WO2013148516A1 - Nuclear magnetic resonance testing for organics and fluids in source and reservoir rock - Google Patents

Abstract

A method of analyzing a geologic sample by NMR is provided for detecting the presence of organic solid and semi-solid materials and saturating fluids by comparing the results of a baseline NMR pulse sequence test with the results of a magnetization transfer contrast NMR pulse sequence test wherein the magnetization transfer NMR test involves a first application of a magnetization transfer contrast preparation RF pulse sequence sufficient to decrease or destroy magnetization coherence attributable to any organic solid and semisolid materials that might be in the geologic sample followed by application of the baseline NMR pulse sequence.

Description

NUCLEAR MAGNETIC RESONANCE TESTING FOR ORGANICS AND FLUIDS IN

SOURCE AND RESERVOIR ROCK

Inventor: Kathryn E. Washburn

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority based on US Provisional Application Serial No. 61/615,250 filed 03-24-2012, which is incorporated herein by reference.

FIELD OF THE INVENTION

[0002] This invention relates generally to apparatus, compositions and methods for nondestructive reservoir and source rock analysis by NMR.

BACKGROUND OF THE INVENTION

[0003] Without limiting the scope of the invention, its background is described in connection with existing non-destructive analysis of core or cuttings samples as well as downhole reservoir analysis. Unconventional reservoirs for hydrocarbons including coalbed methane, tar sands, heavy oil, oil shale, oil bearing shale and shale gas are becoming increasingly important global energy resources. However, it has been difficult to estimate the quantities of potentially recoverable hydrocarbons in these reservoirs.

[0004] Characterizing the organic content is important for evaluating unconventional resources as current research indicates the organic matter holds a significant amount of the hydrocarbons present. Current methods of evaluating Total Organic Content ("TOC") are destructive. Carbon determination using LECO (LECO Corporation, St. Joseph, MI, US) or Carmhograph (Wosthoff, Bochum, Germany) apparatus estimate the organic content of the rock by removing carbonates for the samples by acid treatment and then heating the samples to high temperatures (1500°C for LECO, 1100°C for Carmhograph) under air or an oxygen atmosphere. There are also programmed pyrolysis methods to characterize source rocks such as Rock Eval ("RE") or Source Rock Analysis (SRA) that involve programmed temperature heating of a small sample in an inert atmosphere to volatilize extractable and unextractable organic matter in the sample and measure the production of different chemical species at different temperatures. These techniques are used to quantify the amount of hydrocarbons in the sample, the organic matter present, and the type of organic matter present. There has also been some work to chemically isolate components in organic shales.

[0005] The naturally occurring, solid, insoluble organic matter that occurs in source rocks is termed kerogen. Kerogen, which is a mixture of organic compounds, is typically high molecular weight and insoluble in normal organic solvents. Thus, to analyze kerogen alone, harsh chemicals such as hydrofluoric acid are required to dissolve away the rock matrix and isolate the constituent kerogen. Kerogen type helps predict the type of hydrocarbon that can be found in the reservoir. Type I kerogens are derived from mainly algal sources and highly likely to generate oil. Type II kerogens are of mixed terrestrial and marine source material that can generate waxy oil, while Type III kerogen derived from woody terrestrial source material that typically generates gas. Type IV is inert and is not expected to produce any appreciable type of hydrocarbon. Also, given the heterogeneity within shales and the small sample volumes used for these measurements, the organic content from a sample sent for source rock analysis or similar characterization may not be representative of another, nearby sample. This can make correlation between the geochemistry and different measurements difficult.

[0006] The naturally occurring soluble organic matter that occurs in source rocks is termed bitumen. Bitumen may be present in small quantities in immature source rocks. When rocks undergo catagenesis, the kerogen is cracked into smaller molecules to produce additional quantities of bitumen. These molecules tend to be less viscous than kerogen and can be a precursor to oil generation. As thermal maturity increases, bitumen cracks into smaller, stiffer molecules. In addition, in samples with high thermal maturity, significant quantities of pyrobitumen may be present. Pyrobitumen differs from normal thermobitumen in that, while it is a product of cracked kerogen, it is not extractable from the rock via solvents.

[0007] There has also been some work using spectroscopic techniques like Fourier Transform Infrared Spectroscopy ("FTIR") and Raman Spectroscopy to both quantify and characterize organic content in core samples. FTIR can be performed on only a small fraction of a sample and measurements may not adequately reflect heterogeneity in the sample, again leading to problems in correlation of results. These measurements also typically require the powdering of the samples.

[0008] Nuclear magnetic resonance ("NMR") measurements have been used extensively in the oil industry for characterization of geological formations. These measurements have been made both down the bore hole and in the laboratory on retrieved samples. Commonly, NMR is used to estimate the total pore volume (total porosity) of the rock and has been assumed to be a matrix independent methodology. Other petrophysical techniques (e.g., density porosity measurements) are required to determine if the rock is sandstone, limestone, etc. to be able to calibrate porosity. Nuclear magnetic resonance finds porosity by measuring the amount of hydrogen present in the sample. The intensity of the NMR signal is proportion to the amount of hydrogen present in the sample. By calibrating the intensity of the resulting NMR signal to the signal intensity of known fluids, the porosity of the system can be found. In this analysis it is assumed that the matrix contains little hydrogen itself. Studies have shown good correspondence of porosity calculated from the NMR signal to other porosity methods as shown in Figure 1.

[0009] There has been some high field strength NMR work conducted on kerogens and bitumens. The studies have been mostly using the chemical shift phenomenon in NMR. Chemical shift is the change of measured frequency of the NMR signal due to the change in electron distributions around a nucleus due to bonding with other atoms. Because the ¹³C nucleus has a much larger chemical shift than 1H, many of these studies have focused on the NMR spectra of carbon. However, due to the weak NMR signal of the carbon nucleus, cross- polarization is frequently performed to improve the signal to noise ratio of the measurement, which improves signal quality. Cross polarization increases the signal strength of the carbon by transferring magnetization from 1H nuclei, which have a much stronger NMR signal, to the weaker ¹³C. When the radio frequency ("RF") pulse of the stronger and weaker nuclei are matched during a long matched RF pulse where γ^χΒ^* = γ^γ , magnetization from the stronger nuclei enhances the signal of weaker nuclei if they are coupled, up to a factor of y ^x / y ^y where X is the stronger nuclei. To perform cross-polarization, a very high power probe that can operate on two frequency channels is necessary.

[0010] In geological samples with organics, the strong interactions between the hydrogen in the organics leads to a broad linewidth which obscures the underlying structure from chemical shift. To overcome the broadening, a specialized set of equipment is necessary. By placing the sample at a precise angle with the applied magnetic field and rapidly spinning it, the dipolar interactions between molecules can be decreased and some cases averaged down to effectively zero, such that the underlying chemical shift in the spectra can be observed. The high field work has mostly been concerned with the chemical structure of the organics, such as the alkane and aromatic content of the materials, and not with quantifying them. These measurements often require specialized equipment, such as the rotors for magic angle spinning and specialized probes for the cross-polarization. Given the mechanical difficulty in spinning the samples to the desired rate, the sample volume usually needs to be limited to a small amount of sample, typically powdered. This again runs into the problem of addressing both heterogeneity in the sample and non-destructiveness. In many cases, these measurements also require separating out the kerogen from the rock, which loses the nondestructive advantage.

[0011] In addition to an estimate of total porosity, NMR has been used to estimate a pore size distribution. However, as to use of NMR to detect and characterize the presence of organic material, the signal from organic materials may decay too quickly to be measured by NMR technologies commonly used in core analysis and well logging. Some of the latest technology may be able to capture some of the signal from organics, but in many cases, distinguishing the solids from the liquids in the sample can be difficult due to overlapping signals and uncertainty in interpretation due to non-unique signal responses of different constituents.

[0012] What is needed is a non-destructive, bulk measurement that could more accurately identify and estimate organic content, the properties of the organic content, and fluids present in both the organic and inorganic pore space while preserving the expensive, valuable sample for multiple further measurements. From the foregoing it is apparent the there is a need in the art for non-destructive methods and can be used to analyze rock samples for reservoir analysis. Provided herein are novel methods and apparatus for nondestructive source and reservoir rock analysis by nuclear magnetic resonance.

