Surface Nuclear Magnetic Resonance (SNMR)

Basic Concept

The surface nuclear magnetic resonance (SNMR) method is a nonintrusive technique that is becoming increasingly more valuable in near-subsurface investigations. Nuclear magnetic resonance (NMR), which is also called magnetic resonance imaging (MRI), magnetic resonance sounding (MRS), and proton magnetic resonance, has been applied to many research fields. By recording magnetic resonance signals from proton-containing fluids (e.g., water, hydrocarbons), NMR methods directly estimate the volume and spatial distribution of such fluids.

Surface nuclear magnetic resonance data can be used to quantitatively estimate the porosity of fluid-filled pores, moisture content in the vadose zone, and pore size distribution. SNMR data analysis using empirical relations also allows for the estimation of hydraulic conductivity (K), which is a property vital to hydrogeological studies and models. Though introduced to geophysics as a hydrocarbon exploration technique, the NMR method is being widely applied to hydrogeological and environmental investigations.

Theory

The surface nuclear magnetic resonance method exploits the NMR-phenomenon exhibited by the magnetic nuclear spins of the protons within hydrogen atoms when perturbed in a magnetic field. An SNMR measurement involves three stages that are repeated multiple times: 1) equilibrate, 2) perturb, and 3) measure. Equilibrium conditions (i.e., introduction of background magnetic field (B₀) into the subsurface) are established using magnets within the tool. Though it is not necessary to know the parameters of B₀, it is typically assumed to be homogeneous.

The protons, which are polarized in alignment with the background magnetic field (B₀) during equilibrium, are subjected to an oscillating magnetic field (i.e., radio frequency (RF) pulses). The proton-perturbing RF pulses are produced by a high-voltage alternating current (AC) that is transmitted at the Larmor frequency (i.e., the resonant frequency of hydrogen nuclei). The RF pulse forces reorientation of the macroscopic magnetic moments of protons into a transverse plane (B₁) that is typically normal to B₀ (Yaramanci and Müller‐Petke, 2009).

This excitation causes the protons to generate an electromagnetic (EM) response signal. The perturbation is then terminated, and the protons are free to relax. Relaxation involves the precession of the protons about B₀, where the proton spin axes rotate around B₀ as their orientations gradually become more parallel to B₀. This creates a measurable magnetic field (i.e., free induction decay (FID)) that approaches the background over time and is used to interpret critical hydrologic properties (Dunn and others, 2002; Behroozmand and others, 2015).

The FID induces a voltage within a surface detection coil that records the signal (i.e., transverse decay (T2)) over time. The initial T2 strength is proportional to the amount of water in the measurement zone. The T2 decays with a timing that can be related to the size(s) of the pores in which water resides. Alternatively, T2, which is always less than T2, can be used to represent the FID measurement. Pulse sequences (e.g., CPMG) can be used to correct for the dephasing caused by magnetic field inhomogeneities and yield a T2 decay (Behroozmand and others, 2015).

In saturated conditions where pores are completely filled will fluid, the SNMR measurement represents total porosity. Otherwise, it represents water content as a volumetric percentage of material within the measurement zone. The SNMR signal is sensitive to the presence of water and its pore-surface interactions, which are related to pore surface reflexivity and size. Thus, water content can be separated into bound- and mobile-water using T₂ decay times, the cutoff of which is 31-33 milliseconds (Straley and others, 1997).

The relaxation time of bound water in small pores is shorter primarily due to effects of localized magnetic fields that are generated by unpaired electrons adjacent to pore surfaces. Unpaired electrons of certain paramagnetic atoms (e.g., manganese, iron) are attracted to the negatively charged clay-mineral surfaces that are abundant within less permeable formations. Mobile water in large pores have more space to precess and, due to the lack of interference by surface processes, take longer to relax (Straley and others, 1997; Hertrich, 2008).

