CN101359057B - Method for detecting gas reservoir by using attenuation information of central frequency changing along with incident angle - Google Patents

Abstract

The invention relates to gas reservoir detection by using attenuation information of central frequency changing along with incident angle in geophysical exploration of petroleum. Collecting a trace set after time difference correction; converting the data in the medium offset range domain into an angle domain, and performing frequency compensation on the angle gather to eliminate distortion; calculating the instantaneous frequency spectrum and instantaneous center frequency of a sampling point to form a center frequency angle gather, and extracting a zero-angle center frequency attribute and a gradient attribute; and superposing and averaging the central frequency channels within a certain angle range, and determining fluid abnormity in pores and detecting oil-gas change of formation lithology by using the frequency difference of the attribute channels and the negative abnormal value of the gradient. The method for extracting the frequency anomaly attenuation information closely related to the fluid in the pores from the pre-stack seismic data does not need prior information and does not depend on logging data, so that the sensitivity of gas reservoir identification is improved, the uncertainty in the conventional attenuation method is reduced, and the gas reservoir identification is accurate.

Description

A kind of centre frequency following variation of angle of incidence information of utilizing is carried out the method that gas reservoir detects

Affiliated field

The invention belongs to the hydro carbons detection technique of geophysical prospecting for oil, specifically is that a kind of centre frequency following variation of angle of incidence information of utilizing is carried out the method that gas reservoir detects.

Background technology.

Carry out natural gas pool identification according to seismic data, except the most frequently used AVO analytical approach, the attenuation Characteristics of gas reservoir reflected seismic information just more and more is subject to people's attention.Practice is measured and is shown, for compression P ripple, when propagating in gas-bearing reservoir, the decay of high frequency composition is serious, and energy is by strong absorption (promptly having very little quality factor q).Therefore, whether quality factor q can be used as differentiates in the stratum indicator of gassiness, can utilize it to detect gas-bearing formation.

Utilize decay factor (absorption coefficient) to carry out the existing ripe method of gas reservoir detection, the method that adopts is to compose than method wavelet energy absorbing methods (Sheriff etc., 1995 usually; Lichman etc., 2004).Spectrum is a kind of linear fit method of frequency field than method, and the error of its estimation is subjected to parameter influences such as window length, frequency band range bigger, and the vertical resolution of this method is lower.The frequency attribute of signal also is to carry out attenuation analysis instrument commonly used, and the spectral factorization method is the normal method of using.Be published in the instantaneous spectrum analytical technology that is used for hydrocarbon detection (Castagna etc. 2003) of " Leading Edge " in 2003, utilize signal to detect gas reservoir and obtain effect preferably in actual applications in the energy distribution of different frequency.The decay characteristics of signal can be described by on average measure (Barnes, 1993) of the frequency spectrum/power spectrum of signal in frequency field.Quan etc. 1997 propose the seismic attenuation method of estimation that moves based on centre frequency and it are applied in the seismic data of chromatography between well.In the steam flood project that improves the oil recovery of recovery ratio method, Hedlin etc. (2002) are applied to the centre frequency of signal in the time shift seismic data, the centre frequency difference of the seismic data of utilization two different times phase measuring is discerned fluid as the decay abnormality mark.

Current these decay assessment technologies mainly are to utilize the poststack seismic data, and known by the propagation equation of viscoelastic wave: the damping capacity of signal is relevant with quality factor and travel-time.In quality factor one regularly, damping capacity was controlled by the travel-time.Therefore, there are two defectives in the post-stack attenuation assessment technology: the one, and the poststack record is to connect between the time receiving certainly, and signal is the shortest the hourage in thin layer, and the damping capacity of signal is not high, thereby attenuation attributes is to the susceptibility reduction of thin layer; The 2nd, synergistic effect makes the attenuation attributes of estimating have error, we know the not hourage difference of signal in thin layer of common-range gather, damping capacity is also different, damping capacity a long way is bigger than the damping capacity of shortcut, stack makes the damping capacity of signal in different earthquake road add together, is an average attenuation amount, decays when estimating with it, will inevitably cause error, the effect that influences gas reservoir identification and detect.

Summary of the invention

The object of the invention is to provide a kind of and extracts and the closely related attribute of pore fluid from seismologic record, improves the centre frequency following variation of angle of incidence information of utilizing of oil and gas detection susceptibility and carries out the method that gas reservoir detects.

