CN100543496C - Pulse neutron bispectrum saturation logging method - Google Patents

Abstract

The invention relates to a pulse neutron dual-spectrum saturation logging method, comprising the following steps: 1) collecting logging data, 2) performing depth-time two-dimensional filtering on the thermal neutron time spectrum and the captured gamma time spectrum, and performing two-dimensional filtering on the captured gamma Depth-energy two-dimensional filtering of horse energy spectrum and natural gamma energy spectrum, 3) normalize the filtered data, 4) multi-scale decomposition of the normalized data, 5) decompose 6) display well logging images and well logging curves, and 7) perform geological interpretation; with one downhole, thermal neutron time spectrum and captured gamma time spectrum can be collected simultaneously under the same environment, so that the two The advantages of the neutron lifetime logging method are complementary, and the scope of adaptation to the environment is expanded. The comprehensive processing of the bispectrum can obtain more accurate calculation results and improve the ability to identify oil, gas and water layers.

Description

The two spectrum of pulsed neutron saturation logging method

Technical field

The present invention relates to a kind of pulsed neutron life-span logging method.

Background technology

Neutron life time log, be in well, to be the neutron beam of 14MeV by the sequential emitted energy of setting with pulsed neutron generator, the high fast neutron of energy loses portion of energy and is converted into the neutron of medium energy and follows non-bullet gamma radiation through inelastic scattering earlier, change thermal neutron into through the further slowing down of elastic scattering again, thermal neutron is captured gradually and produces captures gamma radiation.The mean lifetime τ in the stratum is relevant with formation lithology, reservoir water salinity, factor of porosity and water saturation for thermal neutron.When stratum lithology, reservoir water salinity and factor of porosity are known maybe can survey the time, measure oil, gas, water saturation that neutron lifetime can be obtained the reservoir.

The main method of current measurement stratum neutron lifetime is from the suitable distance of neutron source one or two gamma detector to be installed, and the record capture gamma-ray is at the counting in each time road, i.e. gamma time spectrum.Count the macroscopic capture cross section ∑ that time dependent slope is obtained the neutron lifetime τ on stratum or is inversely proportional to it by the time road.The another kind of method of measuring the stratum neutron lifetime is to measure thermal neutron gate time spectrum with neutron detector, is asked the thermal neutron lifetime or the macroscopic capture cross section on stratum by time road counting attenuation rate.

Summary of the invention

The present invention proposes the two spectrum of a kind of pulsed neutron saturation logging method, with one being set with thermal neutron detector, capturing the downhole probe of gamma detector and natural gamma detector, synchronously or select to measure thermal neutron time spectrum, capture the gamma time spectrum, capture gamma spectra and natural gamma spectra.Wherein two spectrums are meant thermal neutron time spectrum and capture the gamma time spectrum that constitute the core information of this method, other is a supplementary.

Thermal neutron time spectrum only reflects fast neutron slowing down and the captive process of thermal neutron, do not comprise and natural gamma, neutron activation and capture the yield of gamma and the information of gamma ray and surrounding medium interaction process, and capture all information that the gamma time spectrum comprises neutron and gamma and surrounding medium interaction overall process.The information of two class data carry is had complementary advantages on preferred yardstick, comprehensive description is done on the stratum, can more fully reflect the characteristic on stratum, more accurately identification oil, gas, water layer.

The two spectrum of a kind of pulsed neutron saturation logging method, its special character is that it may further comprise the steps:

1] gather log data:

In cased well, use the two spectrum of pulsed neutron saturation degree well logger continuous acquisition log datas; Described log data comprises thermal neutron time spectrum, captures the gamma time spectrum, the thermal neutron gross-count rate, capture gamma gross-count rate, natural gamma gross-count rate.

2] to thermal neutron time spectrum, capture the gamma time spectrum and carry out the degree of depth-time 2-D filtering, do the degree of depth-energy two-dimensional filtering to capturing gamma spectra and natural gamma spectra:

The described degree of depth-time 2-D filtering comprises vertical filtering of Depth Domain and in the horizontal filtering of time domain; The described degree of depth-energy two-dimensional filtering comprises vertical filtering of Depth Domain and in the horizontal filtering in energy territory; The method of described vertical filtering and laterally filtering comprises Kalman filtering, multiple spot smothing filtering or multi-scale filtering; Obtain a pair of smooth multi index option curve of decay in time in each depth point after the filtering, in the degree of depth-time domain, obtain two two-dimensional arrays, constitute the two frame of digital images of reflection formation properties with change in depth; And in the degree of depth-energy territory, obtain two two-dimensional arrays, constitute the two frame of digital images of reflection formation properties with change in depth;

3] filtered data are carried out normalized:

The step of described normalized comprises the normalization log value that calculates each sampled point:

$f\′=\frac{f-f_{\min }}{f_{\max }-f_{\min }}\×100$

F in the formula _MinBe the minimum value of logging trace, f _MaxBe the maximal value of logging trace, f is the sampling number certificate of logging trace;

4] data after the normalized are carried out multiple dimensioned decomposition:

The method of the multiple dimensioned decomposition of described horizontal time shaft time spectrum comprises multi index option fitting process and wavelet decomposition method, and the method for the multiple dimensioned decomposition of described vertical degree of depth axle mainly refers to the wavelet decomposition method;

5] decomposed data is reconstructed and merges:

Described log data is reconstructed with merging is included in reconstruct and the fusion of carrying out time spectrum, life spectrum, ∑ spectrum, power spectrum and similar and non-similar logging trace on raw data, parameter, three levels of decision-making respectively, and concrete steps comprise:

5.1] based on the very big restructing algorithm of little mode, select low frequency coefficient weighting, the high frequency coefficient bigger fusion rule that takes absolute value for use, similar or multiclass log data are carried out wavelet reconstruction obtain merging wavelet pyramid;

