RU2453831C2 - Method for conducting petrophysical investigations on large-diameter rock samples in field conditions - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: analysed sample is put into an inductance coil; magnetisation caused by magnetic moments of hydrogen nuclei of rock pore fluid is created; by passing polarisation current through the coil, free precession measuring signals of said megnetisation in the Earth's magnetic field are excited; measurement results are stored and signal parameters are determined, wherein the optimum value of the reception frequency band is selected based on the condition: ΔF≥2/0.14πT_2min, where T_2min is the least time constant of the free precession measuring signals; digital recording of wave forms of the measuring signals is carried out and the obtained information is processed in the frequency domain after optimum digital filtration; for each of the stored signals of the pore fluid of the analysed rock sample, phase shift is performed, which turns the signal phase into the initial phase of the first of the stored signals, wherein the value of the phase shift is determined by measuring the phase and frequency of the signal from a sample of perfluorinated oil additionally placed in the inductance coil and multiplying the phase shift of signals from the perfluorinated oil by the ratio of values of the gyromagnetic ratios of fluorine and hydrogen nuclei.

EFFECT: high accuracy of estimating free fluid index values of reservoir rocks and ensuring rapid analysis of core samples with diameter of up to 120 mm or higher in field conditions.

2 cl

Description

This invention relates to the field of petrophysical studies of rocks using large diameter core samples conducted in the field when studying oil and gas fields using the method of nuclear magnetic resonance (NMR) in the Earth’s magnetic field.

The invention is intended for use in the search, exploration and development of oil and gas deposits, from which, in order to increase the geological representativeness and reliability of determining the petrophysical characteristics of rocks, core samples with a diameter of 80-120 mm are selected.

Known / 1 / various methods of petrophysical studies of rocks by core. These methods are characterized by duration, subjectivity, complexity and are carried out in laboratory conditions. As a result of this, the well-known methods provide belatedly necessary for solving geological and geophysical problems petrophysical information. In addition, the long time between coring and its investigation leads to unreliable results, since the physicochemical properties of the rocks have time to change significantly.

Known / 2 / method for NMR studies in an artificial magnetic field, allowing express analysis of rock samples. However, in practice, this method allows you to explore rock samples with a diameter of not more than 40 mm.

The closest analogue (prototype) of the proposed method is an NMR method for studying rock samples in the Earth’s magnetic field / 3 /, which uses the accumulation method in combination with integration for the purpose of recording signals. The high uniformity of the earth's field allows us to study rock samples of almost any diameter. In the known method, the test rock sample is placed in an inductor, with which the sample is pre-magnetized, and then a free precession (SP) signal of the hydrogen nuclei of the pore liquid of the test rock sample is excited and received with a narrow receiving strip. The measured signal is amplified, amplitude detection is carried out and integrated in several time windows. The amplitudes of the signals are accumulated by averaging a fixed number of measurements. For more efficient accumulation, optimal integration times are selected, which depend on the value of the signal attenuation constant. Using the amplitude values thus obtained, the envelope of the SP signal is extrapolated to the moment of its excitation and the initial amplitude is determined, which is proportional to the main parameter being determined - the free fluid index (ISF). ISF is a measure of the amount of fluid that can move freely in the pore space of the rock, that is, the effective porosity of the rock.

This method has several disadvantages:

1. The use of a narrow receiving frequency band causes a large "dead time" in the measurements and distortion of the spectral characteristics of signals of short duration.

раз меньше значения, потенциально достижимого с помощью когерентного детектора.2. With the initial signal-to-noise ratio ≥3, the signal-to-noise ratio at the output of the amplitude detector is

times less than the value potentially achievable with a coherent detector.

3. When the signal to noise ratio <3, the signal is suppressed by interference; in addition, the measurement error increases due to a shift in the mathematical expectation of the result.

4. The accumulation of the same number of measurements leads to the fact that the relative measurement error for samples of different porosity will be different, namely, it is approximately inversely proportional to the value of the ISF.

5. Loss of information about the frequency and phase of the measured signals does not allow the use of noise-tolerant processing in the frequency domain and coherent accumulation in order to increase the accuracy of estimates in conditions of low signal to noise ratio.