The present invention is achieved through the following technical solutions:

Concrete steps comprise:

(1) excite and write down seismic event, earthquake data is routinely carried out high-fidelity and is handled, and formation can be directly used in the road collection of amplitude after with the normal-moveout correction of offset distance mutation analysis;

It is the relative Changing Pattern that keeps amplitude that described high-fidelity is handled, with reflection, the decay characteristics that reflects that seismic event is propagated in the stratum.

Described high-fidelity is handled or is referred to that the frequency range of signal and dominant frequency are constant, with the original frequency feature of holding signal.

(2) utilize following formula that offset distance numeric field data in the step (1) is transformed in the angle domain, obtain angular-trace gather,

$\sin \mathrm{\θ}=\frac{{\mathrm{xv}}_p}{v_{\mathrm{rms}}^2{(t_0^2+\frac{x^2}{v_{\mathrm{rms}}^2})}^{1/2}}---\left(1\right)$

In the formula, x is a geophone offset, and Vp is the interval velocity on stratum, t ₀Be the zero-offset whilst on tour, Vrms is the root-mean-square velocity on stratum;

(3) compensation of angular-trace gather working frequency is distorted to eliminate the signal frequency that is caused owing to the NMO stretching effect;

Described distortion is meant that seismic signal is along with incident angle increases and the artificial relaxation phenomenon of frequency reduction.

Described frequency compensation is meant carries out the frequency division amplitude compensation to the earthquake reflective data on the different angles territory, adopts following steps:

The amplitude variations that the amplitude drawing coefficient approximate formula based on normal moveout correction below utilizing is proofreaied and correct different offset distances makes statistical average spectral amplitude coupling zero-offset road a long way,

${\mathrm{\α}}_x=\frac{\∂t}{{\∂t}_0}\≈\frac{t_0}{t}(1-\frac{2(t-t_0)}{V_{\mathrm{NMO}}}\frac{{\∂V}_{\mathrm{NMO}}}{{\∂t}_0})---\left(2\right)$

α in the formula _xBe compressibility coefficient, t ₀Be the zero-offset round trip travel-time, t is the round trip travel-time, V _NMOBe the normal moveout correction root-mean-square velocity;

Utilizing following formula that seismic signal is done with the angle is that the compressed transform of parameter is eliminated offset distance far away, wide-angle incident seismic signal because the frequency reducing phenomenon that nmo stretching caused,

W in the formula _θBe the back wavelet that stretches, w is the zero-offset wavelet, and θ is a reflection angle, and t is the travel-time, and ω is an angular frequency;

(4) all seismic traces of road, diagonal angle collection calculate the instantaneous spectrum of each sampled point, utilize following formula to calculate the instantaneous centre frequency that each seismologic record is concentrated in the road, angle, form centre frequency angle Dao Ji,

$f_c=\frac{{\∫}_0^{\∞}\mathrm{fP}\left(f\right)\mathrm{df}}{{\∫}_0^{\∞}P\left(f\right)\mathrm{df}}---\left(4\right)$

F in the formula _cBe centre frequency, P (f) is the power spectrum of signal;

(5) utilize following formula to concentrate and extract zero degree centre frequency attribute f from road, centre frequency angle _c(0 °) and gradient attribute G (1/Q, t),

f _c(θ)＝f _c(0°)+G(1/Q，t)w(sin ²θ)，(5)

F in the formula _c(θ) be the centre frequency of angle θ, and gradient attribute G (1/Q, t) 1/Q reciprocal with travel-time t and stratum quality factor q is directly proportional;

Described centre frequency is except that outside the Pass having with incident angle, also with vertical relevant from the quality factor q that connects from travel-time t that receives and stratum.

Intercept f in the following formula _c(0 °) and gradient attribute G (1/Q t) can be referred to as centre frequency with the incident angle change to attributes, and the leaching process of these two attributes is found the solution following system of equations:

θ in the formula _i, i=1,2 ..., n is the incident angle that obtains road, angle collection geological data in the step (2);

(6) by centre frequency angle Dao Ji, utilize the method for part angle stack, promptly the centre frequency trace-stacking in a certain angular range is asked on average together, extract the centre frequency attribute road of far away, nearly angle

With

(7) utilize the attribute road

And f _cFrequency difference between (0 °), and the negative anomaly value of gradient attribute G determine that the fluid oil gas unusual and the detection formation lithology in the hole changes.

