Nuclear Magnetic Resonance T2 Distribution Research Articles

Objective-Driven Solid-Surface-Roughness Characterization for Enhanced Nuclear-Magnetic-Resonance Petrophysics

Summary Typically, smooth solid surfaces of reservoir rocks are assumed in formation evaluation, such as nuclear-magnetic-resonance (NMR) petrophysics and reservoir-wettability characterization through contact-angle measurements. Measuring the degree of surface roughness (R), or smoothness, and evaluating its effects on formation evaluation are topics of much research. In this paper, we primarily focus on details in characterizing solid-surface roughness and its applications in NMR pore-size analysis. R can be measured by contact techniques and noncontact techniques, such as stylus profilometer, atomic-force microscopy, and different kinds of optical measurements. Each technique has different sensitivities, measurement artifacts, resolutions, and field of view (FOV). Intuitively, although a finer resolution measurement provides the closest account of all surface details, the correspondingly small FOV might compromise the representativeness of the measurement, which is particularly challenging for charactering heterogeneous samples such as carbonates. To balance the FOV and measurement representativeness, and to minimize artifacts, laser scanner confocal microscopy (LSCM) is selected in this study. Results for the more than 27 rock samples tested indicate that rocks of similar rock types have similar R-values. Grainy limestones have relatively higher R-values compared with dolostones, consistent with the dolostone’s crystallization surface features. Muddy limestones have smoother surfaces, resulting in the lowest R-values among the rocks studied. For sandstones, R varies with clay types and content. For rocks containing two distinct minerals, two R-values are observed from the R profiles, which for these rock types justifies the use of two NMR surface relaxivity (ρ2) parameters for determining the pore-size distribution (PSD) from the NMR T2 distribution. The novelty here is the integration of LSCM and NMR to obtain an NMR PSD relevant for permeability, capillary pressure, and other petrophysical parameters. Typically, ρ2 is calibrated using the total surface area from Brunauer-Emmett-Teller (BET; Brunauer et al. 1938) gas adsorption, but this underestimates the NMR pore size because of surface-roughness effects. In our novel approach, we use R measured from LSCM to correct ρ2 for surface-roughness effects, and thereby obtain the NMR pore size more relevant for permeability and other petrophysical parameters. We then compare the roughness-corrected NMR PSD against pore size from microcomputed tomography (micro-CT) scanning (which is roughness independent). The good agreement between roughness-corrected NMR and micro-CT pore sizes in the micropore region validates our new technique, and highlights the importance of surface-roughness characterization in NMR petrophysics.

A new method to improve the NMR log interpretation in drilling mud-invaded zones: A case study from the Brazilian Pre-salt

The Santos Basin hosts several giant carbonate oilfields in the Pre-salt province and is one of the most important hydrocarbon discoveries in the past years. These carbonate reservoirs present exceptional reservoir quality with a challenging depositional and diagenetic history that produced complex and heterogeneous pore systems. As one of the most effective tools for in-situ porosity and permeability evaluation, the nuclear magnetic resonance (NMR) well log plays an essential role in the Pre-salt petrophysical evaluation and its signal interpretation must be performed carefully. This work proposes a new approach to detect and improve the interpretation of zones with unreal NMR logging tool readings in areas heavily affected by drilling mud invasion in caverns/vugs, which causes a T2 distribution shift towards lower relaxation times. This work analyses well log data from 28 wells in the Sapinhoá Field, supported by special core analyses (SCAL) and routine core analyses (RCAL), as well as plugs, sidewall cores and thin sections data. The proposed approach generates synthetic NMR Free Fluid and T2 distributions with minimum error and replaces the faulty data, consisting in five main steps, (1) computing new curves to be used as inputs; (2) detecting the zones affected by the invasion; (3) water saturation calculation by the saturation height modeling method; (4) calculating a synthetic NMR free fluid curve based on a relationship between a NMR-derived water saturation and the saturation height model; and (5) using the Random Forests algorithm to estimate a synthetic T2 distribution. The proposed method has a minimized error and resulted an increase of up to 10% thickness in the pay zone and a reduction of up to 9% in water saturation average, which may represent enormous amounts of oil depending on the reservoir area. This study provides a new method, which can be used in the improvement of petrophysical interpretations in highly heterogeneous carbonate reservoirs and thus reducing the exploratory and development risks. • Drilling mud invades vug pores and results in misleading NMR readings; • A new methodology is proposed to create a synthetic NMR free fluid and T2 distribution with minimum error; • An increase of 10% in pay thickness and a reduction of 9% in average water saturation was achieved in the Pre-salt reservoirs; • The methodology is proper for the petrophysical evaluation of highly heterogeneous carbonate reservoirs

