CN104793245A - Method for recognizing gas reservoirs by utilizing wavelet phase features - Google Patents

Abstract

The invention relates to a method for recognizing gas reservoirs by utilizing wavelet phase features. The method includes the steps: building a plural convolution model; calculating according to the plural convolution model to acquire a time frequency energy spectrum and a time frequency phase spectrum; selecting reservoir effective seismic response starting main frames f1 and f2, and stacking complex reflection coefficient within a frequency range from f1 to f2 along the frequency direction to acquire a reservoir-related complex reflection coefficient spectrum; performing modulus square and anti-tangent operation on the reservoir-related complex reflection coefficient to respectively acquire a frequency abnormality spectrum and a phase abnormality spectrum; utilizing the frequency abnormality spectrum and the phase abnormality spectrum for gas reservoir recognition. The method can be used for detecting gas reservoirs with high attenuation features, capability of describing thin reservoirs can be improved effectively, and solution multiplicity of explaining can be lowered. The method can be widely applied in the process of oil-gas exploration.

Description

A kind of method utilizing sub-wave phase feature identification gas reservoir

Technical field

The present invention relates to a kind of earthquake detecting method of hydrocarbon, particularly about a kind of method utilizing sub-wave phase feature identification gas reservoir be applied in petroleum prospecting process.

Background technology

Rock physical modeling and exploration practices show, ripple, containing when propagating in fluid media (medium), can be subject to diectric attenuation impact, and high frequency comparatively low frequency be more easily attenuated, therefore, prior art often uses low-frequency anomaly section to carry out hydrocarbon prediction.But in reality exploration, this method often can lose efficacy, because water layer also can make seismic event produce decay and produce the spectral response similar to hydrocarbon reservoir.In fact, decay and frequency dispersion not only can change sub-wave frequency, and sub-wave phase is changed, but phase characteristic is never utilized.Main cause is familiar with clear not enough to the phase response feature that reservoir causes at present, and lack the technological means asking for time-varying wavelet phase information from seismic section.

A time-domain signal can be decomposed into the two-dimensional function of time and frequency by earthquake Spectral Decomposition Technique, is widely used in frequency division seismic interpretation field.Conventional spectral factorization method has Instant Fourier Transform, Gabor transformation, continuous wavelet transform (CWT) and S-transformation etc.These methods play very important effect on reservoir prediction and hydrocarbon indication always.But these methods exist two deficiencies: first, low temporal resolution limits method and is portraying the application on thin layer; Secondly, be difficult to obtain sub-wave phase information, which has limited the application of phase place in hydrocarbon indication.

Summary of the invention

For the problems referred to above, the object of this invention is to provide a kind of method utilizing sub-wave phase feature identification gas reservoir, the gas reservoir that the method can be used for having high decay characteristics detects, and effectively promotes to portray ability to thin reservoir, reduces the multi-solution explained.

For achieving the above object, the present invention takes following technical scheme: a kind of method utilizing sub-wave phase feature identification gas reservoir, and it comprises the following steps: 1) set up a wavelet storehouse be made up of the zero phase Ricker wavelet of different frequency k=1,2 ... K, K are the wavelet number participating in calculating, and t represents the time; By wavelet storehouse multiple wavelet storehouse w is become by Hilbert _k(t), and set up plural convolution model; 2) time-frequency energy spectrum F is calculated according to plural convolution model _stfwith time-frequency phase spectrum 3) the initial dominant frequency f of the effective seismic response of reservoir is chosen ₁and f ₂, by f ₁to f ₂complex reflection coefficient between frequency range obtains the relevant complex reflection coefficient spectrum of reservoir along frequency direction superposition; 4) relevant to reservoir complex reflection coefficient the quadratic sum arctangent cp cp operation of delivery obtains frequency anomaly spectrum and phase anomaly spectrum respectively; 5) frequency anomaly spectrum and phase anomaly spectrum is utilized to carry out gas pool identification, recognition methods is as follows: when work area to be identified have drilling information time, when frequency anomaly spectrum is non-vanishing, if when phase anomaly spectrum is identical or close with the phase response feature comprised in drilling information, then judge that there is gas reservoir in this work area; When work area to be identified shakes data only, there is frequency anomaly and reservoir top reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area; In work area to be identified, when the reflection at reservoir top and bottom cannot be distinguished, then by its comprehensively response identify, namely there is frequency anomaly and reservoir reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area.

