WO2011133859A1

WO2011133859A1 - Nmr quantification of the gas resource in shale gas reservoirs

Info

Publication number: WO2011133859A1
Application number: PCT/US2011/033575
Authority: WO
Inventors: Richard F. Sigal
Original assignee: The Boards of Regents of the University of Oklahoma
Priority date: 2010-04-23
Filing date: 2011-04-22
Publication date: 2011-10-27

Abstract

The present invention provides analytical techniques for determining the quantity of natural gas within tight, very low porosity formations. The method of the present invention uses NMR logging data and compares the logged data to laboratory developed data based on core samples from the formation or core samples from a similar formation. Additionally, the present invention provides the ability to determine the amount of gas contained within the organic and inorganic pores within the formation.

Description

NMR Quantification of the Gas Resource in Shale Gas Reservoirs

Back round of the Invention

[001] Shale reservoirs, such as the Barnett shale of North-central Texas, contain a significant volume of natural gas. Due to the nature of the formation, traditional methods for measuring gas quantities do not provide satisfactory results. Shale formations typically have very low porosities, contain organic material, and have complex mineral matrices including conductive clay minerals. Due to the very low permeability, tests to determine in situ gas pressure are not practical. Additionally, acquisition of water samples suitable for determining salinity of pore fluid is very difficult. As a result, standard logging methods to determine porosity, resistivity and water saturation are not suitable for calculating gas quantities.

[002] Current methods for evaluating gas within shale formations use locally valid procedures built on correlations to core measurements to obtain water saturation and porosity values and subsequently to calculate gas volume. These correlations do not provide the gas mass, i.e. the moles of gas, within the formation as this depends on pressure. Prior to the current invention, quantification of gas within shale formations depended on assuming a gas pressure. The pores within the shale formation are in the nanometer size range. The pores within a shale formation may have inorganic mineral walls or organic (kerogen) walls. Within the pores having inorganic mineral walls, capillary pressure effects could make the gas pressure significantly higher than the hydrostatic pressure. Due to the hydrophobic nature of the organic pores, these pores do not contain water and will have a propensity to preclude invasion of water from water based drilling muds. The organic material may also not be stress supporting. Therefore, one would not expect the gas pressure within the organic, hydrophobic pores to have a gas pressure equal to the hydrostatic pressure. Rather, the gas pressure could be equal to, greater, or less than the hydrostatic pressure. Therefore, without independent gas pressure measurements one cannot obtain the total gas in the reservoir or target formation just by knowing its volume.

[003] NMR logging of wells has been practiced for a number of years. The NMR logging tool provides a direct measure of the number of hydrogen atoms within the zone of investigation of the NMR signal. The general nature and operation of the NMR logging tool is well known to those skilled in the art and will not be addressed herein. Rather, the focus of the present invention relates to the novel use of NMR logging to determine not only the presence of natural gas but also the quantity of natural gas within a reservoir or target formation. Further more, the present invention provides a method for quantifying the amount of gas contained in the organic pores and the inorganic pores. Since these pores have very different properties, one must distinguish the amounts of gas in each pore type in order to provide a proper simulation of future production.

[004] The difficulties posed by shale formations to traditional logging methods also play a beneficial role in NMR logging. The tight, low porosity nature of shale reduces the depth of invasion by drilling mud. Additionally, within shale formations, much of the gas is located in very small organic pores where the gas wets the walls. The hydrophobic nature of these pores also opposes invasion of water from water based drilling mud The resulting shallow invasion zone coupled with improvements in NMR logging tools, i.e. improvements in depth of investigation and the ability to vary investigation depth, allows the NMR tool to "see" beyond the invasion zone. By adapting these improvements, the present invention utilizes the NMR logging tool to determine the presence of gas and to provide an independent measure of the number of moles of methane within the reservoir or target formation.

Summary of the Invention

[005] The present invention provides a method for detecting and quantifying hydrocarbon within a shale formation. In the method of the current invention a borehole is drilled into at least a portion of a shale formation. The formation is then logged by passing a logging tool through the borehole. The logging tool includes at least an NMR logging tool. Additionally, the method obtains a core sample from said formation. NMR analysis of core sample determines the water relaxation curve for said core sample. Additionally an NMR relaxation curve for at least a portion of the formation is provided by the NMR logging tool. The water relaxation curve obtained from NMR analysis of said core sample is subtracted from the relaxation curve obtained from logging at least a portion of said formation to determine the quantity of hydrocarbon within said formation.

[006] Additionally, the present invention provides a method for determining the NMR Ti/T₂ ratio corresponding to hydrocarbons within a shale formation. According to the method of the present invention a borehole is drilled into a portion of a shale formation. Logging of the borehole includes the use of at least an NMR logging tool. The resulting log allows the operator to determine NMR relaxation curves for Ti, T₂. Subsequently, the method plots the formation Ti/T₂ ratio. Additionally, the method of the current invention obtains and conducts NMR analysis of a core sample from the same or similar shale formation to determine the NMR water Ti/T₂ ratio for the core sample. By comparing the formation ΊΥΤ₂ ratio to the core sample Ti/T₂ ratio one determines the contribution of hydrocarbon in the formation to the formation NMR T1/T2 ratio.