As a result of this, when studying reservoir rocks of a complex structure using this method, the error in determining the value of the ISF reaches 50% or more.

Thus, the known method does not provide sufficient accuracy in the determination of the parameters of the SP signals, especially for small signal-to-noise ratios.

In various fields of technology where it is necessary to determine the parameters of weak signals, for example, in NMR spectroscopy / 4 /, a method is known for coherently accumulating periodic or quasiperiodic signals with the same frequency to increase the signal-to-noise ratio. This method consists in phase shifting each of the received signals by a certain amount, so that after the phase shift the signals coincide or differ by an integer number of periods, and summing the signals thus corrected. However, the direct application of coherent accumulation to determine the parameters of weak SP signals in the earth's field is impossible due to the fact that in the known methods the initial phase of the measured signals varies randomly with each measurement cycle. Such random shifts of the initial phase of the measured signals, as practice shows, reach values of 180 °. The distribution law of this random variable is unknown, therefore, with the number of accumulations up to 5, the relative errors in determining the signal parameters can exceed 50%. Random changes in the initial phase are caused by the combined action of several physical and technical reasons, among which one can indicate the instability of the characteristics of switching devices, excitation and receiving circuits, etc. The exclusion of the influence of these factors and the achievement of stability of the initial phase presents significant technical difficulties. Random shifts of the initial phase can be compensated by the phase shift of each of the accumulated signals, if the initial phase and frequency of each of these signals are known. However, in the case of weak SP signals, it is impossible to determine their phase and frequency with sufficient accuracy.

The purpose of the present invention is to improve the accuracy of estimating ISF values of reservoir rocks and to provide operational research of core samples with a diameter of up to 120 mm and more in the field. To achieve this goal, the proposed method uses:

1. A fundamentally new measurement method based on digital recording of waveforms of SP signals in the Earth’s magnetic field.

, где

- величина постоянной времени затухания сигнала СП. Из опыта ЯМР исследований образцов пород из разных месторождений страны было найдено минимальное значение

. В соответствии с этим была выбрана ширина приемной полосы частот, равная 340 Гц.2. The optimal width of the receiving frequency band, providing the maximum signal to noise ratio and eliminating distortion of the frequency spectra of the measured signals. The optimal width of the receiving frequency band was selected from the condition:

where

- the constant value of the decay time of the signal SP. The minimum value was found from the experience of NMR studies of rock samples from different fields of the country

. In accordance with this, a width of the receiving frequency band of 340 Hz was selected.

3. A sample of perfluorinated oil as a reference signal source.

4. A coherent method of accumulation of measured SP signals, in which a phase shift is carried out for each accumulated pore fluid signal of the studied rock samples, leading the signal phase to the initial phase of the first of the accumulated signals. The magnitude of the phase shift is determined by measuring the phase and frequency of the signal from the perfluorinated oil and multiplying the phase shift of the signal from the perfluorinated oil by the ratio of the gyromagnetic ratios of the fluorine and hydrogen nuclei. The required number of repeated measurements is determined on the basis of a predetermined allowable relative measurement error ε and the value of ISF ₁ measured in the first measurement cycle in accordance with the formula

where α is the coefficient determined during the calibration of the equipment.

5. Processing of digital measurement information in the frequency domain in combination with optimal digital filtering.

6. Automation of core research and data processing in real time.

7. Conducting studies of rock samples immediately after selection in a state of natural saturation.

The essence of the proposed method is as follows.

A test rock sample and a sample of perfluorinated oil are placed in an inductor. In this case, the following operations are performed cyclically. A polarization current is passed through the coil, with the help of which magnetization is created, due to the magnetic moments of the hydrogen nuclei of the pore liquid and the fluorine nuclei of the perfluorinated liquid. By quickly turning off the polarization current, the signal to be measured from the pore liquid and the signal from the perfluorinated liquid are simultaneously excited, the frequencies of these signals being different due to the difference in the gyromagnetic ratios of the hydrogen and fluorine nuclei, and the initial phases of these signals coincide due to the simultaneous excitation.