Described step 2) in, described time-frequency energy spectrum F _{s, tf}with time-frequency phase spectrum computing method are as follows: arranged by plural convolution model as linear forms: s=Wr+n, wherein, and W=(W ₁w ₂w _k), k=1,2 ... K, W _krepresent multiple wavelet w _kconvolution matrix; R=(r ₁r ₂r _k) ^t, T representing matrix twiddle operation; L is carried out to linear forms plural number convolution model ₁norm regularization:

$\underset{r}{\arg \min }\frac{1}{2}{\left|\right|\mathrm{Wr}-s\left|\right|}_2^2+\mathrm{\λ}{\left|\right|r\left|\right|}_1,\mathrm{\λ}>0---\left(1\right)$

Wherein, for L2 norm, || || ₁for L ₁norm, λ is the regularization parameter regulating degree of rarefication; Formula (1) is obtained complex reflection coefficient r by sparse inversion Algorithm for Solving, time-frequency energy spectrum F is obtained respectively to the quadratic sum arctangent cp cp operation of complex reflection coefficient r delivery _{s, tf}with time-frequency phase spectrum

Described step 4) in, described frequency anomaly spectrum and phase anomaly spectrum are:

${\hat{F}}_{s,\mathrm{stack}}={\left|{\hat{r}}_{\mathrm{stack}}\right|}^2,$

Wherein, for frequency anomaly spectrum, for phase anomaly spectrum, for the complex reflection coefficient spectrum that reservoir is relevant.

The present invention is owing to taking above technical scheme, and it has the following advantages: 1, the time-frequency spectrum resolution that obtains of traditional spectral factorization method is lower, and particularly temporal resolution is low, is difficult to portray thin reservoir; The present invention is by sparse inversion method, and the time-frequency distributions temporal resolution obtained is high, improves to portray ability to thin reservoir.2, the method existed at present can only extract the frequency information of wavelet from geological data, is difficult to extract its phase information; The invention provides a kind of spectral factorization method accurately can asking for time-varying wavelet frequency and phase information.3, only consider frequecy characteristic when present stage, spectral factorization was applied in hydrocarbon indication, but moisture reservoir usually produces similar frequency response with hydrocarbon-containifirst reservoir, strong by means of only frequency information identification multi-solution; Reflection wavelet phase characteristic is used for gas pool identification by the present invention, reduces the multi-solution explained.4, owing to only utilizing ripple propagating containing in fluid media (medium) the decay caused in prior art, the phase place change that the present invention proposes decay relevant does not occur over just in the process of wave traveling, also occurs in the interface that there is impedance and difference in attenuation.In interface, resistance difference only causes the change in polarity of reflection wavelet, and difference in attenuation be cause sub-wave phase to rotate key factor.Therefore, the gas reservoir that sub-wave phase variation characteristic can be used for having high decay characteristics detects.The present invention can extensively apply in petroleum prospecting process.

Accompanying drawing explanation

Fig. 1 is that wavelet storehouse of the present invention builds schematic diagram, the wherein real wavelet storehouse that is made up of the zero phase Ricker wavelet of different frequency of Fig. 1 (a); Fig. 1 (b) converts by carrying out Hilbert to real wavelet storehouse the multiple wavelet storehouse obtained, and in figure, solid line is multiple wavelet storehouse real part, and dotted line is multiple wavelet storehouse imaginary part;

Fig. 2 is that fitted signal time-frequency spectrum of the present invention decomposes the schematic diagram differentiating thin layer ability, wherein the middle fitted signal 3 of Fig. 2 (a) is for being used for the fitted signal calculated, the combination of fitted signal 1 and fitted signal 2, be used for characterize different-thickness stratum top at the bottom of reflection; The time-frequency energy spectrum that Fig. 2 (b) obtains for Using Continuous Wavelet Transform; The time-frequency energy spectrum that Fig. 2 (c) obtains for complex spectral resolution method;

Fig. 3 is fitted signal complex spectral resolution example schematic diagram of the present invention, wherein, the matching seismic signal that Fig. 3 (a) is made up of the Ricker wavelet of different frequency and phase place, wavelet frequency is followed successively by 60Hz from top to bottom, 40Hz, 20Hz, 30Hz and 30Hz, phase place is followed successively by 0 ° ,-90 °, 45 ° ,-180 ° and-180 °; The time-frequency energy spectrum that Fig. 3 (b) and figure (c) obtains for Using Continuous Wavelet Transform and time-frequency phase spectrum; Fig. 3 (d) and Fig. 3 (e) are respectively the time-frequency energy spectrum and time-frequency phase spectrum that complex spectral resolution method obtains;