[007] Still further the present invention provides a method for quantifying the volume of hydrocarbon in said formation using a hydrocarbon relaxation curve. According to this method, a borehole is drilled into a shale formation and the formation subsequently logged by passing a logging tool through the borehole. The logging tool includes an NMR logging to suitable for determining NMR relaxation curves for at least one of Ti or T₂. Additionally, the logging step determines total water volume within the formation region corresponding to the NMR relaxation curves. Subsequently, the logged Ti or T₂ relaxation curves are processed to obtain T] or T₂ spectra corresponding to combined formation hydrocarbon and water spectra. Additionally, the method obtains a core sample from the formation or a similar formation and conducts an NMR analysis of said core sample. The NMR values from the core sample are used to estimate the relaxation spectra for water and hydrocarbon within the formation. The estimated relaxation spectra is compared to relaxation curves obtained from the logged data. Using the logged value for water in the formation the method constrains the water relaxation spectra to a total amplitude corresponding to the separately logged volume of water thereby constructing an estimated relaxation curve. By calculating the difference between the constructed estimated relaxation curve and the relaxation curve measured by the logging tool one will obtain final formation hydrocarbon and water relaxation curves by iteratively changing the estimated curves until the formation relaxation curves substantially fit the core sample relaxation curves. The resulting final hydrocarbon relaxation curve provides a determination of the volume of hydrocarbon in said formation.

[008] Yet further, the present invention provides a method for detecting and quantifying natural gas within a shale formation comprising. According to this method, a borehole is drilled into a shale formation. A core sample is obtained from the formation or a substantially similar formation. The volume of water per cubic centimeter of the core sample is determined. Additionally, the formation is logged through the borehole by passing a logging tool including at least an NMR unit through the borehole. The NMR log determines the total number of hydrogen atoms contained per cubic centimeter of formation. The number of hydrogen atoms contributed by methane within the formation is determined by subtracting the number of hydrogen atoms contributed by water within said formation as determined by the core sample. Subsequently, the moles of methane per cubic centimeter within said formation is calculated using the chemical formula of methane.

Brief Description of the Figures

[009] Figure 1 depicts a relaxation curve for a shale core sample containing water.

[0010] Figure 2 depicts a relaxation curve for a shale core sample containing water and methane gas.

[0011] Figure 3 depicts a spectral curve for water based on processing of the relaxation curve from figure 1.

[0012] Figure 4 depicts a normalized spectral curve based on processing of the relaxation curve from Figure 2.

[0013] Figure 5 depicts a spectral curve for gas only within the core sample calculated by subtracting the water relaxation curve of Figure 1 from the relaxation curve of Figure 2 and processing the resulting curve.

[0014] Figure 6 depicts a T synthetic relaxation spectrum curve built from water and brine measurements on a single crushed organic shale sample and a bulk oil NMR measurement.

[0015] Figure 7 depicts the Ί and T₂ NMR relaxation curves for brine in contact with a crushed shale core sample.

[0016] Figure 8 depicts the T_t and T₂ NMR relaxation curves for dodecane in contact with a crushed shale core sample. Detailed Disclosure of the Preferred Embodiments

[0017] Those skilled in the art are familiar with the use of NMR Ti and T₂ relaxation curves to identify different fluids. The NMR signal in reservoir shale will include a gas signal, a water signal, and perhaps a liquid hydrocarbon signal. These signals can be separated and characterized by exploiting differences in their diffusion coefficient spectra (D), and Ti and T₂ relaxation spectra. The newest methods of acquiring NMR logging data can characterize all three.

[0018] Therefore, the new NMR analysis provided by the present invention enables characterization of the amount and types of fluids in the reservoir. Just as in the case of conventional reservoirs, the present invention may be adapted in many implementations to identify and quantify the gas in place,

[0019] In conventional formations, the identification of gas using NMR logging relies on two basic assumptions: (1) the gas within the formation will relax as a bulk fluid; and, (2) the gas will have diffusivity determined by the bulk diffusion rate. These assumptions follow from the fact that the measurement senses the invaded zone which causes the gas to be located in the center of the pores. (When the pore walls are inorganic material, gas is always the non-wetting fluid.) The assumptions also require that the pores are fairly large to satisfy the diffusion assumption. These characteristics then define an NMR signature for gas response in a space defined by the Ti, T₂ and diffusion spectra. This distinct signature permits the separation of the gas and water NMR signals. The standard literature provides numerous ways to implement these procedures in practice.