Applications

The most basic SNMR survey setup collects a one-dimensional dataset and may be referred to as a magnetic resonance sounding (MRS). A two-dimensional SNMR survey, which is a type of profiling commonly referred to as magnetic resonance tomography (MRT), uses multiple overlapping detection loops with variable offset. Both MRS and MRT operate with a single loop of shielded wire placed on the surface that is designed to induce and measure the NMR signal.

Each signal received represents the superimposed magnetic effect of all transverse decay from multiple proton spins in a specific subsurface measurement volume. The measurement depth may be changed by applying various pulse sequences that focus the pulse moments at a specific depth range. Field parameters (e.g., loop size, shape, distance) affect the orientation and amplitudes of the electromagnetic fields distributed throughout the subsurface and are considered during survey design (Yaramanci and Müller‐Petke, 2009; Hertrich, 2008).

Like most electromagnetic geophysical methods, SNMR data are extremely sensitive to electrical and magnetic variations and noise sources. SNMR data are affected by the local characteristics of Earth’s magnetic field (i.e., strength and inclination), erroneous macro/microscopic magnetic fields, and distribution of electrical conductivity and magnetic susceptibility. Thus, prior to SNMR data collection, a comprehensive knowledge of site parameters proves beneficial for setting acquisition parameters, data analysis, and interpretation.

SNMR is unable to completely and accurately measure the signals produced by water within very tight pores because of their very early signal decay times. New technologies are being implemented to reduce the “dead time” between transmission and measurement that inhibits the detection of such types of water. Currently, however, very small pore structures with high-water content (e.g., clay) may appear to possess very low or zero water (Yaramanci and Müller‐Petke, 2009; Hertrich, 2008).

Even so, there exist ways to help mitigate all these effects in both the data collection and processing phases. Though there is still progress being made to maximize application, SNMR data have the potential to provide vital information about hydrogeologic properties. Empirical relations have been derived that relate NMR porosity and relaxation constants to hydraulic conductivity (K). The measured results of bound versus mobile water are used to estimate K using established empirical equations and site-specific parameters.

Hydraulic conductivity estimates are made using two unit-dependent equations: the Schlumberger-Doll research (SDR) equation (Kenyon and others, 1988) and the sum of echoes (SOE) equation (Allen and others, 2000). The SDR equation uses the measured values of total porosity (ϕ) and the mean log T₂ (MLT₂). The SOE equation uses the summed amplitudes of the echoes in the T₂ decay for each depth interval (Hertrich, 2008). Using these and other relations, the SNMR method has been successfully implemented in the following applications:

Water content (in vadose zone)/porosity (in saturated zone) characterization
Hydraulic conductivity estimation
Aquifer property evaluation
Vadose-zone and infiltration studies
Soil-moisture experiments
Constraining/ parameterizing groundwater flow models
Contamination studies/ environmental impact assessments
Environmental remediation
Hydrocarbon prospecting
Geotechnical assessments

Examples/Case studies

Becker, M.W., Pelc, M., Mazurchuk, R.V., and Spernyak, J., 2003, Magnetic resonance imaging of dense and light non‐aqueous phase liquid in a rock fracture: Geophysical Research Letters, v. 30, no. 12, p. 1646-1649, doi:10.1029/2003GL017375.

Abstract: Magnetic resonance (MR) imaging was used to observe the flow of dense (FC‐75) and light (dodecane) non‐aqueous phase liquids (NAPLs) through a water saturated dolomite fracture. Dynamic two‐phase behavior was influenced by (1) buoyancy of the NAPL relative to the aqueous phase, (2) fracture aperture distribution, and (3) alteration of wettability by long‐term presence of NAPL phase. MR imaging was capable of characterizing the fracture geometry and the fluid flow, but was limited by outlet flow conditions in the sample and acquisition times. This method permits observation of two‐phase flow under natural wettability and matrix porosity, providing significant advantages over plastic or glass replicas.