Fig. 4 (a) ~ Fig. 4 (b) is that numerical simulation of the present invention decays the schematic diagram of the phase change reason caused, wherein, medium on the left of interface is low relative to right side medium wave impedance, attenuation degree is high, and in Fig. 4 (a) and Fig. 4 (b), solid line represents transmitted wave on the interface that only resistance difference exists and reflection wave; Dotted line in Fig. 4 (a) represents the transmittance and reflectance ripple only existed on difference in attenuation interface; In Fig. 4 (b), dotted line represents the transmittance and reflectance ripple simultaneously existed on resistance difference and difference in attenuation interface;

Fig. 5 (a) ~ Fig. 5 (d) is real data hydrocarbon indication example schematic diagram of the present invention, wherein, Fig. 5 (a) is for crossing the seismic section of prospect pit in target area, to throw logging trace be water saturation, gas-bearing formation and the water layer position of drilling well announcement and geologic interpretation indicate in the drawings; The frequency anomaly section for hydrocarbon indication that Fig. 5 (b) obtains for Using Continuous Wavelet Transform; The frequency anomaly section that Fig. 5 (c) obtains for using the complex spectral resolution method of the present invention's proposition; The time-frequency phase spectrum abnormal profile that Fig. 5 (d) asks for for complex spectral resolution method.

Embodiment

Below in conjunction with drawings and Examples, the present invention is described in detail.

As shown in Fig. 1 ~ Fig. 5 (d), the invention provides a kind of method utilizing sub-wave phase feature identification gas reservoir, it comprises the following steps:

1) a wavelet storehouse be made up of the zero phase Ricker wavelet of different frequency is set up k=1,2 ... K, K are the wavelet number participating in calculating, and t represents the time; By wavelet storehouse multiple wavelet storehouse w is become by Hilbert _k(t), and it is as follows to set up plural convolution model:

Wherein, s represents seismologic record, w _krepresent the multiple wavelet of kth in the frequency band participating in calculating, its dominant frequency and phase place be respectively f and r _kseismologic record neutron deficiency dominant frequency information f and phase information is carried for correspondence complex reflection coefficient; K represents the total number of wavelet participating in calculating, and * represents convolution operation, and n represents random noise.

2) by step 1) in the plural convolution model set up arrange as following linear forms:

s＝Wr+n， (2)

Wherein, W=(W ₁w ₂w _k), k=1,2 ... K, W _krepresent multiple wavelet w _kconvolution matrix; R=(r ₁r ₂r _k) ^t, T representing matrix twiddle operation;

L is carried out to formula (2) ₁norm regularization, as follows:

$\underset{r}{\arg \min }\frac{1}{2}{\left|\right|\mathrm{Wr}-s\left|\right|}_2^2+\mathrm{\λ}{\left|\right|r\left|\right|}_1,\mathrm{\λ}>0---\left(3\right)$

Wherein, for L2 norm, || || ₁for L ₁norm, λ is the regularization parameter regulating degree of rarefication;

Formula (3) is obtained complex reflection coefficient r by sparse inversion Algorithm for Solving, time-frequency energy spectrum F is obtained respectively to the quadratic sum arctangent cp cp operation of complex reflection coefficient r delivery _{s, tf}with time-frequency phase spectrum as follows:

3) the initial dominant frequency f of the effective seismic response of reservoir is chosen ₁and f ₂, by f ₁to f ₂complex reflection coefficient between frequency range obtains the relevant complex reflection coefficient spectrum of reservoir along frequency direction superposition, as follows:

${\hat{r}}_{\mathrm{srack}}={\mathrm{\Σ}}_{f_1}^{f_2}r\left(f\right)\mathrm{df},---\left(5\right)$

Wherein, for the complex reflection coefficient spectrum that reservoir is relevant; f ₁and f ₂determine according to real data background band and reservoir significant response frequency band.

4) relevant to reservoir complex reflection coefficient the quadratic sum arctangent cp cp operation of delivery obtains frequency anomaly spectrum and phase anomaly spectrum respectively, as follows:

${\hat{F}}_{s,\mathrm{stack}}={\left|{\hat{r}}_{\mathrm{stack}}\right|}^2,$

Wherein, frequency anomaly is composed, for phase anomaly spectrum; The process utilizing multiple wavelet storehouse to extract frequency and phase information from seismologic record is called complex spectral resolution.

5) utilize frequency anomaly spectrum and phase anomaly spectrum to carry out gas pool identification, recognition methods is as follows:

When work area to be identified have drilling information time, when frequency anomaly spectrum non-vanishing time, namely during frequency anomaly, if phase anomaly spectrum identical or close with the phase response feature comprised in drilling information time, then judge that there is gas reservoir in this work area;

When work area to be identified shakes data only, there is frequency anomaly and reservoir top reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area;

In work area to be identified, when the reflection at reservoir top and bottom cannot be distinguished, then by its comprehensively response identify, namely there is frequency anomaly and reservoir reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area.

