RU2704671C1 - Method of determining viscosity of heavy oil by nuclear magnetic resonance in porous space of a collector and free volume - Google Patents

Abstract

FIELD: oil industry.

SUBSTANCE: use: to determine properties of formation fluids, specifically heavy oil viscosity. Summary of invention consists in the fact that obtaining representative samples of heavy oil by high-speed centrifugation from oil-saturated core sampled from the depth of interest, further measurement of spin-spin relaxation rate of obtained representative samples of heavy oil by pulse nuclear magnetic resonance using a Carr-Parcell-Meibum-Gill sequence on laboratory equipment, followed by processing of amplitude-relaxation characteristic, including obtaining a spectrum of spin-spin relaxation times using the Tikhonov regularization method, determining the average geometrically weighted spectrum relaxation time in interval of 0.01–20 ms, using the obtained average geometrically weighted relaxation time in the ratio for calculating viscosity with known constant coefficients.

EFFECT: technical result is possibility of high-accuracy determination of oil viscosity in 500–100,000 mPa*s range based on nuclear magnetic resonance method both with the help of laboratory investigations, and using borehole NMR logging tools in strong field.

5 cl, 10 dwg

Description

FIELD OF THE INVENTION

The invention relates to methods for determining the properties of reservoir fluids, namely, the viscosity of heavy oil based on the processing and interpretation of measurement data of nuclear magnetic relaxation of heavy oil in the pore space of the reservoir rock or in free volume, and is intended for use in the oil industry for the study of reservoirs with the aim of increasing Efficiency of development of heavy and extra-viscous oil fields. Also, the invention can be used in scientific and design support, research work in the field of geology and oil field development based on the NMR characteristics of the selected core, extracted oil, as well as nuclear magnetic field data in a strong field. The research results can be used in geological and hydrodynamic modeling to select a method and method for developing and operating a heavy oil reservoir.

State of the art

Currently, there is an active industrial development and preparation for commissioning the largest heavy oil deposits in the Republic of Tatarstan. Productive deposits containing the main reserves and resources of heavy oil and natural bitumen are located at a relatively shallow depth (from 50 to 300 m). Despite the relatively shallow depth of the deposits, these deposits are difficult to recover due to the high viscosity of the oil contained in the reservoir and the impossibility of its production without imparting fluidity, for example, due to steam injection. For the effective application and improvement of existing production technologies, a more detailed study of the geological and geochemical conditions for the occurrence of these deposits, as well as the properties of hydrocarbon fluids located in them, is necessary.

It is believed that heavy oil deposits are characterized by heterogeneity of viscosity of the saturating fluid both in depth and in strike, and the viscosity can vary in the range of several orders of magnitude. However, to date, there is no reliable method for determining this heterogeneity and assessing the viscosity of oil directly in reservoir conditions. Existing measurement methods do not allow determining the value of oil viscosity without its extraction from the reservoir rock. Of great practical interest is the use of the method of nuclear magnetic resonance (NMR) to describe the characteristics of the fluid saturating the reservoir with heavy oil directly under reservoir conditions. NMR logging is the only technology that allows continuous assessment of the distribution of oil viscosity under natural bedding conditions directly in the pore space of the reservoir rock. The most widely used methods for interpreting NMR data are based on measuring the characteristics of nuclear magnetic relaxation of reservoir rocks saturated with various fluids, processing and analyzing the resulting spectra of the distribution of relaxation times, and establishing time cutoffs. At the same time, using various empirical and theoretical models, they make assumptions about the properties of rocks and their saturating fluids. Obtaining information on the spatial distribution of oil of various viscosities will provide an opportunity to optimally design a system for developing deposits, monitor its condition and conduct optimal geological and technical measures to increase current and accumulated production.

There are a variety of instruments in the oil industry through which the rheological characteristics of formation fluids can be determined. There are various methods of hydrodynamic, geophysical and other studies of reservoir fluids in the process of developing productive boards, among which there are both direct and indirect methods for determining the rheological properties of oil.