[0020] In contrast to conventional formations, tight organic shale formations have small organic pores and low water saturation for the inorganic pores containing gas. In the gas containing organic pores, the NMR Ti and T₂ relaxation times are controlled by surface relaxation. Additionally, the diffusion coefficient will be much smaller than the bulk diffusion value. As a result, the signature of gas response in the NMR spectrum will be very different from the signature usually used to identify the presence of gas, i.e. the signature discussed above with regard to conventional porous formations. Laboratory experiments have confirmed these distinguishing features. The present invention utilizes these distinguishing features to determine the presence and quantity of gas within the formation. [0021] An example of a conventional implementation used to identify the presence of gas in a conventional, high permeability reservoir uses two NMR measurements. In one measurement, the NMR signals are taken in such a way that the time to polarize the nuclei of the hydrogen atom is less than one second, hi the second measurement, several seconds are allowed for the polarization. Bulk Ti time determines the time needed to polarize hydrogen atoms. Since the first NMR measurement is allowed less than one second for polarization, this measurement will not detect gas as the bulk T_\ time for the hydrogen atom of a gas molecule is several seconds. The second, longer polarization measurement is sufficient to produce a gas signal. The difference between the signals, i.e. the first signal subtracted from the second signal, represents the signal corresponding to hydrogen atoms on methane. The procedure used in high penneability reservoirs will not work in organic shale formations. NMR measurements obtained from organic shale formation have short T] times. As a result, the gas signal will be contained in both measurements and the difference between the selected signals will not isolate the gas signal.

[0022] The current invention provides a method for characterizing the gas signature in low penneability shale formations. To provide the basic data necessary to use the present invention, organic shale samples from producing shale gas reservoirs have been studied in the laboratory. A number of the samples were preserved so as to retain as closely as possible the water content they had in the reservoir. In the studies NMR T₂ measurements were made using a bench top system. This NMR system was run without an external magnetic gradient, thereby minimizing diffusion relaxation. Further, under these conditions the Ti and T₂ spectra are expected to be similar in shape but the T_\ spectrum will be shifted to later time. If the T₂ spectrum shows strong surface relaxation than the Tj will also show strong surface relaxation.

[0023] The laboratory experiments were carried out in two stages. For stage one, the NMR measurement was performed on a sample that contained no gas. To simulate in situ pressure conditions, the sample was housed in a pressure vessel and placed under 5000 psi confining pressure with the temperature controlled to match that of the NMR magnet temperature. The pressure vessel containing the sample was inserted in the NMR measurement system. The NMR data set obtained from the NMR measurement was processed according to standard methods to yield the water signature for the preserved water containing core sample. Figure 1 depicts the T₂ relaxation curve of a water containing core sample and Figure 3 depicts the spectral analysis obtained from processing the relaxation curve according to standard methods. In Figure 3, line 10 represents the incremental water porosity, line 20 represents the cumulative water porosity. Peak 30 in line 10 represents water in a slightly larger set of pores.

[0024] To obtain the water plus gas NMR signature, the core was methane saturated by injecting a measured amount of methane into the sample at 4000 psi pore pressure. For the relaxation curve depicted in Figure 2, 0.01 moles of methane were injected into the core. A second NMR measurement was then performed. Processing this data set using standard processing methods, for preserved samples, will yield the NMR signature of the combined methane and water system. This processed spectral curve is provided by Figure 4. In Figure 4, line 12 represents the incremental water plus gas porosity, line 22 represents the cumulative water plus gas porosity and peak 32 in line 12 represents the gas porosity within a stress fracture created during recovery of the core sample.

[0025] To separate out the methane signature, the relaxation curve from the gas free measurement (Figure 1) is subtracted from the relaxation curve (Figure 2) of the gas saturated sample. Reprocessing of the difference signal provides a gas signature suitable for determining the quantity of gas within the formation. The reprocessed signal is depicted in Figure 5. The area underneath the first peak in the incremental curve 14 corresponds to the gas quantity within the pores. The subsequent peak 34 in this curve corresponds to gas within cracks possibly resulting from the core retrieval process. The quantity of gas represented by peak 34 results from visible stress release cracks. Certain formations may have pores which will yield signals in the approximate region of the stress crack signal 34. These formations can be characterized as needed and the resulting signal included in the calculation of gas quantity. Normally, the value associated with the stress crack signal 34 will be disregarded.

[0026] In Fig. 5, the calculated gas quantity contained in the NMR signal is

0.0083 moles (the total area under the cumulative curve 24 from beginning to end),

1. e. the difference based on this method is less than 17%. The difference between this and the measured amount of gas injected is close to the expected amount of adsorbed gas, as measured by Langmuir adsorption isotherms on other samples from the same core. This may account for the difference but the difference is also within instrumental and equipment error. Thus, the laboratory work demonstrates the ability of the present . invention to detect and calculate the quantity of gas within shale formations.

[0027] The processed curves shown in Figures 3 and 4 are obtained from Figures 1 and 2 by inverting the relaxation curve into a sum of 64 curves each being a single decaying exponential curve. The points on the incremental curves in Figures 3, 4 and 5 represent the amplitudes for the 64 decaying curves. The cumulative curves 20, 22 and 24 are the partial sum of the points of the incremental curves. To provide porosity values, the NMR amplitudes of Figures 3 and 5 have been calibrated to provide the detected number of hydrogen atoms. Using the number of hydrogen atoms provided by water, one can easily calculate the mass of the water. Using the mass and density of water, one can determine volume and subsequently the water porosity of the sample by dividing the water volume by the bulk volume of the core sample. A similar calculation can be made to determine gas porosity using the number of hydrogen atoms provided by methane.

