Determination Of Rock Properties Research Articles

Effects of chemical osmosis on swelling clays and the elastic properties of the Pierre shale with its implications for oil recovery

The main objective in this study is to experimentally investigate how two types of shales different in swelling characteristic and organic carbon content behave in terms of the mechanical and infiltration properties when being subjected to changing osmosis with varying brine salinity under in-situ stress conditions. This research also aims to determine the swelling tendency of thePierre shale and replicate the effect of selection of brine salinity on oil recovery in liquid-rich shales. A series of coupled measurements from the triaxial experiments at reservoir conditions was conducted on Pierre shale core samples to determine rock properties and behavior under fluid pore pressure intrusion. The main use of the measured properties is to demonstrate the impact of fluid transport induced by low salinity water injection accompanied by osmosis, clay swelling, and rock elasticity change on oil recovery from rock matrix. The Young's modulus, bulk modulus, and Poisson's ratio of the Pierre shale were determined from measurements on the two types of Pierre shale samples obtained from an outcrop having high 65 wt% of smectite content and well core having lean clay content and 3.5 wt% TOC (total organic carbon). The first two experiments were performed on the outcrop samples comprising of similar characteristics of 65 wt% smectite with negligible TOC while the last experiment was performed on the well core sample containing negligible smectite content with 3.5 wt% TOC. The membrane efficiencies of the shale samples were experimentally determined to range from 5% to 28% depending on clay content and stress. The outcrop shale showed 1% of swelling that is significantly higher than 0.07% of swelling developed in the well core shale lean in clay content. One of the most important findings of this research is that the membrane coefficient representing the stress sensitivity of membrane efficiency of organic-rich well core shale was found to be 11.5%/kpsi which is significantly higher than that of the swelling shale ranging from 2.2 to 3.5%/kpsi. This is interesting because the approximately 5.5 times stiffer matrix of organic-rich shale shows the significantly higher stress sensitivity in terms of the membrane efficiency. The Young's modulus of swelling shale declined with water saturation as well as the water transfer driven by osmotic pressure with low salinity brine. In contrast, the well core shale showed significant increase in the dynamic Young's modulus. The experimental data were used to evaluate the mass exchange between the fractures and rock matrix using a coupled fluid flow and geomechanics model accounting for shale swelling and osmosis transport to evaluate how fluid and rock interaction affect the oil recovery from the rock matrix of the Pierre shale during low salinity water injection (LSWI). A 25% increase in oil recovery factor was determined after considering the effects of changing membrane efficiency due to clay swelling resulting from LSWI. The modeling results therefore suggest that the variation of membrane efficiency due to swelling can be an important factor affecting the oil recovery from matrix block during LSWI enhanced oil recovery.

Reconstructing high fidelity digital rock images using deep convolutional neural networks

Imaging methods have broad applications in geosciences. Scanning electron microscopy (SEM) and micro-CT scanning have been applied for studying various geological problems. Despite significant advances in imaging capabilities, and image processing algorithms, acquiring high-quality data from images is still challenging and time-consuming. Obtaining a 3D representative volume for a tight rock sample takes days to weeks. Image artifacts such as noise further complicate the use of imaging methods for the determination of rock properties. In this study, we present applications of several convolutional neural networks (CNN) for rapid image denoising, deblurring and super-resolving digital rock images. Such an approach enables rapid imaging of larger samples, which in turn improves the statistical relevance of the subsequent analysis. We demonstrate the application of several CNNs for image restoration applicable to scientific imaging. The results show that images can be denoised without a priori knowledge of the noise with great confidence. Furthermore, we show how attaching several CNNs in an end-to-end fashion can improve the final quality of reconstruction. Our experiments with SEM and CT scan images of several rock types show image denoising, deblurring and super-resolution can be performed simultaneously.