For example, there is a method for predicting the viscosity of reservoir fluids from NMR relaxation results (T ₁ and T ₂ ) and diffusion coefficients based on empirical correlations (see [1] Morriss et al., SPWLA Annual Transaction, pp. 1-24, June 19 22, 1944; [2] Freedman et. Al., SPE Journal (75325), December 2001; [3] Lo et. Al., SPE Journal (77264), March 2002). Correlations through empirically determined constants relate logarithmic mean distributions to viscosity. The accuracy of the viscosity values predicted from these correlations is limited by the following factors:

1) the detailed form of distributions is not taken into account;

2) the empirical constants used in the correlations are not universal and can vary for different types of oil and devices;

3) the accepted form for the correlation equation is not precisely defined;

4) empirical correlations were obtained for oil samples with viscosities up to 1000 cps;

5) The influence of pore space for various types of oil is not taken into account.

A method is also known for determining viscosity by measuring spin-spin relaxation of oil in a free volume or in the pore space of a reservoir rock under laboratory conditions (see [4] Ahmed K. et al. Viscosity Predictions of Viscous Oil from a Kuwait Oil Field by Low- field NMR // SPE International Heavy Oil Conference and Exhibition. - Society of Petroleum Engineers, 2014.), as well as a method for determining oil viscosity based on two-dimensional relaxation maps T ₂ -D (see [5] RF application No. 2005117079, IPC G01N 1/00, published on November 20, 2006). However, these methods can be applied with varying degrees of confidence only to oil, with a viscosity below 1000 mPa⋅s. This method is more resource-intensive and not always possible to use, especially when measuring nuclear magnetic logging.

A close analogue is a method for determining the viscosity of oil through downhole sampling and fluid analysis tools, such as Schlumberger MDT. Using a specialized tester, you can obtain information about the type and characteristics of reservoir fluids using various modules in the sensors located in the device. However, the number of measurements is limited by the number of modules, which makes it impossible to obtain a continuous distribution of viscosity values in the reservoir.

Also known is a method for determining the properties of formation fluids surrounding an underground well, namely, the viscosity of oil, by measuring nuclear magnetic resonance. The measurements are carried out by a modular sensor on moving oil samples collected in the reservoir by means of a downhole sampler (see [6] US patent No. 6111408, IPC G01R 33/44, G01V 3/32, publ. 08/29/2000; [7] RF patent 2367981, IPC G01V 3/32, published on September 20, 2009). Analogs of this kind have several disadvantages. First of all, the need to obtain a moving sample of formation fluid, which, in the case of geological and chemical conditions of occurrence of heavy oil reservoirs without chemical or thermal effects on the formation, is unlikely. A side effect of this effect is an irreversible change in the physicochemical properties and structure of the formation fluid, and an unsatisfactory degree of contamination of the crude formation oil sample, which may contain hydrocarbon-based drilling mud filtrate.

The closest analogue of the proposed method (prototype) by methodological essence is a method for determining the viscosity of oil in a pore medium using a nuclear magnetic resonance relaxometer, including measuring the spin-lattice relaxation time and the amplitude of the liquid NMR signal in free volume using the results of which the oil viscosity is calculated ( see [8] Aut. St. USSR No. 1339440, IPC G01N 11/00, publ. 09/23/1987).

The presented method allows to accurately determine the viscosity of oil in the range of 500 - 100,000 mPa * s based on the method of nuclear magnetic resonance using laboratory tests, and using borehole NMR tools in a strong field, which is associated with a different approach to processing and interpretation of data.

SUMMARY OF THE INVENTION

The objective of the claimed invention is to develop a method for determining the viscosity of heavy oil, including in reservoir reservoir conditions, which eliminates the disadvantages of analogues and prototype, as well as the possibility of using both in measurements on laboratory equipment, and in the so-called downhole cable logging devices nuclear magnetic resonance.

The technical result is to increase the efficiency of development of heavy oil deposits confined to terrigenous reservoirs according to the results of the development of a method for determining the viscosity of oil based on the application of the NMR method.

The problem is solved by determining the relationship of the magnetic relaxation and rheological characteristics of oil, and the technical result is achieved by using the interpretation of data from borehole NMR logs in a strong field, which allows you to obtain detailed information about the distribution of the viscosity of heavy oil over the section of the reservoir.