[0028] In Figures 3 and 5, the cumulative curves 20 and 24 show the water and gas porosities respectively. The sum of the two is the total core porosity. Thus, according to Figure 3, the core sample had a water porosity of 3.85%. According to Figure 5, the core had a methane porosity of 5.1 %. In Figure 4, the cumulative curve 22 has been normalized to one, as porosity can not be defined from the data until fluids with different hydrogen molecules per cc have been separated out.

[0029] In the laboratory tests, measurement of the T₂ relaxation time of the gas free sample required five hours. The second, T₂ measurement required three days. These long times are due to the very small volumes of gas in the sample. Logging procedures in the downhole environment will not require this time frame. Logging tools investigate much larger volumes thereby reducing the time to acquire data by many orders of magnitude. Thus, the present invention enables quantification measurements directly from logging operations.

[0030] As discussed above, the alteration in T₂ relaxation time results from the surface wetting of the organic pore surface by the methane gas. By wetting the surface, the hydrogen atoms do not respond to changes in the NMR magnetic field in the same manner as methane in a bulk situation. The relaxation time is represented by the following equations. In contrast to conventional formations where surface relaxation of gas is ignored, the following equations clearly indicate that surface relaxation within the organic pores of shale formations plays a significant role in the NMR data. Published studies have also shown that even for pores where the gas is the non-wetting fluid its relaxation contains a surface relaxation component at low water saturations. However, standard analytical techniques ignore this characteristic as the logging tool's zone of investigation corresponds to the invaded zone. In general, if a low viscosity fluid has a surface relaxation term, the surface relaxation term will be much faster than the bulk relaxation term. Also for the magnetic field gradient typically found in logging tools and methane diffusivity at its bulk value, the relaxation produced by diffusion is considerably slower than the experimentally observed relaxation value. As a result the surface relaxation term will dominate the relaxation spectrum (this has been confirmed for this case by the laboratory experiments).

1/T₂ = 1/T₂surface +l/T₂Bulk +l/T₂diffusion

1/Ti = 1/Ti surface +l/TiBulk

1/T₂ surface = p_s S V controls wetting fluid relaxation,

S, surface area,

V, volume,

p_s, surface relaxivity - Surface relaxivity captures the strength of the relaxation induced by the magnetic fields of molecules at the pore wall surface.

l/T₂Bulk Controls non-wetting fluid relaxation l/T₂Bulk = 1/TiBulk for fluids

l/T₂diffusion = D(g*G*TE)²/12

D diffusion coefficient

g Gyro magnetic ratio

G magnetic field gradient

[0031] The above referenced laboratory tests did not have an external gradient.

As a result diffusion had a minimal influence on the NMR signal. While there can be diffusion relaxation due to internal gradients, the generally accepted range of values for these gradients combined with the TE value (NMR echo time) used in these measurements results in a slow diffusion relaxation contribution compared to the observed relaxation rate.

[0032] The present invention may be implemented in the downhole environment in several ways. The following examples will detail the currently preferred applications of the present invention for practice in the field.

[0033] The first three examples assume that the formation is a shale formation and that drilling has been carried out using water based mud. While the present invention may be used with oil based mud, results will be enhanced by use of water based muds. Even though the invasion zone is minimal within shale formations, oil from oil based muds will more readily invade the organic pores of the shale. This invasion will force scans further into the reservoir in order to characterize the gas within the shale. Additionally, the following examples assume that mud pressure is kept to a minimum in order to minimize the depth of the invade zone.

[0034] NMR logging measurements may be carried out using conventional methods such as either logging while drilling (LWD) or subsequent to borehole formation by use of a wireline tool. LWD methods will further minimize the degree of mud invasion prior to NMR logging; however, wireline tools generally have greater measurement capacity. For example, some LWD tools measure only T_s.

[0035] The number of logging runs will be determined on a case by case basis considering the nature of the formation as well as the configuration and capacities of the logging tool. Although continuous logging measurements are desired to reduce operational costs, the logging run may include station measurements to provide a few high signal to noise measurements. These stationary measurements can be used to enhance the calibration of the logging run and subsequent results.

[0036] As in the laboratory example above, calculation of the quantity of gas within the shale formation will require extraction of the gas signal from the combined gas and water signal. As would be familiar to those skilled in the art of NMR and other downhole logging operations, there are a wide of variety of logging sequences that might be used to obtain the underlying data used in the practice of the present invention to separate out the gas and water signals. There are also generally multiple ways to process the data. The final outcome of most of these methods would be represented by curves similar to those found in Figures 3, 4, and 5. [0037] As noted above, the present invention relates to the interpretation and application of the results of the logging runs. The methods for carrying out well logging are well known to those skilled in the art and will not be discussed herein. In general, those skilled in the art of logging formations will recognize that constraints imposed by well bore and formation conditions will determine the number of logging runs and the sequences of the runs.

[0038] In addition to determining the volume of gas during NMR logging operations, the present invention may be readily adapted to measure the volume of oil, i.e. liquid hydrocarbons, during NMR logging operations within shale reservoirs. The method of quantifying liquid hydrocarbons found within the organic pores of a shale reservoir will be quantified in a manner similar to gaseous hydrocarbons. Under these conditions, water beyond the invaded zone the water will still have the fastest T] and T₂ relaxation times. Liquid hydrocarbons in the organic pores will have a surface relaxation component; therefore, liquid hydrocarbons in the pores will relax faster than bulk oil.