Predicting the impact of hydrocarbon saturation on T2 distribution curve of NMR logs – A case study

The Nuclear Magnetic Resonance (NMR) is a valuable technique that provides estimates of various rock and fluid properties, including porosity, permeability, and pore size distribution. The decay or relaxation time of NMR signals is directly related to the pore size, but altering fluid in the pore spaces would also change the T 2 distribution curve. This change is due to the different relaxation time of hydrocarbon and water. This study focuses on predicting NMR T 2 distribution curves at different water saturations using a single T 2 saturation at any given water saturation. To do so, NMR measurements were performed on 9 plugs each at different saturations. Two general equations were derived. The first equation is related to the water response, while the second equation is related to the oil response. These two equations were calibrated and verified using core data. The coefficient of determination (R 2 ) for the first equation is equal to 0.592, while it is equal to 0.975 for the second equation. T 2 distribution curves depend on the amount and the distribution of the fluids within the pore structure. The findings of this study can help to increase the accuracy of the petrophysical properties derived by NMR logging. • A new model is proposed to predict NMR's T2 distribution curve in the oil bearing rocks. • NMR measurements were investigated in 9 plugs individually in different saturations in an oil field. • Two general equations were derived. To validate the equations, five T2 distribution curves which are equivalent to the T2 distribution curve of 100% water saturated rock were constructed by using 5 experimental T2 distributions.

A new method for predicting capillary pressure curves based on NMR echo data: Sandstone as an example

The mercury injection capillary pressure (MICP) curves have been widely used for the evaluation of the pore structure. In practical application, the MICP curves are usually obtained by the mercury injection experiment, but the experimental measurement can not obtain the MICP curves of the reservoir continuously. Since the pore size distribution obtained from the nuclear magnetic resonance (NMR) transverse relaxation time (T2) distribution is related to the pore-throat size distribution obtained from the MICP curve, so in previous studies, the MICP curves were predicted based on the NMR T2 distribution. However, the NMR T2 distribution obtained by the inversion of the NMR echo data has uncertainty, which affects the prediction accuracy of the MICP curves. In this paper, a new method for predicting the MICP curves of sandstone based on NMR echo data is proposed for the first time. Multiple characteristic parameters of the NMR echo data are calculated. The relationship between the parameters and the mercury saturation (Shg) is established at each capillary pressure point to predict the MICP curves. And the parameters of the MICP curves and the physical parameters are selected to create a reservoir classification index. The results show that the method for predicting the MICP curves of sandstone based on NMR echo data has high prediction accuracy and strong stability, and the reservoir classification index provides an accurate classification of sandstone reservoirs.

Experimental study of pore characteristics and radon exhalation of uranium tailing solidified bodies in acidic environments.