The measured signals after the necessary amplification without detection are converted with a sampling time Δt into a digital sequence U _m (t _m ), where t _m = t _m + mT / N, t _m is dead time, T is the recording time, N = T / Δt is the number of recorded points of the signal, m = 0, 1, 2, ..., N-1.

The sampling time Δt is selected from the Nyquist condition: Δt <1/2 / f _max , where f _max is the upper boundary of the frequency band of the measured signals.

As a result of the fast Fourier transform (FFT), a discrete spectrum of the original signal is obtained with a frequency step of 1 / T. The spectrum coefficients of the source signal are determined by the well-known / 4 / expressions

1≤k≤2N,

and the original signal itself is represented as a polynomial

Using a comb of spectral components of the signal constructed with a step of 1 / T Hz, after optimal digital filtering, Lorentz spectral curves of the measured and reference signal of the form are obtained

where Δω = ω-ω ₀ . The signal parameters are obtained from the corresponding spectra using the following expressions

where Δf _0.7 is the width of the spectral line at a level of 0.707 from | G _max |.

, представлено в таблице 1.Comparison of the signal-to-noise ratio obtained by a single measurement of rock samples with ISF = 10% using the prototype and the proposed method for different values

are presented in table 1.

Table 1 Comparison of the signal-to-noise ratio for the prototype and the proposed method

, мс

, ms twenty thirty 40 fifty 60 80 one hundred 150 200 Signal to noise ratio for prototype 1,2 2.9 5,6 7.8 10,4 15.1 18.9 27.5 34.8 The signal-to-noise ratio for the proposed method 8.3 12,2 15.4 17.9 20.8 24.9 28.1 36,4 42.6

. Для наиболее характерного диапазона значений

это превышение составляет (200-400)%. В результате этого точность определения ИСФ исследуемых образцов пород с помощью заявляемого способа повышается в 2-4 раза по сравнению с прототипом.These data show that the signal-to-noise ratio for the proposed method exceeds the corresponding signal-to-noise values for the prototype in the entire range of values

. For the most characteristic range of values

this excess is (200-400)%. As a result of this, the accuracy of determination of the ISF of the studied rock samples using the proposed method is increased by 2-4 times in comparison with the prototype.

In the case of small values of the signal-to-noise ratio, which are observed when measuring rock samples with ISF = (1-5)%, a coherent accumulation of the measured signals is performed. Since the initial phase of the measured signals from the pore liquid of the rock sample under study varies randomly with each measurement cycle, a phase shift is carried out for each of the accumulated signals, bringing the signal phase to the initial phase of the first of the accumulated signals, which is necessary for the coherent summation of signals. The magnitude of this shift on a time scale is determined by the known relation:

Δt = -Δφ / 2πf, where Δφ is the phase shift in radians, f is the signal frequency in Hz. In the considered case of accumulation of signals from hydrogen nuclei of the pore fluid, the phase shift for the i-th accumulated signal is determined by the formula

_{Δt i = - (φ 0Hi -φ} 0H1) / 2πf H, where φ _0Hi - initial phase i-th signal hydrogen, Δt _i - shift time scale i-th signal combination needed for the accumulated signals in phase with the first signal , f _H is the frequency of the signal of free hydrogen precession. In the case of weak signals, it is impossible to determine their frequency and initial phase with sufficient accuracy, which is necessary for coherent summation. At the same time, the signal from fluorine nuclei of perfluorinated oil has a sufficiently high intensity due to the high concentration of fluorine nuclei. Therefore, the phase shift required for the coherent summation of signals from the pore fluid of the test rock sample is determined by measuring the frequency and initial phase of the signal from perfluorinated oil, considered as a reference signal. Recalculation of the phase shift of the reference signal by the phase shift of the signal from the pore fluid of the test sample is carried out by multiplying the phase shift of the reference signal by the ratio of the gyromagnetic ratios of fluorine nuclei and hydrogen. The last operation is based on a well-known physical fact: the frequencies of the SP signals for the nuclei of various elements in the same magnetic field are referred to as the magnitudes of the gyromagnetic ratios of these nuclei. The procedure for measuring the frequency and initial phase of signals from perfluorinated oil and conversion is as follows.