[0039] As discussed above, the inorganic pores within the formation will be water wetting. Therefore, oil found in the inorganic pores will have Ti and T₂ NMR relaxation times similar to the relaxation rate of bulk oil.

[0040] When water based mud is used, oil will not be displaced from the organic pores found in the invaded zone. However, the water component of the drilling mud will displace gas and oil from inorganic pores located within the invaded zone.

[0041] Therefore, quantification of hydrocarbons within the reservoir will require a measurement that senses beyond the invaded zone. At least two depths of NMR measurements are required in order to quantify the hydrocarbon content in organic pores and the hydrocarbon content in inorganic pores.

[0042] Figures 7 and 8 depict the measurements on a crushed core sample. In Figure 7, line 52 represents the Ti curve and line 54 the T₂ curve. In Figure 8, line 56 represents the Ti curve and line 58 the T₂ curve. According to Figures 7 and 8 the T|/T₂ ratio serves to identify the water and hydrocarbon signals. Figure 8 shows a dodecane T]/T₂ ratio of 3.84, Figure 7 a water Ti/T₂ ratio of 2.25. Measurements on other crushed organic shale core with dodecane and crude oil, and on this core with crude oil systematically show the same result. Methane also has shown a T]/T₂ ratio similar to that seen for liquid hydrocarbons measured on crushed samples. The difference in these ratios implies that there is greater separation between hydrocarbon and water in a Ti spectrum measurement. Therefore, one can use the Ti measurements to separate the water and hydrocarbon components more easily than T₂ measurements.

[0043] By doing a combined analysis of both the T₂ and the Ti spectra and utilizing the difference in the Ti/T₂ ratio between the hydrocarbon and the water, which provides a signature for each component, a more robust identification of the hydrocarbon and water components is possible. Figure 4 shows a T₂ spectrum typical of what a logging tool records. As can be seen, additional processing is needed to separate the water and gas signal. On the other hand the synthetic Ti curve in Figure 6 shows a clear separation between the water (peaks 36 and 38) and hydrocarbon (peaks 42 and 44) signals due to the hydrocarbon signal being shifted to relatively longer times in Ti space as compared to the water signal. A comparison of the Ti and T₂ spectrum curves will confirm and better specify which part of the Tj curve corresponds to the smaller Ti/T₂ ratio (i.e. water), and which part corresponds to the larger ratio so represents the gas or liquid hydrocarbon signal.

[0044] With reference to the Ti synthetic relaxation spectrum curve of Figure 6, the depicted spectrum represents the NMR Ti relaxation spectrum for a crude oil and water saturated sample constructed from measurements on a crushed organic shale sample. It represents what a NMR Ti logging measurement would record from signal beyond the invaded zone for a liquid hydrocarbon case. The small first peak 36, which represents the fastest relaxing signal, and the second peak 38 are from the water. The third peak 42 is from liquid hydrocarbon or gas in the organic pores. The last peak 44 in this case is the bulk oil response in the inorganic pores.

[0045] Published studies of gas in small inorganic pores show it has a significant surface relaxation so the signal from gas in the inorganic pores would overlap with the signal from the organic pores. Because it is a crushed sample the surface to volume of fluid ratio is not the same as for a saturated whole sample where fluid only exists in the pores. Therefore, the Ti relaxation is slower than the Tj for fluid in the pores of a whole sample, except for fluid in the inorganic pores, which as it relaxes at the bulk relaxation rate, would have the same i relaxation. As a result, the first two large peaks occur at slower times than they would for the whole plug. However, the ratio of these relaxation times depends on the oil and water surface relaxivities. Therefore, the relative position is the same permitting one to calculate the volume of liquid hydrocarbon in both the inorganic and organic pores, and the volume of water either using the Ti spectrum curve or in a more robust way the T_f and T₂ spectra. For the gas case, separation of gas and water can be done using the Ti spectrum or more robustly using both, but separation of gas into organic and inorganic pores requires performing the analysis on both NMR data from the invaded zone and the un-invaded zone. A robust calculation method using Ti and T₂ will be described that applies to gas or liquid hydrocarbons in the invaded or un-invaded zone.

[0046] The robust calculation will be carried out by constructing a mapping function from a T₂ spectrum (that would be similar to that in Figure 4) to a Tj spectrum (that would be similar to that in Figure 5). The mapping will be constrained to satisfy the differences in Ti/T₂ ratios, and that liquid hydrocarbon in inorganic pores will have a longer T| and T₂ relaxation times than any other component of the system. The mapping will work with the cumulative spectral curves to start. For NMR logging data, the amplitudes are normalized so that if the NMR signal consisted only of water it would give the true volume of the water. For NMR data of satisfactory data quality, the amplitude of the Ti and T₂ cumulative curves at long relaxation times will be the same. The amplitudes of the cumulative curves at very short relaxation times will be close to zero. They will then increase until they flatten out at some relaxation time. If those times are the same for Tj and T₂ this shows the presence of liquid hydrocarbon in water wet inorganic pores with a Tj/T₂ ratio of 1. The portion of the cumulative curves with the smallest Ts/T₂ ratio and the fastest relaxation time represents the water.