Evaluation and prediction of the rock static and dynamic parameters

Determination of rock properties as materials, foundations and sites of civil projects, is one of the priorities. This study aimed to assess the relationship between static elastic modulus (Es) and dynamic elastic modulus (Ed) and to estimate static properties and shear wave velocity (Vs) using simple regression (SR), support vector regression (SVR), multivariate linear regression (MVLR) and artificial neural network (ANN) methods based on compressional wave velocity and physical properties. For this purpose, first geomechanical characteristics of 80 specimens of the limestone rocks from the Asmari and Ilam formations in Karun 4 (K4) and Karun 2 (K2) dam sites, in southwestern Iran were measured. Then, data related to the various studies from different parts of the world were collected and a global relationship was presented. The average Ed obtained from the various relationships of different researchers was equal to 19.90 GPa, which is less than the average Ed of the present study (31.20 GPa). According to the most accurate fit, the presented relationship between Es and Ed was power. The analysis of all model hypotheses by MVLR showed that it is possible to estimate the static and dynamic properties. Predicting dynamic and static parameters of the limestone rocks using Sigmoid transfer functions and Hyperbolic tangents and various training rules showed that the Sigmoid transfer function and Levenberg-Marquardt training law have the best performance in predictions. Comparison of the methods performance in estimating Vs and static properties showed that the SVR has higher accuracy than other methods.

Lithofacial analysis and possibilities for prediction of properties on geophysical research and seismic exploration data by methods of machine learning

The success of a development strategy for any field depends on the degree of knowledge of the geological structure of its main reservoirs. As the area is drilled out, the concept of the structure of the hydrocarbon accumulation is refined, but in the case of a complex structure of the void space of the reservoirs and the lithological heterogeneity of the section over the area, geological uncertainties and risks during the subsequent placement of wells remain high. For these reasons, one of the main problems in hydrocarbon production is predicting rock types and the distribution of fluids throughout the reservoir away from wells, since the determination of rock properties is a major source of uncertainty in reservoir modeling studies [1, 2]. The proposed project will demonstrate algorithms based on machine learning methods that allow predicting the distribution of lithology and the uncertainty of lithofacies variability in the section.

Investigation of the availability of a new point load test device in characterization of rocks

Difficulties in determining rock properties have led to the developed and increased use of index test methods predicting them. Index test methods are mostly simple, cheap, and easy to apply but there are some restrictions due to these specifications. The most used method to determine the strength values of rocks indirectly is the point load index. The main aim of this study is primarily investigating the usability of modified test device instead of classical test device. For this, laboratory tests were carried out on rocks with different strength values (3 igneous, 1 metamorphic, 3 sedimentary). The point load index tests were carried on 15 different classical test devices and on a modified test device which the limitations of the classical test device were eliminated. Analysis of the obtained results was carried out by the simple regression method. It was determined the modified test device can be used reliably instead of the classical test devices. Besides, while determining the strength tests the stress distributions on the samples were examined with the finite element method.

Non-destructive techniques for linking methodology of geochemical and mechanical properties of rock samples

Inadequate conventional core data, increasing cost of native sampling techniques, economic downturn in the oil sector, sparse distribution of modern well log suites in horizontal wells, and data acquisition challenges are few among the reasons driving innovation in the use of drill cuttings for determining rock properties and adequate characterization of reservoir units. In this paper, we propose a rock strength proxy based on the correlation between rock strength and elemental geochemistry. The technique uses a combination of X-ray fluorescence and rebound impulse hammer measurements to establish links between the geochemical and mechanical properties of the sampled rocks. We identified five chemo-mechanical facies (CMF) that are formation specific. Furthermore, a cross-validation approach was used to build a predictive model based on acquired elemental and impulse hammer measurements. Consequently, the model was used to predict elastic property and results were encouraging with R2 of 0.83, 0.87, and 0.89 for training, validation, and testing respectively. This proposed method was developed using core plugs retrieved from subsurface formations. It will be useful for examples in rapid assessment of drill cuttings for the determination of rock strength, especially where the only available data is elemental geochemical data.

Radial profiling of near-borehole formation velocities by a stepwise inversion of acoustic well logging data

The radial heterogeneity of near-wellbore formation is frequently encountered in petroleum exploration and production. It usually manifests as the variations of formation wave velocities in radial position. Mapping radial variations of formation velocities is of great significance in identifying invaded zones and determining rock properties, which are valuable for engineering measures, such as fracturing, wellbore stability evaluation and optimal well completion. A stepwise inversion scheme is developed to image the radial profiles of formation P- and S-wave velocities. Instead of determining the formation velocity variation and its corresponding radial position simultaneously as traditional methods do, the inversion procedure is divided into two steps: the velocity array is determined by semblance processing of contiguous receiver pairs of acoustic array data, and then the thickness of the layer (radial position) is obtained based on ray theory. The inversion results from Step 1 could be used to guide the operation and decide whether Step 2 is necessary. In addition, calculation on the arrival time of waves is not required in this inversion method, thus mitigating the inaccuracy in processing field data with noises. Firstly, the wave field characteristics of heterogeneous formation models have been analysed, indicating that both amplitude and arrival time differences are caused by heterogeneity. The waveforms are then processed using the stepwise inversion method to estimate the radial profiles of formation wave velocities, where the factors influencing the inversion procedure are also studied. It is shown that, longer TR spacing is required for a specific investigation depth with the decreasing of velocity ratio; the resolution of the velocity inversion profile degrades with the increasing number of receiver pairs. The modelling-based inversion results and the application to field data indicate that, in addition to profiling formation P-wave velocity, our proposed methodology also has encouraging application in mapping formation S-wave velocity. The efficiency and accuracy of the stepwise inversion method enables real-time implementation with fast calculation speed, thus providing useful information of monopole acoustic well logging for engineering.