The problem is solved, and the technical result is achieved due to the proposed method for determining the viscosity of oil, including obtaining representative samples of heavy oil by high-speed centrifugation from oil-saturated core samples taken from the depth of interest, subsequent measurement of the spin-spin relaxation rate of the obtained representative samples of heavy oil using pulsed nuclear magnetic resonance using Carr-Parcell-Maybum-Gill sequence on laboratory equipment, with by processing the amplitude-relaxation characteristic, which includes obtaining a spectrum of spin-spin relaxation times using the Tikhonov regularization method, determining the geometrically weighted average relaxation time of the spectrum in the range 0.01–20 ms, using the obtained geometrically average relaxation time in the ratio for calculating the viscosity with previously known constant coefficients.

Also, the technical result is achieved due to the fact that the measurement of the spin-spin relaxation rate on laboratory equipment is carried out on a naturally-saturated core selected from a depth of interest.

The technical result is also achieved due to the fact that the spin-spin relaxation rate is measured using a sequence of radio frequency pulses 180 ⁰ -KPMG and the subsequent construction of a two-dimensional map T ₁ -T ₂ with the selection of the spin-spin relaxation spectrum from the data array.

The task and technical result are also achieved by the second variant of the method for determining oil viscosity, which includes continuous measurement of the spin-spin relaxation rate by a nuclear magnetic logging tool using the productive interval of the well, followed by processing of the amplitude-relaxation characteristics, including obtaining a spectrum of spin-spin relaxation times using the regularization method Tikhonova, determination of the geometrically weighted average relaxation time on the spectrum interval 0.01 - 2 0 ms., The use of the obtained geometrically weighted average relaxation time in the ratio for calculating the viscosity with predefined correlation coefficients.

The technical result is achieved due to the fact that the device nuclear magnetic logging tool in the field of subterranean wells is measured RF pulse sequence 180 ⁰ -KPMG followed by construction of a two-dimensional map T ₁ -T ₂ at intervals of depths of interest to determine the type of fluid and the mobility its component based on the established correlation dependences of the ratio T ₁ / T ₂ on the viscosity of heavy oil.

Brief Description of the Drawings

The figure 1 shows the ratio of the maximum and minimum times of spin-spin relaxation depending on the viscosity of the oil.

Figure 2 shows the distribution of spin-spin relaxation times T ₂ for various combinations of rock wettability and fluid properties.

The figure 3 shows the dependence of the coefficient of dynamic viscosity of heavy oil under thermobaric conditions of the reservoir on the viscosity of oil under standard conditions.

4 is a graphical representation of the KPMG sequence (Carr-Purcell-Maybum-Gill, in the original: Carr-Purcell-Meiboom-Gill - CPMG).

Figure 5 shows the spectra of spin-spin relaxation times obtained by various methods of processing the initial data on the decline.

Figure 6 shows the contribution of the terms of the empirical formula to the calculated value of the dynamic viscosity coefficient under standard conditions.

The figure 7 shows the correspondence in numerical values of the dynamic viscosity coefficients obtained by NMR and on a viscometer under standard conditions.

The figure 8 shows the distribution of the viscosity of heavy oil along the trunks of pairs of horizontal wells in the area of heavy oil deposits.

The figure 9 shows an example of a comparison of the main indicators of the development of deposits by PGD technology, taking into account information about the heterogeneity of the distribution of oil viscosity in the reservoir and without it: a) fluid production; b) steam injection; c) oil production.

The figure 10 shows a comparison of the main indicators of the development of deposits by PGD technology, taking into account information about the heterogeneity of the distribution of oil viscosity over the reservoir and without it (cumulative production, steam-oil ratio, water-oil factor).