5.2] wavelet pyramid after merging is carried out wavelet inverse transformation realization data reconstruction at all levels;

5.3] adopt entropy, average and variance as judgment criteria, fused data is carried out quantitative evaluation: average, variance data reflection peak information; Information entropy data reflection spatial detail information; Y-PSNR, related coefficient reflection spectrum information;

6] show log picture and logging trace

Described log picture and logging trace comprise with at least one group in hypograph and the curve:

The degree of depth-time coloured image: on each depth point, get a level and smooth multi index option die-away curve after the filtering, in the digital picture that obtains on the tested section on the width of cloth degree of depth-time 2-D plane; This array is carried out visualization processing,, obtain the coloured image of width of cloth reflection depth of stratum-temporal characteristics with look post demarcation signal amplitude;

The degree of depth-the life-span or the degree of depth-∑ coloured image: neutron lifetime τ and thermal neutron macroscopic capture cross section ∑ are inversely proportional to, and τ distributes and is easy to change into the ∑ distribution; Through time-life-span or time-conversion in territory, cross section, on each sampled point, all obtain a τ or ∑ distribution curve, in the digital picture that obtains on the tested section on a width of cloth degree of depth-life-span or the degree of depth-∑ two dimensional surface; With look post demarcation signal amplitude, obtain τ or the ∑ distribution coloured image of width of cloth reflection formation characteristics with change in depth;

Described logging trace comprises: thermal neutron and capture gamma gross-count rate, characteristic time door and can window counting rate and ratio, thermal neutron lifetime and macroscopic capture cross section; Selected through type well check, by multiple dimensioned decomposition obtain to oil saturation or the good branch discharge curve of lithology resolution characteristic; Selected through type well check, by multiple dimensioned data rebuild and merge obtain to oil saturation or good reconstruction and the blend curve of lithology resolution characteristic; Uncased hole and real-time oil, gas, water saturation curve;

7] carry out geologic interpretation:

Described step of carrying out geologic interpretation comprises at least a with in hypograph and the curve:

The degree of depth-time spectrum: neutron and capture gamma counting rate in time die-away curve on different depth point and can be divided into well, excessive, three periods of stratum, the high more then thermal neutron lifetime of the attenuation rate of stratomere counting rate is short more, and oil saturation is low more;

Utilize the foundation of life spectrum or ∑ spectrum as differentiation low mineralization water layer and remaining oil gas-bearing formation: life-span τ distributes more can reflect the overall picture on stratum than single τ value, asks the stratum neutron lifetime more can reflect the statistical property of stratum nuclear parameter with τ distribution characteristics peak or the selected interior mean value of gate-width Δ τ; With the general volume-based model of well logging interpretation, can obtain water saturation S by the ∑ distribution _WDistribution plan, water saturation S _WDistribution plan can more properly depict profit and distribute;

Described step 2] in carry out multi-scale filtering step comprise:

1] chooses Orthogonal Wavelets with certain tight supportive, symmetry and flatness;

2] select small echo and definite its to decompose level N, then data are carried out N layer wavelet decomposition, that is:

On certain yardstick i, to given nuclear logging burst $\begin{array}{cc}x(i,k)\∈V_i\⋐l^2\left(Z\right)& (k\∈Z)\end{array}$ ,, obtain the smooth signal of (low-frequency range) on the thick yardstick by the low-pass filter that an impulse response is h (k)

x _V(i-1，k)∈V _i-1

$x_V(i-1,l)=\underset{k}{\mathrm{\Σ}}h(2l-k)x(i,k)$

((i is that the Hi-pass filter of g (k) obtains the detail signal of (high band) on the thin yardstick by an impulse response k) to signal x by x for i, " detail signal " k) lost in low-pass filter

x _D(i-1，k)∈D _i-1

$x_D(i-1,l)=\underset{k}{\mathrm{\Σ}}g(2l-k)x(i,k)$

Subscript D represents x, and (i is k) at detail signal space D _I-1On projection;

3] carry out the multiple dimensioned decomposition that above-mentioned steps can realize the N layer one by one, obtain 2N different frequency band, comprise N high-frequency signal and N low frequency signal;

4] select a threshold value to carry out the soft-threshold quantification treatment to the 1st layer of each floor height frequency coefficient to the N layer; By the N layer low frequency coefficient of wavelet decomposition with through the 1st layer of high frequency coefficient after the quantification treatment to the N layer, carry out the reconstruct of log data, only comprise useful information in the reconstruction signal;

Described step 5] step of the very big restructing algorithm of medium and small mode comprises:

1] establishing the signal that logging instrumentation receives is f (x), and the Morlet continuous wavelet transform is defined as:

$W_f(a,b)=<f,{\mathrm{\&psi;}}_{\mathrm{ab}}>=\frac{1}{\sqrt{\left|a\right|}}\underset{-\&infin;}{\overset{+\&infin;}{\&Integral;}}f\left(x\right)\mathrm{\&psi;}\left(\frac{x-b}{a}\right)\mathrm{dx}$

In the formula: the Morlet wavelet basis function is

$\mathrm{\&psi;}\left(x\right)=e^{-{\mathrm{<figure-callout\; id="2"\; label="x"\; filenames="C200710018164E00201.png,C200710018164E00231.png"\; state="\{\{state\}\}">x</figure-callout>}}^2/2}\&CenterDot;e^{i{\mathrm{\&omega;}}_{0}x}$

A and b are respectively scale factor and shift factor;

2] signal is carried out wavelet analysis, can obtain the big coefficient of discharge of the assessing signal of different scale a, promptly corresponding wavelet coefficient W in the different spaces section _f(a, b); And represent with the mode of chromatogram, in chromatogram, represent the size of coefficient with change in color;

Wavelet coefficient curve when 3] obtaining a certain yardstick a, this curve can intuitively show the similarity degree between wavelet coefficient and the analyzed signal, and can make wavelet coefficient curve and some parameter have certain correlativity by the size of control a;

The two spectrum of described pulsed neutron saturation degree well logger comprises well logger housing 5 and the pulsed neutron generator in housing 1, thermal neutron detector 2, captures gamma detector 3 and natural gamma detector 4; Wherein the distance between thermal neutron detector 2 and the pulsed neutron generator 1 is 300-400mm; The distance of capturing between gamma detector 3 and the pulsed neutron generator 1 is 450-550mm; Distance between natural gamma detector 4 and the pulsed neutron generator 1 is 700-1000mm.

