CN111537544B - Improve nuclear magnetic resonance T 2 Conversion method for spectral characterization of pore size distribution precision of tight reservoir - Google Patents

Abstract

The invention provides a method for improving nuclear magnetic resonance T ₂ The conversion method for the spectral characterization of the precision of the pore size distribution of the tight reservoir comprises the following steps: the method comprises the following steps: performing nuclear magnetic resonance test on a compact reservoir sample of saturated formation water to obtain nuclear magnetic resonance T ₂ A spectrum; step two: performing a constant-speed mercury-pressing experiment on the dry and compact reservoir sample to obtain mercury-pressing pore size distribution data; step three: carrying out sectional processing on the mercury intrusion pore size distribution data; step four: obtaining T using a cylindrical pore model ₂ The conversion relation with the aperture size; step five: respectively using the conversion relation in the step four to obtain the T of each section ₂ Conversion method from aperture size to obtain complete T ₂ A conversion method for spectral characterization of tight reservoir pore size distribution. The method improves the accuracy of pore size distribution obtained by nuclear magnetic resonance experiments.

Description

A Transformation Method for Improving the Accuracy of NMR T2 Spectrum Characterization of Pore Size Distribution in Tight Reservoirs

technical field

The invention relates to reservoir prediction technology, in particular to a nuclear magnetic resonance characterization method for the pore structure of tight reservoirs, and belongs to the field of petroleum exploration and development.

Background technique

Compared with conventional reservoirs, one of the main characteristics of tight reservoirs is the small pore throat space, which makes it very difficult for fluids to migrate in them. Therefore, accurate evaluation of the pore structure of tight reservoirs is one of the prerequisites for evaluating the potential, recoverability _and production capacity of tight oil and gas resources. Nuclear magnetic resonance is a commonly used testing method. is the pore radius. The advantage of this is that (1) combined with centrifugation or drying means, the pore size distribution of the bound fluid can be clarified; (2) combined with the relaxation spectrum of the manganese-saturated sample, the pore size distribution occupied by the oil phase can be clarified; (3) combined with the constant velocity pressure Mercury experiments can clarify throat control of bound fluid or residual oil.

其中，n为常数。但该式不能通过数学推导得到，同时缺少物理意义。并且，通常的转化方法是基于实测的压汞孔径分布数据，如果实测数据在某一进汞饱和度段内分布密集，则转换公式会更多的考虑这部分权重，使得实测数据稀疏段的拟合程度较差现，导致利用T₂转换为孔径分布的效果较差。Often the conversion of _T2 to a pore size distribution is an experimental approach in conjunction with mercury injection. The conventional conversion method is derived through the formula: T ₂ =Cr _t , wherein, C is the conversion coefficient, and _rt is the pore size obtained by mercury injection. The above method has related records in CN104634718A, CN106249306A and other prior art. However, the problem with this method is that the linear relationship between T ₂ and r _t cannot fit them well. It may be due to the complex pore throat structure and smaller pore size distribution, which is more obvious in tight reservoirs. For this reason, some researchers have obtained empirical formulas based on conventional methods:

Among them, n is a constant. But this formula cannot be obtained through mathematical derivation, and lacks physical meaning at the same time. Moreover, the usual conversion method is based on the measured mercury injection pore size distribution data. If the measured data is densely distributed in a certain mercury saturation range, the conversion formula will take this part of the weight into consideration, so that the simulated data for the sparse segment of the measured data The degree of integration is poor, resulting in poor conversion of _T2 to pore size distribution.

Therefore, the present invention proposes a new method of transforming the pore size distribution by using _T2 , and on the basis of the usual derivation, the cylindrical pore model is also fully considered. On the one hand, it shows that the linear relationship is not good for transversal relaxation time T ₂ to convert the pore size distribution, on the other hand, the proposed model is more in line with the physical meaning. In addition, the processing of the mercury injection pore size distribution data by the present invention can help improve the accuracy of the pore size distribution obtained through the nuclear magnetic resonance experiment, and relatively accurately evaluate the potential of tight oil and gas resources.

Contents of the invention

The purpose of the present invention is to provide a conversion method that has certain physical significance and helps to improve the accuracy of characterization of tight reservoir pore size distribution by nuclear magnetic resonance _T2 spectrum, so as to evaluate the potential of tight oil and gas resources relatively accurately.

