METHOD FOR CHARACTERIZING GAS-BEARING RESERVOIR BASED ON LOGGING CONSTRAINT

Abstract
The invention relates to petroleum seismic exploration, and more specifically to a method for gas-bearing reservoir characterization using logging information. In the method, logging information is used as a constraint to indirectly characterize the distribution range of the gas-bearing reservoir by determining the upper and lower boundaries. In addition, this method enables the automatic determination of the optimal calculation parameters according to the characteristics of input data, allowing for more accurate results.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of priority from Chinese Patent Application No. 201811388601.9, filed on Nov. 21, 2018. The content of the aforementioned application, including any intervening amendments thereto, is incorporated herein by reference in its entirety.


TECHNICAL FIELD

The present invention relates to petroleum seismic exploration, and more particularly to a method for gas-bearing reservoir characterization based on logging constraint.


BACKGROUND

In recent years, the exploration for natural gas in China has mainly focused on offshore deep-water oil-gas reservoir; however, since the deep-water oil-gas reservoir is usually susceptible to structure and reservoir heterogeneity, it often has complicated geological conditions, resulting in great difficulties in identification. As recorded in seismic data, the top interface of a gas-bearing reservoir usually shows strong negative reflection, which may be actually caused by other factors such as the tuning thickness of a formation. Therefore, how to use the seismic data to perform the gas-bearing reservoir prediction, especially a high gas saturation reservoir, to reduce the prediction multiplicity, improve the prediction accuracy and reduce the risks of exploration and development, has always been one of the important research contents in the geophysical exploration.


At present, it has been demonstrated by petrophysics experiments that a saturated fluid-containing pore medium reservoir may cause low-frequency large amplitudes and phase delay at a reflection event, so that these features may facilitate the identification of the abnormity of the saturated fluid-containing reservoir in low-frequency energy. Therefore, the low-frequency information in the seismic data is of great significance to the exploration of the gas-bearing reservoir. Since the viscosity of a multiphase saturated fluid is directly determined by gas saturation or oil saturation, the viscosity contained in the fluid mobility parameters of the reservoir fluid may be used to reflect the gas saturation or the oil saturation. Meanwhile, the low-frequency seismic reflection coefficient is a function of the fluid mobility (defined as a ratio of the permeability of the reservoir to the viscosity coefficient of the fluid), which enables that there is a relationship established between the low-frequency information of seismic reflection and the fluid mobility of the reservoir fluid. Therefore, it is feasible to apply the sensitive frequency band of seismic data to invert the fluid mobility of the reservoir fluid to perform the gas-bearing reservoir prediction, especially a high gas saturation reservoir. Chen et al. (2012) proposed a method for calculating the fluid mobility of reservoir fluid using low-frequency seismic information, and this method obtains excellent results in applications of 3D offshore and onshore seismic data process.


However, in some offshore deep-water exploration areas, especially for the anomalous surpressure reservoirs, the fluid mobility of the reservoir fluid calculated according to the low-frequency seismic information is not correlated well with the gas saturation indicated by logging. In addition, some important parameters in the above method are required to be selected based on the experience of a geophysicist or by a trial-and-error method, which has certain subjectivity, affecting the detection for the gas-bearing reservoir.


SUMMARY

This invention provides a method for gas-bearing reservoir characterization based on logging constraint to overcome the difficulties in the seismic detection for gas-bearing reservoirs.


The invention provides a method for gas-bearing reservoir characterization based on logging constraint, including:


(1) inputting a seismic record of a 3D seismic data set, and smoothing two ends of the seismic record to obtain the smoothed seismic record x(t);


(2) substituting the smoothed seismic record x(t) to the following formula to obtain the optimal fluid mobility interface curve FB:







FB
=


[




SP
opt





ω
opt



]

2


;




where ωopt is optimal frequency parameter, and SPopt is energy of optimal instantaneous amplitude spectrum; SPopt is calculated according to the following formula:






SP
opt=[real (Sopt)]2+[imag(Sopt)]2;


where Sopt is optimal instantaneous amplitude spectrum; real (g) represents the real part of Sopt, imag(g) represents the imaginary part of Sopt;


Sopt is calculated according to the following formula:








S
opt

=

ift


[

X
×

exp


(


2






π
2



f
2




α
opt
2



ω
opt

2






β
opt





)



]