Above-mentioned log data also comprises captures gamma spectra, natural gamma spectra and well temperature, pressure and casing coupling data.

The two spectrum of above-mentioned pulsed neutron saturation degree well logger can be operated under NTS pattern, CTS pattern, three kinds of measurement patterns of DTS pattern, and described NTS pattern is only gathered thermal neutron time spectrum; Described CTS pattern-only gather and capture the gamma time spectrum; Described DTS pattern can be gathered thermal neutron time spectrum simultaneously and be captured the gamma time spectrum.

Distance between above-mentioned thermal neutron detector 2 and the pulsed neutron generator 1 is 350mm; The distance of capturing between gamma detector 3 and the pulsed neutron generator 1 is 500mm; Distance between natural gamma detector 4 and the pulsed neutron generator 1 is 900mm.

Above-mentioned thermal neutron detector 2 is ³The He counter tube; The described gamma detector 3 of capturing is scintillation spectrometer; Described natural gamma detector 4 is a scintillation spectrometer.

The two spectrum of above-mentioned pulsed neutron saturation degree well logger comprises thermometer, pressure gauge and casing collar locator (CCL).

The advantage of the inventive method is: once go into the well and can gather thermal neutron time spectrum simultaneously under equivalent environment, capture the gamma time spectrum, two kinds of neutron life time log methods are had complementary advantages, enlarged accommodation to environment, two spectrums are done overall treatment can obtain more precise calculation result, improved recognition capability oil-gas-water layer.

Description of drawings

Fig. 1 is the two spectrum of an a kind of pulsed neutron fluid saturation well logger structural representation of realizing the inventive method, wherein: the 1-pulsed neutron generator, the 2-thermal neutron detector, 3-captures gamma detector, 4-natural gamma detector, 5-well logger housing.

Fig. 2 is multiple dimensioned decomposition-fusion-reconstruction synoptic diagram, provides the decomposition of data yardstick, fusion and the process of reconstruction that obtain from two detectors.

Fig. 3 is thermal neutron counting rate chromatogram, neutron lifetime curve (τ) and the macroscopic capture cross section curve of slab bridge area, huge port producing well.Research well section is a 3050-3120m degree of depth section, and the down-hole is a steel sleeve, this well sandstone macroscopic capture cross section curve.Research well section is a 3050-3120m degree of depth section, and the down-hole is a steel sleeve, and this well sandstone water layer factor of porosity is 18-23%, and reservoir water salinity is 7.091g/l.Provide thermal neutron counting rate chromatogram among the figure, and neutron lifetime curve (τ) and macroscopic view capture capture cross-section curve (SIGMA), on behalf of the thermal neutron counting rate, the variation that color is turned left from the right side in the chromatogram be worth from the low value to the height.From chromatogram as can be seen, 3101-3106m well section, the decay in time of thermal neutron counting rate is the slowest, 3059-3061m well section and 3072-3074m well section, the thermal neutron decay is slowly.With GR among the figure (natural gamma) and the contrast of SP (spontaneous potential) curve, No. 23 floor and No. 25 floor are the reservoir as can be seen.No. 28 floor SP is reactionless, but GR is shown as low value.From chromatogram as can be seen, above-mentioned 3 stratum are the reservoir, and the thermal neutron counting rate is decayed slowly in time.And the mud stone layer is strong to the ability of capturing of thermal neutron, and the thermal neutron decay is fast, and Show Color changes soon on chromatogram, as 3098-3100m well section.In addition, compare, can judge the reservoir oiliness by chromatogram and SIGMA curve.Oil saturation is high more, and thermal neutron decay in time is slow more, and the thermal neutron counting rate that shows as the reflection formation characteristics on chromatogram extends back, and therefore judges that No. 28 floor oil saturation are higher, is consistent with the explanation conclusion.

Fig. 4 is the neutron lifetime spectrum that the huge port-Y well thermal neutron counting rate obtains after the multi index option match, and the degree of depth of sampled point is followed successively by 3103.5m, 3057m and 3088.8m.According to explaining conclusion, these depth points respectively with oil reservoir, the oil-containing water layer is corresponding with the mud stone layer, calculates formation macro capture cross-section with selected time gate and is followed successively by 10.29c.u., 19.8c.u. and 31c.u..When carrying out the multi index option match, access time, the road was 300-1170 μ s, reflected mainly that promptly the thermal neutron counting rate of formation information carries out match.The form of the thermal neutron lifetime spectrum that obtains on three sampled points is obviously different, and the neutron lifetime of the life spectrum main peak correspondence of depth point 3103.5m, 3057m and 3088.8m is respectively 316.2278 μ s, 227.5846 μ s and 163.7894 μ s; Corresponding formation macro capture cross-section is respectively 14.3884c.u., and 19.9926c.u. and 27.7796c.u. conform to substantially with the ∑ value of previous calculations gained.The thermal neutron lifetime of oil reservoir is the longest, and the thermal neutron lifetime of mud stone is the shortest, and the thermal neutron lifetime of oil-water common-layer is between between the two.This explanation can be differentiated mud stone layer, oil-water-layer and oil reservoir according to the neutron lifetime spectrum.