Abstract: For characterizing water flow in the vadose zone, the water retention curve (WRC) of the soil must be known. Because conventional WRC measurements demand much time and effort in the laboratory, alternative methods with shortened measurement duration are desired. The WRC can be estimated, for instance, from the cumulative pore size distribution (PSD) of the investigated material. Geophysical applications of nuclear magnetic resonance (NMR) relaxometry have successfully been applied to recover PSDs of sandstones and limestones. It is therefore expected that the multiexponential analysis of the NMR signal from water‐saturated loose sediments leads to a reliable estimation of the WRC. We propose an approach to estimate the WRC using the cumulative NMR relaxation time distribution and approximate it with the well‐known van‐Genuchten (VG) model. Thereby, the VG parameter n, which controls the curvature of the WRC, is of particular interest, because it is the essential parameter to predict the relative hydraulic conductivity. The NMR curves are calibrated with only two conventional WRC measurements, first, to determine the residual water content and, second, to define a fixed point that relates the relaxation time to a corresponding capillary pressure. We test our approach with natural and artificial soil samples and compare the NMR‐based results to WRC measurements using a pressure plate apparatus and to WRC predictions from the software ROSETTA. We found that for sandy soils n can reliably be estimated with NMR, whereas for samples with clay and silt contents higher than 10% the estimation fails. This is the case when the hydraulic properties of the soil are mainly controlled by the pore constrictions. For such samples, the sensitivity of the NMR method for the pore bodies hampers a plausible WRC estimation.

Haber-Pohlmeier, S., Vanderborght, J., and Pohlmeier, A., 2017, Quantitative mapping of solute accumulation in a soil‐root system by magnetic resonance imaging: Water Resources Research, v. 53, no. 8, p. 7469-7480, doi:10.1002/2017WR020832.

Abstract: Differential uptake of water and solutes by plant roots generates heterogeneous concentration distributions in soils. Noninvasive observations of root system architecture and concentration patterns therefore provide information about root water and solute uptake. We present the application of magnetic resonance imaging (MRI) to image and monitor root architecture and the distribution of a tracer, GdDTPA2− (Gadolinium‐diethylenetriaminepentacetate) noninvasively during an infiltration experiment in a soil column planted with white lupin. We show that inversion recovery preparation within the MRI imaging sequence can quantitatively map concentrations of a tracer in a complex root‐soil system. Instead of a simple T1 weighting, the procedure is extended by a wide range of inversion times to precisely map T1 and subsequently to cover a much broader concentration range of the solute. The derived concentrations patterns were consistent with mass balances and showed that the GdDTPA2− tracer represents a solute that is excluded by roots. Monitoring and imaging the accumulation of the tracer in the root zone therefore offers the potential to determine where and by which roots water is taken up.

Ioannidis, M.A., Chatzis, I., Lemaire, C., and Perunarkilli, R., 2006, Unsaturated hydraulic conductivity from nuclear magnetic resonance measurements: Water Resources Research, v. 42, no. 7, 6 p., doi:10.1029/2006WR004955.

Abstract: Gravity‐driven drainage of water from a column of glass beads of uniform size is studied using nuclear magnetic resonance (NMR). The evolution of proton magnetization and its spin‐spin relaxation time is measured as a function of drainage time at different locations within the column. On the basis of these measurements a model for calculating water relative permeability directly from relaxation time data at various saturations, originally proposed by Chen et al. (1994), is successfully tested for the first time.

Legchenko, A., Ezersky, M., Camerlynck, C., Al-Zoubi, A., Chalikakis, K., and Girard, J-F., 2008, Locating water-filled karst caverns and estimating their volume using magnetic resonance soundings: Geophysics, v. 73, no. 5, p. 1SO-Z88, doi:10.1190/1.2958007.