;




where X is derived from Fourier transform of x(t); f is a vector constructed by the index numbers of sampling points; αopt and βopt are optimal adjustment parameters, ift(g) represents the inverse Fourier transform;


based on a one-dimension vector ω=[4,5,L, ω.], where ωu, is an upper limit frequency; and a one-dimension vector α=[0.5, 0.6,L,1.5]; and setting the subscripts of the elements in the vector α to be i=1, 2,L,11, optimal parameters ωopt, αopt and βopt are determined according to the following steps:


a) for α(i), setting the subscripts of the elements in the vector ω to be j=1, 2,L, n, and establishing the following objective function for each element ω(j) of ω:






T(α(i), ω(j))=−(ƒ12);


where η1=sum(fbn), η2=sum(−g), sum(g) represents a summation operation;


finding out β, which maximizes the value of objective function T(α(i), ω(j)), by using a one-dimension grid search method; storing β into β(j) of a one-dimensional vector β, and storing the corresponding value of the objective function into T(j) of a one-dimension vector T;


b) finding out the maximum value T (maxid) of T, and finding out the corresponding ω(maxid) and β(maxid) at the location of subscript (maxid); and storing T(maxid), ω(maxid), α(i) and β(maxid) into the i-th row of a two-dimension vector Ω with eleven rows and four columns;


c) setting i=i+1, if i>12, performing step d), otherwise repeating step a) and step b) again;


d) finding out the maximum value in the first column of Ω, and the last three values in the row corresponding to the subscript of the maximum value are the optimal frequency parameter ωopt and the optimal adjustment parameters αopt and βopt, respectively;


in the objective function, fbn is the normalization of the fluid mobility interface curve fb, and fb is calculated according to the following formula:







fb
=


[



SP




ω


(
i
)




]

2


;




where SP is energy of instantaneous amplitude spectrum, which is calculated according to the following formula:






SP=[real(S)]2+[imag(S)]2;


where S is instantaneous amplitude spectrum, which is calculated according to the following formula:







S
=

ift


[

X
×

exp


(

-


2






π
2



f
2




α


(
i
)





ω


(
j
)



2

β





)



]



;




in the objective function, g is calculated as follow:





g=fbngIG;


where the symbol g represents the Hadamard product operation;


a threshold a is determined according to characteristics of input seismic data and logging information, and IG is calculated as the following formula:






{







I
G



(


I
G

<
a

)


=
0








I
G



(


I
G


a

)


=
1









or






{






I
G



(


I
G


a

)


=
1








I
G



(


I
G

>
a

)


=
0










(3) normalizing the optimal fluid mobility interface curve FB to obtain FBn, and standardizing x(t) to obtain xz(t);


(4) determining an upper boundary point set and a lower boundary point set of the gas-bearing reservoir according to extrema on the curve xz(t) and FBn, and then determining indication results of the gas-bearing reservoir by using the upper boundary point set and the lower boundary point set according to the following steps:


a) determining other threshold b according to the information about gas saturation in the logging information, picking up extrema bigger than the threshold b on the curve FBn, and the corresponding index numbers of these extrema on the curve FBn form a contrast point set CP;


b) standardizing the x(t) to obtain xz(t), and picking up extrema on the xz(t), and the corresponding index numbers of these extrema on the curve xz(t) form an information point set I;


c) matching the information point set I with the contrast point set CP to form the candidate point set CA;


d) selecting the maximum point on the curve FBn, finding out a point corresponding to the index number of the maximum point from the candidate point set CA and taking the point as a reference point, taking the reference point as a boundary to divide CA into an upper candidate point set CAL and a lower candidate point set CAB, sorting points in the upper candidate point set CAL in descending order, and sorting points in the lower candidate point set CAB in ascending order;


e) for the point in the upper candidate point set CAL and the lower candidate point set CAB, if its corresponding value in xz(t) is smaller than the average of xz(t), putting the point into the upper boundary point set UB; otherwise if its corresponding value in xz(t) equal or bigger than the average of xz(t), putting the point into the lower boundary point set LB; finally, there is an one-to-one match between the upper boundary point set UB and the lower boundary point set LB;


f) calculating the normalized gas-bearing reservoir indication GI of the input seismic record according to the following formula:







GI


(
j
)


=

{





1
,





UB


(
i
)



j


LB


(
i
)








0
,



others



;






where j is the subscript of an element in GI, and i is the subscript of an element in UB and LB;


g) performing inverse proportional transformation on GI to obtain the gas-bearing reservoir indication result of the input seismic record;


(5) repeating steps (1) to (4) until all the seismic records in the input 3D seismic data set are completely processed to obtain the gas-bearing reservoir indication results of the whole 3D seismic data set.


a plurality of wells are used as constraints to obtain a plurality of sets of the optimal frequency parameters ωopt, and the optimal adjustment parameters αopt and βopt, each set of the optimal parameters is used to calculate the gas-bearing reservoir indication results of the input seismic data set, and the respective calculation results are superimposed to obtain a more accurate result.