[0047] To estimate the Tj/T₂ of the water, one identifies on the T₂ cumulative plot the relaxation at which the curve has clearly risen out of the short time early signal. Examination of the incremental curve will help in this analysis. Then identify the corresponding point on the Ti cumulative curve. These points represent the starting point of the water signal. These two points from the Ti and T₂ cumulative curves provide the initial estimate of the water T_\/ T₂ ratio. Similarly, one obtains an initial estimate of the gas ratio or liquid hydrocarbon ratio in the organic pores by finding the corresponding times on Ti and T₂ cumulative curve where both begin to flatten before the rise produced by the any presence of liquid hydrocarbon in the inorganic pores. Start by using the estimated Tj/T₂ ratio for water to reconstruct the Ti curve from the T₂ curve. At some point in time the reconstructed Ti curve will begin rising at too early a time. Continue the reconstruction with the estimated hydrocarbon ratio until the bulk response part of the curve is reached where a ratio of 1 is applied.

[0048] This first reconstruction will include some water mapped with the hydrocarbon ratio so there will be some deviation between the reconstructed curve and the measured curve. By allowing some of the intermediate cumulative curve to be water and some hydrocarbon a closer match will be achieved. Several iterations may be needed to obtain a satisfactory reconstruction. In this process the values of the Ti/T₂ ratios will also be refined. This reconstruction can be thought of an inversion problem. In this case a wide number of well-known inversion methods could be used to find the appropriate values for the Ti/T₂ ratios of the water and hydrocarbon and the quantity of each. The knowledge of the constraint that the bulk ratio is one and the water ratio is smaller than the hydrocarbon ratio provides stabilization for the inverse. The existence of core data such as shown in Figures 7 and 8 can also be used to constrain the inverse.

[0049] in each of the following examples, the interpretation of the data relies upon the fact that the calibrated NMR measurement from a log reflects the number of hydrogen atoms in the fluid per unit of volume. Thus, knowledge of the molecular formulas for water and methane allows one to convert the NMR data to moles of water and moles of gas. Since we know water has a density of approximately 1 g/cc, we can also use the volume of water in the calculations. Since we do not know the gas density, the method permits extraction of the moles of gas but not the volume of gas from the NMR measurement. However, if other logging tools are used sufficient to determine formation porosity, then the gas volume may be subsequently calculated using the gas laws.

Example 1

[0050] The current preferred method for fluid identification from NMR logging is to acquire the sets T\, T₂, and D (diffusion spectra). According to one embodiment of the present invention, the sets T_l5 T₂, and D (diffusion spectra), or Ti and T₂ or Tj and D, will be extracted from logging runs as discussed above. If all logging sequences have been carried out to extract all three spectra, the resulting spectra will be plotted in three-dimensional space. In general, different fluids occupy different locations in this space. For example the discussed differences in Ti/T₂ ratio will clearly show up on such a plot. The laboratory experimental data that has been obtained will help to identify the signal corresponding to each fluid. If only two of the spectra have been obtained then two-dimensional cross plots using the respective signals can be used in the same way.

[0051] As known to those skilled in the art, general diffusivity of water is expected to be several orders of magnitude smaller than the diffusivity of gas. Thus, even though the small pore space will reduce the diffusivity of both fluids, the difference between the two will still enhance the separation between the gas and water signals in the resulting two or three dimensional cross plots. With reference to laboratory experimental data on cores from the same formation, one skilled in the art will be able to interpret the data presented in such cross plots. Subsequently, once the gas signal is separated, it will directly provide the moles of gas/cc in the shale formation. For liquid hydrocarbon the compositional analysis will provide the conversion from amounts of hydrogen to amounts of hydrocarbon. Thus, this method of the current invention enables gas quantification without determining the formation porosity or in situ formation pressure.

[0052] Thus, in order to calculate the gas quantity within a shale formation in accordance with the method of Example 1, one will initially obtain NMR logging data. This logging data will reflect the combination of gas or liquid hydrocarbons with water. As such, the relaxation curve will be similar to that depicted in Figure 2. The relaxation curve will be converted to spectral plots for gas and water similar to that depicted in Figure 4. The separation provided by the two or three-dimensional cross plots will be used to obtain a gas or liquid hydrocarbon spectrum similar to Figure 5 from this spectrum the total quantity of gas or liquid hydrocarbon will be calculated.

Example 2

[0053] In this embodiment of the present invention, the logging process will obtain either Ti or T₂ relaxation values. The resulting relaxation curves will appear similar to that depicted in Figure 2. The relaxation curve will be processed to obtain Ti or T₂ spectra of the combined gas or oil/water spectra. The resulting spectral curve will appear similar to that depicted in Figure 4.

[0054] Quantifying the amount of gas or oil in the reservoir or target formation according to this method will require a comparison of the logged NMR data to general NMR signatures for water and gas/oil obtained from core samples. Additionally, data relating to total water volume within the formation or target formation obtained from other logging tools will be utilized.