Embodiment

Core technology of the present invention is to gather thermal neutron and capture gamma dual-time spectrum in cased well with the two spectrum of pulsed neutron saturation degree well logger, obtains degree of accuracy high residue oil gas and water saturation through overall treatment.Specifically comprise:

1, data acquisition

Formulate arrangement and method for construction according to well, geologic condition and well logging purpose, select the well logging pattern.The two spectrum of pulsed neutron saturation degree well loggers can be selected three kinds of measurement patterns for use: NTS pattern-only gather thermal neutron time spectrum; CTS pattern-only gather and capture the gamma time spectrum; DTS pattern-gather thermal neutron time spectrum simultaneously and capture the gamma time spectrum, promptly two spectrum drainage patterns.With the two spectrum of pulsed neutron saturation degree well logger in cased well, gather thermal neutron time spectrum, capture the gamma time spectrum, capture gamma spectra, natural gamma spectra and auxiliary datas such as well temperature, pressure and casing coupling.Also can comprise the thermal neutron gross-count rate, capture the gamma gross-count rate, the natural gamma gross-count rate, downhole probe is on each depth point, after the neutron emission stops, with the helium of installing in the instrument string-3 thermal neutron detector, write down the thermal neutron in each time road by the sequential of setting and count, be i.e. the thermal neutron time spectrum.The road in time road is wide, and promptly the default value of sampling interval is 30 μ s, can regulate automatically or artificial the setting according to the thermal neutron lifetime variation range on tested section stratum.When gathering thermal neutron time spectrum, be installed in the flicker gamma detector in the same instrument string, capture the gamma time spectrum with identical sequential record.Thermal neutron time spectrum is identical with the sampling interval of capturing the gamma time spectrum, but the gamma detector spacing is long thicker in spatial domain mesoscale.That gathers simultaneously captures gamma spectra and natural gamma spectra, can be in order to dividing the stratum, discern lithology, to do the simple elements analysis and compare etc. with the open-hole log curve, and for the processing and the explanation of time spectrum provides supplementary.The two spectrum of pulsed neutron saturation degree well logger comprises well logger housing 5 shown in the accompanying drawing 1 and the pulsed neutron generator in housing 1, thermal neutron detector 2, captures gamma detector 3 and natural gamma detector 4.Wherein: the distance between thermal neutron detector 2 and the pulsed neutron generator 1 is 300-400mm; The distance of capturing between gamma detector 3 and the pulsed neutron generator 1 is 450-550mm; Distance between natural gamma detector 4 and the pulsed neutron generator 1 is 700-1000mm.Thermal neutron detector 2 is ³The He counter tube; Capture gamma detector 3 and be scintillation spectrometer; Natural gamma detector 4 is a scintillation spectrometer.Temperature, pressure gauge and casing collar locator (CCL) also have been installed in the instrument.

2, data filtering and conversion

Downhole probe moves in well, collects one group of thermal neutron time spectrum in each depth point and captures the gamma time spectrum.To whole measuring well section, each time spectrum all is a degree of depth-time bidimensional array, and each raw data that each element in the array promptly directly measures all comprises statistical error.The initial spectrum that collects will be through handling and conversion could be eliminated physics that error also therefrom obtains the stratum and preserve parameter.

(1) each time spectrum is done the degree of depth-time 2-D filtering, do energy-degree of depth two dimension-filtering capturing gamma spectra and natural gamma spectra.The method of filtering comprises Kalman filtering, multi-point fitting filtering, multiple spot smothing filtering or multi-scale filtering.

The initial statistical property of discrete Kalman's estimated state equation, detection equation and variable is respectively:

X(k+1)＝A(k+1，k)X(k)+G(k+1，k)U(k)+Γ(k+1，k)W(k) (7)

Z(k)＝C(k)X(k)+Y(k)+V(k) (8)

$\left\{\begin{array}{c}E\left[X\left(t_0\right)\right]=m_0\\ P\left(t_0\right)=E\left\{[X\left(t_0\right)-m_0]{[X\left(t_0\right)-m_0]}^T\right\}\end{array}\right.---\left(9\right)$

In the formula, X (k) is the n dimension state vector of control procedure, and U (k) is r dimension control vector, and W (k) is that average is zero P dimension white noise vector, and A (k) is n * n matrix, and B (k) is n * r matrix, and F (k) is n * p matrix, and Z (k) is m dimension matrix.

Kalman's estimation is exactly a criterion in accordance with regulations, from observation series Z (0), and Z (1) ..., Z (k) finds out the linear optimal of X (j) and estimates

Make valuation

And the variance of error is minimum between the X (j).

The Kalman estimates to be divided three classes: j〉k is called prediction (or extrapolation) problem, promptly from observation series Z (0), Z (1) ..., Z (k) finds out the linear optimal of X (k+1) and estimates

J=k is called filtering (or estimation) problem, promptly from observation series Z (0), and Z (1) ..., Z (k) finds out the linear optimal of X (k) and estimates

J＜k is called smoothly (or interpolation) problem, promptly from observation series Z (0), and Z (1) ..., Z (k) finds out the linear optimal of X (k-1) and estimates

Discrete system Kalman optimum prediction estimate equation is

$\hat{X}(k+1|k)=A(k+1,k)\hat{X}\left(k\right|k-1)+G(k+1,k)U\left(k\right)+$ (10)

$K\left(k\right)[Z\left(k\right)-C\left(k\right)\hat{X}\left(k\right|k-1)-Y\left(k\right)]$

And Kalman's optimal filtering estimate equation is

$\hat{X}(k+1|k+1)=\hat{X}(k+1|k)+K(k+1)\&times;$ (11)

$[Z(k+1)-C(k+1)\hat{X}(k+1|k)-Y\left(k\right)]$

Promptly obtain the optimum prediction estimated value earlier

Again actual observation to the predicted value of the observed reading that obtains before arriving with its of Z (k+1)