For a free precession signal of the form U ₀ exp (-βt) cos (φ ₀ + 2πf ₀ t), where U ₀ is the initial amplitude, β is the signal attenuation rate, φ ₀ is the initial phase, f ₀ is the center frequency, and the signal amplitude spectrum has a maximum at a frequency equal to the center frequency of the signal. The value of the phase-frequency characteristic at this frequency, taking into account the condition β << f ₀ is

Since the initial phases of both signals coincide, and the frequencies are rigidly connected through a constant coefficient of 0.9409, the phase shift for the i-th accumulated signal is determined by the formula

The required number of repeated measurements used for coherent accumulation of signals is determined on the basis of a given permissible relative error in measuring signals from the studied rock samples ε. The value of this error should not be exceeded irrespective of the value of the ISF of the studied rock sample. For this, after the first measurement of the ISF of the studied rock sample, the value αISF ₁ ε is compared with 1 and if the indicated value is less than 1, the number of repeated measurements is calculated by the formula

Rounding to the integer value of N is taken in excess. The use of this formula is based on the following. The value of the ISF is proportional to the amplitude of the signal SP. Therefore, with the stationary noise of the equipment, there is a proportionality between the signal-to-noise ratio and the value of the IFF. The proportionality coefficient α is established as a result of the calibration of the equipment. In turn, the relative measurement error is equal to the noise / signal ratio, which, with coherent accumulation, decreases inversely with the square root of the number of accumulations.

After performing a certain number of repeated measurements, they are coherently summed and the resulting signal is scaled according to the number of repeated measurements.

The proposed method, using coherent accumulation of weak signals, provides improved measurement accuracy compared to the known method by reducing the relative measurement error for the same initial signal-to-noise ratio and the same number of accumulations. Thanks to the adaptive control of the number of repeated measurements, depending on the measured value of the ISF, a relative measurement error is ensured for all studied rock samples not higher than the specified permissible value.

The error in determining the amplitude of the signal is calculated by the formula

where Δ ₁ is the standard deviation, Δ ₂ is the bias of the mathematical expectation.

, а

, где σ² - дисперсия шума, u - амплитуда сигнала. Для предлагаемого способа

, а

.For prototype

, but

where σ ² is the noise variance, u is the signal amplitude. For the proposed method

, but

.

раз, а смещение математического ожидания в N раз. Следовательно, ошибка измерений составит:

. Таким образом, относительная погрешность измерений, равная ε=Δ/u, при накоплении N измерений составит:With the accumulation of N measurements, the standard deviation increases in

times, and the shift in mathematical expectation is N times. Therefore, the measurement error will be:

. Thus, the relative measurement error equal to ε = Δ / u, with the accumulation of N measurements will be:

for prototype

for the proposed method

From the above formulas it is seen that the relative error for the proposed method is always less, as shown by the data presented in table 2.

Feasibility study of the application of this invention is to increase the accuracy, efficiency and objectivity of determining the values of the ISF of reservoir rocks of oil and gas and to provide for the study of large diameter core in the field with a significant reduction in labor and time.

Information sources

1. Dobrynin V.M., Wendelstein B.Yu., Kozhevnikov D.A. Petrophysics. Moscow. "Bosom". 1991 year

2. USSR author's certificate No. 1073654 of 02.15.84. Bulletin of inventions No. 6. Cl. G01N 24/08.

3. Relaxometer - equipment for the study of rock samples by free precession. Orlov G.L., Danevich V.I., Mazniashvili G.B., Sadikhov D.M. "Izv. universities. Ser. Oil and gas ”, 1972, No. 9, pp. 91-93.