Abstract: Magnetic resonance sounding (MRS) is a geophysical technique developed for groundwater exploration. This technique can be used for investigating karst aquifers. Generally, the study of a karst requires a 3D field setup and corresponding multichannel data-acquisition instruments. Now only single-channel MRS equipment is available; i.e., the time needed for a 3D MRS field survey is multiplied by a factor of four or five. Where karst caverns are natural hazards, as in the Dead Sea coastal area at Nahal Hever, Israel, even an approximate localization of potentially dangerous zones and a corresponding estimation of the hazard dimensions are useful. We studied numerically the accuracy of MRS estimations of the volume of different 3D targets aroundNahal Hever, shifting a 3D target inside the MRS loop and calculating the volume-estimation errors for each target position. The calculations covered targets of different sizes. The size and position of a target being unknown factors in a field survey, the numerical data were considered as random values to be analyzed statistically. Using a 1D approximation of the MRS solution and assuming a 100-×100-m2 MRS loop, the volume of a 3D target under Nahal Hever conditions is estimated within a ±75% error when the target is smaller than the MRS loop, and within a ±50% error when the target size is about the same as the MRS loop. The lower threshold of karst-cavity detection with MRS is about 6500m3. For such estimation, only one sounding is required.

Parsekian, A.D., Grosse, G., Walbrecker, J.O., Müller‐Petke, M., Keating, K., Liu, L., Jones, B.M., and Knight, R., 2013, Detecting unfrozen sediments below thermokarst lakes with surface nuclear magnetic resonance: Geophysical Research Letters, v. 40, no. 3, p. 535-540, doi:10.1002/grl.50137.

Abstract: A talik is a layer or body of unfrozen ground that occurs in permafrost due to an anomaly in thermal, hydrological, or hydrochemical conditions. Information about talik geometry is important for understanding regional surface water and groundwater interactions as well as sublacustrine methane production in thermokarst lakes. Due to the direct measurement of unfrozen water content, surface nuclear magnetic resonance (NMR) is a promising geophysical method for noninvasively estimating talik dimensions. We made surface NMR measurements on thermokarst lakes and terrestrial permafrost near Fairbanks, Alaska, and confirmed our results using limited direct measurements. At an 8 m deep lake, we observed thaw bulb at least 22 m below the surface; at a 1.4 m deep lake, we detected a talik extending between 5 and 6 m below the surface. Our study demonstrates the value that surface NMR may have in the cryosphere for studies of thermokarst lake hydrology and their related role in the carbon cycle.

Posadas, A., Quiroz, R.A., Tannús, A., Crestana, S., and Vaz, C.M.P., 2009, Characterizing water fingering phenomena in soils using magnetic resonance imaging and multifractal theory: Nonlinear Processes in Geophysics, v. 16, no. 1., p. 159-168, doi:10.5194/npg-16-159-2009.

Abstract: The study of water movement in soils is of fundamental importance in hydrologic science. It is generally accepted that in most soils, water and solutes flow through unsaturated zones via preferential paths or fingers. This paper combines magnetic resonance imaging (MRI) with both fractal and multifractal theory to characterize preferential flow in three dimensions. A cubic double-layer column filled with fine and coarse textured sand was placed into a 500 gauss MRI system. Water infiltration through the column (0.15 x 0.15 x 0.15 m(3)) was recorded in steady state conditions. Twelve sections with a voxel volume of 0.1 x 0.1 x 10 mm(3) each were obtained and characterized using fractal and multifractal theory. The MRI system provided a detailed description of the preferential flow under steady state conditions and was also useful in understanding the dynamics of the formation of the fingers. The f (alpha) multifractal spectrum was very sensitive to the variation encountered at each horizontally-oriented slice of the column and provided a suitable characterization of the dynamics of the process identifying four spatial domains. In conclusion, MRI and fractal and multifractal analysis were able to characterize and describe the preferential flow process in soils. Used together, the two methods provide a good alternative to study flow transport phenomena in soils and in porous media.

Yaramanci, U., Lange, G., and Hertrich, M., 2002, Aquifer characterisation using Surface NMR jointly with other geophysical techniques at the Nauen/Berlin test site: Journal of Applied Geophysics, v. 50, no. 1-2, p. 47-65, doi:10.1016/S0926-9851(02)00129-5.