Confocal laser scanning microscopy and automated petrographic image analysis in different rock types: Two-dimensional images capillary pressure curves estimation and three-dimensional porosity reconstruction

In the past, determination of rock properties using image analysis relied upon petrographic transmitted-light images, but with limited success because of a lack of resolution and restricted computer processing power. A new technique that employs confocal laser scanning microscopy (CLSM) can be considered complementary to laboratory measurements and applicable to several samples, saving time and money and requiring only a limited amount of rock sample for analysis. We have studied several types of rocks with CLSM and fluorescent dye–impregnated thin sections. The two-dimensional scans of each thin section images is an area of 12 mm2, with a pixel size of 0.198 μm and were used to simulate capillary pressure curves for pore bodies and pore throats. The CLSM technique also enables three-dimensional (3-D) visualization of the rock porosity. The studied rock samples were taken from diverse oil and gas field reservoirs: case A, a conventional sandstone (15.1% porosity, 29.8 md permeability); case B, a tight sandstone (3.7%, 0.02 md); case C, an oolitic carbonate (9.6%, 0.1 md); case D, a rhodolithic algal carbonate (19.8%, 43.7 md); case E, dolomitized carbonate (17%, 21.7 md); and case F, a naturally fractured carbonate (2.4%, 0.6 md). Our results confirm that the CLSM technique can be applied to rocks of contrasting porosity and permeability to obtain computed synthetic capillary pressure curves faster than with conventional measurement methods. The technique quantifies different pore-body and pore-throat sizes and distributions, with the added ability to visualize 3-D porosity and to extract from thin section analysis petrologic properties.

Challenges and Prospects of Digital Core-Reconstruction Research

The simulation of various rock properties based on three-dimensional digital cores plays an increasingly important role in oil and gas exploration and development. The accuracy of 3D digital core reconstruction is important for determining rock properties. In this paper, existing 3D digital core-reconstruction methods are divided into two categories: 3D digital cores based on physical experiments and 3D digital core stochastic reconstructions based on two-dimensional (2D) slices. Additionally, 2D slice-based digital core stochastic reconstruction techniques are classified into four types: a stochastic reconstruction method based on 2D slice mathematical-feature statistical constraints, a stochastic reconstruction method based on statistical constraints that are related to 2D slice morphological characteristics, a physics process-based stochastic reconstruction method, and a hybrid stochastic reconstruction method. The progress related to these various stochastic reconstruction methods, the characteristics of constructed 3D digital cores, and the potential of these methods are analysed and discussed in detail. Finally, reasonable prospects are presented based on the current state of this research area. Currently, studies on digital core reconstruction, especially for the 3D digital core stochastic reconstruction method based on 2D slices, are still very rough, and much room for improvement remains. In particular, we emphasize the importance of evaluating functions, multiscale 3D digital cores, multicomponent 3D digital cores, and disciplinary intersection methods in the 3D construction of digital cores. These four directions should provide focus, alongside challenges, for this research area in the future. This review provides important insights into 3D digital core reconstruction.

Studying the correlation of rock properties with P-wave velocity index in dry and saturated conditions