Compare, if both are unequal then right

Revise, estimate to obtain optimal filtering

Thermal neutron time spectrum of gathering in arbitrary depth point and the counting of capturing i time road in the gamma time spectrum are comprising the different j=1 of neutron lifetime ..., the contribution and the statistical error of n kind medium can be expressed as multi index option function (12).The time spectrum that records in each depth point during well logging can be done real-time Kalman filtering, with the statistical error that comprises in the counting of suppressing each time road.Same time road also comprises statistical error between the data of different depth point, can obtain one group of smooth curve in the vertical through low-pass filtering.The elapsed time-Depth Domain two dimension Kalman filtering, what describe at the time spectrum of each depth point will be a smooth multi index option curve, and time-Depth Domain is seen on the plane, is a two-dimensional array, can depict one group of smooth time gate curve that changes with stratum character.The window counting is expressed as (12) formula, and becomes (13) formula after the filtering in the time of collect on a depth point i before filtering.

$N_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_j\exp [-\frac{t}{{\mathrm{\&tau;}}_j}]+{\mathrm{\&epsiv;}}_i---\left(12\right)$

$N_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{a}_j\exp [-\frac{t}{{\mathrm{\&tau;}}_j}]---\left(13\right)$

In the formula, die-away time t=i Δ t, i=1 ..., m is the location, Δ t is wide.τ _jBe the neutron lifetime of j kind medium, a _jBe the amplitude of j kind component at start channel, ε _iIt is the statistical error in the counting.

In even infinite medium, (13) formula can be reduced to

N(t)＝aeXp(-t/τ) (14)

Both sides take the logarithm a straight line equation, then can obtain the inverse of thermal neutron lifetime τ by the slope of straight line.

The formula that current logging community uses is following two exponential forms

N(t)＝a ₁eXp(-t/τ ₁)+a ₂eXp(-t/τ ₂) (15)

As well medium neutron lifetime τ ₁＜＜τ ₂The time, can obtain the neutron lifetime τ on stratum at the back segment of time domain ₂

Filtered data are through normalized, so that the yardstick of information decomposes and reconstruct.

(2) filtered time spectrum is done conversion from the time domain to the age field

The time spectrum data that the time road counting of being represented by (13) formula constitutes, can be write as matrix equation:

Y＝AP (16)

In the formula,

Y _m×1＝(N ₁，N ₂，Λ，N _m) ^T (17)

P _n×1＝(p ₁，p ₂，ΛΛp _n) ^T (18)

Matrix (17) is a count matrix; Matrix (18) is the start channel counting of the pairing counting component of each life value of the Inversion Calculation of wanting, and obtains the shared share of each component thus, and promptly τ distributes.τ value in the life spectrum is the series of layouting in the preassigned life-span, can be in stratum thermal neutron minimum-maximum life interval evenly reconnaissance, also can adopt 2 power power to layout.Matrix (16) may be overdetermination, positive definite or owe fixed, this depend on that the number of measuring road counting of trying to achieve and life spectrum cloth counts how much.Find the solution formula (16) and also just realized the multi index option inverting of life spectrum.This is a non-well-posed problem on mathematics, can adopt the method for iteration to find the solution.

τ distributes more can reflect the overall picture on stratum than single τ value, ask the stratum neutron lifetime more can reflect the statistical property of stratum nuclear parameter with τ distribution characteristics peak or the selected interior mean value of gate-width Δ τ.

(3) filtered signal is done multiscale analysis

The nuclear logging curve being done wavelet analysis, can change time or degree of depth one dimension function into yardstick-time or yardstick-degree of depth two-dimensional function, is the different component of yardstick with signal decomposition.

On certain yardstick i, to given logging signal sequence

(k ∈ Z) is that the low-pass filter of h (k) can obtain the smooth signal on the thick yardstick by an impulse response

x _V(i-1，k)∈V _i-1 (20)

$x_V(i-1,l)=\underset{k}{\mathrm{\&Sigma;}}h(2l-k)x(i,k)---\left(21\right)$

(i, " detail signal " k) lost in low-pass filter can (i be that the Hi-pass filter of g (k) obtains by an impulse response k) to signal x by x

x _D(i-1，k)∈D _i-1 (22)

$x_D(i-1,l)=\underset{k}{\mathrm{\&Sigma;}}g(2l-k)x(i,k)---\left(23\right)$

Subscript D represents x, and (i is k) at detail signal space D _I-1On projection.

Like this, (i k) just is broken down into two components of high and low frequency to the signal x on yardstick i.The nuclear logging data are carried out the multiple dimensioned decomposition of N layer, can obtain 2N different frequency band, wherein comprise N high frequency band and N low-frequency band.Low frequency on the various yardsticks and high-frequency signal are observed, analyzed and compare, can determine the signal of each yardstick and the correlativity of ground sheaf space and physical parameter, low frequency signal can reflect the stratum profile information that comprises in the logging trace preferably, and high-frequency signal can reflect the details on stratum.

Time spectrum, τ distribution, tale, time period counting, various nuclear parameter and reservoir parameter all can carry out yardstick and decompose with the change curve of the degree of depth.As time spectrum and τ distribution are done multiple dimensioned decomposition, can select the yardstick that is subjected to borehole environment to influence minimum and can reflects stratum neutron lifetime variation.

3] filtered data are carried out normalized

The step of described normalized comprises the normalization log value that calculates each sampled point:

$f\&prime;=\frac{f-f_{\min }}{f_{\max }-f_{\min }}\&times;100---\left(5\right)$

F in the formula _MinBe the minimum value of logging trace, f _MaxBe the maximal value of logging trace, f is the sampling number certificate of logging trace.

4] data after the normalized are carried out multiple dimensioned decomposition, laterally the method for the multiple dimensioned decomposition of time shaft time spectrum comprises multi index option fitting process and wavelet decomposition method, and vertically the method for the multiple dimensioned decomposition of degree of depth axle mainly refers to the wavelet decomposition method.