Above-mentioned steps 5) in, gas pool identification supposes that original incident seismic wavelet is the zero-phase wavelet close to Ricker wavelet form, if incident wavelet itself has the phase rotating of certain angle, after incident wavelet phase place should being deducted from phase spectrum abnormal profile, be used further to gas reservoir prediction.

Below by embodiment, the method for sub-wave phase feature identification gas reservoir that utilizes of the present invention is described further.

Embodiment 1: fitting data example

As shown in Figure 2, this embodiment is the example utilizing the inventive method to differentiate thin layer ability.Fig. 2 (a) be by the Ricker wavelet at different time interval to the fitted signal formed, be used for characterize different-thickness stratum top at the bottom of reflection.The time-frequency energy spectrum of Fig. 2 (b) for being obtained by CWT method.When stratum is thicker (60ms interval), CWT method can be distinguished at the bottom of top, stratum and reflect; Along with zone thickness thinning (<30ms interval), CWT time-frequency spectrum well can not distinguish the reflection of the end, top.The time-frequency energy spectrum that Fig. 2 (c) obtains for complex spectral resolution method, can find out that the temporal resolution of the complex spectral resolution method of invention is very high.

As shown in Figure 3, be a matching seismic signal example be made up of the Ricker wavelet of different frequency and phase place.As shown in Fig. 3 (a), the wavelet frequency of fitted signal is followed successively by 60Hz, 40Hz, 20Hz, 30Hz and 30Hz from top to bottom, and phase place is followed successively by 0 ° ,-90 °, 45 ° ,-180 ° and-180 °.The time-frequency energy spectrum that Fig. 3 (b) obtains for CWT method, represents the level of resolution of conventional spectral factorization.The time-frequency phase spectrum that Fig. 3 (c) obtains for CWT method, is therefrom difficult to extract the sub-wave phase information relevant with decay.Fig. 3 (d) and Fig. 3 (e) are respectively the time-frequency energy spectrum and time-frequency phase spectrum that earthquake complex spectral resolution method obtains.Earthquake complex spectral resolution obtains time-frequency distributions and has very high temporal resolution, and required result is consistent with fitted signal truth.Matching example shows, the advantage of complex spectral resolution method is not only to have high time frequency resolution feature, but also is the phase information that can calculate time-varying wavelet exactly.

As shown in Fig. 4 (a), Fig. 4 (b), the phase place change that numerical simulation demonstrates decay relevant does not occur over just in the process of wave traveling, also occurs in the interface that there is impedance and difference in attenuation.Resistance difference only causes the change in polarity of reflection wavelet, and difference in attenuation be cause sub-wave phase to rotate key factor.Usual gas reservoir and cap rock have obvious resistance difference and difference in attenuation, therefore, by phase information Gas Reservoir Prediction.

Embodiment 2: real data example

As shown in Fig. 5 (a) ~ Fig. 5 (d), this embodiment is a real data example.Fig. 5 (a) is for crossing the seismic section of prospect pit in target area, and the logging trace that section is thrown is water saturation curve.Indicate in the drawings according to the gas-bearing formation of prospect pit announcement and geologic interpretation and water layer position.The frequency anomaly section for hydrocarbon indication that Fig. 5 (b) obtains for CWT method, temporal resolution, lower than seismic resolution, cannot accurately be portrayed thin layer position; The frequency anomaly section that Fig. 5 (c) obtains for using complex spectral resolution method in this paper, has very high temporal resolution, and consistent with the upper position of stratum disclosed of well logging.Gas-bearing reservoir has response in frequency anomaly section, but water layer also has response on frequency anomaly section, so this can disturb gas reservoir to predict, causes the multi-solution of explanation.The time-frequency phase spectrum abnormal profile that Fig. 5 (d) asks for for complex spectral resolution method, as we can see from the figure, there is some difference on phase anomaly section for gas-bearing formation and water layer response, reduces the multi-solution only utilizing frequency anomaly to carry out hydrocarbon indication.