The pore characteristics and radon exhalation of uranium tailings solidified in an acid environment were investigated in this study. Tailings from the beach of a uranium tailing reservoir in the acid rain area of Central China were selected as samples and solidified with cement, slag powder (GGBS), metakaolin (MK), or slag powder and metakaolin (GM), then immersed in simulated acid rain solution for 60 days. The transverse relaxation time T2 distribution and porosity of each solidified sample before and after immersion were measured by nuclear magnetic resonance (NMR) and the cumulative radon concentration before and after immersion was measured by a RAD7 radon meter. The experimental results show that the nuclear magnetic resonance T2 distribution curve shifts to the left, the peak amplitude decreases, and the pores in the sample gradually shrink as the admixture content increases. The porosity and radon exhalation rate of solidified samples also appear to decrease gradually as admixture content increases; a quadratic function relationship was observed between porosity and radon exhalation rate. The pore size and effective pore volume of solidified samples increase as immersion time increases, while the radon exhalation rate increases and the pore volume gradually increases. The results of this study may provide a sound theoretical basis for the solidification treatment of uranium tailings in engineering practice.

Qualitative and quantitative characterization of multiple factors that influence movable fluid saturation in lacustrine deep-water gravity-flow tight sandstones from the Yanchang Formation, southern Ordos Basin, China

Fluid mobility is one of the most important factors in evaluating the potential for recovering tight oil. However, quantitative effects of multiple factors that influence movable fluid saturation in different pore-throat combinations and their relationships with lithofacies in tight sandstones, especially for lacustrine deep-water gravity-flow deposits, remain controversial due to the strong heterogeneity and complex pore structure of tight reservoirs. Core samples obtained from the Upper Triassic Yanchang Formation in southern Ordos Basin were evaluated by using a variety of techniques, including nuclear magnetic resonance (NMR), scanning electron microscopy (SEM), X-ray diffraction (XRD), pressure-controlled porosimetry (PCP), rate-controlled porosimetry (RCP), impregnated thin sections, and helium porosity and nitrogen permeability measurements. Based on NMR T2 distributions under water-saturated and centrifugal conditions, four types of pore-throat combinations are identified by using the relaxation time thresholds Ts1 and Ts2 from NMR: large intergranular pore-dominated pore-throat combinations (LIP), small intergranular pore-dominated pore-throat combinations (SIP), intragranular pore-dominated pore-throat combinations (IAP), and micropore-dominated pore-throat combinations (MP). Among them, SIP is the dominant pore space favorable for fluid flow. Movable fluid saturation in SIP pore space varies among different lithofacies. The best fluid mobility commonly occurs in fine-grained, cross-bedded sandstones (Sc), whereas the worst fluid mobility usually appears in siltstones to very fine-grained sandstones (Ss). The quantitative effects of a total of 41 factors that influence movable fluid saturation of different pore-throat combinations were investigated through the analysis of the Pearson correlation matrix. The results demonstrate that porosity and permeability mainly affect the movable fluid saturation in the pore space of SIP and MP. In addition, maximum mercury intrusion saturation from PCP (Smax) and tortuosity (λ) are the critical pore structure parameters affecting the movable fluid saturation in the pore space of SIP and MP, whereas the average pore throat radius ratio (η) is the critical factor affecting the movable fluid saturation in LIP pore space. Overall, movable fluid saturation in different pore-throat combinations is characterized mainly by different microscopic pore structure parameters. A general pore network model for different lithofacies with different fluid mobility is established to facilitate assessment of the heterogeneity of tight sandstones and to further guide hydrocarbon exploration and development in similar lacustrine deep-water depositional settings.

Applications of nuclear magnetic resonance (NMR) logging in tight sandstone reservoir pore structure characterization

Tight sandstone reservoir evaluation and characterization faced great challenge by using conventional well logging data due to the complicated pore structure. To improve tight sandstone reservoir identification, the pore structure should be first characterized. In this study, using the tight Chang 8 Formation of Pengyang Region, west Ordos Basin as an example, 20 core samples were drilled for laboratory nuclear magnetic resonance (NMR) and mercury injection capillary pressure (MICP) experiments. A model, which was used to construct capillary pressure (Pc) curves from NMR data, was proposed, and the corresponding models were established based on classified power function (CPF) method to classify formations into three types. Based on these relationships, the NMR T2 distributions were transformed as pseudo Pc curves and pore throat radius distributions. After these relationships were extended into field applications, consecutive pseudo Pc curves were acquired, and the pore structure evaluation parameters and permeability were also predicted. Comparisons of predicted parameters with core-derived results illustrated the reliability of our proposed model and method.

