal the true relationship between pore structure and internal gradients so that Δχ can be approximated from the correlation maps. This limit is met most easily at low field. It provides information about porous media, which is expected to benefit the oil industry, in particular NMR well logging.

A multi-dimensional experiment for characterization of pore structure heterogeneity using NMR

In a liquid saturated porous sample the spatial inhomogeneous internal magnetic field in general depends on the strength of the static magnetic field, the differences in magnetic susceptibilities, but also on the geometry of the porous network. To thoroughly investigate how the internal field can be used to determine various properties of the porous structure, we present a novel multi-dimensional NMR experiment that enables us to measure several dynamic correlations in one experiment, and where all of the correlations involve the internal magnetic field and its dependence on the geometry of the porous network. (Correlations: internal gradient - pore size, internal gradient - magnetic susceptibility difference, internal gradient - longitudinal relaxation, longitudinal relaxation - magnetic susceptibility difference.) It is always a spatial average of the internal magnetic field, or one of the related properties, that is measured, which is important to take into consideration when analyzing the obtained results. We demonstrate how these correlations can be an indicator for pore structure heterogeneity, and focus in particular on how the effect from spatial averaging can be evaluated and taken into account in the different cases.

Real-time oil-saturation monitoring in rock cores with low-field NMR

Nuclear magnetic resonance (NMR) provides a powerful suite of tools for studying oil in reservoir core plugs at the laboratory scale. Low-field magnets are preferred for well-log calibration and to minimize magnetic-susceptibility-induced internal gradients in the porous medium. We demonstrate that careful data processing, combined with prior knowledge of the sample properties, enables real-time acquisition and interpretation of saturation state (relative amount of oil and water in the pores of a rock). Robust discrimination of oil and brine is achieved with diffusion weighting. We use this real-time analysis to monitor the forced displacement of oil from porous materials (sintered glass beads and sandstones) and to generate capillary desaturation curves. The real-time output enables in situ modification of the flood protocol and accurate control of the saturation state prior to the acquisition of standard NMR core analysis data, such as diffusion-relaxation correlations. Although applications to oil recovery and core analysis are demonstrated, the implementation highlights the general practicality of low-field NMR as an inline sensor for real-time industrial process control.

Gint2D-T2 correlation NMR of porous media

The internal magnetic field gradient induced in porous media by magnetic susceptibility differences at material interfaces impacts diffusion measurements in particular at high magnetic field and can be used to probe the pore structure. Insight about the relationship between pore space and internal gradient G(int) can be obtained from 2D Laplace NMR experiments. When measuring distributions of transverse relaxation times T(2) in fluid filled porous media, relaxation and diffusion in internal gradients arise simultaneously and data are often interpreted with the assumption that one or the other parameter be constant throughout the sample. To examine this assumption we measure correlations of the distributions of G(int)(2)D and T(2) by 2D Laplace NMR for three different kinds of samples, glass beads with different bead diameters saturated with water, glass beads filled with oil and water, and a wet mortar sample. For the first two samples the cases where either the internal gradient or diffusion dominates were examined separately in order to better understand the relationship between G(int) and D. These results are useful for assessing the impact of internal gradients and diffusion in unknown samples, such as the mortar sample. The experiments were performed at different magnetic field strengths corresponding to 300 MHz and 700 MHz (1)H Larmor frequency to identify the impact of the magnetic field on the internal gradient. Subsequently, spatially resolved Gint(2)D-T(2) maps were obtained to study the sample heterogeneity.

Contributed review: nuclear magnetic resonance core analysis at 0.3 T

Nuclear magnetic resonance (NMR) provides a powerful toolbox for petrophysical characterization of reservoir core plugs and fluids in the laboratory. Previously, there has been considerable focus on low field magnet technology for well log calibration. Now there is renewed interest in the study of reservoir samples using stronger magnets to complement these standard NMR measurements. Here, the capabilities of an imaging magnet with a field strength of 0.3 T (corresponding to 12.9 MHz for proton) are reviewed in the context of reservoir core analysis. Quantitative estimates of porosity (saturation) and pore size distributions are obtained under favorable conditions (e.g., in carbonates), with the added advantage of multidimensional imaging, detection of lower gyromagnetic ratio nuclei, and short probe recovery times that make the system suitable for shale studies. Intermediate field instruments provide quantitative porosity maps of rock plugs that cannot be obtained using high field medical scanners due to the field-dependent susceptibility contrast in the porous medium. Example data are presented that highlight the potential applications of an intermediate field imaging instrument as a complement to low field instruments in core analysis and for materials science studies in general.