In sum, the present invention and prior art have obvious difference, its be mainly following some: 1) the time-frequency spectrum resolution that obtains of traditional spectral factorization method is lower, and particularly temporal resolution is low, is difficult to portray thin reservoir; The present invention is by sparse inversion technology, and the time-frequency distributions temporal resolution obtained is high, improves the ability of portraying (as shown in Figure 2) to thin reservoir.2) method existed at present can only extract the frequency information of wavelet from geological data, is difficult to extract its phase information; The invention provides a kind of spectral factorization method (as shown in Figure 3) accurately can asking for time-varying wavelet frequency and phase information.3) only consider frequency anomaly when present stage spectral factorization is applied in the detection of hydro carbons, but moisture reservoir usually produces similar frequency response with hydrocarbon-containifirst reservoir, strong by means of only frequency information identification multi-solution; Reflected phase will feature is used for gas pool identification by the present invention, reduces the multi-solution (as Suo Shi Fig. 5 (a) ~ Fig. 5 (d)) explained.4) people only utilize ripple propagating containing in fluid media (medium) the decay characteristics caused usually, present invention demonstrates that the phase place change decaying relevant does not occur over just in the process of wave traveling, also occur in the interface that there is impedance and difference in attenuation.Interface resistance difference only causes the change in polarity of reflection wavelet, and difference in attenuation be cause sub-wave phase to rotate key factor.Therefore, the gas reservoir that sub-wave phase change can be used for having high decay characteristics detects (known as Suo Shi Fig. 4 (a), Fig. 4 (b)).

The various embodiments described above are only for illustration of the present invention; each step all can change to some extent; on the basis of technical solution of the present invention, all improvement of carrying out separate step according to the principle of the invention and equivalents, all should not get rid of outside protection scope of the present invention.

Claims (3)

1. utilize a method for sub-wave phase feature identification gas reservoir, it comprises the following steps:

1) a wavelet storehouse be made up of the zero phase Ricker wavelet of different frequency is set up k=1,2 ... K, K are the wavelet number participating in calculating, and t represents the time; By wavelet storehouse multiple wavelet storehouse w is become by Hilbert _k(t), and set up plural convolution model;

2) time-frequency energy spectrum F is calculated according to plural convolution model _{s, tf}with time-frequency phase spectrum

3) the initial dominant frequency f of the effective seismic response of reservoir is chosen ₁and f ₂, by f ₁to f ₂complex reflection coefficient between frequency range obtains the relevant complex reflection coefficient spectrum of reservoir along frequency direction superposition;

4) relevant to reservoir complex reflection coefficient the quadratic sum arctangent cp cp operation of delivery obtains frequency anomaly spectrum and phase anomaly spectrum respectively;

5) utilize frequency anomaly spectrum and phase anomaly spectrum to carry out gas pool identification, recognition methods is as follows:

When work area to be identified have drilling information time, when frequency anomaly spectrum non-vanishing time, if phase anomaly spectrum identical or close with the phase response feature comprised in drilling information time, then judge that there is gas reservoir in this work area;

When work area to be identified shakes data only, there is frequency anomaly and reservoir top reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area;

In work area to be identified, when the reflection at reservoir top and bottom cannot be distinguished, then by its comprehensively response identify, namely there is frequency anomaly and reservoir reflection wavelet phase spectrum angle between 120 ° to 200 °, then judge that there is gas reservoir in this work area.

2. a kind of method utilizing sub-wave phase feature identification gas reservoir as claimed in claim 1, is characterized in that: described step 2) in, described time-frequency energy spectrum F _{s, tf}with time-frequency phase spectrum computing method are as follows:

Plural convolution model is arranged for linear forms:

s＝Wr+n，

Wherein, W=(W ₁w ₂w _k), k=1,2 ... K, W _krepresent multiple wavelet w _kconvolution matrix; R=(r ₁r ₂r _k) ^t, T representing matrix twiddle operation;

L is carried out to linear forms plural number convolution model ₁norm regularization:

$\underset{r}{\arg \min }\frac{1}{2}{\left|\right|\mathrm{Wr}-s\left|\right|}_2^2+\mathrm{\&lambda;}{\left|\right|r\left|\right|}_1,\mathrm{\&lambda;}>0---\left(1\right)$

Wherein, for L2 norm, || || ₁for L ₁norm, λ is the regularization parameter regulating degree of rarefication;

Formula (1) is obtained complex reflection coefficient r by sparse inversion Algorithm for Solving, time-frequency energy spectrum F is obtained respectively to the quadratic sum arctangent cp cp operation of complex reflection coefficient r delivery _{s, tf}with time-frequency phase spectrum

3. a kind of method utilizing sub-wave phase feature identification gas reservoir as claimed in claim 1 or 2, is characterized in that: described step 4) in, described frequency anomaly spectrum and phase anomaly are composed and are:

${\hat{F}}_{s,\mathrm{stack}}={\left|{\hat{r}}_{\mathrm{stack}}\right|}^2,$

Wherein, for frequency anomaly spectrum, for phase anomaly spectrum, for the complex reflection coefficient spectrum that reservoir is relevant.