The implementation of the invention

Based on the systematic selection of representative samples of super-viscous and heavy oil from oil-saturated core taken from the depth of interest, high-speed centrifugation followed by measurement of the rheological (viscosity) and magnetic relaxation (NMR) properties of the oil samples extracted from the core, establish empirical relationships between the rheological properties and magnetic relaxation characteristics of oil, moreover, to estimate the oil yield index, the geometric mean dependence is used weighted time ki its spin-spin relaxation time on this indicator. When observing the spin-spin relaxation of T ₂ hydrogen nuclei in hydrocarbon chains, the relaxation time depends on the molecular mass of the molecule, on its formation, as well as on the restriction of motion (porous medium) and on the amount of paramagnetic impurities in contact with the hydrocarbon molecule. It has been established that the lithological type of the rock and the volume of its pore space with an increase in the viscosity of saturating oil according to a power law decrease their influence on the spread in the values of the spin-spin relaxation time, and, starting from the dynamic viscosity of oil equal to 400 mPa⋅s, the ratio maximum and minimum times of spin-spin relaxation tends to unity (Fig. 1).

It was established that the pore space of the reservoir rock and paramagnetic impurities in natural concentrations do not affect the spin-spin relaxation time in the case of heavy oil (Fig. 2). Thus, the distribution of spin-spin relaxation times of heavy oil reflects the distribution of the mobilities of its components. Based on the mobility distribution, in turn, fluid viscosity can be determined.

In one embodiment, a method for determining the viscosity of heavy oil based on measuring the spin-spin relaxation rate of representative representative samples of heavy oil by NMR is provided. The measurements are carried out on oil in free volume extracted from the core by the WCC method (high-speed centrifugation method), or by oil-saturated core. Based on the mathematical processing of the measured relaxation curve, a spectrum of relaxation times is obtained using the Tikhonov regularization method, after which the geometrically weighted average relaxation time is calculated in the spectral range 0.01 - 20 ms.

Based on the obtained value of the geometrically weighted average time of spin-spin relaxation using a mathematical transformation, the value of the coefficient of dynamic viscosity of heavy oil is calculated. The viscosity value is determined within the confidence interval of the established correlation - 20%.

In another embodiment, a method is proposed for determining the viscosity of heavy oil based on measurements from a well’s production interval of the spin-spin relaxation rate by an NMR tool. The measured decays of the spin-spin relaxation by the NMR logging tool are converted into spectra using the A.N. regularization method Tikhonov and determine the geometrically weighted average time of spin-spin relaxation as an integral NMR characteristic of the studied region. Based on the obtained value of the geometrically weighted average time of spin-spin relaxation, using a mathematical transformation to determine the viscosity of the fluid located in the measuring region of the NMR log with an accuracy exceeding 80%.

In one embodiment of the invention, when drilling an appraisal well in an unexplored superviscous or heavy oil field or in areas of a field not affected by in-situ combustion or chemical attack, core sampling is performed from the production interval, followed by its isolation to avoid losses of saturating fluid through evaporation (GOST 12071-2014 Soils. Sampling, packaging, transportation and storage of samples). Candidate wells are selected to conduct NMR studies and extract oil samples from the core, taking into account their location, which ensures maximum coverage of the dome area of the reservoir, but not less than 1 well per 5 km ² , and the thickness of the productive interval of at least 10 meters.

Core sampling for obtaining oil samples from oil-saturated core is carried out from the roof to the sole with an interval equal to 1/3 of the productive formation thickness in order to obtain a sample containing the widest range of viscosity values. Using high-speed centrifugation, we obtain at least 3 samples of heavy oil from the oil-saturated core of each candidate well, followed by dehydration, in the presence of mineralized formation water or an emulsion in the oil sample by thermodynamic separation. On a viscometer, we measure the values of the dynamic viscosity coefficients of the obtained samples of heavy oil under standard and thermobaric reservoir conditions corresponding to this particular field, and based on the measurements we establish an empirical relationship between them (Fig. 3).

For each sample, we measure one spin-spin (T ₂ ) relaxation time using the KPMG pulse sequence (Carr-Parsell-Maybum-Gill - a sequence of electromagnetic pulses (see Fig. 4)) to obtain a relaxation decay. A long series of RF pulses called the KPMG sequence, the key characteristic of which is the alternation of the polarity of the received signal to eliminate possible artifacts caused by the operation of electronic circuits. Further, a series of radio frequency pulses acts on the magnetic moments of hydrogen nuclei, causing them to rotate 90 °, and then precess around the direction of the constant magnetic field. In the interval between the sent pulses of the hydrogen nucleus, the chemical components of the formation fluids generate radio frequency echo signals received and measured by the antenna of the NMR device. The time interval between successive pulses when an echo appears is indicated by TE. During the measurement cycle using the KPMG sequence of the hydrogen nucleus of the chemicals that make up the formation fluids, detectable radio frequency echo signals are generated with a frequency equal to their excitation frequency. The amplitude of the echo signals decreases with an exponentially varying speed, characterized by relaxation times T ₂ depending on the size distribution of the pores, the properties of the formation fluids, the mineralogical composition of the formation, and molecular diffusion.