[0055] Once this data has been obtained, a mathematical inversion will be performed to determine the gas/oil quantity. Mathematical inversion techniques of this type are well known to those skilled in the art. The inversion can be performed using either the resulting spectral data, or by calculating the expected relaxation decay from the gas/oil and water spectra used. For the second approach the initial estimate for the water and gas spectra will be used to calculate the initial estimate for the water and gas relaxation curves.

[0056] The sum of these curves is the initial estimated relaxation curve. The estimated relaxation curve is compared to the recorded logging data. The initial estimated relaxation spectra will be constructed from laboratory measurements performed on core samples. The relaxation spectra will be similar to Figures 3 and 5, which provide type curves for each relaxation. The water relaxation spectrum is further constrained to have a total amplitude given by the separately obtained volume of water. The inversion proceeds by calculating the difference between the constructed estimated relaxation curve and the relaxation curve measured by the logging tool. The final gas/oil and water relaxation curves are obtained by iteratively changing the estimated curves until the difference has been made acceptably small. In order to stabilize this process the iterated water and gas relaxation curves are constrained so that they do not become too different from the laboratory type curves. That is the objective function that is minimized in the inversion process contains, not only a measure of the difference between the measured relaxation curve and the estimated relaxation function, but also a measure of the difference between the estimated water spectra and the water type curve spectra and a measure of the difference between the estimated gas/oil spectra and the gas/oil type curve spectra. The discussed functionalities, manipulations and mathematical operations performed on the data are well known to those skilled in the art. Example 3

[0057] In this example, the method of the present invention will examine core samples to provide the volume of water per cubic centimeter for the reservoir or target formation. Alternatively, water volumes may be obtained from joint interpretation of other logging tools according to conventional procedures. Subsequently, the NMR logging data will be used to determine the total number of hydrogen atoms contained per cc in the formation. The water volume then provides the number of hydrogen atoms contributed by the water. The difference between this value and the total number of hydrogen atoms in the formation is the number of hydrogen atoms provided by the methane or other hydrocarbons in the formation. Using the chemical formula of methane or a compositional analysis for the hydrocarbons present in the reservoir one then calculates the moles of hydrocarbon/cc.

[0058] This procedure can be illustrated for methane using the data that is plotted in Figures 3, 4 and 5. NMR data is conventionally calibrated so the amplitudes represent the number of grams of water detected if the signal is assumed to come from water. In Figure 3, the number of grams of water detected is 0.225 grams. In Figure 4 the grams of water equivalent is 0.543 grams. Subtracting 0.225 from 0.543 gives 0. 18 grams as the amount of gas measured in gram equivalent of water. The amount of gas can also be directly obtained from Figure 5. Using only Figure 5, the amount of gas is calculated to be 0.299 grams. The difference between these measurements is 0.019 grams or 6%. This difference is within the errors of the experimental measurements.

Example 4

[0059] This example demonstrates the ability to determine the amount of gas in the NMR signal attributable to the inorganic pores and the organic pores within a shale formation. According to this embodiment of the present invention, a shallow NMR measurement will be taken to observe only the invaded zone around the borehole, i.e. that region invaded by the drilling mud. hi this region, gas or oil within the inorganic pores will be displaced by water from the drilling mud. However, gas or oil within the hydrophobic organic pores will not be displaced. Thus, using any of the above techniques, one will be able to isolate the gas or oil signal and determine the quantity of hydrocarbon within the invaded zone. Typically, the total gas is reported as moles or grams per cubic centimeter. However, any mass measurement is appropriate.

[0060] Subsequently, a deeper NMR measurement will be taken as discussed above. The resulting NMR signal will be used to determine the total gas/oil within the examined zone. By subtracting the hydrocarbon quantity within the invaded zone (the quantity of hydrocarbons in the organic pores) from the hydrocarbon quantity within the non-invaded zone, one determines the amount of hydrocarbon in the inorganic pores. One would assume a general consistency within the formation for the mass of hydrocarbon on a per cubic centimeter basis within the organic pores. Thus, any additional amount of hydrocarbon within the non-invaded zone will be attributable to the inorganic pores within the shale.

Example 5

[0061] To demonstrate the ability to determine the volume of liquid hydrocarbons within a shale formation, and to demonstrate that hydrocarbons have a larger T[/T₂ ratio than water, NMR laboratory tests were carried out on crushed core samples saturated with dodecane, crude oil, and brine An example of these measurements is shown in Figures 7 and 8. With reference to Figures 7 and 8, the core sample saturated with brine had a Ίι to T₂ ratio of 2.25. When saturated with dodecane, a liquid hydrocarbon at room temperatures, the NMR relaxation curves produced a Tt to T₂ ratio of 3.84. As described above, to determine the volume of liquid hydrocarbon in the shale, one will construct a mathematical mapping function using the plots of the T_t and T₂ curves. The Ti/T₂ ratios for the Ti and T₂ curves will separate the respective curves for liquid hydrocarbons and water. As demonstrated by Figure 8, this part of the curve will correspond to the hydrocarbon component. Subsequently, deleting that portion of the curve contributed by Tj/T₂ ratio of approximately 2 to 2.3 will provide one with the mass of liquid hydrocarbons in the formation. The separation could also be achieved by any of the methods described in the proceeding examples. In this synthetic curve constructed from the crushed sample data, Figure 6, it is clear that a reasonable separation could be achieved from the T| spectrum curve by itself. Once the parts of the curve have been identified as gas/oil and water the cumulative spectrum curve will give the amount of water. The amount of gas or liquid is determined by subtracting the water cumulative curve from the total cumulative curve and then using the_compositional analysis for the hydrocarbons present in the reservoir to convert the density of hydrogen atoms to hydrocarbon amounts.