SUMMARY OF THE INVENTION

[0013] Provided herein are methods of analyzing geologic samples by NMR for the characterization of organic solid and semi-solid materials and saturating fluids in the sample. The methods rely on a comparison between the results of a baseline NMR pulse sequence test run on the geologic sample and a magnetization transfer contrast NMR pulse sequence test. The magnetization transfer contrast is conducted by first applying a preparation RF pulse sequence to the geologic sample to decrease magnetization attributable to any organic solids or semi-solid materials in the sample then waiting a variable time for any transfer of magnetization between sample fluids or potentially semi-solids that have retained their magnetization and the decreased magnetization of the organic solids and semi-solids before performing the same baseline NMR pulse sequence on the geologic sample. The variable time will be determined empirically for a given sample. The results from the baseline NMR pulse sequence test are compared with the same NMR pulse sequence preceded by magnetization transfer contrast preparation pulse(s) to observe transfer or non-transfer of magnetization between any organic solids and semi-solids and fluids present in the geological sample and thereby determine characteristics of organic materials and saturating fluids in the geologic sample. The baseline NMR pulse sequence test can be conducted before or after the magnetization transfer contrast NMR pulse sequence test.

[0014] The organic solids or semi-solid materials detectable by the method may be in the form of kerogen, bitumen, heavy oil or waxes. In some embodiments the method is performed in a laboratory or wellsite setting while in other embodiments the method is performed downhole.

[0015] The geologic samples to be tested can be in treated or untreated states including: native, native resaturated with brine, oil or another fluid, cleaned and dried, cleaned and resaturated with brine, oil or another fluid, cleaned, resaturated with brine and then desaturated with air, cleaned, resaturated with brine and then desaturated with oil, cleaned, resaturated with brine and then desaturated with oil and aged, and any of the above states pressurized with gas, either wet or dry.

[0016] The analysis can be conducted under conditions selected from one or more of: ambient temperature and pressure, elevated temperature and elevated pressure, elevated temperature and ambient pressure, elevated pressure and ambient temperature, cooled temperature and ambient pressure, cooled temperature and elevated pressure, any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states without permanently altering any chemical or structural features of the core sample, and any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states and permanently alter some chemical or structural features of the core sample.

[0017] In certain embodiments the magnetization transfer preparation RF pulse sequence can be a binomial pulse sequence or can be a series of binomial pulse sequences. Alternatively the preparation RF pulse sequence can be an off-resonance pulse. [0018] A number of different NMR pulse sequences can be utilized including a Carr-Purcell- Meiboom-Gill (CPMG) spin echo train; a spin echo pulse sequence; a solid echo sequence; a solid echo train sequence, a free induction decay pulse sequence, an inversion recovery sequence, a saturation recovery sequence, a quantum filter measurement sequence; and an internal gradient measurement sequence.

[0019] In certain embodiments the baseline NMR pulse sequence is a combination two or more different NMR pulse sequences including a CPMG spin echo train, a spin echo pulse, a solid echo sequence, a solid echo train sequence, a free induction decay pulse sequence, an inversion recovery sequence, a saturation recovery sequence, a diffusion measurement, a quantum filter measurement sequence, and an internal gradient measurement sequence.

BRIEF DESCRIPTION OF THE DRAWINGS

[0020] For a more complete understanding of the present invention, including features and advantages, reference is now made to the detailed description of the invention along with the accompanying figures:

[0021] Figure 1 illustrates the correspondence of porosity calculated from the NMR signal using existing techniques to other porosity methods.

[0022] Figure 2A represents a spin echo pulse. Figure 2B represents a solid echo. Figure 2C presents a spin echo pulse and a series of spin echoes to produce a CPMG sequence pulse.

[0023] Figures 3A - D show the rotation of the magnetization vector for different types and phase of pulses.

[0024] Figures 4A and B present a simplified illustration of the effect of binomial pulses on liquids and solids. Magnetization of liquids (4 A) is left relatively unchanged while dephasing destroys the magnetization coherence from solids (4B) leaves their resulting signal reduced or gone.

[0025] Figure 5 depicts the effects of an off resonance pulse upon the magnetization of solids versus liquids.

[0026] Figure 6 illustrates a binomial pulse sequence with a variable wait time combined with an FID measurement.

[0027] Figure 7 illustrates two sets of binomial pulses combined with a CPMG pulse sequence. [0028] Figure 9 illustrates a set of binomial pulses combined with a T1-T2 correlation measurement.

[0029] Figure 10A shows the results of a CPMG pulse vs Binomial vs Tl edited in organic oil-bearing shale.

[0030] Figure 10B shows the results of a CPMG pulse vs Binomial CPMG in a tight carbonate with minimal organics expected.

[0031] Figure 11 shows the results of a CPMG pulse vs Binomial vs Tl edited in oil shale.

[0032] Figure 12 shows the results of a T1-T2 correlation measurement.

[0033] Figure 13 shows results of a T1-T2 binomial correlation measurement with magnetization transfer contrast.

[0034] Figure 14 shows one example of a stepwise method of magnetization transfer contrast for reservoir and source rock analysis.

DETAILED DESCRIPTION OF THE INVENTION

[0035] While the making and using of various embodiments of the present invention are discussed in detail below, it should be appreciated that the present invention provides many applicable inventive concepts which can be employed in a wide variety of specific contexts. The specific embodiment discussed herein are merely illustrative of specific ways to make and use the invention and do not delimit the scope of the invention.

[0036] The present inventor appreciated that in unconventional resources, many previous assumptions made about NMR are no longer true. First of all, in some geological formations, such as shales, it is no longer a correct assumption that the matrix contains little or no hydrogen. In many organic shales, it is believed a significant proportion of the porosity exists in pores in the organic matrix rather than the conventionally perceived inorganic matrix. However, in many cases, it has been appreciated by the present inventor that some of signal from the organics relaxes so quickly that it cannot be accurately measured with current technology, in part because the pulse lengths and dead times of existing equipment are too long. The portion of the organic signal that does not decay too quickly to be measured will add signal intensity and short T2 times and potentially long and short Tl times. Instead of the T2 and Tl signal reflecting the fluid filled pores, it now contains contributions from the matrix. [0037] Based on the composition of the organics and the saturating fluids and measurements undertaken by the present inventor, interactions between solids and fluids have been detected and the methods have been adapted to characterization of source and reservoir rock by NMR. Unlike kerogen separation, LECO, Carmhograph, SRA, or Rock Eval, this technique is nondestructive, non-invasive, does not necessitate the crushing of the sample or separation of the organics from the rock matrix. The technique is a bulk measurement, unlike methods such as FTIR. Although high field resolution or specialized NMR equipment may be employed, such is not required in accordance with certain embodiments. The technique disclosed herein relies on an NMR measurement of the hydrogen atoms, without involving other nuclei such as carbon. Given the weak NMR signal of carbon, relying on hydrogen measurements alone will allow a more rapid measurement than compared to techniques that need to use carbon. It also allows for the use of standard equipment, as opposed the multifrequency probes that are needed for measurements than involve two different nuclei, such as ¹H-¹³C. In contrast to high field NMR work that has been primarily narrow line spectroscopy, in certain embodiments disclosed herein, NMR relaxometry is a focus. In certain embodiments, measurements are focused on dipolar coupling, which is the phenomenon that high field NMR work often attempts to remove from measurements.

[0038] To perform a magnetization transfer contrast technique as disclosed herein, at least two measurements are needed. One is a baseline NMR measurement using a without the magnetization transfer contrast section preceding it, compared with the same baseline NMR measurement preceded by a magnetization transfer contrast section. The baseline NMR test records the response of the sample to a given pulse sequences. Suitable pulse sequences can be determined empirically for the same but may include, without limitation, a Carr-Purcell- Meiboom-Gill (CPMG) pulse echo train, a spin echo pulse, a solid echo, a solid echo train, a free induction decay pulse sequence, an inversion recovery sequence, a saturation recovery sequence, a diffusion measurement, a quantum filter measurement sequence, an internal gradient measurement sequence and combinations thereof including a T1-T2 correlation.

[0039] The baseline NMR measurement results are compared with the signals obtained when baseline NMR measurement is preceded by a preparation pulse or pulses directed to induce magnetization transfer contrast to see how the signal is transformed due to the magnetization transfer contrast section. There is a variable wait time between the magnetization transfer preparation pulse sequence and the subsequent NMR measurement to allow time for magnetization transfer between the solids, semi-solids and fluids through dipolar coupling and chemical exchange. The optimum wait time is sample dependent and will need to be determine empirically. The wait time should not be so long that the magnetization transfer contrast preparation has been lost due to Tl relaxation. It may be desirable to run several different measurements with different magnetization transfer contrast parameters to obtain a more complete picture of the interactions in the system.