A novel method to estimate resistivity index of tight sandstone reservoirs using nuclear magnetic resonance logs

The relationship between resistivity index and water saturation (I-Sw) is very important in formation evaluation via petrophysical data. Low permeability and complicated pore structure make it more difficult to determine I-Sw relationship through rock's electrical property measurements. Nuclear magnetic resonance (NMR) logs are alternative ways to obtain these key parameters. To establish a model to estimate resistivity index from NMR logs, 17 cores from the study area were selected for experiments. The helium porosity, air permeability, NMR T2 distribution, and resistivity of cores were measured. I-Sw relationship was evaluated accordingly. Given these variables, Archie formula, J function, Schlumberger-Doll-Research (SDR) model, and correlation between capillary pressure curve and T2 distribution (Pc-T2) were used to derive a novel method. It aims to calculate resistivity index from T2 time, porosity, and geometric mean of T2 time. The experimental data were applied and the corresponding model was established by multiple statistical regression method. Resistivity index is successfully predicted by T2 distribution. Then, the established model was tested by the experimental data not involved in the modeling. The validation shows that the model is highly precise. Finally, the model was applied to formation evaluation for a field case with satisfactory results. The proposed method and model provide a quantitative basis to analyze and predict resistivity index by well log data. They are of great significance for the accurate determination of oil and gas saturation of tight sandstone reservoirs.

The rock-fabric/petrophysical characteristics and classification of the micropores hosted between the calcite and dolomite crystals

For the substantial hydrocarbon potential and academic significance, focus of interest has been directed towards the microporous carbonate. Compared with the documented shallow burial microporous carbonate, carbonate with a depth deeper than 6000 m present, of course, different characteristics due to the long-term influences of the strong cementation and compaction. With the deepening of exploration, it is therefore essential to improve the understanding about the deep burial microporous carbonate. In this paper, an integrated petrographic and petrophysical approach, which combines thin section, high-resolution field emission scanning electron microscopy (FESEM) observations, and nuclear magnetic resonance (NMR) measurements, is chosen to study the deep to ultradeep carbonate in the Shunnan area, Tarim Basin, NW China.Results show that all the studied carbonate samples consist entirely of micropores with a radius less than 10 μm, belonging to the strictly microporous carbonate. These micropores exist between the calcite and dolomite microcrystals. To reliably calculate the permeability based on NMR data, a newly introduced factor P aiming to adjust the T2 cut-off value derived from the relaxation curve is applied to enhance the Coates model. Values of P are found to correlate well with the FFI (free fluid index) of the bioclastic and bioclasts-free detrital limestones through different quantitative relationships, respectively. Permeabilities calculated from the modified Coates model are in good agreements with those measured by the pulse-decay method (R2 = 0.93, 0.67). Considering the existing classifications of pore systems are not applicable to the peculiar intercrystalline micropores within the deep burial microporous carbonate, a noval classification scheme is thus explored by integrating the rock-fabric and petrophysical properties. Micropores hosted within the subhedral to anhedral microcrystals with distinct boundaries are the main reserving space of the microbial lime-/dolostone, which lead to relatively higher porosities (averaging 3.41%) than the detrital limestone (1.99%) and dolostone (2.47%), and an identical right-skewed trimodal NMR T2 distribution. Micropores related with the coalescent polyhedral to anhedral microcrystals, that are always glued together to form fusing boundaries, are mainly located within the micrite matrix and allochem of the detrital limestone. Certain relationships among porosity, permeability and FFI/BVI also exist for the bioclastic and bioclasts-free detrital limestones, respectively. Micropores within the dolostone exist between the rhombic dolomite crystals. This proposed classification of micropores comprehensively reflects the depositional, diagenetic textures and physical properties of the deep burial microporous carbonate. The particularity of microbial lime-/dolostone is highlighted.