4. Farrar T.S., Becker E.D. Pulse and Fourier transform NMR spectroscopy. M., "World", 1973.

What is claimed:

1. NMR method for studying full-sized samples of large-diameter rocks with natural saturation in the field, based on digital measurement of waveforms of free precession waveforms of hydrogen nuclei of pore rock fluid in the Earth’s magnetic field, which improves the accuracy, speed and objectivity of determining the free fluid index (ISF ), which consists in the fact that:

a) a test rock sample and a sample of perfluorinated oil are placed in an inductor;

b) a polarizing constant magnetic field is applied to the test rock sample and the perfluorinated oil sample by passing a direct current through an inductor;

c) excite SP signals from hydrogen nuclei of the pore fluid of the test rock sample and from fluorine nuclei of perfluorinated oil by quickly turning off the polarization field;

d) select the sampling time Δt from the Nyquist condition: Δt <1 / 2f _max , where f _max is the upper boundary of the frequency band of the measured signals;

d) transform the measured signals after necessary amplification without detection into a digital sequence U _m (t _m ) with a sampling time Δt, where t _m = t _m + mT / N, t _m is dead time, T is the recording time, N = T / Δt is the number of recorded signal points, m = 0, 1, 2, ..., N-1.

e) determine the discrete spectrum of the source signal in steps of 1 / T in frequency using the fast Fourier transform (FFT);

g) carry out optimal digital filtering using the constructed comb of spectral components of the signal;

h) determine the parameters of the signal from the hydrogen nuclei of the pore fluid of the test rock sample from its frequency spectrum using the following expressions:

,

,

,

,

- постоянная времени затухания сигнала,

- ширина спектральной линии по уровню 0,707 от |G_max|;Where

- signal decay time constant,

- the width of the spectral line at a level of 0.707 from | G _max |;

i) determine the value of the ISF of the studied rock sample by the formula: ISF ₁ = U ₀ K, where K is the calibration coefficient;

j) calculate the value αISF ₁ ε, where ε is the specified permissible relative measurement error, α is the coefficient determined during the calibration of the equipment, and compared with 1;

k) determine the required number of repeated measurements when αISF ₁ ε <1, according to the formula: N = (αISF ₁ ε) ^-2 -1, rounding to the integer value of N is taken in excess;

m) perform N repeated measurements;

m) determine the initial phase of the reference signals from the perfluorinated sample according to their phase-frequency characteristic at the center frequency of the spectrum of fluorine nuclei: φ _0F = φ (f ₀ );

n) carry out phase shifts of the accumulated signals from the pore fluid of the test rock sample by Δt values determined by the formula:

Δt _i = - (φ _0Fi -φ _0F1 ) 0.9409 / 2πf _F ;

o) produce a coherent summation of the accumulated signals and scale the resulting signal according to the number of repeated measurements;

p) determine the ISF value of the studied rock sample according to the results of coherent signal accumulation: ISF = U _0p K, where U _0p is the amplitude of the resulting signal.

2. The method according to claim 1, characterized in that for the measurement of the signals of the SP apply the optimal width of the receiving frequency band equal to 340 Hz.

3. The method according to claim 2, characterized in that for digital recording of waveforms of the measured SP signals, a sampling time of 80 μs is selected.

Claims (2)

1. NMR method for studying full-sized large-diameter rock samples with natural saturation in the field, including placing the test sample in an inductor, creating magnetization due to the magnetic moments of the hydrogen nuclei of the pore rock fluid, by passing a polarization current through the coil, exciting the measured free precession signals this magnetization in the magnetic field of the Earth, the accumulation of measurement results and determination of signal parameters, characterized in that, in order to increase the measurement accuracy, the optimum value is selected receiving frequency band from the condition:

where

- the smallest value of the decay time constant of the measured SP signals, digitally register the waveforms of the measured signals and process the received information in the frequency domain after optimal digital filtering, for each of the accumulated signals of the pore fluid of the test rock sample, a phase shift is carried out, leading the signal phase to the initial phase the first of the accumulated signals, and the magnitude of the phase shift is determined by measuring the phase and frequency of the signal from by thallium in inductance of the sample and multiplying the perfluorinated oil phase shift signals from the perfluorinated oil at the ratio of the gyromagnetic ratios of the nuclei of hydrogen and fluorine. 2. The NMR research method according to claim 1, characterized in that the number of repeated measurements required for accumulation is determined based on a predetermined allowable relative measurement error ε and measured in the first measurement cycle of the free fluid index ISF ₁ according to the formula

where α is the coefficient determined during the calibration of the equipment.