[0040] The processes disclosed herein rely on a certain type of magnetization transfer as described in more detail herein. There are two main methods of coupling that could lead to magnetization transfer in rock systems. The first is through dipolar coupling. In dipolar coupling, the nuclei of two spins in close proximity influence each other. The strength of the dipolar coupling is related to the two nuclei undergoing coupling, and their distance from each other and their angle with the applied magnetic field. It is independent of the frequency of the applied magnetic field. When spins are coupled, they sometimes will exchange magnetization with each other, it is referred to as "Spin Diffusion." The term spin diffusion gets used indiscriminately, and is often used when referring to "molecular diffusion of spin- bearing molecules", leading to confusion of the two effects. The former refers to exchange of magnetization between coupled spins. The later refers to the random movement of molecules. Also, when there are significant paramagnetics in the sample, spin diffusion around these can lead these to act as magnetization "sinks." As used herein the term "spin diffusion" refers to exchange of magnetization between coupled spins.

[0041] The second is chemical exchange. Some molecules that have a tendency to exchange their hydrogen atoms with one another. When the hydrogen molecules exchange, their magnetization also exchanges. The likelihood of this occurring increases with the acidity of the system, which is given by an acid dissociation constant, a pKa value. Molecules, like methane, with a very high pKa value (~55) are unlikely to exchange their hydrogen atoms. Water is has a pKa value of ~ 12, which makes it a likely candidate for exchange. Organic molecules such as kerogen and bitumen that contain acid functional groups (COOH, NO, SO₃H, etc.) have a low pKa and are also likely to exchange. The problem of chemical exchange averaging the NMR response between different species within a sample is a well- known issue in areas of NMR research in protein structure and macromolecule characterization, but has been of little concern previously for NMR in oilfield exploration. [0042] There are two other types of coupling where magnetization transfer may occur called j-coupling and NOE. In the J-coupling type of coupling or scalar coupling, the two nuclei interact through their electrons. This coupling is much weaker than dipolar coupling, and will tend to be overwhelmed if dipolar coupling is also present. The Nuclear OverHauser Exchange ("NOE") is the second form of potential coupling. Unlike the other forms of coupling, which occur through bonds between atoms, the NOE occurs through space, so no direct bond is required. These two types of couplings are expected to be much weaker than the dipolar coupling and chemical exchange.

[0043] When there is coupling going on between spins of different environments, this can affect the measured signal in a variety of ways. If the exchange is very rapid compared to the rate of measurement, this will lead to a single averaged value. If the exchange is slow, the two components will appear as the two distinct regimes. If the exchange is on the order of the measurement, this will lead to an incomplete averaging of the two environments.

[0044] Magnetization transfer occurs naturally in a system where dipolar coupling, chemical exchange, j-coupling or NOE coupling exists. However, we are unable to observe this transfer directly because in a normal NMR measurement the sample constituents all have magnetization to exchange, so the transfer is not observed. To observe the magnetization transfer, we need to induce a contrast. This is done by getting rid of the magnetization associated with one or more of the constituents in the sample before making an NMR measurement. Then, as the magnetization is transferred between the sample constituents, this will affect the resulting signal of an associated NMR measurement, such as a Tl or T2 distribution.

[0045] An NMR test starts by placing a sample into or near a static magnetic field (B0). The strength of this applied magnetic field can vary between equipment. For illustrative purposes, some nuclei can be thought to act like little bar magnets, and their magnetic moment will align along an applied magnetic field. The measurement is made by exciting the aligned nuclei away from this equilibrium. This is done with a radio frequency (RF) pulse (also called an oscillating field or the Bl field) produced by a "probe." The probe is usually a coil of wire although other geometries may be used and the geometry is not of particular importance to the measurement. The maximum measured signal is obtained when the nuclei are rotated 90° away from the applied static field. A pulse that rotates the magnetization 90° is referred to a 90° pulse. The rotation caused by an RF pulse is a combination of power and duration of the RF pulse and this varies from equipment to equipment. The magnetic moment of the nuclei precess around the applied magnetic field and are referred to as "spins". The rate of precession is a combination of the type of nuclei under observation, the applied static or constant field (H₀), and then nuclear and electric interactions the nuclei in the sample are undergoing. This precession induces a voltage in the coil. A change of magnetic field will induce a change of current in a coil, leading to a voltage. There is a dead period after an RF pulse where magnetization cannot be measured because there is also "ringing" in the coil due to the produced RF pulse. The length of this dead period is also equipment dependent.

[0046] There are two ways the NMR signal relaxes after excitation, Tl and T2 relaxation. Tl relaxation time (also called spin-lattice relaxation or longitudinal magnetic relaxation) describes the time taken the magnetic moment of nuclei to align along the applied magnetic field after first being placed in the magnetic field or the time required to regain longitudinal magnetization following an RF pulse. Tl is determined by interactions between the resonating protons and their environment ("lattice") that allow the energy absorbed by the protons during resonance to be dispersed in the lattice.

[0047] The second process is called T2 relaxation (also known as spin-spin relaxation or transverse relaxation). T2 is a measure of how long the precessing spins take to go from a coherent state to a disordered state. T2 decay is due to magnetic interactions that occur between the spins, each other and their environment. In contrast to Tl interactions, T2 interactions involve a phase change and thus a loss of coherence between spins rather than a transfer of energy. Initially after excitation, all the spins precess in unison around the applied magnetic field. As times goes on, the spins interact with each other and their environment and get out of sync. The time it takes for the spins to lose their coherency and lose order in their precession around the applied magnetic field is the T2 time. Tl and T2 relaxation events occur simultaneously but the Tl must be longer than or equal to the T2 time. In some cases, such as solids, the Tl time is significantly longer than the T2 time. There is an additional type of T2 relaxation called T2* relaxation. This is additional dephasing of the spins due to time invariant magnetic field inhomogenieties, caused by things such as magnet inhomogeneity, dipolar coupling, and chemical shift. Unlike T2, T2* relaxation is reversible and can be reversed with a range of different pulse sequences such as a spin echo or a solid echo. [0048] When fluids are inside a porous material, such as rocks that may act as a reservoir or source rock, the relaxation time of the bulk fluid is enhanced by contact with pore surfaces. This will hold true for both Tl and T2 relaxation processes. In smaller pores, the fluid will encounter the pore surface more frequently than in larger pores. This means that fluid inside small pores has a bulk relaxation time that is quicker than that of fluid inside large pores. The general basic relation to correlate for the measured Tl or T2 is given by:

1 s

T ⁼ A 2 ~~

M

^Al,2 v ^V

Where S and V are pore surface and volume, T_lj2 is relaxation time for longitudinal and transverse magnetization respectively and ?_lj2 is the surface relaxivity.

[0049] The surface relaxivity of the system is how effective the surface is in enhancing the relaxation of the fluid and depends on the lithology of the rock and the amount of paramagnetics in the samples. In addition, the T2 relaxation will have additional influence from the presence of internal gradients as defined by the following equation:

1 1 S /D²Gte

— = + — +

T ^A2 T ¹2Bulk V ^¥ 1 ⁱ2^l

Where y is the gyromagnetic ratio for the nuclei under observation, D is the diffusion coefficient, and G is the constant magnetic field gradient where T2 is measured with either a CPMG pulse sequence of inter-echo spacing TE or a variable TE Spin Echo sequence.

[0050] The standard measurement for measuring Tl relaxation is called "Inversion recovery." In certain embodiments of the present invention, the baseline NMR pulse sequence includes application of an inversion recovery sequence to the geologic sample. Inversion recovery starts with an initial 180° pulse to invert the magnetization to lie along the negative Z-axis. The magnetization is allowed to relax for a time Tl-tau. Then a second 90° pulse is used to place the magnetization in the XY plane. The inversion recovery may be measured from the Free Induction Decay (FID), off a spin echo, or solid echo. If the signal is measured off of an FID, the NMR signal is recorded after the dead time after the 90° pulse. If the signal is measured off of a spin echo, a 180° pulse will be performed at a time tau (τ) after the 90° pulse and the signal is measured from the resulting echo that occurs at a time tau (τ) after the 180° pulse. If the signal is measured off of a solid echo, another 90° pulse will be performed at a time tau after the first 90° pulse and the signal is measured from the resulting echo that occurs at a time tau (τ) after the second 90° pulse. The time Tl-tau is varied from short to long values to adequately measure the range of possible Tl relaxation times in the sample. There is another method called saturation recovery that is used to measure Tl, but it is less preferred.