[0062] In sum, the present invention provides suitable embodiments for readily determining hydrocarbon quantities in shale formations using currently available MR logging tools.

[0063] Other embodiments of the current invention will be apparent to those skilled in the art from a consideration of this specification or practice of the invention disclosed herein. Thus, the foregoing specification is considered merely exemplary of the current invention with the true scope and spirit of the invention being defined by the following claims.

Claims

I claim:

1. A method for detecting and quantifying hydrocarbon within a shale formation comprising:

drilling a borehole at least into a portion of a shale formation; logging the borehole using at least an NMR logging tool;

obtaining a core sample from said formation;

conducting NMR analysis of said core sample to determine the water relaxation curve for said core sample;

from the logging of said formation with said NMR tool, obtaining a relaxation curve for at least a portion of said formation;

subtracting the water relaxation curve obtained from NMR analysis of said core sample from the relaxation curve obtained from logging at least a portion of said formation and determining the quantity of hydrocarbon within said formation.

2. The method of claim 1, wherein said step of drilling utilizes water based drilling mud.

3. The method of claim 1, wherein said formation is a shale formation having organic pores.

4. The method of claim 1, wherein said hydrocarbon is methane.

5. The method of claim 1, further comprising the step of infusing said core sample with a brine solution prior to conducting said NMR analysis.

6. The method of claim 1, wherein said step of logging said formation obtains at NMR measurements from at least two separate depths.

7. The method of claim 4, wherein said step of logging the borehole further determines the porosity of the formation.

8. The method of claim 7, further comprising the step of calculating the volume of methane in said formation.

9. A method comprising:

drilling a borehole at least into a portion of a shale formation; logging the borehole using at least an NMR logging tool and determining NMR relaxation curves for Tj, T₂ and plotting the formation Ti/T₂ ratio;

obtaining a core sample from said formation;

conducting NMR analysis of said core sample to determine the water Ti/T₂ ratio for said core sample;

comparing the formation Ti/T₂ ratio to the core sample T]/T₂ ratio to determine the contribution of hydrocarbon to the formation T1/T2 ratio.

10. The method of claim 9, further comprising obtaining NMR relaxation curves during logging from at least two different depths into the formation.

11. A method comprising:

drilling a borehole at least into a portion of a shale formation; logging the borehole using at least an NMR logging tool and determining NMR relaxation curves for at least one of Ti or T₂;

logging the borehole to determine total water volume within the formation region corresponding to the NMR relaxation curves;

processing the logged Ti or T₂ relaxation curve to obtain Tj or T₂ spectra corresponding to combined formation hydrocarbon and water spectra;

obtaining a core sample from said formation;

conducting NMR analysis of said core sample and using the NMR values from the core sample to estimate the relaxation spectra for water and hydrocarbon within the formation;

compare the estimated relaxation spectra to relaxation curves obtained from the logged data;

constraining the water relaxation spectra to a total amplitude corresponding to the separately logged volume of water thereby constructing an estimated relaxation curve;

calculating the difference between the constructed estimated relaxation curve and the relaxation curve measured by the logging tool; obtaining final formation hydrocarbon and water relaxation curves by iteratively changing the estimated curves until the formation relaxation curves substantially fit the core sample relaxation curves; and,

quantifying the volume of hydrocarbon in said formation using said final hydrocarbon relaxation curve.

12. A method for detecting and quantifying natural gas or oil within a shale formation comprising:

drilling a borehole at least into a portion of a shale formation; obtaining a core sample from said formation;

calculating the volume of water per cubic centimeter in said core sample logging said formation with an NMR tool and determining the total number of hydrogen atoms contained per cubic centimeter of formation;

using the calculated volume of water per cubic centimeter of core sample to determine the number of hydrogen atoms contributed by water within said formation and the number of hydrogen atoms contributed by methane within said formation; and,

using the chemical formula of methane to calculate the moles of methane per cubic centimeter within said formation.

13. The method of claim 12, wherein said step of drilling includes the use of a water based drilling mud and said step of logging with an NMR tool measures only the region of said shale formation invaded by said drilling mud.

14. The method of claim 13, further comprising the step of taking a second NMR measurement to determine the number of hydrogen molecules per cubic centimeter of formation beyond the region of the shale formation invaded by said drilling mud and calculating the moles per cubic centimeter of methane in the region of said shale formation invaded by said drilling mud and in the region beyond said drilling mud.