Abstract: The quite new technique of Surface Nuclear Magnetic Resonance (SNMR) has been extensively tested on the test site Nauen near Berlin to yield the geometry, water content and hydraulic conductivity of the aquifer. The test site is composed of an unconfined aquifer consisting of Quaternary sands with glacial till beneath. It is a very favourable site for assessing the suitability and performance of joint geophysical methods for groundwater exploration. Complementary measurements to SNMR were conducted with Ground Penetrating Radar (GPR), 1D-complex resistivity soundings, i.e. Spectral Induced Polarisation (SIP), 2D-geoelectrics and refraction seismics. Laboratory measurements of porosities, grain size distributions and Nuclear Magnetic Resonance (NMR) decay times were carried out on core samples, and hydraulic conductivities were also derived in order to control and interpret the results of field measurements. The SNMR method allowed the detection of the aquifer beyond any doubt and the determination of the approximate aquifer geometry. The aquifer water content found by SNMR fits very well with the independent measurements on core samples. Hydraulic conductivities derived from decay times are well in range with those from laboratory measurements. GPR allowed a very reliable determination of the aquifer geometry. This information, incorporated into inversion of geoelectric data, led to an improved determination of aquifer electrical resistivity. The estimation of water content by GPR and geoelectrics, even under the favourable conditions in Nauen, is by far not as reliable as that by SNMR. Obtaining information about hydraulic conductivity is possible only with SNMR. Thus, in combination with other geophysical methods, SNMR allows a much more detailed and reliable assessment of aquifers than what was possible with other surface geophysical methods before. In fact, it is, by far, the only method that allows direct detection of water and reliable estimations about water content. It is expected that SNMR will turn out to be a valuable and powerful tool in applied geophysics for groundwater exploration.

References

Allen, D., Flaum, C., Ramakrishnan, T.S., Bedford, J., Castelijns, K., Fairhurst, D., Gubelin, G., Heaton, N., Minh, C.C., Norville, M.A., Seim, M.R., Pritchard, T., and Ramamoorthy, R., 2000, Trends in NMR Logging: Oilfield Review, v. 12, no. 3, p. 2-19.

Behroozmand, A.A., Keating, K., and Auken, E., 2015, A Review of the Principles and Applications of the NMR Technique for Near-Surface Characterization: Surveys in Geophysics, v. 36, no. 1, p. 27–85, doi:10.1007/s10712-014-9304-0.

Costabel, S. and Yaramanci, U., 2013, Estimation of water retention parameters from nuclear magnetic resonance relaxation time distributions: Water Resources Research, v. 49, no. 4, p. 2068-2079, doi:10.1002/wrcr.20207.Dunn, K.J., Bergman, D.J. and Latorraca, G.A. (2002) Nuclear Magnetic Resonance Petrophysical and Logging Applications. Oxford, UK: Elsevier Science, 293p.

Dunn, K.-J., Bergman, D.J., and LaTorraca, G.A., 2002, Nuclear Magnetic Resonance Petrophysical and Logging Applications: Oxford, UK, Elsevier Science, 312 p.

Hertrich, M., 2008, Imaging of groundwater with nuclear magnetic resonance: Progress in Nuclear Magnetic Resonance Spectroscopy, v. 53, no. 4, p. 227-248, doi:10.1016/j.pnmrs.2008.01.002.

Kenyon, W.E., Day, P.I., Straley, C., and Willemsen, J.F., 1988, A Three-Part Study of NMR Longitudinal Relaxation Properties of Water-Saturated Sandstones: Society of Petroleum Engineers Formation Evaluation, v. 3, no. 3, 15 p., doi:10.2118/15643-PA.

Straley, C., Rossini, D., Vinegar, H., Tutunjian, P., and Morris, C., 1997, Core Analysis By Low Field NMR: The Log Analyst, Society of Professional Well-Log Analysts, v.38, no.2, p. 84-94.

Yaramanci, U. and Müller‐Petke, M., 2009, Surface nuclear magnetic resonance – A unique tool for hydrogeophysics: The Leading Edge, v. 28, no. 10, p. 1240-1247, doi:10.1190/1.3249781.