5] the two spectrum of nuclear logging data being done multi-scale information merges

Information fusion be meant utilize computer technology the observation information of some sensors of obtaining is chronologically being analyzed automatically, is being optimized under certain criterion comprehensive, the information process that carries out to finish required decision-making and estimation task.The core of information fusion is meant carries out multi-level, many-sided, multi-level processing to the data from a plurality of sensors, thereby produces new significant information, and this fresh information be any single-sensor can't obtain.The all the sensors that adopts in the system all must be in aspect compatibilities such as sampling rate, observation field and scopes, and they also must provide the information of complementation to improve the quality of inference.The repeated measures signal of sensor can improve the reliability of system by redundant information.

Multiple dimensioned data fusion as the linear combination of selected component, can be carried out on raw data, physical parameter and reservoir parameter, three levels of decision-making.Can be from the data that similar detector is gathered in fusion at all levels, the data of non-similar detector collection can merge on parameter and decision-making level.

The two spectrum of pulsed neutron saturation degree well logger can record thermal neutron and capture two time spectrums of gamma, and note is data N1 and data N2 respectively, sees Fig. 2.To time spectrum, thermal neutron with capture gamma detector and can regard similar as.This two column data is carried out yardstick decompose, promptly raw data is carried out low, high-pass filtering respectively, be decomposed into two subdatas that contain the different frequency composition.A is the low frequency approximation signal, and d is a detail of the high frequency.Again as required, the low frequency subdata is repeated top process, realize that progressively the small echo turriform of this two column data is decomposed.Carry out fusion treatment at each decomposition layer then, obtain merging wavelet pyramid, again the wavelet pyramid after merging is carried out wavelet inverse transformation and just can obtain required reconstruct data N3.

The water saturation S that neutron life time log, carbon-to-oxygen ratio spectrum logging, resistivity logging obtain _W1, S _W2And S _W3Can do multiple dimensioned decomposition and fusion at parametric degree.

The data fusion effect can adopt three class statistical parameters to carry out A+E: first kind reflection peak information, as average, variance; Second class reflection spatial detail information is as information entropy; The 3rd class reflection spectrum information is as Y-PSNR, related coefficient.Signal to noise ratio (S/N ratio) is high more, illustrates that syncretizing effect is good more.

The step of reconstruct and fusion comprises:

5.1] based on the very big restructing algorithm of little mode, select low frequency coefficient weighting, the high frequency coefficient bigger fusion rule that takes absolute value for use, similar or multiclass log data are carried out wavelet reconstruction obtain merging wavelet pyramid.The step of the very big restructing algorithm of described little mode comprises:

5.1.1] to establish the signal that logging instrumentation receives be f (x), the Morlet continuous wavelet transform is defined as:

$W_f(a,b)=<f,{\mathrm{\&psi;}}_{\mathrm{ab}}>=\frac{1}{\sqrt{\left|a\right|}}\underset{-\&infin;}{\overset{+\&infin;}{\&Integral;}}f\left(x\right)\mathrm{\&psi;}\left(\frac{x-b}{a}\right)\mathrm{dx}$

(6)

In the formula: the Morlet wavelet basis function is

$\mathrm{\&psi;}\left(x\right)=e^{-x^2/2}\&CenterDot;e^{i{\mathrm{\&omega;}}_{0}x}---\left(7\right)$

A and b are respectively scale factor and shift factor.

5.1.2] signal is carried out wavelet analysis, can obtain the big coefficient of discharge of the assessing signal of different scale a, promptly corresponding wavelet coefficient W in the different spaces section _f(a, b); And represent with the mode of chromatogram, in chromatogram, represent the size of coefficient with change in color;

5.1.3] wavelet coefficient curve when obtaining a certain yardstick a, this curve can intuitively show the similarity degree between wavelet coefficient and the analyzed signal, and can make wavelet coefficient curve and some parameter have certain correlativity by the size of control a.

5.2] wavelet pyramid after merging is carried out wavelet inverse transformation realization data reconstruction at all levels.

5.3] adopt entropy, average and variance as judgment criteria, fused data is carried out quantitative evaluation: average, variance data reflection peak information; Information entropy data reflection spatial detail information; Y-PSNR, related coefficient reflection spectrum information.

6] the Data Post achievement shows and explains

6.1] degree of depth-time coloured image: on each depth point, get a level and smooth multi index option die-away curve after the filtering, in the digital picture that obtains on the tested section on the width of cloth degree of depth-time 2-D plane.This array is carried out visualization processing,, obtain the coloured image of width of cloth reflection depth of stratum-temporal characteristics, as the foundation of qualitative identification water layer and residue oil reservoir with look post demarcation signal amplitude.Equally, life spectrum and power spectrum also are the digital pictures on the two dimensional surface, also can be shown as coloured image.

6.2] degree of depth-life-span or the degree of depth-∑ coloured image: neutron lifetime τ and thermal neutron macroscopic capture cross section ∑ be inversely proportional to, τ distributes and is easy to change into ∑ and distributes.Through time-life-span or time-conversion in territory, cross section, on each sampled point, all obtain a τ or ∑ distribution curve, in the digital picture that obtains on the tested section on a width of cloth degree of depth-life-span or the degree of depth-∑ two dimensional surface.With look post demarcation signal amplitude, obtain τ or the ∑ distribution coloured image of width of cloth reflection formation characteristics with change in depth.With the general volume-based model of well logging interpretation, can obtain water saturation S by the ∑ distribution _WDistribution plan.Water saturation S _WDistribution plan can more properly depict profit and distribute.