In order to achieve the above object, the embodiment of the present invention provides the following technical solution: a conversion method for improving the precision of pore size distribution of tight reservoirs characterized by nuclear magnetic resonance _T2 spectrum, comprising the following steps:

Step 1: Carry out nuclear magnetic resonance test on the tight reservoir sample saturated with formation water, and obtain the nuclear magnetic resonance _T2 spectrum;

Step 2: Carry out a constant-speed mercury intrusion experiment on dry tight reservoir samples to obtain mercury intrusion pore size distribution data;

Step 3: Perform segmentation processing on the mercury injection pore size distribution data;

Step 4: Use the cylindrical pore model to obtain the conversion relationship between _T2 and the pore size;

Step 5: Use the conversion relationship in step 4 to obtain the conversion method between T ₂ and pore size of each segment of the data in step 3, so as to obtain a complete T ₂ spectrum to characterize the conversion method of tight reservoir pore size distribution .

Further, the subsection processing method in step 3 is: in the order of small aperture to large aperture, calculate the difference between the aperture size data of different test points, and select the aperture size data whose aperture difference is within an order of magnitude range as a data segment.

Further, the conversion relationship in step 4 is:

The relationship between _T2 , pore volume and pore fluid volume expressed by KST model:

The volume of a thin layer of fluid in a cylindrical pore model is:

The surface area of a cylindrical pore is:

S ^t =2πr _t ·l (3)

Put formula (2) and formula (3) into formula (1) to get

make

Among them, T ₂ represents the transverse relaxation time, T _2s represents the particle surface relaxation, V _s represents the pore fluid volume, V represents the pore volume, V _s ^t represents the volume of the thin layer fluid in the cylindrical pore model, h represents the thickness of the thin layer fluid , r _t represents the pore radius, l represents the cylindrical height, S ^t represents the surface area of cylindrical pores, ρ ₂ is the relaxation rate; S/V is the pore specific surface; F _s is the shape factor, dimensionless, for spherical pores , F _s =3; for columnar pipes, F _s =2. The relaxation rate ρ ₂ and the pore shape factor F _s can be approximately regarded as constants, so C is a constant value;

Bringing (5), (6) and (7) into (4), the conversion relationship between _T2 and aperture size can be obtained:

Further, the C and h values of each section can be obtained by the least squares principle by each section of data, thereby obtaining the conversion method between _T2 and aperture size of each section;

Further, a conversion method to obtain the complete _T2 spectrum to characterize the pore size distribution of tight reservoirs.

Compared with the prior art, the present invention has the following characteristics and advantages:

1. The present invention fully considers the pore structure model of the rock itself, changes the plane physical model commonly used by NMR into a cylindrical pore model consistent with the mercury injection experiment, and improves the pore obtained by the NMR data _T2 and the mercury injection data The accuracy of the correspondence between radii.

2. The present invention fully considers the characteristics of the mercury intrusion data distribution, uses segmental fitting, and reduces the fitting error caused by the data itself.

Description of drawings

Figure 1 shows the segmentation processing of mercury injection pore size distribution data

Figure 2 shows the conversion process from the plane physical model to the cylindrical pore model

Fig. 3 is the fitting relation of _T obtained by conversion method of the present invention and pore radius

Fig. 4 is the comparison result of the pore size distribution calculated by the conversion method of the present invention and the mercury injection pore size distribution

Figure 5 is the fitting relationship between _T2 and pore radius obtained by empirical method

Figure 6 is the comparison result of the pore size distribution calculated by using the empirical method and the mercury injection pore size distribution

Figure 7 is the fitting relationship between T _and pore radius obtained by conventional methods

Fig. 8 is the comparison result of the pore size distribution calculated by the conventional method and the mercury injection pore size distribution

Detailed ways

The details of the present invention can be understood more clearly with reference to the accompanying drawings and the description of specific embodiments of the present invention. However, the specific embodiments of the present invention described here are only for the purpose of explaining the present invention, and should not be construed as limiting the present invention in any way. Under the teaching of the present invention, the skilled person can conceive any possible modification based on the present invention, and these should be regarded as belonging to the scope of the present invention.

Example 1

Taking four tight sandstone samples from the west of the Yishan slope in the Ordos Basin, China as an example; the nuclear magnetic resonance and constant-speed mercury injection experiments were completed at the Institute of Fluid Mechanics, Chinese Academy of Sciences.

RecCore2500 low-field nuclear magnetic resonance instrument was used in the nuclear magnetic resonance experiment. The resonance frequency is 2.38MHz, the number of echoes is 2048, the number of scans is 128, the waiting time is 5000ms, the echo interval is 0.6ms, and the gain is 50. Vacuumize the rock sample at 120°C for at least 24 hours, weigh it and soak it in simulated formation water (total salinity 25000mg/L) for at least 24 hours, remove the surface water of the rock sample with slightly wet filter paper, weigh the rock sample and The first NMR measurement yielded a _T2 spectrum.