[0051] A preferred method of measuring the T2 relaxation is using a particular spin echo train sequence called the Carr-Purcell-Meiboom-Gill (CPMG) sequence, named after the creators. The sequence uses an initial 90° pulse to excite the system. After the spins begin to dephase, the spins start interacting with each other and their environment to lose coherency in their precession. Some of this loss of unison is random, while another part is due to local magnetic field inhomogeneities. The loss of coherency because of time independent magnetic field inhomogeneity is called T2* relaxation and because the local variants in the field inhomogeneity are not random, they can be refocused. To do this, a 180° RF pulse is applied at a time tau after the initial 90° pulse. The spins then start refocusing until they regain coherency at a time tau after the 180° pulse to produce what is referred to as a spin echo as depicted in Figure 2. However, some of the signal intensity is lost due to the random true, underlying T2 relaxation. After an echo is produced, the magnetization begins to dephase again. By applying yet another 180° pulse at time 2 x tau, the magnetization can be refocused again. This can be repeated until the magnetization has completely decayed away as depicted in Figure 2. By measuring the intensity of each echo and applying an inversion, the raw data can be converted to a distribution of T2 times within the sample.

[0052] There is another, less preferred method of measuring T2 where only a single 180° pulse is used and the time between the initial 90° pulse and the second 180° pulse is varied. When there is molecular diffusion through magnetic gradients, this will lead to loss in addition to the underlying T2 magnetization. These magnetic gradients can arise from gradients in the applied B0 field, as with logging tools, or due to magnetic susceptibility differences between the rock matrix and saturating fluids. While some applications make use of this effect, in general it is preferred to keep the pulse sequence echo time (TE) as short as possible to avoid this influence on the measured signal.

[0053] There is another way to refocus magnetization lost called a "solid echo sequence." The sequence uses an initial 90° pulse to excite the system. After the spins begin to dephase, the spins start interacting with each other and their environment to lose coherency in their precession. If the loss of coherency is due to homonuclear dipolar coupling, they can be refocused with a solid echo. To do this, a 90° RF pulse is applied at a time tau after the initial 90° pulse. The spins that have dephased due to homonuclear dipolar coupling then start refocusing until the regain coherency at a time tau after the 90° pulse to produce what is referred to as a solid echo as depicted in Figure 2B. However, some of the signal intensity is lost due to the random true, underlying T2 relaxation. In addition, the solid echo is only able to completely refocus magnetization dephased due to homonuclear dipolar coupling for isolated spin pairs. The solid echo will be less effective at refocusing magnetization lost due to homonuclear dipolar for three or more coupled spins. After an echo is produced, the magnetization begins to dephase again. By applying yet another 90° pulse at time 2 x tau, the magnetization can be refocused again. This can be repeated until the magnetization has completely decayed away, creating a "solid echo train." With the solid echo train sequence, multiple solid echoes are performed in succession instead of just one. By measuring the intensity of each solid echo and applying an inversion, the raw data can be converted to a distribution of T2 times within the sample.

[0054] In certain embodiments the baseline NMR pulse sequence includes application of a "free induction decay" pulse sequence to the geologic sample. A free induction decay pulse sequence is characterized by measurement of the NMR signal after the dead time of the probe from a single 90° pulse.

[0055] Alternatively the baseline NMR pulse sequence can include a "saturation recovery sequence" to measure Tl, which is characterized by numerous 90° degree pulses in short succession followed by a wait time to allow Tl relaxation. The number of 90° pulses needed to adequately saturate the system is determined empirically for a given sample. The series of 90° pulses is performed to saturate the magnetization of the entire system to an excited state. A period of time Tl-tau is waited to allow the excited magnetization to relax via Tl relaxation. After the time Tl-tau another 90° pulse is applied. The saturation recovery signal can be measured from the Free Induction Decay (FID), off a spin echo, or solid echo. If the signal is measured off of an FID, the NMR signal is recorded after the dead time after the 90° pulse. If the signal is measured off of a spin echo, a 180° pulse will be performed at a time tau (τ) after the 90° pulse and the signal is measured from the resulting echo that occurs at a time tau (τ) after the 180° pulse. If the signal is measured off of a solid echo, another 90° pulse will be performed at a time tau (τ) after the first 90° pulse and the signal is measured from the resulting echo that occurs at a time tau (τ) after the second 90° pulse. This is performed for a range of different Tl-tau values, ranging from short to long times, to adequately characterize the range of Tl relaxation in the sample.

[0056] In other embodiments the baseline NMR pulse sequence includes application of a "quantum filter measurement" sequence to the geologic sample. Common quantum filter measurements are the double quantum filter and the triple quantum filter. These are used to filter out all signal except those arising from a double quantum coherence or triple quantum coherence respectively. In practice, quantum filters can be designed to filter for signal from a quantum coherence of any given rank. A double quantum filter begins by a 90° pulse to place the magnetization in the XY plane. The system is allowed to evolve for a time tau (τ) in the presence of dipolar and j -couplings, which will lead to the creation of double quantum coherences. After a time tau, a second 90° pulse is applied. A 180° pulse may be inserted at time tau/2 to refocus dephasing but is not required. Finally, a third 90° pulse places the magnetization into the XY plane for measurement of the NMR signal. This measurement is repeated several times and the resulting signals are added together. The phase of the pulses are altered between the measurements such that when the signals are added together, only signal arising from the double quantum coherences remains and other signal cancels out. For other types of quantum filters, the angles of the pulses may be different.

[0057] In still other embodiments the baseline NMR pulse sequence includes application of an "internal gradient measurement" sequence to the geologic sample. An internal gradient measurement sequence involves application of a 90° pulse to place the magnetization into the XY plane. After the 90° pulse, there is a fixed period of time. After this fixed period of time, the NMR signal is measured. In this fixed period of time, there are a varied number of 180° pulses, from as low one to as many as possible that can be performed with the equipment at hand. The decrease in signal as the number of 180° pulses decreases is used to encode for the internal gradients present in the system.

[0058] In still other embodiments the baseline NMR pulse sequence includes measurement of the "diffusion coefficient." There are three main ways to measure the diffusion coefficient. The first is the Pulsed Field gradient Spin Echo experiment. A 90° pulse is used to excite the system. A transient magnetic field gradient is applied for a time δ to the system to impart a phase shift to the precessing spins. The gradient is turned off. There is a period of time to allow the spin bearing molecules to diffuse and a 180° pulse is applied. There is a period of time to allow the spin-bearing molecules to diffuse. A transient magnetic field gradient is applied for a time δ to the system to again impart a phase shift to the precessing spins. The second gradient will serve to refocus the magnetization from spin bearing molecules that have not moved since the first transient gradient, but will not complete refocus the magnetization from spin-bearing molecules that have moved since the first. This is repeated with an increased gradient strength to increase attenuation of signal. The decrease in signal as the transient gradient is applied is used to encode for the diffusion coefficients of the constituents in the sample.

[0059] The second way to measure diffusion is with the Pulsed Field Gradient Stimulated Echo. A 90° pulse is used to excite the system. A transient magnetic field gradient is applied for a time δ to the system to impart a phase shift to the precessing spins. The gradient is turned off. There is a brief period of time and a 90° pulse is applied to store the magnetization along the z-axis. There is a period of time to allow the spin-bearing molecules to diffuse. Another a 90° pulse pulse is applied to return the magnetization to the XY plane. A transient magnetic field gradient is applied for a time δ to the system to again impart a phase shift to the precessing spins. The second gradient will serve to refocus the magnetization from spin bearing molecules that have not moved since the first transient gradient, but will not complete refocus the magnetization from spin-bearing molecules that have moved since the first. This is repeated with an increased gradient strength to increase attenuation of signal. The decrease in signal as the transient gradient is applied is used to encode for the diffusion coefficients of the constituents in the sample. The third way is the constant gradient variable echo spacing method. Here, a constant gradient is applied during the course of the measurement. A 90° pulse is used to excite the system. There is a wait time tau (τ) and a 180° pulse is used to refocus the magnetization. The value of tau (τ) is varied and the decrease in signal with increased tau is measured to determine the diffusion coefficient of the constituents in the sample.