BRIEF DESCRIPTION OF THE DRAWINGS


FIG. 1 is a section passing through well A of high gas saturation reservoir indication result in the target area according to an embodiment of the invention; where the abscissa indicates trace number; and the ordinate indicates time in second (i.e. s);



FIGS. 2A-2D show the comparison between the normalized high gas saturation reservoir indication result and the logging information of well A;



FIG. 2A shows five seismic traces nearby well A, where a borehole-side trace of well A is marked by arrow, the abscissa indicates trace number, the ordinate indicates time(s);



FIG. 2B shows a logging gas saturation curve, where the abscissa indicates logging gas saturation (%), the ordinate indicates time(s);



FIG. 2C shows a normalized high gas saturation reservoir indication result (solid line) and the optimal fluid mobility interface curve (dash line), where the ordinate indicates time(s);



FIG. 2D shows a logging interpretation curve, where the abscissa indicates the numerical codes, the ordinate indicates time(s);



FIG. 3 is another section passing through well B of high gas saturation reservoir indication result in the same target area according to an embodiment of the invention; where the abscissa indicates trace number; the ordinate indicates time(s);



FIGS. 4A-4D show the comparison between a normalized high gas saturation reservoir indication result and the logging information of well B;



FIG. 4A shows five seismic traces near well B, where a borehole-side trace of well B is marked by an arrow, the abscissa indicates trace number; the ordinate indicates time (s);



FIG. 4B shows a logging gas saturation curve, where the abscissa indicates logging gas saturation (%), the ordinate indicates time(s);



FIG. 4C shows a normalized high gas saturation reservoir indication result (solid line) and the optimal fluid mobility interface curve (broken line), where the ordinate indicates time(s);



FIG. 4D shows a logging interpretation curve, where the abscissa indicates numerical code, the ordinate indicates time(s);



FIG. 5 is a horizontal slice extracted at the target interval, which shows a 3D high-gas saturation reservoir indication result in the target area.





DETAILED DESCRIPTION OF EMBODIMENTS

The method for gas-bearing reservoir characterization based on logging constraint according to the present application is now described below.


(1) inputting a seismic record of a 3D seismic data set, and smoothing two ends of the seismic record to obtain the smoothed seismic record x(t);


(2) substituting the smoothed seismic record x(t) to the following formula to obtain the optimal fluid mobility interface curve FB:







FB
=


[




SP
opt





ω
opt



]

2


;




where ωopt is optimal frequency parameter, and SPopt is energy of optimal instantaneous amplitude spectrum; SPopt is calculated according to the following formula:






SP
opt=[real(Sopt)]2+[imag(Sopt)]2;


where Sopt is optimal instantaneous amplitude spectrum; real(g) represents the real part of Sopt, imag(g) represents the imaginary part of Sopt;


Sopt is calculated according to the following formula:








S
opt

=

ift


[

X
×

exp


(

-


2


π
2



f
2




α
opt
2



ω
opt

2






β
opt






)



]



;




where X is derived from Fourier transform of x(t); f is a vector constructed by the index numbers of sampling points; αopt and βopt are optimal adjustment parameters, ift(g) represents the inverse Fourier transform;


based on a one-dimension vector ω=[4, 5,L, ωu], where ωu, is an upper limit frequency; and a one-dimension vector α=[0.5, 0.6,L,1.5]; and setting the subscripts of the elements in the vector α to be i=1,2,L,11, optimal parameters ωopt, αopt and βopt are determined according to the following steps:


a) for a(i), setting the subscripts of the elements in the vector ω to be j=1, 2,L,n, and establishing the following objective function for each element ω(j) of ω:






T(α(i), ω(j))=−(ƒ12)


where ƒ1=sum(fbn), ƒ2=sum(−g), sum(g) represents a summation operation;


finding out β, which maximizes the value of objective function T(α(i), ω(j)), by using a one-dimension grid search method; storing β into β(j) of a one-dimensional vector β, and storing the corresponding value of the objective function into T(j) of a one-dimension vector T;


b) finding out the maximum value T(maxid) of T, and finding out the corresponding ω(maxid) and ⊕(maxid) at the location of subscript (maxid); and storing T(maxid), ω(maxid), α(i) and β(maxid) into the i-th row of a two-dimension vector Ω with eleven rows and four columns;


c) setting i=i+1, if i>12, performing step d), otherwise repeating step a) and step b) again;


d) finding out the maximum value in the first column of Ω, and the last three values in the row corresponding to the subscript of the maximum value are the optimal frequency parameter ωopt and the optimal adjustment parameters αopt and βopt, respectively;


in the objective function, fbn is the normalization of the fluid mobility interface curve fb, and fb is calculated according to the following formula:







fb
=


[



SP




ω


(
i
)




]

2


;




where SP is energy of instantaneous amplitude spectrum, which is calculated according to the following formula:






SP=[real(S)]2+[imag(S)]2;


where S is instantaneous amplitude spectrum, which is calculated according to the following formula:








S
opt

=

ift


[

X
×

exp


(

-


2


π
2



f
2




α


(
i
)





ω


(
j
)



2





β





)



]



;




in the objective function, g is calculated as follow:





g=fbngIG;


where the symbol g represents the Hadamard product operation;


a threshold a is determined according to characteristics of input seismic data and logging information, and IG is calculated as the following formula:






{







I
G



(


I
G

<
a

)


=
0








I
G



(


I
G


a

)


=
1









or






{






I
G



(


I
G


a

)


=
1








I
G



(


I
G

>
a

)


=
0










(3) normalizing the optimal fluid mobility interface curve FB to obtain FBn, and standardizing x(t) to obtain xz(t);


(4) determining an upper boundary point set and a lower boundary point set of the gas-bearing reservoir according to extrema on the curve xz(t) and FBn, and then determining indication results of the gas-bearing reservoir by using the upper boundary point set and the lower boundary point set according to the following steps:


a) determining other threshold b according to the information about gas saturation in the logging information, picking up extrema bigger than the threshold b on the curve FBn, and the corresponding index numbers of these extrema on the curve FBn form a contrast point set CP;


b) standardizing the x(t) to obtain xz(t), and picking up extrema on the xz(t), and the corresponding index numbers of these extrema on the curve xz(t) form an information point set I;


c) matching the information point set I with the contrast point set CP to form the candidate point set CA;


d) selecting the maximum point on the curve FBn, finding out a point corresponding to the index number of the maximum point from the candidate point set CA and taking the point as a reference point, taking the reference point as a boundary to divide CA into an upper candidate point set CAL and a lower candidate point set CAB , sorting points in the upper candidate point set CAL in descending order, and sorting points in the lower candidate point set CAB in ascending order;


e) for the point in the upper candidate point set CAL and the lower candidate point set CAB, if its corresponding value in xz(t) is smaller than the average of xz(t), putting the point into the upper boundary point set UB; otherwise if its corresponding value in xz(t) equal or bigger than the average of xz(t), putting the point into the lower boundary point set LB; finally, there is an one-to-one match between the upper boundary point set UB and the lower boundary point set LB;


f) calculating the normalized gas-bearing reservoir indication GI of the input seismic record according to the following formula:







GI


(
j
)


=

{





1
,





UB


(
i
)



j


LB


(
i
)








0
,



others



;






where j is the subscript of an element in GI, and i is the subscript of an element in UB and LB;


g) performing inverse proportional transformation on GI to obtain the gas-bearing reservoir indication result of the input seismic record;


(5) repeating steps (1) to (4) until all the seismic records in the input 3D seismic data set are completely processed to obtain the gas-bearing reservoir indication results of the whole 3D seismic data set.


a plurality of wells are used as constraints to obtain a plurality of sets of the optimal frequency parameters ωopt, and the optimal adjustment parameters αopt and βopt each set of the optimal parameters is used to calculate the gas-bearing reservoir indication results of the input seismic data set, and the respective calculation results are superimposed to obtain a more accurate result.