Estimation of physical and mechanical properties of rocks plays crucial role in designing of mining and civil engineering projects. Determination of these properties using direct methods i.e., laboratory determination is very expensive and time consuming in most cases despite having the high accuracy. Moreover, preparing core samples to perform direct tests is extremely difficult and needs double precision. For this reason, indirect and non-destructive methods can be utilized as suitable alternatives to determine rock properties. In this paper, the relationships between the different physico-mechanical properties of intact rocks with P-wave velocity have been investigated. For this purpose, rock samples of schist, phyllite, and sandstone were firstly collected from drilling operation in Marivan Azad dam site of Iran. Then, different physio-mechanical properties of core samples including uniaxial compressive strength, modulus of elasticity, porosity, Poisson's ratio, slake durability index, density, water absorption and P-wave velocity were determined in dry and saturated conditions in the laboratory. Based on the statistical analysis, linear, logarithmic, polynomial, exponential and power equations between the above parameters and P-wave velocity were investigated. An equation with very high determination coefficient (R2) was proposed as best equation to determine each of the above properties based on the P-wave velocity index. To validate the proposed empirical equations, obtained results were compared with the measured data which was not incorporated in equations development. Finally, comparison from the obtained results showed that physical and mechanical properties of rock have higher correlation with the P-wave velocity in saturated and dry conditions, respectively.

Estimating rock properties using sound signal dominant frequencies during diamond core drilling operations

In many engineering applications such as mining, geotechnical and petroleum industries, drilling operation is widely used. The drilling operation produces sound by-product, which could be helpful for preliminary estimation of the rock properties. Nevertheless, determination of rock properties is very difficult by the conventional methods in terms of high accuracy, and thus it is expensive and time-consuming. In this context, a new technique was developed based on the estimation of rock properties using dominant frequencies from sound pressure level generated during diamond core drilling operations. First, sound pressure level was recorded and sound signals of these sound frequencies were analyzed using fast Fourier transform (FFT). Rock drilling experiments were performed on five different types of rock samples using computer numerical control (CNC) drilling machine BMV 45 T20. Using simple linear regression analysis, mathematical equations were developed for various rock properties, i.e. uniaxial compressive strength, Brazilian tensile strength, density, and dominant frequencies of sound pressure level. The developed models can be utilized at early stage of design to predict rock properties.

A Realistic Transport Model with Pressure-Dependent Parameters for Gas Flow in Tight Porous Media with Application to Determining Shale Rock Properties

A nonlinear transport model for single-phase gas flow in tight porous media is developed. The model incorporates many important physical processes that occur in such porous systems: continuous flow, transition flow, slip flow, Knudsen diffusion, adsorption and desorption into and out of the rock material, and a correction for high flow rates. This produces a nonlinear advection–diffusion type of partial differential equation with pressure-dependent model parameters and associated compressibility coefficients, and highly nonlinear apparent convective flux (velocity) and apparent diffusivity. A key finding is that all model parameters should be kept pressure dependent for the best results. An application is to the determination of rock properties, such as porosity and permeability, by history matching of the simulation results to data from pressure-pulse decay tests in a rock core sample (Pong et al. in ASME Fluids Eng Div 197:51–56, 1994).

Digital Rock Physics Aplications: Visualisation Complex Pore and Porosity-Permeability Estimations of the Porous Sandstone Reservoir

Reservoir rock containing oil and gas generally has high porosity and permeability. High porosity is expected to accommodate hydrocarbon fluid in large quantities and high permeability is associated with the rock’s ability to let hydrocarbon fluid flow optimally. Porosity and permeability measurement of a rock sample is usually performed in the laboratory. We estimate the porosity and permeability of sandstones digitally by using digital images from μCT-Scan. Advantages of the method are non-destructive and can be applied for small rock pieces also easily to construct the model. The porosity values are calculated by comparing the digital image of the pore volume to the total volume of the sandstones; while the permeability values are calculated using the Lattice Boltzmann calculations utilizing the nature of the law of conservation of mass and conservation of momentum of a particle. To determine variations of the porosity and permeability, the main sandstone samples with a dimension of 300 × 300 × 300 pixels are made into eight sub-cubes with a size of 150 × 150 × 150 pixels. Results of digital image modeling fluid flow velocity are visualized as normal velocity (streamline). Variations in value sandstone porosity vary between 0.30 to 0.38 and permeability variations in the range of 4000 mD to 6200 mD. The results of calculations show that the sandstone sample in this research is highly porous and permeable. The method combined with rock physics can be powerful tools for determining rock properties from small rock fragments.

Introduction to Digital Rock Physics and Predictive Rock Properties of Reservoir Sandstone

Rock properties analysis (porosity, permeability, elastic modulus, and wave velocity) of the rock is important to note as one of the methods to determine the characteristics of the reservoir rock. Rock properties can calculated in conventional (laboratory), indirect (inversion of seismic waves), and digital computation (Digital Rock Physics). This paper will introduce and discuss the digital calculation/simulation and empirical equation to predict the value of the rock properties from reservoir sandstone. The data used is the samples of the data sandstone core (reservoir) subsurface in an oil field. The research method is to combine the data from a thin layer, a digital image of rocks in three-dimensional (μ-CT-Scan), and empirical approaches of the equations of permeability on rocks and Lattice Boltzmann equation. Digital image of a scanned using μ-CT-Scan used to determine value rock properties and pore structure at the microscale and visualize the shape of the pores of rock samples in 3D. The method combined with rock physics can be powerful tools for determining rock properties from small rock fragments.