Fluid in described definite hole oil gas unusual and the detection formation lithology changes: on far away, nearly angle centre frequency section, the center frequency value of gas-bearing reservoir reduces, and difference is big, and the center frequency value difference of non-gas-bearing reservoir is then little.

Fluid in described definite hole oil gas unusual and the detection formation lithology changes: on the centre frequency gradient profile that is obtained with the inverting of incident angle variation formula by centre frequency, the centre frequency Grad of gas-bearing reservoir is unusual high negative value, but not the centre frequency Grad of gas-bearing reservoir then shows as less negative value.

The present invention directly from the pre-stack seismic data, extract with hole in the method for the closely-related frequency anomaly dampening information of fluid do not need prior imformation, do not rely on well-log information, improved the susceptibility of gas-bearing formation identification, reduced the uncertainty in the conventional damped system, gas reservoir identification accurately.

Description of drawings

Fig. 1 is a centre frequency following variation of angle of incidence analytical technology process flow diagram of the present invention;

Fig. 2 a and 2b concentrate road, the centre frequency angle composite section figure that extracts from the other CDP of the well road of gas well (the WC1 well of Fig. 3) and non-gas well (the YL2 well of Fig. 3), and angular range is at 1-33 °.

Fig. 3 a is the centre frequency attribute zero degree sectional view that the CFVA inverting obtains;

Fig. 3 b is the angle centre frequency sectional view far away that the centre frequency road of 22-30 ° of the centre frequency attribute angle that obtain of CFVA inverting superposes and obtains;

Fig. 3 c is the centre frequency gradient profile figure that extracts;

Fig. 4 is the centre frequency sectional view that utilizes post-stack seismic data to extract.

Embodiment

Common midpoint (CDP) the normal moveout correction road collection of the present invention from keeping through the prestack frequency at first will carry out frequency compensation to the CDP road collection of having done after normal moveout correction is handled, and this is because the NMO stretching effect causes the frequency of seismic signal to produce distortion; From seismic channel set, extract centre frequency then and change (CFVA) road collection with incident angle; Carry out fluid identification from the centre frequency that collection utilization in CFVA road is derived with incident angle variation formulas Extraction zero degree, centre frequency gradient road and far away, nearly angle centre frequency attribute road.

Centre frequency following variation of angle of incidence analytical technology has higher requirement to the pre-stack seismic DATA PROCESSING, should not adopt the processing module that the frequency content of seismic signal is had transformation in the processing procedure, that is to say that the object of being analyzed must be the CDP normal moveout correction road collection that keeps through the prestack frequency.

Concrete steps of the present invention comprise:

(1) excites and writes down seismic event, earthquake Data Processing is routinely carried out the high-fidelity processing that amplitude keeps, but the frequency of signal is not made the original frequency feature of any modification processing with holding signal, and formation can be directly used in the road collection of amplitude after with the normal-moveout correction of offset distance mutation analysis;

(2) utilize following formula that offset distance numeric field data in the step (1) is transformed in the angle domain, obtain angular-trace gather,

$\sin \mathrm{\θ}=\frac{{\mathrm{xv}}_p}{v_{\mathrm{rms}}^2{(t_0^2+\frac{x^2}{v_{\mathrm{rms}}^2})}^{1/2}}---\left(1\right)$

In the formula, x is a geophone offset, and Vp is the interval velocity on stratum, t ₀Be the zero-offset whilst on tour, Vrms is the root-mean-square velocity on stratum;

(3) compensation of angular-trace gather working frequency is distorted to eliminate the signal frequency that is caused owing to the NMO stretching effect,, this distortion shows the seismic signal artificial relaxation phenomenon that frequency reduces along with the incident angle increase.Frequency compensation is meant carries out the frequency division amplitude compensation to the earthquake reflective data on the different angles territory, adopts following steps:

1. the amplitude variations that the amplitude drawing coefficient approximate formula based on normal moveout correction below utilizing is proofreaied and correct different offset distances makes statistical average spectral amplitude coupling zero-offset road a long way,

${\mathrm{\α}}_x=\frac{\∂t}{{\∂t}_0}\≈\frac{t_0}{t}(1-\frac{2(t-t_0)}{V_{\mathrm{NMO}}}\frac{{\∂V}_{\mathrm{NMO}}}{{\∂t}_0})---\left(2\right)$

α in the formula _xBe compressibility coefficient, t ₀Be the zero-offset round trip travel-time, t is the round trip travel-time, V _NMOBe the normal moveout correction root-mean-square velocity.