Classification of tight sandstone reservoirs based on NMR logging

The traditional reservoir classification methods based on conventional well logging are inefficient for determining the properties, such as the porosity, shale volume, J function, and flow zone index, of the tight sandstone reservoirs because of their complex pore structure and large heterogeneity. Specifically, the method that is commonly used to characterize the reservoir pore structure is dependent on the nuclear magnetic resonance (NMR) transverse relaxation time (T2) distribution, which is closely related to the pore size distribution. Further, the pore structure parameters (displacement pressure, maximum pore-throat radius, and median pore-throat radius) can be determined and applied to reservoir classification based on the empirical linear or power function obtained from the NMR T2 distributions and the mercury intrusion capillary pressure curves. However, the effective generalization of these empirical functions is difficult because they differ according to the region and are limited by the representative samples of different regions. A lognormal distribution is commonly used to describe the pore size and particle size distributions of the rock and quantitatively characterize the reservoir pore structure based on the volume, mean radius, and standard deviation of the small and large pores. In this study, we obtain six parameters (the volume, mean radius, and standard deviation of the small and large pores) that represent the characteristics of pore distribution and rock heterogeneity, calculate the total porosity via NMR logging, and classify the reservoirs via cluster analysis by adopting a bimodal lognormal distribution to fit the NMR T2 spectrum. Finally, based on the data obtained from the core tests and the NMR logs, the proposed method, which is readily applicable, can effectively classify the tight sandstone reservoirs.

NMR-Responsive Paramagnetic [M-EDTA] (M = Fe3+, Mn2+, Cu2+) Complexes to Differentiate T2-Distribution Signals of Crude Oil and Brine

Nuclear magnetic resonance (NMR) well logging tools are being commercially utilized for the characterization of various fluids confined within reservoir rocks. Conventional NMR T2-distribution measurements are unable to quantify the actual fractions of crude oil and brine simultaneously present in the reservoir rocks. Herein, we reported NMR-responsive paramagnetic [M-EDTA] (M = Fe3+, Mn2+, Cu2+) complexes having the capability to differentiate NMR T2-distribution signals coming from crude oil and brine by accelerating the water (1H) relaxation. The formation of these [Fe-EDTA]−, [Mn-EDTA]2–, and [Cu-EDTA]2– complexes was explored by UV–visible spectroscopy, and their chemical stability in brine solution was monitored by measuring the change in percent transmittance and backscattering intensities with time. Spin–spin (T2) relaxation signals and T2-distribution spectra of various carbonate and sand packs were recorded. The separation of NMR T2-distribution signals was achieved using the optimized concentra...

Sensitive parameters of NMR T2 spectrum and their application to pore structure characterization and evaluation in logging profile: A case study from Chang 7 in the Yanchang Formation, Heshui area, Ordos Basin, NW China

Abstract The evaluation of the pore structure of tight sandstone reservoirs has a significant influence on the effective exploration and development of tight sandstone oil. Laboratory measurements including scanning electron microscopy (SEM), mercury-injection capillary pressure (MICP), gas adsorption, nuclear magnetic resonance (NMR), and Nano-CT can provide detailed pore structure data. However, only NMR results can be used to evaluate the pore structure in a logging profile due to NMR logging. In this study, the relationships between NMR T2 distribution, pore structure, and pore size of tight sandstone in Chang 7 of the Yanchang Formation, in the Heshui area of the Ordos Basin, are analyzed based on NMR principles. The characteristics of the NMR T2 distribution of different rock samples are analyzed; the sensitive parameters of the NMR T2 spectrum are proposed. These parameters are then used to classify pore structure types in tight sandstone reservoirs and divide the active layers in the logging profile of the study area. The results indicate that different pore size structures have different distributions, and that the NMR T2 spectrum can highlight the difference in the pore structure type and pore size distribution using sensitive parameters such as T2P2 (the value of T2 corresponding to the last peak of the bimodal NMR T2 spectrum) and TDM (the mean value of T2 relaxation time obtained by weighted average method). As the pore structure of the rock samples in the study area worsened, their porosity and permeability worsened, T2P2 and TDM decreased, and the displacement pressure and NMR irreducible water saturation increased. The results of classifying the pore structure types obtained in the logging profile are helpful in evaluating the effectiveness of the reservoir and in the broad application of NMR logging in terms of understanding pore structure.