Preliminary processing of the experimental data includes the conversion of the attenuation curves of the longitudinal and transverse magnetizations to the spectrum of spin-spin (T ₂ ) and spin-lattice (T ₁ ) relaxation times, respectively.

The problem of finding the spectrum of relaxation times, based on the experimentally obtained relaxation damping curve, is inherently incorrectly posed and there are several known approximate methods for solving it. This means that there is no one correct solution. The most common solution is the inverse Laplace transform, however, for all its advantages, this algorithm has several disadvantages. Thus, low resolution in the time domain is critical in the analysis of heavy oil. As an alternative way to obtain the spectrum, we proposed to use the regularization algorithm, namely, A.N. Tikhonov. Finding a spectrum is, in fact, an equation of the following form:

The generalized algorithmic course of solving equation (1):

1. A⋅u = F

2. A ^* ⋅A⋅u = A ^* ⋅F

3.α⋅u + A ^* ⋅A⋅u = A ^* ⋅F

4. α⋅u _{n, α} + A ^* ⋅A⋅u _{n, α} = α⋅u _{n-1, α} + A ^* ⋅F

5. u _{n, α} = (A ^* ⋅A + α) ^-1 ⋅ (α⋅u _{n-1, α} + A ^* ⋅F)

A_n,m=||a_ij||_n,m=[exp(-T_i/T_2j)]_n×m where A is the matrix of coefficients a _ij , where

A _{n, m} = || a _ij || _{n, m} = [exp (-T _i / T _2j )] _{n × m}

A * is the complex conjugate matrix A; T _i - times of measuring the decay of the transverse magnetization; T _2j — spin-spin relaxation times in the spectrum measurement range; u is a one-dimensional matrix of populations of components with times T _2j ; F is a one-dimensional matrix of amplitude values f _i , where fi are the signal amplitudes at time T _i ; α is the regularization parameter; u _{n, α} is the matrix of amplitudes U at the n-th iteration step and the regularization parameter α, where n is the number of the iteration step.

The last stage is iterable and it is required to repeat it until the spectrum satisfies the requirements. Preliminary calculations indicate a higher resolution of this method for calculating the spectra of relaxation times (Fig. 5).

The spectrum of the spin-spin relaxation time is calculated using the A.N. regularization method Tikhonov. Exclude from further processing relaxation times of more than 20 ms. We calculate the geometric mean weighted relaxation time based on the formula:

where A _i is the amplitude of each time spectrum of the spin-back relaxation T _2i .

The coordination of the values of the dynamic viscosity coefficients determined by direct measurement on a viscometer and by the method of nuclear magnetic resonance is achieved on the basis of the empirically obtained two-term formula (3), which relates the spin-spin relaxation time (T ₂ ) and the dynamic viscosity coefficient. This dependence generally looks like:

where T _2GM is the average geometrically weighted time of the spectrum of spin-spin relaxation, A and B are coefficients depending both on the properties of the physical system and on the parameters of the equipment.

The technical complexity of measuring the spin-spin relaxation of heavy oil is associated with a large volume fraction of components having relaxation times inaccessible to observation both on laboratory NMR relaxometers and on nuclear magnetic logging instruments. Thus, most of the important information about the real mobility distribution of the components of the heavy oil is lost. The higher the viscosity, the greater the proportion of unregistered components becomes, in order to compensate for the loss of useful information, an additional term is introduced into the empirical formula for determining viscosity based on NMR characteristics of heavy oil. Such an approach shows a high accuracy in determining viscosity both in laboratory NMR studies and when using downhole NMR equipment.