Example I: Magnetization Transfer Testing of Core Samples

[0060] To perform a magnetization transfer contrast technique as disclosed herein, at least two measurements are needed. One is a regular baseline NMR pulse sequence measurement without the magnetization transfer contrast section preceding it, and another in which the same baseline NMR pulse sequence is preceded by a preparation RF pulse sequence that will enable detection of magnetization transfer events. The two measurements are then compared to see how the signal is transformed due to the magnetization contrast section. There is a variable wait time between the magnetization transfer preparation pulse sequence and the subsequent NMR measurement to allow time for magnetization transfer between the solids, semi-solids and fluids through dipolar coupling and chemical exchange. The optimum wait time is sample dependent and will need to be determined empirically. The wait time should not be so long that the magnetization transfer contrast preparation has been lost. It may be desirable to run several different measurements with different magnetization transfer contrast parameters, including multiple wait times, to obtain a more complete picture of the interactions in the system.

[0061] The measurement starts by taking a geological sample consisting of a rock matrix and a potentially saturated pore space. The sample may be in any of the following:

• native state or fresh state( which may or may not contain invasion from drilling fluids);

• native or fresh state resaturated (flushed or pressure saturated) with brine;

• native or fresh state resaturated (flushed or pressure saturated) with oil;

• native or fresh state resaturated (flushed or pressure saturated) with another fluid;

• cleaned and dried;

• cleaned and resaturated with brine;

• cleaned and resaturated with oil;

• cleaned and resaturated with another fluid;

• cleaned, resaturated with brine, desaturated with oil and aged; and

• any of the above states pressurized with gas, either wet or dry.

[0062] The sample may be a regular cylinder as commonly used in rock cores or core plugs. Dimensions are typically 1" by 1.5" or 1.5" by 2", but exact sample dimension may vary significantly from this depending on the coring method used to obtain the samples. While a cylindrical plug is preferred, irregular shaped samples can also be used. In addition, crushed samples can also be used. These may be crushed during sampling, such as cuttings obtained from a well, or afterwards, by crushing a core or core plug.

[0063] The samples can be run under any of the following conditions:

• ambient temperature and pressure;

• elevated temperature and pressure;

• elevated temperature and ambient pressure;

• elevated pressure and ambient temperature;

• cooled temperature and ambient pressure;

• cooled temperature and elevated pressure; • any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states without permanently altering any chemical or structural features of the core sample; and

• any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states and permanently alter some chemical or structural features of the core sample

[0064] To run NMR samples at elevated pressure and/or modified temperature, a specialized

NMR cell must be used. This cell must be made of a material that will not interfere with the applied BO magnetic field. A confining sleeve that also does not contribute to the NMR signal will be used in the pressure cell. A confining fluid that does not contribute to the NMR signal can also be used to exert the confining stress.

[0065] The sample is then placed in the proximity of the magnet to obtain the B0 magnetic field. This may be inside a magnet, next to a magnet for the case of the one sided magnetic geometry, in the earth's magnetic field, or the magnet can be placed in a sample in the case of an inside out probe.

[0066] In one tested embodiment, a 2 Mhz Magritek rock core analyzer was employed that applied a magnetic field of 0.05T, which is the commonly used frequency in core analysis. Any number of magnet vendors, such as Bruker, Oxford, Agilent, or even "home built" NMR systems could potentially be used for the measurement. The exact magnetic field strength is also not vital. The field could range from as weak as the earth's magnetic field up to the highest strengths of available superconducting magnet (~1 Ghz). The magnetization transfer contrast is not magnetic field dependent, but in certain embodiments the range 0.5 Mhz to 20 Mhz is utilized. The signal to noise ratio gets weak under 0.5 Mhz such that measurements will take extended periods of time to run and the potential of artifacts to the measurement due to internal gradients are greater above 20 Mhz.

[0067] The sample is placed in the proximity of the magnetic field long enough to allow the magnetic to come to equilibrium aligned along the applied field. This time is minimally 3 times the maximum Tl of the sample. In certain embodiments a time of approximately 10 times the maximum Tl of the sample is employed. For many samples, 10-20 seconds is ample time.

[0068] The system is tuned and matched to ensure the best excitation, as the excitation frequency and reflected power of the probe may shift when loaded with many type of samples and sample holders as compared to the frequency and reflected power produced when the probe is empty. This is performed by running and adjusting tuning capacitors until the produced signal by the probe is optimized. Tuning is complete when the frequency produced by the probe is within an acceptable frequency distance from the main resonance frequency. Matching is complete when the reflected power is to an acceptable minimum level.

[0069] Pulses at 90° and 180° are optimized for the particular system of magnet, probe, and sample. The combination of pulse length and pulse power that produces the maximum signal is established as the 90° pulse. The combination of pulse length and power that produces a minimum signal is established as the 180° pulse. Typically when making a measurement, either the power is kept fixed and the length of the pulse is varied to get the different pulses, or the length of the pulse is kept constant and the power is varied. In certain embodiments, a 180° pulse that is twice the power of a 90° pulse is employed, as this keeps the bandwidth used to excite the sample constant and is less likely to introduce issues given the broad linewidth which often occurs in geological samples. If other pulse values are needed, 45°, 30°, etc. the necessary pulse length or power is calculated from the empirically obtained 90° and 180° pulses.

[0070] In certain embodiments, an initial magnetization transfer contrast preparation is combined with a CPMG sequence. A loss of signal at short T2 times in the magnetization transfer contrast experiment versus the regular CPMG sequence is seen for samples with a significant amount of organics. This loss of signal is not seen for samples that do not contain significant amounts of organics. The loss is not a Tl -effect, as a measurement of a Tl -edited CPMG with the same wait times as the binomial CPMG does not show the same loss. Figure 10A illustrates the results of a CPMG pulse vs Binomial vs Tl edited in organic oil-bearing shale. For both the regular CPMG and the CPMG with magnetization transfer contrast using binomial pulses, the pulse lengths were 12 microseconds (μβ) with a power of -13.5 dB for the 90 degree pulses and -7.5 dB for the 180 degree pulses. The TE was 60 and 8000 echoes were used. Two sets of binomial pulses, 4 pulses in total, were used to perform the magnetization transfer contrast measurement. No wait time beyond the dead time of the probe, 15 μβ, was used. The results were inverted using an inverse Laplace transform based on the Butler Reed Dawson algorithm. See Butler JP, Reeds JA, Dawson SV. Estimating Solutions of First Kind Integral Equations with Non-negative Constraints and Optimal Smoothing. SIAM Journal on Numerical Analysis, 18(3)(1981) 381- 397. The type of algorithm used for the inverse Laplace transform is not crucial and other types of algorithms for inversion are acceptable.

[0071] In performing the method as depicted in Figure 14, a standard baseline NMR pulse sequence is run such as for example the depicted CPMG sequence. In Figure 14, it is depicted that the standard baseline pulse sequence is run first although the baseline sequence can be run following a test sequence including magnetization transfer as long as the sample has returned to its native magnetization state prior to the baseline test. In this first standard CPMG NMR pulse sequence, the decay of the initial magnetization of the sample is due to T2 relaxation. A static magnetic field BO is applied to the sample. This magnetic field can range from the strength of the Earth's magnetic field up to the 25T produced by the strongest superconducting magnets. Ideally, the magnetic field is homogeneous across the volume of the sample, though the technique will still function in an inhomogeneous field. There is a wait period to allow the magnetic moments of the spins to align along the applied static magnetic field BO. This is usually allowed to be 3 to 10 times the Tl relaxation time of the sample. For the majority of samples 15-30 seconds should be adequate to allow the system to come to equilibrium with the applied magnetic field, though times as short as 1-3 seconds may be adequate in some samples. The relaxation time is determined empirically for the given sample. Then, a 90° radio frequency (RF) pulse is applied to the system. The combination of RF pulse duration and power sufficient to produce a rotation of the magnetization vector 90° is previously determined. The RF pulse is tuned to the main resonance frequency of the magnet. This resonance frequency is a combination of magnetic field strength and the nuclei under investigation. There is a wait time tau and then a 180° RF pulse is applied to the system. The 180° pulse can either be twice the duration or power of the 90° pulse, though it is preferred to keep the duration the same and double the power in inhomogeneous systems. At a time tau after the 180° pulse is applied, there will occur a responsive NMR signal referred to a spin echo. The intensity of this spin echo is recorded. At a time two times (x) tau after the initial 180° pulse, a second 180° pulse is applied. Another spin echo will form at a time tau after the second 180° pulse and the intensity of this spin echo is also recorded. The 180° pulses are repeatedly applied and the intensity of the spin echoes recorded until there are no resulting spin echoes, only noise. After this, the system is allowed to come to equilibrium again with the static applied magnetic field B0. If the signal to noise (S/N) ratio of the measurement is not adequate, the sequence of 90° RF pulses followed by the series of 180° RF pulses are repeated and added to the initial measurement. This is repeated until an acceptable S/N ratio is reached. After the above described standard CPMG NMR pulse sequence has been run, the system is allowed to come to equilibrium with the applied static magnetic field BO again.