6.3] described logging trace comprises: thermal neutron and capture gamma gross-count rate, characteristic time door and can window counting rate and ratio, thermal neutron lifetime and macroscopic capture cross section; Selected through type well check, by multiple dimensioned decomposition obtain to oil saturation or the good branch discharge curve of lithology resolution characteristic; Selected through type well check, by multiple dimensioned data rebuild and merge obtain to oil saturation or good reconstruction and the blend curve of lithology resolution characteristic; Uncased hole and real-time oil, gas, water saturation curve.

The method of carrying out geologic interpretation comprises:

7.1] degree of depth-time spectrum: neutron and capture gamma counting rate in time die-away curve on different depth point and can be divided into well, excessive, three periods of stratum, the high more then thermal neutron lifetime of the attenuation rate of stratomere counting rate is short more, and oil saturation is low more.

7.2] utilizing life spectrum or the conduct of ∑ spectrum to distinguish the foundation of low mineralization water layer and remaining oil gas-bearing formation: life-span τ distributes more can reflect the overall picture on stratum than single τ value, ask the stratum neutron lifetime more can reflect the statistical property of stratum nuclear parameter with τ distribution characteristics peak or the selected interior mean value of gate-width Δ τ; With the general volume-based model of well logging interpretation, can obtain water saturation S by the ∑ distribution _WDistribution plan, water saturation S _WDistribution plan can more properly depict profit and distribute.

Claims (6)

1, the two spectrum of a kind of pulsed neutron saturation logging method is characterized in that it may further comprise the steps:

1] gather log data:

In cased well, use the two spectrum of pulsed neutron saturation degree well logger continuous acquisition log datas; Described log data comprises thermal neutron time spectrum, captures the gamma time spectrum, the thermal neutron gross-count rate, capture gamma gross-count rate, natural gamma gross-count rate;

2] to thermal neutron time spectrum, capture the gamma time spectrum and carry out the degree of depth-time 2-D filtering, do the degree of depth-energy two-dimensional filtering to capturing gamma spectra and natural gamma spectra:

The described degree of depth-time 2-D filtering comprises vertical filtering of Depth Domain and in the horizontal filtering of time domain; The described degree of depth-energy two-dimensional filtering comprises vertical filtering of Depth Domain and in the horizontal filtering in energy territory; The method of described vertical filtering and laterally filtering comprises Kalman filtering, multiple spot smothing filtering or multi-scale filtering; Obtain a pair of smooth multi index option curve of decay in time in each depth point after the filtering, in the degree of depth-time domain, obtain two two-dimensional arrays, constitute the two frame of digital images of reflection formation properties with change in depth; And in the degree of depth-energy territory, obtain two two-dimensional arrays, constitute the two frame of digital images of reflection formation properties with change in depth;

3] filtered data are carried out normalized:

The step of described normalized comprises the normalization log value that calculates each sampled point:

$f\&prime;=\frac{f-f_{\min }}{f_{\max }-f_{\min }}\&times;100$

F in the formula _MinBe the minimum value of logging trace, f _MaxBe the maximal value of logging trace, f is the sampling number certificate of logging trace;

4] data after the normalized are carried out multiple dimensioned decomposition:

The method of the multiple dimensioned decomposition of described horizontal time shaft time spectrum comprises multi index option fitting process and wavelet decomposition method, and the method for the multiple dimensioned decomposition of described vertical degree of depth axle mainly refers to the wavelet decomposition method;

5] decomposed data is reconstructed and merges:

Described log data is reconstructed with merging is included in reconstruct and the fusion of carrying out time spectrum, life spectrum, ∑ spectrum, power spectrum and similar and non-similar logging trace on raw data, parameter, three levels of decision-making respectively, and concrete steps comprise:

5.1] based on the very big restructing algorithm of little mode, select low frequency coefficient weighting, the high frequency coefficient bigger fusion rule that takes absolute value for use, similar or multiclass log data are carried out wavelet reconstruction obtain merging wavelet pyramid;

5.2] wavelet pyramid after merging is carried out wavelet inverse transformation realization data reconstruction at all levels;

5.3] adopt entropy, average and variance as judgment criteria, fused data is carried out quantitative evaluation: average, variance data reflection peak information; Information entropy data reflection spatial detail information; Y-PSNR, related coefficient reflection spectrum information;

6] show log picture and logging trace

Described log picture and logging trace comprise with at least one group in hypograph and the curve:

The degree of depth-time coloured image: on each depth point, get a level and smooth multi index option die-away curve after the filtering, in the digital picture that obtains on the tested section on the width of cloth degree of depth-time 2-D plane; This array is carried out visualization processing,, obtain the coloured image of width of cloth reflection depth of stratum-temporal characteristics with look post demarcation signal amplitude;

The degree of depth-the life-span or the degree of depth-∑ coloured image: neutron lifetime τ and thermal neutron macroscopic capture cross section ∑ are inversely proportional to, and τ distributes and is easy to change into the ∑ distribution; Through time-life-span or time-conversion in territory, cross section, on each sampled point, all obtain a τ or ∑ distribution curve, in the digital picture that obtains on the tested section on a width of cloth degree of depth-life-span or the degree of depth-∑ two dimensional surface; With look post demarcation signal amplitude, obtain τ or the ∑ distribution coloured image of width of cloth reflection formation characteristics with change in depth;

Described logging trace comprises: thermal neutron and capture gamma gross-count rate, characteristic time door and can window counting rate and ratio, thermal neutron lifetime and macroscopic capture cross section; Selected through type well check, by multiple dimensioned decomposition obtain to oil saturation or the good branch discharge curve of lithology resolution characteristic; Selected through type well check, by multiple dimensioned data rebuild and merge obtain to oil saturation or good reconstruction and the blend curve of lithology resolution characteristic; Uncased hole and real-time oil, gas, water saturation curve;

7] carry out geologic interpretation:

The described geologic interpretation that carries out comprises at least a with in hypograph and the curve:

The degree of depth-time spectrum: neutron and capture gamma counting rate in time die-away curve on different depth point and can be divided into well, excessive, three periods of stratum, the high more then thermal neutron lifetime of the attenuation rate of stratomere counting rate is short more, and oil saturation is low more;