Origin and formation mechanisms of low oil saturation reservoirs in Nanpu Sag, Bohai Bay Basin, China

Abstract Low oil saturation reservoirs are increasingly encountered in superimposed sedimentary basins, however, the uncertainty about the origin of the low oil saturation becomes a major factor impeding further hydrocarbon exploration. Routine core analysis, (fluorescence) thin section, scanning electron microscope (SEM) and nuclear magnetic resonance (NMR) measurements were used to understand the characteristics and formation mechanisms of the low oil saturation reservoirs in the Paleogene Dongying Formation in the Nanpu sag, Bohai Bay Basin, East China. Irreducible water saturation was determined from sealed coring oil saturation testing and NMR logs. Wide ranges of porosity and permeability were observed, and the samples with moderate reservoir quality have the best oil bearing property. Low oil saturation oil layers are characterized by low oil saturation ( 40%), with only oil being produced whereas no water or only minor amounts of water being produced. The low oil saturation oil-water reservoirs refer to the low oil saturation layers in which both oil and water will be produced during oil testing. Irreducible water saturation increases with total clay content, with the authigenic clays (kaolinite, illite and mixed layer illite/smectite) blocking pore spaces, and resulting in high irreducible water saturations. Fluorescence thin sections combined with NMR measurements show that the NMR T2 (transversal relaxation time) distribution is uni-modal, with large intergranular pores being rare. Almost the entire pore-size range is oil bearing for the low oil saturation oil reservoirs. The large intergranular pores are fluorescence-free, while the intragranular dissolution pores and micropores emit fluorescence in the low oil saturation oil-water reservoirs. The NMR T2 distribution of low oil saturation oil-water reservoirs is bi-modal and right-skewed, and the main oil-bearing NMR pore size distribution ranges from 0.1 to 100 ms. High irreducible water content was encountered in the low oil saturation oil reservoirs and low oil saturation oil-water reservoirs, and only part of the large intergranular pores are oil saturated, and in some cases only the intragranular pores are saturated with oils forming low oil saturation oil reservoirs. The results help improve the knowledge of characteristics and formation of the low oil saturation oil layers, and could provide insights in enhancing oil recovery.

Neural network modeling of in situ fluid-filled pore size distributions in subsurface shale reservoirs under data constraints

Subsurface nuclear magnetic resonance (NMR) logs acquired in the wellbore environment are sensitive to fluid-filled pore size distribution, fluid mobility, permeability, and porosity in the near-wellbore reservoir volume. NMR response of a formation layer is processed to extract the T2 distribution, which approximates the fluid-filled pore size distribution. NMR logs are acquired in limited number of wells due to financial and operational challenges, which adversely affects reservoir characterization. We developed two neural-network-based machine learning techniques, long short-term memory (LSTM) network and variational autoencoder with a convolutional layer (VAEc) network, to process the ‘easy-to-acquire’ formation mineral and fluid saturation logs to generate synthetic NMR T2 distributions in the absence of ‘hard-to-acquire’ NMR T2 distribution log. Both the predictive models are trained and tested on limited wireline log measurements randomly selected from a 300-ft depth interval of the Bakken shale formation. Synthesis performances of LSTM and VAEc models in terms of R2 are 0.78 and 0.75, respectively. Noise is inevitable in logging data due to the complex wellbore and formation conditions. Notably, both the predictive models robustly synthesize the fluid-filled pore size distributions in the presence of 50% noise in input logs and 30% noise in training T2 data. The performance of the proposed methodology improves with access to larger volume of training data from other formation types. The proposed method is critical to the synthesis of in situ fluid-filled pore size distributions in shale formations under data constraints due to financial and operational challenges.