2. utilizing following formula that seismic signal is done with the angle is that the compressed transform of parameter is eliminated offset distance far away, wide-angle incident seismic signal because the frequency reducing phenomenon that nmo stretching caused,

W in the formula _θBe the back wavelet that stretches, w is the zero-offset wavelet, and θ is a reflection angle, and t is the travel-time, and ω is an angular frequency;

(4) all seismic traces of road, diagonal angle collection calculate the instantaneous spectrum of each sampled point, utilize following formula to calculate the instantaneous centre frequency that each seismologic record is concentrated in the road, angle, form centre frequency angle Dao Ji (as shown in Figure 2),

$f_c=\frac{{\∫}_0^{\∞}\mathrm{fP}\left(f\right)\mathrm{df}}{{\∫}_0^{\∞}P\left(f\right)\mathrm{df}}---\left(4\right)$

F in the formula _cBe centre frequency, P (f) is the power spectrum of signal;

(5) utilize the formula of inventing below to concentrate and extract zero degree centre frequency attribute f from road, centre frequency angle _c(0 °) and gradient attribute G (1/Q, t) (as Fig. 3 c),

f _c(θ)＝f _c(0°)+G(1/Q，t)w(sin ²θ)，(5)

F in the formula _c(θ) be the centre frequency of angle θ, (1/Q, t) 1/Q's gradient G reciprocal with travel-time t and stratum quality factor q is directly proportional.The centre frequency of signal is except that outside the Pass having with incident angle, also with vertical relevant from the quality factor q that connects from travel-time t that receives and stratum.At zone thickness one regularly, quality factor is more little, the centre frequency of signal with incident angle reduce fast more, thereby the centre frequency of angle far away has bigger susceptibility to the quality factor on stratum, (1/Q t) has reflected the reduction degree of centre frequency with incident angle to gradient G.Intercept f in the following formula _c(0 °) and gradient G (1/Q, t) can be referred to as centre frequency is the CFVA attribute with the incident angle variation, the leaching process of these two attributes can be converted into finds the solution following system of equations:

θ in the formula _i, i=1,2 ..., n is the incident angle that obtains road, angle collection geological data in the step (2);

(6) by centre frequency angle Dao Ji, utilize the method for part angle stack, promptly the centre frequency trace-stacking in a certain angular range is asked on average together, extract the centre frequency attribute road of far away, nearly angle

With

(as Fig. 3 a and 3b);

(7) utilize the attribute road

And f _cFrequency difference feature between (0 °), and whether gradient attribute G exist bigger negative anomaly value, we just can analyze the variation of the unusual and formation lithology of fluid in the hole, carry out oil and gas detection.Its centre frequency reduced all the time when seismic event was propagated in the stratum, owing to have very low quality factor during formation gas bearing, according to formula 5 as can be known, on far away, nearly angle centre frequency section (as Fig. 3 a and 3b), the center frequency value difference of gas-bearing reservoir can be bigger, but not the center frequency value difference of gas-bearing reservoir is then less, although they also reduce.Therefore, utilize the difference characteristic of the centre frequency attribute of far and near angle can detect gas reservoir.Correspondingly, on the centre frequency gradient profile (as Fig. 3 c) that obtains with the inverting of incident angle variation formula by centre frequency, the centre frequency Grad of gas-bearing reservoir shows as unusual high negative value, but not the centre frequency Grad of gas-bearing reservoir then shows as less negative value, thereby big negative anomaly value is indicated the existence of gas reservoir usually on the gradient attribute.

Centre frequency f _c(θ) be the even power function of the sine function of incident angle.Along with the increase in travel-time, the centre frequency of signal can progressively reduce, and the amplitude of reduction is relevant with the quality factor on stratum.Along with the increase centre frequency reduction of incident angle, quality factor is more little, and centre frequency is fallen soon more.