Looking for Oil From More Angles

Seismic Data New generation of seismic promises big advances Selling new offshore seismic survey methods can be reduced to two goals: “We want more of everything at less cost.” That maxim was offered by Andrew Long, chief scientist of geoscience and engineering, imaging, and engineering for Petroleum Geo- Services, one of the players in this intensely competitive, economically depressed, technology-driven business. While the science behind using sound waves to image underground formations is baffling to outsiders, the two avenues for doing it better are not. There are improvements in data acquisition covering sound sources, receivers, and survey methods. And there is the mathematics and computer power needed to process and extract more information from the constantly growing amount of data gathered. Faster, Better Seismic Surveys by Making Many Sounds at Once The pace of a seismic survey can be measured by the number of shots taken per day. Huge onshore surveys have shown that allowing 10 or more sound sources working at the same time without any coordination can drastically reduce the time needed to gather data over a large area, and get better data as well. “One of the biggest surprises is we have gotten almost universally better data from our simultaneous source surveys,” said Craig Beasley, a Schlumberger Fellow and chief geophysicist at WesternGeco. More data are gathered from more angles for a more accurate picture of what is below, said Beasley, who explained that one flashlight can help you see in a dark basement, but it is better with two and more is better. The Hunger for the Lowest Frequencies Takes Data Gathering in New Directions The word “broadband” is used to sell a lot of what is new in offshore seismic. It can mean different things depending on who is speaking. But most often, it is applied to things used to gather scarce signals at the lowest frequencies. Gathering more low-frequency data is the “holy grail’ in broadband, said Craig Beasley, chief geophysicist at Western- Geco, which is part of Schlumberger. They are valuable for determining rock properties and imaging deep formations. Seemingly small gains can be big. “Moving down from “3 to 1.5 (Hz) does not sound like a lot. But in octaves, it is a whole other octave,” he said. Intractable Problem and Big Payoff Sell a Seismic Acquisition Innovation The Mad Dog field was a major discovery for BP in the late 1990s, but developing it looked like a high-risk proposition. The problem was the field was an early discovery beneath the thick layer of salt covering much of the deepwater US Gulf of Mexico. Shooting seismic through the salt resulted in murky images that increased the risk of development wells that produced little or nothing, which could turn a 4-billion-bbl field into a marginal investment. BP needed to solve the problem because it had leased a large number of deepwater blocks where the industry was drilling a growing number of dry holes due to poor subsurface imaging. “It was pretty common knowledge in the industry that you have this problem. But what are you going to do about it?” said John Etgen, a distinguished advisor in seismic imaging at BP. The problem is salt distorts sound waves as they travel down and are reflected back up. The salt is like broken eyeglasses. It acts as “big, very complex lens distorting the sound waves.”

A new technique for lithology and fluid content prediction from prestack data: An application to a carbonate reservoir

One of the leading challenges in hydrocarbon recovery is predicting rock types/fluid content distribution throughout the reservoir away from the boreholes because rock property determination is a major source of uncertainty in reservoir modeling studies. Spatial determination of the lateral and vertical heterogeneities has a direct impact on a reservoir model because it will affect the property distributions. An inappropriate determination of the facies distribution will lead to unrealistic reservoir behavior. Because these data can take different forms (lithologs, cuttings, and for seismic, poststack, and prestack attributes) and have different resolutions, the manual integration of all the information can be tedious and is sometimes impractical. We developed a new neural network-based methodology called democratic neural network association (DNNA). The DNNA method was trained using lithology logs from wells simultaneously with prestack seismic data. This technique, using a probabilistic approach, aims to find patterns in seismic that will predict lithology distribution and uncertainty.