Inhibitor-wax interactions and PPD effect on wax crystallization: New approaches for GC/MS and NMR, and comparison with DSC, CPM, and rheometry

Abstract Three pour point depressants (PPDs) with different chemistries, and extracted asphaltenes were studied in their influence on wax crystallization in a model system. A new procedure for nuclear magnetic resonance (NMR) was developed to monitor changes in liquid phase as wax precipitated. Similar experiments were used to study n-alkane depletion during precipitation via the use of GC/MS. Additive performance was furthermore evaluated via differential scanning calorimetry (DSC), rheometry, and cross-polarized microscopy (CPM). All additives induced morphological changes to wax crystals, which also affected apparent viscosity and waxy gelling. Some additives improved flow-ability of the waxy model oil significantly, while others showed less pronounced or even adverse effects. Following DSC and NMR results, PPDs with good wax inhibition could delay both the onset of wax crystallization and reduce the amount of precipitated wax. In accordance with that, GC/MS showed additives to delay certain stages in compositional change during wax precipitation. Moreover, efficient PPDs could suppress interactions between liquid and solid waxes that showed in the NMR T2-distribution. It appears that PPD beneficiation greatly influences the mobility of dissolved wax molecules, which would also entail an effect on interactions at the solid-liquid interface of wax crystals. In conclusion, experimental procedures for NMR and GC/MS were extended to study PPD-wax interactions more in-depth, and experimental results show potential for improving the knowledge of wax inhibition.

An effective approach to generate drainage representative capillary pressure and relative permeability curves in the framework of reservoir electrofacies

Representative capillary pressure and relative permeability curves for reservoir rock types are important inputs for a dynamic reservoir simulator. The current study proposes a third-step approach to generate representative capillary pressure and relative permeability curves of reservoir rock types from nuclear magnetic resonance (NMR) log. First, reservoir electrofacies were identified by integration of conventional well logs and NMR log parameters. Employing cluster analysis reservoir rocks were classified into six electrofacies (EF) with reservoir quality being increased from EF1 to EF6. In the second step, the NMR T2-distribution arrays were converted to synthetic drainage capillary pressure curves and the results were validated by available laboratory measured mercury injection capillary pressure curves (MICP). Finally, the relative permeability curves were generated from the NMR-derived MICP data by using the Wyllie and Gardner equations for each individual electrofacies. Comparing the results with laboratory data indicates the effective role of electrofacies control in generating high accuracy capillary pressure and relative permeability curves. The results of this study show that the application of NMR log can further be extended through the generation of drainage capillary pressure and relative permeability data in the framework of electrofacies clustering.

Long Short-Term Memory and Variational Autoencoder With Convolutional Neural Networks for Generating NMR T2 Distributions

Downhole nuclear magnetic resonance (NMR) logs acquired in the borehole environment are valuable for subsurface characterization because they contain information about the pore size distribution, fluid composition, fluid saturation, fluid mobility, formation permeability, and porosity. NMR log acquisition can be challenging due to operational and financial constraints. Recently, NMR T2 distributions of the subsurface were generated by processing conventional well logs using deeplearning neural-network (NN) models. This improves the accessibility to subsurface pore size distributions. We implement two neural-network models, variational autoencoder-based NN with convolutional layers and long short-term memory (LSTM), to generate NMR T2 distributions from formation mineral content and fluid saturation logs. Prediction performance is evaluated for the entire NMR T2 spectrum ranging from 0.3 to 3000 ms as well as for T2 spectra within four bins obtained by dividing the entire spectrum into four equal parts. Each bin represents a specific pore size range. In terms of R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> , both the models have prediction performances above R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.75 for the entire spectrum. The best prediction performance is achieved for the bin of size ranging from 2.7 to 28 ms, representing pore diameters from 10 to 100 nm. The performance of this bin in terms of R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> is 0.78. The LSTM model is highly sensitive to noise in T2 distribution used during the training, and both the models are robust to noise in the conventional input logs.