Utilize the foundation of life spectrum or ∑ spectrum as differentiation low mineralization water layer and remaining oil gas-bearing formation: life-span τ distributes more can reflect the overall picture on stratum than single τ value, asks the stratum neutron lifetime more can reflect the statistical property of stratum nuclear parameter with τ distribution characteristics peak or the selected interior mean value of gate-width Δ τ; With the general volume-based model of well logging interpretation, can obtain water saturation S by the ∑ distribution _WDistribution plan, water saturation S _WDistribution plan can more properly depict profit and distribute;

Described step 2] in carry out multi-scale filtering step comprise:

1] chooses Orthogonal Wavelets with certain tight supportive, symmetry and flatness;

2] select small echo and definite its to decompose level N, then data are carried out N layer wavelet decomposition, that is:

On certain yardstick i, to given nuclear logging burst $x(i,k)\&Element;V_i\&Subset;l^2\left(Z\right)$ (k ∈ Z) by the low-pass filter that an impulse response is h (k), obtains the smooth signal on the thick yardstick

x _V(i-1，k)∈V _i-1

$x_V(i-1,l)=\underset{k}{\mathrm{\&Sigma;}}h(2l-k)x(i,k)$

((i is that the Hi-pass filter of g (k) obtains the detail signal on the thin yardstick by an impulse response k) to signal x by x for i, " detail signal " k) lost in low-pass filter

x _D(i-1，k)∈D _i-1

$x_D(i-1,l)=\underset{k}{\mathrm{\&Sigma;}}g(2l-k)x(i,k)$

Subscript D represents x, and (i is k) at detail signal space D _I-1On projection;

3] carry out the multiple dimensioned decomposition that above-mentioned steps can realize the N layer one by one, obtain 2N different frequency band, comprise N high-frequency signal and N low frequency signal;

4] select a threshold value to carry out the soft-threshold quantification treatment to the 1st layer of each floor height frequency coefficient to the N layer; By the N layer low frequency coefficient of wavelet decomposition with through the 1st layer of high frequency coefficient after the quantification treatment to the N layer, carry out the reconstruct of log data, only comprise useful information in the reconstruction signal;

Described step 5] step of the very big restructing algorithm of medium and small mode comprises:

1] establishing the signal that logging instrumentation receives is f (x), and the Morlet continuous wavelet transform is defined as:

$W_f(a,b)=<f,{\mathrm{\&psi;}}_{\mathrm{ab}}>=\frac{1}{\sqrt{\left|a\right|}}\underset{-\&infin;}{\overset{+\&infin;}{\&Integral;}}f\left(x\right)\mathrm{\&psi;}\left(\frac{x-b}{a}\right)\mathrm{dx}$

In the formula: the Morlet wavelet basis function is

$\mathrm{\&psi;}\left(x\right)=e^{-x^2/2}\&CenterDot;e^{i{\mathrm{\&omega;}}_{0}x}$

A and b are respectively scale factor and shift factor;

2] signal is carried out wavelet analysis, can obtain the big coefficient of discharge of the assessing signal of different scale a, promptly corresponding wavelet coefficient W in the different spaces section _f(a, b); And represent with the mode of chromatogram, in chromatogram, represent the size of coefficient with change in color;

Wavelet coefficient curve when 3] obtaining a certain yardstick a, this curve can intuitively show the similarity degree between wavelet coefficient and the analyzed signal, and can make wavelet coefficient curve and some parameter have certain correlativity by the size of control a;

The two spectrum of described pulsed neutron saturation degree well logger comprises well logger housing 5 and the pulsed neutron generator in housing 1, thermal neutron detector 2, captures gamma detector 3 and natural gamma detector 4; Wherein the distance between thermal neutron detector 2 and the pulsed neutron generator 1 is 300-400mm; The distance of capturing between gamma detector 3 and the pulsed neutron generator 1 is 450-550mm; Distance between natural gamma detector 4 and the pulsed neutron generator 1 is 700-1000mm.

2, the two spectrum of pulsed neutron according to claim 1 saturation logging method, it is characterized in that: described log data also comprises captures gamma spectra, natural gamma spectra and well temperature, pressure and casing coupling data.

3, the two spectrum of pulsed neutron according to claim 1 and 2 saturation logging method, the two spectrum of wherein said pulsed neutron saturation degree well logger is characterised in that: the two spectrum of described pulsed neutron saturation degree well logger can be operated under NTS pattern, CTS pattern, three kinds of measurement patterns of DTS pattern, and described NTS pattern is only gathered thermal neutron time spectrum; Described CTS pattern is only gathered and is captured the gamma time spectrum; Described DTS pattern can be gathered thermal neutron time spectrum simultaneously and be captured the gamma time spectrum.

4, the two spectrum of pulsed neutron according to claim 3 fluid saturation logging method, the two spectrum of wherein said pulsed neutron saturation degree well logger is characterised in that: the distance between described thermal neutron detector 2 and the pulsed neutron generator 1 is 350mm; The distance of capturing between gamma detector 3 and the pulsed neutron generator 1 is 500mm; Distance between natural gamma detector 4 and the pulsed neutron generator 1 is 900mm.

5, the two spectrum of pulsed neutron according to claim 4 fluid saturation logging method, the two spectrum of wherein said pulsed neutron saturation degree well logger is characterised in that: described thermal neutron detector 2 is the 3He counter tube; The described gamma detector 3 of capturing is scintillation spectrometer; Described natural gamma detector 4 is a scintillation spectrometer.

6, the two spectrum of pulsed neutron according to claim 5 fluid saturation logging method, the two spectrum of wherein said pulsed neutron saturation degree well logger is characterised in that: it also comprises thermometer, pressure gauge and casing collar locator (CCL).

Images