FIG. 1 is a section passing through well A of high gas saturation reservoir indication result in the target area according to an embodiment of the invention, while there are the optimal frequency ωopt=16 Hz, and the optimal adjusting parameters αopt=0.7 and βopt=21.11. In this case, the logging gas saturation and the interpretation of well A are used as constraints to obtain the optimal parameters ωopt, αopt and βopt. In the target area, the main object is to find out the distribution range of subsurface high gas saturation reservoirs. Although low gas saturation reservoirs contain gas, they are not profitable for the hydrocarbon exploration in deep water.



FIGS. 2A-D shows the comparison between the normalized indication result of the high gas saturation reservoir and the logging information at well A, while there are the optimal frequency ωopt=16 Hz, the optimal adjusting parameters αopt=0.7 and βopt=21.11. In the logging interpretation curve of this area, the numerical code of the high gas saturation reservoir is defined as 11 and the numerical code of the low gas saturation reservoir is defined as 12. The upper limit frequency ωu in step (2) of the embodiment is set to 100 Hz. According to the logging interpretation curve, the threshold a is set to 12, and IG is calculated according to






{







I
G



(


I
G


a

)


=
1








I
G



(


I
G

>
a

)


=
0




.





The threshold b in step (6) of this embodiment is set to 0.4. It can be seen in the FIG. 2 that the normalized indication result (the solid line in FIG. 2C) of high gas saturation reservoir is highly consistent with the logging gas saturation curve (FIG. 2B) and the logging interpretation curve (FIG. 2D).



FIG. 3 is another section passing through well B of the indication result of high gas saturation reservoir according to an embodiment of the invention, while there are optimal frequency ωopt=16 Hz and the optimal adjusting parameters αopt=0.7 and βopt=21.11. In this case, the logging gas saturation and the interpretation of well A are used as constraints to obtain the optimal parameters ωopt, αopt and βopt. Well B is used as a reference.



FIGS. 4A-D shows the comparison between the normalized indication result of high gas saturation reservoir and the logging information at well B, while there are optimal frequency ωopt=16 Hz , optimal adjusting parameters αopt=0.7 and βopt=21.11. It also can be seen that the normalized indication result (the solid line in FIG. 4C) of high gas saturation reservoir is highly consistent with both the logging gas saturation curve (FIG. 4B) and the logging interpretation curve (FIG. 4D) Thus, the result verifies the accuracy of the invention for the prediction of the distribution range of high gas saturation reservoirs.



FIG. 5 is a horizontal slice extracted at the target interval, which shows a 3D high-gas saturation reservoir indication result in the target area. It can be seen that well A and well B are both located in high gas saturation zones. More importantly, in undrilling zones, the slice may provide reference for determining new drilling locations for profitable wells.


The advantages of the invention are described as follows. 1) The logging information is used as constraints, so that the optimal frequency ωopt and the optimal adjustment parameters αopt and βpt can be automatically determined according to the characteristics of the input data. Thus it does not need to manually adjust the parameters. 2) The distribution range of the gas-bearing reservoirs is indirectly obtained by determining the upper and lower boundaries of the gas-bearing reservoirs, which contributes to a more accurate result. 3) The invention has a high computational efficiency, and thus it can be applied to the large-scale 3D seismic data set.


The above embodiments are merely illustrative of the invention, and various variations can be made to the steps in the above embodiments. Any equivalent changes and improvements made without departing from the spirit of the invention should fall within the scope of the invention.

Claims
  • 1. A method for gas-bearing reservoir characterization based on logging constraint, comprising: (1) inputting a seismic record of a 3D seismic data set, and smoothing two ends of the seismic record to obtain the smoothed seismic record x(t);(2) substituting the smoothed seismic record x(t) to the following formula to obtain the optimal fluid mobility interface curve FB:
  • 2. The method of claim 1, wherein in the use of the logging information as constraints, the distribution range of the gas-bearing reservoir is indirectly characterized by determining upper and lower boundaries of the gas-bearing reservoir, and optimal calculation parameters are automatically determined according to the characteristics of the input data.
  • 3. The method of claim 1, wherein a plurality of wells are used as constraints to obtain a plurality of sets of the optimal frequency ωopt, and the optimal adjustment parameters αopt and βopt, and each set of the optimal parameters is used to calculate the gas-bearing reservoir indication result of the input seismic data set, and the respective calculation results are superimposed to obtain a more accurate result.
Priority Claims (1)
Number Date Country Kind
201811388601.9 Nov 2018 CN national