A new method for measuring shale porosity with low-field nuclear magnetic resonance considering non-fluid signals

Abstract Porosity is a key parameter for the evaluation of potential shale oil and gas resources. Low-field nuclear magnetic resonance (NMR) is a rapid technique for measuring shale porosity without causing damage to the testing samples. Previous NMR methods for porosity characterization are mainly based on the NMR T2 distribution of saturated oil shale, but for shales with high total organic carbon (TOC) and clay mineral contents, not all signals of NMR detection originate from the pore fluid. Nine continental shales in Damintun Sag were selected and subjected to pyrolysis, X-ray diffraction (XRD), mercury injection capillary pressure (MICP), weighing before and after fluid saturation, and NMR tests to improve the accuracy of NMR porosity measurements. According to the NMR T2 distribution and T1–T2 map of shale with different oil contents (original, extracted, and saturated oil states), we revealed that the non-fluid signals (kerogen and structural water) accounted for 15.77%–43.10% (with a mean of 28.87%) of saturated oil shale. A stronger NMR signal intensity was observed for the extracted shale with higher TOC and clay mineral contents. A ΔT2 distribution of oil present in the pores of oil-saturated shale was constructed based on the difference between the oil-saturated shale and extracted shale and was combined with the calibration equation of oil (the relationship between oil volume and its NMR signal intensity) to directly evaluate the porosity. The porosity that was calculated based on the ΔT2 distribution versus the weighing method had an absolute error of ≤0.7%, and the relative error was

An Investigation into Different Correlation Methods between NMR T2 Distributions and Primary Drainage Capillary Pressure Curves Using an Extensive Sandstone Database

Many workers have recognised the link between Nuclear Magnetic Resonance (NMR) derived T2 distributions and pore size distributions in reservoir rocks. This property has been used to develop models to obtain primary drainage capillary pressure curves from T2 distributions. These models often assume that the rocks pore space resembles a simple bundle of capillary tubes. They do not consider the existence of multiple pore body connections and pore body restrictions/throats. The most successful models utilise variable scaling factors to convert T2 times to pore diameters and hence capillary pressure. The variable scaling factor approach recognises the existence of variable surface relaxivity throughout the pore space due to variations in mineralogy and pore topography. This investigation uses SCAL data from the ART NMR Sandstone Rock Catalogue to obtain core calibrated variable scaling factors for 174 reservoir sandstone samples. The depositional environments for these samples include; aeolian, fluvial, coastal and shallow and deep marine. The samples used have a wide variety of mineralogy, diagenetic overprints and cover six orders of magnitude in absolute permeability. Three different methods for obtaining the scaling factors are presented and the relative merits of each discussed. A global model to predict capillary pressure from NMR T2 distributions in reservoir sandstones has been developed using correlations between the variable scaling factors and permeability.

F63IDENTIFICATION OF PERIPHERAL miRNA BIOMARKERS IMPLICATED IN THE DEFAULT MODE NETWORK

Representative capillary pressure and relative permeability curves for reservoir rock types are important inputs for a dynamic reservoir simulator. The current study proposes a third-step approach to generate representative capillary pressure and relative permeability curves of reservoir rock types from nuclear magnetic resonance (NMR) log. First, reservoir electrofacies were identified by integration of conventional well logs and NMR log parameters. Employing cluster analysis reservoir rocks were classified into six electrofacies (EF) with reservoir quality being increased from EF1 to EF6. In the second step, the NMR T2-distribution arrays were converted to synthetic drainage capillary pressure curves and the results were validated by available laboratory measured mercury injection capillary pressure curves (MICP). Finally, the relative permeability curves were generated from the NMR-derived MICP data by using the Wyllie and Gardner equations for each individual electrofacies. Comparing the results with laboratory data indicates the effective role of electrofacies control in generating high accuracy capillary pressure and relative permeability curves. The results of this study show that the application of NMR log can further be extended through the generation of drainage capillary pressure and relative permeability data in the framework of electrofacies clustering.