Considerations on stability of open stopes in 'Jama Bor' underground mine

Dominant ore bodies in 'Jama Bor' underground mine in last few years, from the aspect of production and engaged resources, have a common characteristic that after being mined out, excavated areas remained open. These are following ore bodies: T, T1 and T2. Mining in ore body T had been ended two years ago and the stope was supposed to be backfilled after the end of excavation. However, process of backfilling did not start yet. Applied mining method in currently active ore bodies, T1 and T2, is pillar mining in horizontal slices, with leaving open voids after the end of extraction. Existence of voids in underground mines is always a challenge, regarding stability and safety. In considerations of stability of open stopes in 'Jama Bor', there are several key factors: rock mass properties, stope geometry, support properties and interaction between installed support and surrounding rock. Contemporary approach applied in each phase of the process, including determination of rock properties, mine design, mining operations, support installation and monitoring of stope stability during and after the excavation, enables accurate and reliable defining of stope stability. Also, such approach enables early detection of possible endangerments in stope stability, thus providing possibility for prevention and securing of endangered zones. Since the future of underground mining in RTB Bor is related deep lying ore deposits, bellow current operations, importance of their stability is even bigger.

High-resolution acoustic impedance inversion to characterize turbidites at Marlim Field, Campos Basin, Brazil

Seismic inversion is routinely used to determine rock properties, such as acoustic impedance and porosity, from seismic data. Nonuniqueness of the solutions is a major issue. A good strategy to reduce this inherent ambiguity of the inversion procedure is to introduce stratigraphic and structural information a priori to better construct the low-frequency background model. This is particularly relevant when studying heterogeneous deepwater turbidite reservoirs that form prolific, but complex, hydrocarbon plays in the Brazilian offshore basins. We evaluated a high-resolution inversion workflow applied to 3D seismic data at Marlim Field, Campos Basin, to recover acoustic impedance and porosity of the turbidites reservoirs. The Marlim sandstones consist of an Oligocene/Miocene deepwater turbidite system forming a series of amalgamated bodies. The main advantage of our workflow is to incorporate the interpreter’s knowledge about the local stratigraphy to construct an enhanced background model, and then extract a higher resolution image from the seismic data. High-porosity zones were associated to the reservoirs facies; meanwhile, the nonreservoir facies were identified as low-porosity zones.

Integration of seismic velocity measurements in the context of the CO2 Storage project in the Bonaparte Basin, offshore NW Australia

SUMMARY An accurate seismic velocity model is essential for depth conversion and rock property determination in the context of fluid flow modelling for secure storage of carbon dioxide in the study area. Three types of seismic velocity measurements are available within the study area: velocities derived from stacking of multi-channel reflection seismic data; velocities determined in the process of ray tracing modelling of large offset refraction data acquired by ocean bottom seismographs (OBSs) along the coincident reflection/refraction transect, and velocities from well log measurements. Generally, the OBS derived velocities are faster than the stacking derived velocities in the shallow section down to 2 km depth, with particularly large (exceeding 15 %) discrepancies in the NE part of the study area. High velocities in the OBS model are consistent with well log measurements in the Newby 1 and Flat Top 1 exploration wells. Integration of OBS refraction velocities with stacking velocities through the calibration of both of them against well logs is the best way to construct a reliable velocity model for the whole section. Refraction seismic data coverage offshore Australia is expected to significantly improve as in 2013 Australia will obtain a National Pool of OBSs suitable for multi-scale experiments.

Evaluation of surface nuclear magnetic resonance-estimated subsurface water content

The technique of nuclear magnetic resonance (NMR) has found widespread use in geophysical applications for determining rock properties (e.g. porosity and permeability) and state variables (e.g. water content) or to distinguish between oil and water. NMR measurements are most commonly made in the laboratory and in boreholes. The technique of surface NMR (or magnetic resonance sounding (MRS)) also takes advantage of the NMR phenomenon, but by measuring subsurface rock properties from the surface using large coils of some tens of meters and reaching depths as much as 150 m. We give here a brief review of the current state of the art of forward modeling and inversion techniques.In laboratory NMR a calibration is used to convert measured signal amplitudes into water content. Surface NMR-measured amplitudes cannot be converted by a simple calibration. The water content is derived by comparing a measured amplitude with an amplitude calculated for a given subsurface water content model as input for a forward modeling that must account for all relevant physics.A convenient option to check whether the measured signals are reliable or the forward modeling accounts for all effects is to make measurements in a well-defined environment. Therefore, measurements on top of a frozen lake were made with the latest-generation surface NMR instruments. We found the measured amplitudes to be in agreement with the calculated amplitudes for a model of 100 % water content. Assuming then both the forward modeling and the measurement to be correct, the uncertainty of the model is calculated with only a few per cent based on the measurement uncertainty.