The first term in relation (3) is consistent with the existing empirical dependences for light oil in the general case, where coefficient A, in the general case, depends on temperature, echo signal time, imaginary hydrogen index, etc. The second term is corrective and allows you to compensate for the loss of the signal from the components having short relaxation times inaccessible to registration on the measuring device. Thus, coefficient B does not apply to the physical system, but to measurement parameters, such as the ringing time between pulses and the TE time.

One of the possible methods for determining the values of the coefficients A and B is their direct calibration when using heavy oil extracted from the core by the WCC method, taken from the field where viscosity will be determined by NMR. The calibration process includes the following steps:

• coring from appraisal wells of the field;

• extraction of heavy oil by the WCC method;

• measurement of spin-spin relaxation rate;

• direct measurement of dynamic viscosity coefficients on a viscometer under standard and natural conditions of occurrence of oil;

• processing the results of NMR measurements using the method presented above;

• minimizing the error in determining viscosity by clarifying the coefficients A and B;

• use of the obtained coefficients in further research at this field.

Taking into account the technical features of the GeoSpec2 device, formula (3) will take the following form:

moreover, the first term makes the dominant contribution to the value of the dynamic viscosity coefficient for the viscosity of heavy oil under standard conditions from 1000 to 10,000 mPa · s, and the second term makes more than 10,000 mPa · s (Fig. 6).

As a demonstration of the reliability of the method, a statistical analysis of the results was performed (Fig. 7).

Thus, the developed procedure for matching the values of dynamic viscosity coefficients obtained by various methods allows one to reliably determine the value of the viscosity of heavy oil based on NMR data.

After establishing the coefficients A and B of formula (3) for the studied field under standard conditions, laboratory measurements of the magnetic relaxation properties (spin-spin relaxation times T ₂ ) of core samples taken in the context of the productive interval of appraisal wells with an interval of 0.5 m are carried out. NMR studies select candidate wells, preferably located on the design trajectory of production wells, taking into account their location, ensuring maximum coverage of the reservoir area, and not its 1 well per 3 km ² .

Based on the initial data processing method described above, we obtain the distribution of oil viscosity over the well section under standard conditions. Based on the obtained empirical dependence of the dynamic viscosity coefficients of heavy oil under standard and thermobaric reservoir conditions, we convert the viscosity values obtained on the basis of NMR data to the viscosity values under thermobaric conditions of oil occurrence.

In another embodiment of the invention, NMR measurements are performed in an open wellbore using downhole logging devices lowered on the cable. Based on laboratory studies, we carry out the adjustment of logging tools and determine the coefficients A and B for the studied field and the NMR tool used.

It is recommended to use a logging tool whose TE time does not exceed 1 ms, a higher TE time can introduce large errors in the measurement of spin-spin relaxation of heavy oil due to the impossibility of observing high molecular weight components, the proportion of which can exceed 30%. It is shown that using the proposed method of interpretation and processing of experimental data, it is possible to compensate for the loss of a useful signal from rapidly relaxing components with a fraction of no more than 30%. The most reliable data are obtained using NMR tools with a focused magnetic field with a vertical resolution of not more than 10 cm. Processing a set of relaxation attenuation curves does not differ from processing in the case of measurements on a laboratory setup.

To achieve a technical result, it is necessary to take into account in the geological and hydrodynamic model the heterogeneity of the viscosity distribution over the reservoir.

An example of a specific implementation.

The studied wells are fully processed core data and well logging. For each layer characterized as a reservoir rock, “continuous” values of porosity and oil saturation are obtained. Additionally, according to NMR studies (NMR logging, laboratory NMR studies), stable viscosity dependences on depth are obtained. A geological 3D model is created; it is created in a licensed software package for geological modeling. The first step in creating a 3D model is to create a structural framework for the model. The roof of the structural frame is the roof of the collector of the developed horizon. The bottom of the structural frame is the sole of the reservoir of the developed horizon. The grid of the 3D model is uniform with a step in X and Y of 25 m. The sampling step in Z is 0.2 m.