The constant-speed mercury intrusion experiment adopts the ASPE-730 constant-speed mercury intrusion experiment device produced by Coretest Company, with a contact angle of 140° and a surface tension of 485dyne/cm. The rock sample is prepared into a cylindrical core with a diameter of 1 cm and a length of 1 cm, vacuumed at 120°C for at least 24 hours, soaked in a mercury solution, and mercury is injected into the core at a constant speed of 0.00005ml/min, and the experiment ends when the pressure reaches 6.2055MPa .

Furthermore, considering the distribution of data points will affect the fitting accuracy, this is because the place where the data is more concentrated tends to have a better fitting effect, while the place where the data points are less or the distribution is more scattered is often not satisfactory.

For tight sandstone, there are relatively few large pores and relatively few small pores. At the same time, the pressure points of the mercury injection test are concentrated in the high-pressure area corresponding to the small pores, so reflected in the mercury injection data, it appears as data points with small pores. More, the distribution is more concentrated, while the data points of large apertures are less, the distribution is more scattered.

According to the sequence from small aperture to large aperture, the present invention calculates the difference in the aperture size of the mercury intrusion data at different test points, and the results show that the aperture difference increases sequentially, which indicates that the larger the aperture, the larger the data points in the unit aperture range. The less (as shown in Figure 1, taking sample J-1 as an example). Therefore, starting from the minimum aperture difference value, a corresponding aperture range within an order of magnitude range is selected as a data segment. The segmental conversion of the aperture can avoid the problem that the large aperture data with sparse distribution is less considered in the data fitting process to a certain extent.

In the next step, the conversion relationship between _T2 and pore size is obtained using the cylindrical pore model. The specific process is:

The relationship between _T2 , pore volume and pore fluid volume expressed by KST model:

Replace the plane physical model (thin-layer fluid volume V _s ) in the conventional or empirical method with a cylindrical pore model (as shown in Fig. 2), the volume of the thin-layer fluid in the cylindrical pore is:

The surface area of a cylindrical pore is:

S ^t =2πr _t ·l (3)

Put formula (2) and formula (3) into formula (1) to get

make

Bringing (5), (6) and (7) into (4), the conversion relationship between _T2 and pore radius can be obtained;

In the next step, use the principle of linear least squares to obtain the values of C and h. For the process of obtaining C and h values, refer to the _T2 spectrum lateral relaxation of the rock core disclosed in the inventor's previous study "Effects of pore-throat structure on gas permeability in the tights and stone reservoirs of the Upper Triassic Yanchang formation in the Western Ordos Basin, China" Time conversion pore radius distribution method.

Finally, use the same method to obtain the C and h values of different segments in the same way for the data of other segments, so as to obtain the conversion method of the complete core _T2 spectrum to characterize the pore size distribution. The calculation results are shown in Table 1:

The fitting relationship between _T2 and pore radius obtained by the above method is shown in Figure 3 (taking sample T-6 as an example):

The comparison results of the pore radius distribution calculated by the above method and the mercury injection pore radius distribution are shown in Figure 4 (taking sample T-6 as an example)

Comparative example 1

两侧取对数，采用最小二乘原理得到C和n值。Take the empirical method as an example, namely:

Take the logarithm on both sides, and use the principle of least squares to get the values of C and n.

The calculation results are shown in Table 2:

The fitting relationship between _T2 and pore radius obtained by empirical method is shown in Fig. 5 (taking sample T-6 as an example):

The comparison results of the pore radius distribution calculated by the empirical method and the mercury injection pore radius distribution are shown in Figure 6 (taking sample T-6 as an example):

Comparative example 2

Taking the conventional method as an example, that is, n takes 1, T ₂ =Cr _t , and the value of C is obtained by using the least square principle.

The calculation results are shown in Table 3:

The fitting relationship between _T2 and pore radius obtained by conventional methods is shown in Figure 7 (taking sample T-6 as an example):

The comparison results of the pore radius distribution calculated by the conventional method and the mercury injection pore radius distribution are shown in Fig. 8 (taking sample T-6 as an example):

Contrast Table 1, Table 2 and Table 3, meanwhile, contrast Fig. 3, Fig. 5 and Fig. 7, the result shows, utilizes the conversion method that the present invention obtains, T _Value and aperture fitting coefficient are all more than 0.99, visually fitting The effect is better, obviously superior to the conventional method and the empirical method, indicating that the conversion relationship obtained by the present invention can better predict the pore radius, and reflect the pore size distribution of the nuclear magnetic signal more accurately to a certain extent.