Practice shows: there is tangible difference in the quality factor of reservoir gas-bearing during with gassiness not, and the quality factor of gas-bearing reservoir is distributed between the 5-30, but not the quality factor of gas-bearing horizon then has bigger value (Klimentos, 1995; Sheriff etc., 1995).Therefore, on far away, the nearly angle centre frequency section that is obtained with the inverting of incident angle variation formula by centre frequency, the center frequency value of gas-bearing reservoir differs greatly, but not the center frequency value difference of gas-bearing reservoir is less, although they also reduce; Equally, on the centre frequency gradient profile, the centre frequency Grad of gas-bearing reservoir shows as unusual high negative value, but not the centre frequency Grad of gas-bearing reservoir then shows as less negative value.

Instance graph 1 of the present invention is a centre frequency following variation of angle of incidence analytical technology flow process of the present invention, 2a and 2b concentrate road, the centre frequency angle composite section that extracts from the other CDP of the well road of gas well (the WC1 well of Fig. 3) and non-gas well (the YL2 well of Fig. 3), and angular range is at 1-33 °.As can be seen, the centre frequency of gas-producing well objective interval (the gas-bearing formation top is positioned at t=4115ms) reduces fast with the increase of incident angle, drop to the 16Hz of 33 ° of angles far away from the 35Hz of nearly angle, the low-frequency anomaly that occurs between angle 9-12 ° then since ground roll cause.But not the centre frequency of gas-producing well changes not quite on the angle near, far away of objective interval.

3a is the centre frequency attribute zero degree section that the CFVA inverting obtains, 3 mouthfuls of drilling wells on the survey line, and wherein WC1 is for containing gas well, and YL2 and WS2 are the non-gas well that contains.As can be seen, WC1 well and YL2 well are more approaching in the center frequency value of objective interval, and they are positioned at same colour code scope, are difficult to identify gas-bearing formation.

Fig. 3 b is the angle centre frequency section far away that angle 22-30 ° centre frequency road is superposeed and obtains, as can be seen, compare with the zero degree centre frequency section of Fig. 3 a, frequency values is whole to be reduced, the WC1 well has low-down center frequency value at objective interval, and expanded range is bigger, although and YL2 well objective interval is positioned at the higher position of structure, show as the high value of centre frequency.

Fig. 3 c is the centre frequency gradient profile that extracts, and as can be seen, it is very big in the centre frequency graded of objective interval to contain gas well WC1, non-contain then change near the gas well YL2 well little.

The centre frequency section (Fig. 4) that post-stack seismic data extracts is although also can reflect the frequency decay characteristics of gas well and non-gas well preferably, but do not compare with Fig. 3 b, it is more messy that the entire profile seems, as containing near the gas well WC1, non-gas bearing interval also shows as characteristics of low-frequency, detects to gas-bearing formation and brings multi-solution.

Claims (5)

1. one kind is utilized centre frequency following variation of angle of incidence information to carry out the method that gas reservoir detects, and it is characterized in that:

Concrete steps comprise:

(1) excite and write down seismic event, earthquake data is routinely carried out high-fidelity and is handled, and formation can be directly used in the road collection of amplitude after with the normal-moveout correction of offset distance mutation analysis;

(2) utilize following formula that offset distance numeric field data in the step (1) is transformed in the angle domain, obtain angular-trace gather,

$\sin \mathrm{\θ}=\frac{{\mathrm{xv}}_p}{v_{\mathrm{rms}}^2{(t_0^2+\frac{x^2}{v_{\mathrm{rms}}^2})}^{1/2}}---\left(1\right)$

In the formula, x is a geophone offset, and Vp is the interval velocity on stratum, t ₀Be the zero-offset whilst on tour, Vrms is the root-mean-square velocity on stratum;

(3) compensation of angular-trace gather working frequency is distorted to eliminate the signal frequency that is caused owing to the NMO stretching effect;

(4) all seismic traces of road, diagonal angle collection calculate the instantaneous spectrum of each sampled point, utilize following formula to calculate the instantaneous centre frequency that each seismologic record is concentrated in the road, angle, form centre frequency angle Dao Ji,

$f_c=\frac{{\∫}_0^{\∞}\mathrm{fP}\left(f\right)\mathrm{df}}{{\∫}_0^{\∞}P\left(f\right)\mathrm{df}}---\left(4\right)$