[0072] Next, to allow observation of magnetization transfer, a series of RF pulses is applied to destroy the magnetization coherence belonging to the organic solids or semi-solids in the sample, followed by a variable wait time to allow for magnetization transfer, and then measurement of the CPMG is performed. Under the identical magnetization conditions of the baseline CPMG pulse sequence, a 90° RF pulse is applied to place the magnetization into the transverse plane. There is a variable wait time to allow the solids magnetization to dephase as shown in Figure 6. Then a 90° RF pulse is used to return the magnetization back along the z- axis. In certain embodiments 2 sets of binomial 90° RF pulses are applied as depicted in Figure 7. Depending on the properties of the sample at hand, this series of pulses may be repeated until the decrease in measured signal from solids and semi-solids is at an acceptable level. Depending on the parameters of the number of binomial pulses and wait times between the pulses, some to all of the magnetization coherence from solids and semi-solids may be destroyed. The amount of signal loss related to different binomial pulse parameters may give information regarding the type and properties of the solids and semi-solids. In other embodiments a shaped, off-resonance pulse is used to pre-saturate the magnetization from solids in the sample as shown in Figure 8. Pre-saturation is used to excite the magnetization from solids and semi-solids before the main, baseline NMR measurement such that their signal will not contribute to the subsequent NMR measurement.

[0073] There is a variable wait time to allow magnetization to transfer from fluids to the solids and semi-solids which had destroyed magnetization coherence. This magnetization transfer will occur due to the dipolar coupling or chemical exchange between fluids, semisolids and solids. Depending on how rapidly the magnetization transfer occurs, a short, long, or multiple wait times may be needed to adequately characterize the magnetization transfer. This wait time should not be so long that pre-saturation or destruction of the magnetization coherence of the solids and semi-solids has been lost due to Tl relaxation. Then, a 90° RF pulse is applied to the system. There is a wait time tau and then a 180° RF pulse is applied to the system. At a time tau after the 180° RF pulse is applied, there will occur an NMR signal referred to a spin echo. The intensity of this spin echo is recorded. At a time two x tau after the initial 180° RF pulse, a second 180° RF pulse is applied. Another spin echo will form at a time tau after the second 180° RF pulse and the intensity of this spin echo is also recorded. The 180° RF pulses are repeatedly applied and the intensity of the spin echoes recorded until there are no resulting spin echoes, only noise. After this the system is allowed to come to equilibrium again with the applied magnetic field. If the signal to noise ratio of the measurement is not adequate, the sequence of 90° RF pulses followed by a series of 180° RF pulses are repeated and added to the initial measurement. This is repeated until an acceptable S/N ratio is reached. This measurement may be repeated with additional wait times for magnetization transfer to better understand the dynamics in the system. The results from the two or more measurements are then compared to check for loss of signal or shift of signal between relaxation times.

[0074] The binomial CPMG sequence may show significant loss of signal at short relaxation times compared to the normal CPMG. Some of this loss may be a Tl effect, but there may still be appreciable loss in the signal in the binomial CPMG compared to the Tl -edited CPMG. Figure 10B illustrates the results of a CPMG pulse vs Binomial CPMG in a tight carbonate with minimal organics expected. While there are minor differences in the spectra, the overall signal intensity and resulting distribution are similar. Figure 11 illustrates the results of a CPMG pulse vs Binomial vs Tl edited CPMG in oil shale, which contains significant quantities of organics but little free fluid. There is less difference between the Binomial CPMG and the Tl -edited CPMG, indicating the signal loss may simply be the destruction of signal from solids without significant magnetization transfer.

[0075] By comparison of the standard baseline CPMG measurements and the measurement with magnetization transfer performed, organic solids, semi-solids, and different fluids present in the system can be detected and characterized. If there is no appreciable change in the signals between the two measurements, this means that either there were no appreciable organic solids in the system or there were organic solids in the system but they were too short to measure with the equipment available and no magnetization transfer between the solids and liquids occurred. If there is loss of signal, several situations could be occurring. Firstly, the organic solids could be too short to measure with the equipment at hand and the loss of signal arises from transfer of magnetization from the fluids to the organic solids and semi-solids which have relaxation times too short to measure. Under such results the process can then be repeated under different parameters, such as adjusting the variable wait time to clarify the process of magnetization transfer. Secondly, if the signal loss occurs at short T2 relaxation times, this loss of signal could arise from the destruction of magnetization coherence from the organic solids and semi-solids that is possible to measure with the available equipment. If there is no loss of signal at long relaxation times, this means that despite the destruction of the organic solids and semi-solids magnetization coherence, there is no magnetization transfer between the organic solids and fluids and this can help characterize the fluids and organic solids. This could arise if the organic solids are bitumen or kerogen and the fluids are water, which are expected to have minimal interaction. Alternatively, the organic solids could be very thermally mature, such that they have lost functional groups that are more likely to allow magnetization transfer between organic solids and liquids to occur. If there is loss of signal at both short and long relaxation times, this could be arising from the destruction of the organic solids signal, which are long enough to be measured with the available equipment, and subsequent magnetization transfer from liquids. Variation of the wait time to allow for magnetization transfer will help clarify these different scenarios, to see if there is further shift in the relaxation distributions with further time. In addition, the ratio of the T1-T2 can be analyzed to clarify which components are present. Organic solids sometimes, though not always, will have signals with short T2 relaxation times and long Tl relaxation times that can help identify them.

[0076] In measurement of organics in the system, the magnetization transfer may not only manifest as a loss of short T2 components, but a shift from longer T2 times to the shorter T2 components. In some cases, it may be even possible that the transfer of magnetization from liquids to the organics will lead to a shift from shorter T2 components to longer T2 components. This may be utilized to distinguish solids from liquids when their T2 components are overlapping.

[0077] Magnetization transfer contrast as applied herein is employed by destroying the magnetization coherence from the solid hydrogen bearing components so that they no longer contribute to the measured NMR signal. Normally, even if the signal from solids relax too quickly for measurement due to instrument limitations, they are still excited by the pulses and can interact with other parts of the system. Pre-saturation or magnetization coherence destruction works by exciting one or more constituents of the system before the main measurement so that its magnetization will not contribute to the main measurement. The liquid components, which relax slowly enough to be measured, are relatively unaffected by the pre-saturation of an off-resonance pulse or magnetization coherence destruction from binomial pulses. Then, when there is exchange between the solid components and the liquid components, the liquid components will lose magnetization to the solid components, but having had their magnetization coherence destroyed, the solid components will not have any magnetization to transfer back. This is seen as a decrease of intensity in the measured signal of the main measurement.

[0078] In certain embodiments, the wait times are varied between the pre-saturation and the main measurement to determine how the magnetization distribution and intensity change with wait time; with longer wait time, there is more opportunity for interaction between more weakly coupled constituents or constituents that are further apart. Care is taken to avoid Tl effects and loss of the magnetization coherence destruction.

[0079] However, the distinction between solid and liquid is an incomplete description of many geological systems. In core samples, constituents may exist that are neither full solid nor fully liquid, for example bitumens, asphaltenes, waxes, resins and heavy oils. Magnetization transfer can also exist between the solids and semi-solids, as well as the semisolids and liquids in addition to between solids and liquids. These semi-solids will have different transfer properties and running magnetization transfer contrast measurements with a variety of different parameters may be employed to resolve multiple constituents in the system.

[0080] There two main methods used to perform the destruction of the magnetization coherence of the solid/semi-solid signal. The first is through binomial pulses. The most basic form of the binomial pulses use a series of 90° pulses before the main measurement is performed. This is used to flip the magnetization back and forth between the XY plane and the z-axis, where they are aligned either along, or against the applied magnetic field. In the basic measurement, the phase of the pulses are alternated so that the magnetization simply returns to the z-axis. In certain tests conducted, a basic +x, -x phase alternation as depicted in Figure 9 was employed, though more complicated phase cycles may be alternatively utilized. In other more complicated binomial pulse sequences, other pulse degrees may be used. In these sequences, the flip angle of the measurement may be other values than 90°.