To create 3D model parameters (porosity, oil saturation, viscosity), well data is transferred to a 3D grid. The main direction and parameters of the formation anisotropy are accepted. Set the interpolation radius along the anisotropy axis and the interpolation radius perpendicular to the anisotropy axis. These anisotropy parameters are taken based on the shape and structure of the elevation under consideration. According to RD 153-39.0-047-00, the difference between the parameters of the 3D model and the initial received data should not exceed 5%. To calculate the viscosity by cube, a module with the same anisotropy parameters is used as for the calculation of oil saturation and porosity. According to the obtained cube of viscosity, the dependence of viscosity on depth is visually established.

In order to justify the necessity and relevance of obtaining information on viscosity, the constructed geological model of a super-viscous oil deposit is used, taking into account the heterogeneity of the viscosity of oil in all directions. The figure (Fig. 8) shows the characteristic features of the distribution of oil viscosity in the reservoir, as well as the profiles of pairs of wells located in the center of the reservoir and at the periphery. It is shown that the viscosity of oil depends not only on the absolute elevation of the viscosity measurement, but can also vary along the strike of the formation. In the western part of the reservoir, a higher viscosity of oil is observed compared with the eastern side.

Hydrodynamic modeling of two versions of the models is carried out: in the first, the viscosity is set by a single value over the entire volume of the reservoir; the second model uses data on the heterogeneity of the distribution of viscosity. Then, without changing the position and number of injection and production wells, the optimal steam injection cycle and annual oil production from the studied area of the reservoir were calculated.

The graphs (Fig. 9) show an example of the dynamics of reservoir development parameters, there is a significant difference both in the optimal steam injection program and in oil production over 8 years. The calculation of cumulative oil production, steam-oil ratio and water-oil factor for the entire period of development of the studied section of the heavy oil deposit was also made. As can be seen from the graph (Fig. 10), the use of information on the distribution of viscosity across the reservoir will allow us to clarify the water-oil factor and the oil-vapor ratio throughout the entire development period, and also explains the acceleration of oil production by 20% compared with the forecast model using the average viscosity over the site deposits. In the future, taking into account information on the heterogeneity of the distribution of viscosity will optimize the arrangement of the field by arranging a smaller number of pairs of wells at distances that make it possible to carry out the selection evenly across all the wells of the facility.

Claims (5)

1. A method for determining the viscosity of oil, including obtaining representative samples of heavy oil by high-speed centrifugation from an oil-saturated core taken from a depth of interest, subsequent measurement of the spin-spin relaxation rate of obtained representative samples of heavy oil by pulsed nuclear magnetic resonance using the Carr-Parcell-Maybum sequence -Gilla on laboratory equipment, followed by processing the amplitude-relaxation characteristics, including Acquiring spectra of spin-spin relaxation using Tikhonov regularization, the definition of weighted geometric mean spectrum of the relaxation time in the range of 0.01-20 ms, using the obtained weighted geometric mean relaxation time in the ratio to calculate the viscosity with known constant coefficients. 2. The method according to claim 1, characterized in that the measurement of the spin-spin relaxation rate on laboratory equipment is carried out on a naturally-saturated core selected from a depth of interest. 3. The method according to p. 1-2, characterized in that the spin-spin relaxation rate is measured using a sequence of 180-KPMG RF pulses and the subsequent construction of a two-dimensional map of T ₁ -T ₂ with the allocation of a data array of the spin-spin relaxation spectrum. 4. A method for determining oil viscosity, which includes continuous measurement of the spin-spin relaxation rate from a well’s production interval using a nuclear magnetic logging tool, followed by processing of the amplitude-relaxation characteristics, including obtaining a spectrum of spin-spin relaxation times using the Tikhonov regularization method, determining the geometrically weighted average time relaxation in the spectral interval 0.01–20 ms, using the obtained average geometrically weighted relaxation time in a ratio for calculating viscosity with predetermined correlation coefficients. 5. The method according to p. 4, characterized in that the nuclear magnetic logging tool in the area surrounding the underground well, measures a sequence of radio frequency pulses of 180 ° KPMG, followed by the construction of a two-dimensional map T1-T2 at depth intervals of interest to determine the type of fluid and the degree of mobility of its components based on the established correlation dependences of the ratio T1 / T2 on the viscosity of heavy oil.

Images