F in the formula _cBe centre frequency, P (f) is the power spectrum of signal;

(5) utilize following formula to concentrate and extract zero degree centre frequency attribute f from road, centre frequency angle _c(0 °) and gradient attribute G (1/Q, t),

f _c(θ)＝f _c(0°)+G(1/Q，t)w(sin ²θ)， (5)

F in the formula _c(θ) be the centre frequency of angle θ, and gradient attribute G (1/Q, t) 1/Q reciprocal with travel-time t and stratum quality factor q is directly proportional;

Intercept f in the following formula _c(0 °) and gradient attribute G (1/Q t) can be referred to as centre frequency with the incident angle change to attributes, and the leaching process of these two attributes is found the solution following system of equations:

θ in the formula _i, i=1,2 ..., n is the incident angle that obtains road, angle collection geological data in the step (2);

T connects from the travel-time of receiving certainly for vertical;

Q is the quality factor on stratum;

(6) by centre frequency angle Dao Ji, utilize the method for part angle stack, promptly the centre frequency trace-stacking in a certain angular range is asked on average together, extract the centre frequency attribute road of far away, nearly angle

With

(7) utilize the attribute road

And f _cFrequency difference between (0 °), and the negative anomaly value of gradient attribute G determine that the fluid oil gas unusual and the detection formation lithology in the hole changes;

On far away, nearly angle centre frequency section, the center frequency value of gas-bearing reservoir reduces, and difference is big, and the center frequency value difference of non-gas-bearing reservoir is then little;

On the centre frequency gradient profile that is obtained with the inverting of incident angle variation formula by centre frequency, the centre frequency Grad of gas-bearing reservoir is unusual high negative value, but not the centre frequency Grad of gas-bearing reservoir then shows as less negative value.

2. a kind of centre frequency following variation of angle of incidence information of utilizing according to claim 1 is carried out the method that gas reservoir detects, it is characterized in that: it is the relative Changing Pattern that keeps amplitude that the described high-fidelity of step (1) is handled, with reflection, the decay characteristics that reflects that seismic event is propagated in the stratum.

3. a kind of centre frequency following variation of angle of incidence information of utilizing according to claim 1 and 2 is carried out the method that gas reservoir detects, it is characterized in that: the described high-fidelity of step (1) is handled or is referred to that the frequency range of signal and dominant frequency are constant, with the original frequency feature of holding signal.

4. a kind of centre frequency following variation of angle of incidence information of utilizing according to claim 1 is carried out the method that gas reservoir detects, and it is characterized in that: the described distortion of step (3) is meant that seismic signal is along with incident angle increases and the artificial relaxation phenomenon of frequency reduction.

5. a kind of centre frequency following variation of angle of incidence information of utilizing according to claim 1 is carried out the method that gas reservoir detects, it is characterized in that: the described frequency compensation of step (3) is meant carries out the frequency division amplitude compensation to the earthquake reflective data on the different angles territory, adopts following steps:

The amplitude variations that the amplitude drawing coefficient approximate formula based on normal moveout correction below utilizing is proofreaied and correct different offset distances makes statistical average spectral amplitude coupling zero-offset road a long way,

${\mathrm{\α}}_x=\frac{\∂t}{{\∂t}_0}\≈\frac{t_0}{t}(1-\frac{2(t-t_0)}{V_{\mathrm{NMO}}}\frac{\∂V_{\mathrm{NMO}}}{{\∂t}_0})---\left(2\right)$

α in the formula _xBe compressibility coefficient, t ₀Be the zero-offset round trip travel-time, t is the round trip travel-time, V _NMOBe the normal moveout correction root-mean-square velocity;

Utilizing following formula that seismic signal is done with the angle is that the compressed transform of parameter is eliminated offset distance far away, wide-angle incident seismic signal because the frequency reducing phenomenon that nmo stretching caused,

$w_{\mathrm{\θ}}\left(t\right)=w\left(\frac{t}{\cos \mathrm{\θ}}\right)\stackrel{\mathrm{FT}}{\→}\tilde{w}(\mathrm{\ω}\·\cos \mathrm{\θ})---\left(3\right)$

W in the formula _θBe the back wavelet that stretches, w is the zero-offset wavelet, and θ is a reflection angle, and t is the travel-time, and ω is an angular frequency.

Images