[0081] In one set of tests, 2 sets of binomial pulses (4 pulses in all) were used, though the exact best number will likely vary with the sample depending on the T2* present in the sample. The binomial pulses can be applied immediately one after another or there can be a wait time between sets to allow extra dephasing of the NMR signal with short T2* components. Solid signals, which have very quick dephasing time due to short T2*'s, will be affected by these pulses and but the liquid like components with longer T2*'s will not be affected. A simplified illustration of the process is presented in Figures 4A and 4B.

[0082] In each of the illustrations in Figures 3 A, B and C, magnetization vector begins aligned along the applied magnetic field, BO. The reference frame is chosen such that the Z- axis is aligned along the applied magnetic field. Figure 3A the RF pulse produces a second magnetic field that rotates the magnetization vector. The RF pulse is chosen such that the second magnetic field it produces is aligned with the X-axis. The duration of the pulse is long enough that the magnetization vector rotates 90 degrees to align with the Y-axis. The phase of this RF pulse is said to be a 90x pulse. In Figure 3B the RF pulse produces a second magnetic field that rotates the magnetization vector. The RF pulse is chosen such that the second magnetic field it produces is aligned with the Y-axis. The duration of the pulse is long enough that the magnetization vector rotates 90 degrees to align with the X-axis. The phase of this RF pulse is said to be a 90y pulse.

[0083] Figure 3C the RF pulse produces a second magnetic field that rotates the magnetization vector. The RF pulse is chosen such that the second magnetic field it produces is aligned with the x-axis. The duration of the pulse is long enough that the magnetization vector rotates 180 degrees to align with the Y-axis. The phase of this RF pulse is said to be a 180x pulse. Figure 3D the RF pulse produces a second magnetic field that rotates the magnetization vector. The RF pulse is chosen such that the second magnetic field it produces is aligned with the Y-axis. The duration of the pulse is long enough that the magnetization vector rotates 180 degrees to align with the X-axis. The phase of this RF pulse is said to be a 180y pulse.

[0084] The Figure 4A set of 4 indicates the behavior of the liquid components during pre- saturation or magnetization coherence destruction portion of magnetization transfer. The magnetization vector is rotated into the XY plane. Because the T2* of liquid components is long, there isn't significant dephasing while the magnetization vector is allowed to stay in the XY plane. The magnetization vector is returned to along the z-axis, where its intensity is relatively unchanged from its equilibrium value. The magnetization can then be combined with another type of NMR measurement, e.g. a CPMG or T1-T2 correlation measurement. The Figure 4B set of 4 indicates the dephasing loss of the solid like components during pre- saturation or magnetization coherence destruction of magnetization transfer. The magnetization vector is rotated into the XY plane. Because the T2* of solids is short, there is significant dephasing of the magnetization vector while it is in the XY plane. When the magnetization is returned to the Z-axis, the dephasing has decreased or destroyed the magnetization coherence arising from the solids, such that they will have a reduced or no contribution to an NMR measurement made after pre-saturation, e.g. a CPMG or T1-T2 correlation measurement.

[0085] Figure 5 depicts the relative timing and effects of an off resonance pulse. In samples containing hydrogen in solids or solid like components, the linewidth of the NMR signal will be quite broad. The liquids in the sample will experience a magnetic field similar to that of the applied BO field, but the solid and solid-like components will experience a different magnetic field due to their interactions with one another. This difference may be on the order of tens of kilohertz in typical core samples. By applying an off resonance pulse preceding the desired measurement, the hydrogen that is present in solid or solid like components is excited. The off resonance pulse is assumed to be far enough away from the central frequency that liquids won't be presaturated by it and only the solids will be excited by this and then, due to the efficiency of coupling in solids, the exciting pulse will transfer the saturation to all the solid components, not just the ones at the offset frequency. The sample constituents that experience this pre-saturation will not be excited in the following NMR measurement, allowing magnetization transfer contrast to be observed. A relatively narrow frequency for the off resonance pulse is chosen as not to inadvertently pre-saturate the liquid signal. This can be performed using a longer, standard rectangular pulse, or a shaped pulse may be used to select the frequency. There exist a range of different shaped pulses, sine, Gaussian, etc. that may be employed to create the narrow range. The shape of pulse used to create the off- resonance pulse is not of particular importance.

[0086] In several non-limiting examples, Figure 6 illustrates a Free Induction Delay ("FID") pulse sequence with a leading binomial pulses and variable wait time for dephasing. The magnetization from the system is placed in the XY plane where the magnetization from solids and semi solids are allowed to dephase during the variable wait time. The magnetization is then returned to along the Z-axis. A period of time is waited to allow magnetization transfer between liquids, solids and semi-solids. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is magnetization transfer, the resulting signal will indicate a loss of signal intensity compared to a normal FID measurement. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is no magnetization transfer, the resulting signal will be relatively unchanged compared to a normal FID measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is magnetization transfer, there will be decreased signal for both the solid and liquid components compared to a normal FID measurement. If the signal from the solids and semisolids relaxes slowly enough to be measured with the available equipment and there is no magnetization transfer, the liquids signal will be relatively unchanged compared to a normal FID while the solids signal will be reduced or compared to a normal FID measurement depending on the effectiveness of the pre-saturation at destroying the magnetization coherence from the solids and semi-solids.

[0087] Figure 7 illustrates a CPMG pulse sequence with 2 sets of binomial pulses. The magnetization from the system is placed in the XY plane where the magnetization from solids and semi solids are allowed to dephase. The magnetization is then returned to along the z- axis. A second set of binomial pulse places the magnetization back into the XY plane where the magnetization from the solids and semi-solids continue to dephase. The magnetization is then returned to along the z-axis. A period of time is waited to allow magnetization transfer between liquids, solids and semi-solids. A CPMG pulse sequence is then performed. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is magnetization transfer, the resulting signal will indicate a loss of signal intensity compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is no magnetization transfer, the resulting signal will be relatively unchanged compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is magnetization transfer, there will be decreased signal for both the solid and liquid components compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is no magnetization transfer, the liquids signal will be relatively unchanged compared to a normal CPMG while the solids and semisolids signal will be reduced or compared to a normal CPMG measurement depending on the effectiveness of the pre-saturation at destroying the magnetization from the solids and semisolids.

[0088] Figure 8 illustrates CPMG pulse sequence with leading off resonance shaped pulse. A shaped, off-resonance pulse is used to pre-saturate the magnetization from solids in the sample. A period of time is waited to allow magnetization transfer between liquids, solids and semi-solids through the mechanisms of dipolar coupling and chemical exchange. A CPMG pulse sequence is then performed. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is magnetization transfer, the resulting signal will indicate a loss of signal intensity compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is no magnetization transfer, the resulting signal will be relatively unchanged compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is magnetization transfer, there will be decreased signal for both the solid and semi-solids and liquid components compared to a normal CPMG measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is no magnetization transfer, the liquids signal will be relatively unchanged compared to a normal CPMG while the solids signal will be reduced or compared to a normal CPMG measurement depending on the effectiveness of the pre-saturation at destroying the magnetization from the solids.

[0089] Figure 9 illustrates a pulse sequence Tl - T2 correlation with a binomial leading pulse. The magnetization from the system is placed in the XY plane where the magnetization from solids and semi solids are allowed to dephase during the variable wait time. The magnetization is then returned to along the Z-axis. A period of time is waited to allow magnetization transfer between liquids, solids and semi-solids. A T1-T2 correlation pulse sequence is then performed. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is magnetization transfer, the resulting signal will indicate a loss of signal intensity compared to a normal T1-T2 correlation measurement. If the signal from the solids and semi-solids relaxes too quickly to be measured with the available equipment and if there is no magnetization transfer, the resulting signal will be relatively unchanged compared to a normal T1-T2 correlation measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is magnetization transfer, there will be decreased signal for both the solid and liquid components compared to a normal T1-T2 correlation measurement. If the signal from the solids and semi-solids relaxes slowly enough to be measured with the available equipment and there is no magnetization transfer, the liquids signal will be relatively unchanged compared to a normal T1-T2 correlation while the solids and semi-solids signal will be reduced or compared to a normal T1-T2 correlation measurement depending on the effectiveness of the pre-saturation at destroying the magnetization from the solids.

[0090] The NMR measurements may be strung together, such that a variety of different effects can be measured. A good example is the T1-T2 correlation measurement. A Tl measurement is followed by a T2 measurement, and measurement parameters varied such that the Tl and T2 values present in the system can be correlated to one another.