Further, compare Fig. 4, Fig. 6 and Fig. 8, the result shows, utilize the conversion method that the present invention obtains to obtain the nuclear magnetic resonance T The pore diameter converted from the _spectrum is compared with the constant-speed mercury injection pore-throat radius, the corresponding relationship between the two better. Compared with the pore diameter converted from the NMR T ₂ spectrum obtained by the conventional method and the pore throat radius of the constant-speed mercury injection, there is a large difference in the prediction results of the conventional formula method. However, compared with the pore diameter converted from the NMR _T2 spectrum obtained by the empirical method and the pore-throat radius of the constant-velocity mercury injection, although the small pore diameter part of the prediction result of the empirical formula method has a good corresponding relationship, the corresponding relationship of the large pore diameter part is poor, indicating that The accuracy of the empirical formula method for large aperture prediction is not high, which shows that the present invention processes the mercury injection data in segments, which improves the prediction accuracy, especially for the prediction of large aperture segments.

Therefore, the present invention uses a cylinder model for the conversion relationship between _T2 spectrum and pore size, and at the same time processes the mercury injection data in segments to obtain a conversion method with higher precision for NMR _T2 spectrum to characterize the pore size distribution of tight reservoirs.

The purpose of the detailed explanations for the above-mentioned embodiments is only to explain the present invention so as to better understand the present invention. However, these descriptions cannot be construed as limiting the present invention for any reason, especially, in different The various features described in the embodiments can also be combined arbitrarily with each other to form other embodiments. Unless there is an explicit description to the contrary, these features should be understood as being applicable to any embodiment and not limited to the described embodiment. Way.

Claims (5)

1. Improving nuclear magnetic resonance T ₂ The conversion method for spectral characterization of tight reservoir pore size distribution precision is characterized by comprising the following steps of:

the method comprises the following steps: performing nuclear magnetic resonance test on a compact reservoir sample of saturated formation water to obtain nuclear magnetic resonance T ₂ A spectrum;

step two: carrying out mercury injection experiment on the dry compact reservoir sample to obtain mercury injection pore size distribution data;

step three: carrying out sectional processing on the mercury intrusion pore size distribution data;

step four: obtaining T using a cylindrical pore model ₂ The conversion relation with the aperture size;

step five: respectively using the conversion relation in the step four to obtain the T of each section ₂ Conversion method from aperture size to obtain complete T ₂ A conversion method for spectral characterization of tight reservoir pore size distribution;

the conversion relation obtaining method of the fourth step comprises the following steps:

t expressed using KST model ₂ Relationship between pore volume and pore fluid volume:

the volume of the thin layer fluid in the cylindrical pore model is:

the surface area of the cylindrical pores is:

S ^l ＝2πr _t ·l (3)

substituting the formula (2) and the formula (3) into the formula (1) to obtain

Order to

Wherein, T ₂ Representing transverse relaxation time, T _2s Denotes the relaxation, V, of the particle surface _s Denotes the pore fluid volume, V denotes the pore volume, V _s ^l The volume of the thin layer fluid in the cylindrical pore model is shown, h represents the thickness of the thin layer fluid, r _t Denotes the pore radius, l denotes the cylinder height, S ^l Expressed as the surface area of the cylindrical pores, p ₂ Is the relaxation rate; S/V is the specific surface of the pores; f _s Is a form factor, dimensionless, to spherical pores, F _s =3; for cylindrical pipes, F _s =2; relaxation rate ρ ₂ Pore shape factor F _s Can be approximately regarded as a constant, so C is a constant value;

bringing (5), (6) and (7) into (4) can give T ₂ Conversion relationship with pore radius:

2. the method of claim 1 for increasing nuclear magnetic resonance T ₂ The conversion method for the spectrum characterization of the pore size distribution precision of the tight reservoir is characterized in that the segmented processing method in the step three comprises the following steps: and according to the sequence from the small aperture to the large aperture, performing difference calculation on the aperture size data of different test points, and selecting the aperture size data with the aperture difference within an order of magnitude range as a data segment.

3. The method of claim 1 for increasing nuclear magnetic resonance T ₂ The conversion method for the spectral characterization of the precision of the pore size distribution of the tight reservoir is characterized in that each section of data passes through a least square methodObtaining the C and h values of each segment, thereby obtaining the T of each segment ₂ And the pore size.

4. The method of claim 1 for increasing nuclear magnetic resonance T ₂ The conversion method for spectral characterization of the pore size distribution precision of the tight reservoir is characterized in that the tight reservoir is one of tight sandstone, shale or mudstone.

5. The method of claim 1 for increasing nuclear magnetic resonance T ₂ The conversion method for the spectral characterization of the pore size distribution precision of the compact reservoir is characterized in that the nuclear magnetic resonance in the step one is low-field nuclear magnetic resonance, and the mercury intrusion experiment in the step two is constant-speed mercury intrusion.

Images