[0091] Figure 12 shows inverted data of a T1-T2 correlation measurement, while Figure 13 shows inverted data of T1-T2 correlation measurement with magnetization transfer contrast using binomial pulses. The basic T1-T2 correlation measurement was performed using a combined inversion recovery-CPMG sequence. The pulse lengths were 12 microseconds ("μδ") with a power of -13.5 dB for the 90 degree pulses and -7.5 dB for the 180 degree pulses. The variable wait time of the inversion recovery sequence ranged from 0.03 milliseconds ("ms") to 2 seconds ("s"). The CPMG portion of the sequence used a TE of βθμβ and had 2000 echoes. The binomial T1-T2 correlation with magnetization transfer used the same for the measurement of T1-T2, preceded by two 90 degree pulses and a wait time of 20 between the first of the binomial pulse pair and the second of the pair. Both sets of 2D data were inverted using a maximum entropy method. By comparison of the regular and the measurement with magnetization transfer performed, the organic solids and semi-solids and fluids present in the system can be discerned and initially characterized as previously discussed.

Example II:

[0092] It is noted that this procedure is not limited to combining magnetization transfer with a T2 measurement. In other embodiments, magnetization transfer is combined with one or more of a Free Induction Decay measurement, a Tl measurement, a solid echo measurement, a solid echo train, a diffusion measurement, an internal gradient measurement, and a quantum filter measurement. [0093] In certain embodiments the NMR measurements are strung together, such that several effects are measured concurrently or in series. For example, a T1-T2 correlation measurement may be performed in which a Tl measurement is followed by a T2 measurement, and measurement parameters varied such that the Tl and T2 values present in the system are correlated to one another.

[0094] The different ways of measuring Tl include Inversion Recovery and Saturation Recovery. The different ways of measuring T2 include Spin echo, Solid Echo, a Spin Echo train (such as a CPMG sequence) and a Solid Echo train.

Example III:

[0095] Other properties of organics in a sample may be determined by magnetization transfer behavior. These properties may include organic maturity, type of kerogen, ratio of kerogen to bitumen, types of saturating fluids, wettability effects, surface relaxivity, and pore sizes. In certain embodiments, this technique is utilized downhole.

Example IV:

[0096] Using existing technology, gas hydrates or hydrocarbon ice, suffer from a similar problem as the organic shales. That is the magnetization decays so quickly it cannot be measured by standard lab NMR equipment. In one embodiment, magnetization from hydrates potentially present is pre-saturated and the presence of the gas hydrates is detected through loss of liquid signal through dipolar coupling or chemical exchange.

Example V:

[0097] Bitumen and heavy oil present can be both source rocks and reservoir rocks. In many cases, it is a challenge to determine what part of the signal arises from producible hydrocarbons and hydrocarbons that may be immovable, like bitumen and heavy oil. Magnetization transfer could be used to help determine which part of the NMR signal arises from liquids and which arises from the high viscosity components.

Example VI:

[0098] Wettability alteration of reservoir rock occurs due to the deposition of organic matter on the pore surface. The presence of this organic matter and its effect upon the wettability properties of the material may be determined via NMR. Destruction of the magnetization coherence from this organic material and then the subsequent loss of liquid signal may be detectable as a function of wettability of the rock.

[0099] All publications, patents and patent applications cited herein are hereby incorporated by reference as if set forth in their entirety herein. While this invention has been described with reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. Various modifications and combinations of illustrative embodiments, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to the description. It is therefore intended that the appended claims encompass such modifications and enhancements.

Claims

Claims:

1. A method of analyzing a geologic sample by NMR for the characterization of organic solid and semi-solid materials and saturating fluids comprising:

performing a baseline NMR pulse sequence test on the geologic sample and recording signals from the baseline NMR pulse sequence test;

performing a magnetization transfer contrast NMR pulse sequence test by first applying a preparation RF pulse sequence to the geologic sample to decrease magnetization attributable to any organic solids or semi-solid materials in the sample, waiting a variable time for any transfer of magnetization between sample fluids that have retained their magnetization and the decreased magnetization of the organic solids and semi-solids, and then performing the baseline NMR pulse sequence on the geologic sample and recording signals from the magnetization transfer NMR pulse sequence; and

comparing the results from baseline NMR test with the magnetization transfer NMR pulse sequence to observe transfer or non-transfer of magnetization between any organic solids and semi-solids and fluids present in the geological sample and thereby determine characteristics of organic materials and saturating fluids in the geologic sample.

2. The method of claim 1 wherein the organic solids or semi-solid materials are selected from the group consisting of kerogen, bitumen, heavy oil and waxes.

3. The method of claim 1 wherein the preparation RF pulse sequence is a binomial pulse sequence.

4. The method of claim 1 wherein the preparation RF pulse sequence is a series of binomial pulse sequences.

5. The method of claim 1 wherein the preparation RF pulse sequence is an off-resonance pulse.

6. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence test includes application of a Carr-Purcell-Meiboom-Gill (CPMG) pulse echo train to the geologic sample.

7. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of a spin echo pulse sequence to the geologic sample.

8. The method of any one of claims 1 - 5 wherein baseline NMR pulse sequence includes application of a solid echo sequence to the geologic sample.

9. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of a solid echo train sequence to the geologic sample.

10. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of a free induction decay pulse sequence to the geologic sample.

11. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of an inversion recovery sequence to the geologic sample.

12. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of a saturation recovery sequence to the geologic sample.

13. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of a quantum filter measurement sequence to the geologic sample.

14. The method of any one of claims 1 - 5 wherein the baseline NMR pulse sequence includes application of an internal gradient measurement sequence to the geologic sample.

15. The method of claim 1 wherein the waiting for a variable time to allow for magnetization transfer varies from zero to minutes.

16. The method of claim 1, wherein the method performed is in a laboratory or wellsite setting.

17. The method of claim 1, wherein the method is performed in a downhole environment

18. The method of claim 1, wherein the geologic sample is in a state selected from the states consisting of: native; native resaturated with brine; native resaturated with oil; native resaturated with another fluid; cleaned and dried; cleaned and resaturated with brine; cleaned and resaturated with oil; cleaned and resaturated with another fluid; cleaned, resaturated with brine; desaturated with oil and aged; and any of the above states pressurized with gas, either wet or dry.

19. The method of claim 1, wherein the analysis is conducted under a condition selected from: ambient temperature and pressure; elevated temperature and ambient pressure; elevated pressure and ambient temperature; cooled temperature and ambient pressure; cooled temperature and elevated pressure; any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states without permanently altering any chemical or structural features of the core sample; and any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states and permanently alter some chemical or structural features of the core sample.

20. A method of analyzing a geologic sample by NMR for the presence of organic solid and semi-solid materials and saturating fluids comprising performing a series of analyses comprising:

performing a baseline NMR pulse sequence test on the geologic sample;

performing a magnetization transfer NMR pulse sequence test on the geological sample wherein the magnetization transfer contrast NMR test involves a first application of a magnetization transfer preparation RF pulse sequence sufficient to decrease or destroy magnetization attributable to any organic solid and semi-solid materials that might be in the geologic sample followed by application of the baseline NMR pulse sequence after a variable time sufficient to allow transfer of magnetization between any sample fluids that have retained their magnetization and the decreased magnetization of the organic solids and semi-solids; and

comparing the results of the baseline NMR pulse sequence test and the magnetization transfer NMR pulse sequence test to detect the presence of organic solid and semi-solid materials in the geologic sample.

21. The method of claim 20 wherein the magnetization transfer preparation RF pulse sequence is selected from the group consisting of one or more of: a binomial pulse sequence; a series of binomial pulse sequences; and an off-resonance pulse.

22. The method of claim 21 wherein the analysis is repeated using at least one different magnetization transfer preparation RF pulse sequence.

23. The method of claim 20 wherein the sequential analysis is repeated with a combination of at least two different baseline NMR pulse sequence tests, the tests characterized by application of pulse sequences selected from the group consisting of: a Carr- Purcell-Meiboom-Gill (CPMG) pulse echo train; a spin echo pulse; a solid echo; a solid echo train; a free induction decay pulse sequence; an inversion recovery sequence; a saturation recovery sequence; a diffusion measurement; a quantum filter measurement sequence and an internal gradient measurement sequence.

24. The method of claim 20, wherein the sequential analysis is conducted under a condition selected from: ambient temperature and pressure; elevated temperature and ambient pressure; elevated pressure and ambient temperature; cooled temperature and ambient pressure; cooled temperature and elevated pressure; any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states without permanently altering any chemical or structural features of the core sample; and any of the above pressures and temperatures are applied to levels sufficient to affect the population of spin states and permanently alter some chemical or structural features of the core sample.