System and method for the removal of shallow water multiples using a hybrid multi-channel prediction method

Patent metadata

Field	Value
Publication number	US-9477000-B2
Application number	US-201314107583-A
Country	US
Kind code	B2
Filing date	Dec 16, 2013
Priority date	Jan 11, 2013
Publication date	Oct 25, 2016
Grant date	Oct 25, 2016

Abstract

A system and method are provided for determining shallow water multiples when seismically exploring a geographical area of interest under a body of water. The system and method estimate a multi-channel prediction operator F using a model of water layer related multiples with respect to received and stored seismic data, estimate a travel time of the transmitted seismic wavelets from the one or more sources to each of the plurality of receivers, and then generate water layer primary reflections models using the estimated travel time and Green's function. The system and method then merge the generated water layer primary reflections models with the multi-channel prediction operator F to create a hybrid multi-channel prediction operator F H , and convolute the hybrid multi-channel prediction operator F H with the stored received data to determine a final multiples model.

First claim

We claim:

1. A method for determining a shallow water multiples model, the method comprising:
generating a series of underwater seismic wavelets by one or more sources;
acquiring seismic data from a plurality of receivers, and storing the seismic data;
estimating a prediction operator by a shallow water demultiple (SWD) technique using the acquired receiver seismic data;
generating a multi-channel water layer de-multiple (MWD) primary reflections model using the acquired receiver seismic data; and
merging said SWD prediction operator with said MWD primary reflection model and with said acquired receiver seismic data to determine the shallow water multiples model,
wherein the step of generating a MWD primary reflections model comprises:
estimating a travel time of the transmitted seismic wavelets from one or more sources to each of a plurality of receivers; and
generating the MWD primary reflection model using the estimated travel time and Green's function.

2. The method according to claim 1 , wherein the step of estimating an SWD prediction operator comprises:
using a model of water layer related multiples, M with respect to said acquired receiver seismic data, P; and
determining the prediction operator according to
M=ΔP w−1 P and
M=FP,

wherein the SWD prediction operator F is equivalent to ΔP, a primary response, and w −1 is a source wavelet.

3. The method according to claim 1 , wherein Green's function is evaluated according to the expression
G ( {right arrow over (s)},{right arrow over (r)} ;ω)=∫ G ( {right arrow over (s)},{right arrow over (r)} ;ω) F ′( {right arrow over (s)},{right arrow over (x)},R ) G ( {right arrow over (X)},R ;ω) d{right arrow over (x)}

where {right arrow over (s)} and {right arrow over (r)} are source and receiver locations, respectively, F′ is an auto-picked water bottom event from the multichannel prediction operator F at location {right arrow over (x)}, and ω is frequency.

4. The method according to claim 3 , wherein auto-picking is performed using one or more the following criteria: estimation of the time of the water bottom reflection event, amplitude, amplitude ratio, and neighbouring trace information.

5. The method according to claim 1 , wherein the step of estimating the travel time of the transmitted seismic wavelets from the one or more sources to each of the plurality of receivers comprises:
evaluating a second Green's function according to the following expression

G
=

j
4

×

H
0

(
2
)

⁡

(
kr
)

,

wherein j=√{square root over (−1)}

k
=

ω
c

,

where ω is frequency, c is velocity, r is a distance from a first source to a first receiver, and H 0 (2) is a 0 th order Hankel function of the second kind, and further wherein the variable “kr” is equal to

ω
c

×
r

,

which is equal to ω×t, where t is the travel time, which is the same as distance r divided by velocity c.

6. The method according to claim 1 , wherein the step of merging the SWD prediction operator with the MWD primary reflection model and with acquired receiver data to determine a shallow water multiples model comprises:
merging the generated MWD primary reflection models with the SWD prediction operator to create a hybrid prediction operator; and
convoluting the hybrid prediction operator with the acquired receiver data to determine the shallow water multiples model.

7. The method according to claim 6 , wherein the step of merging the generated MWD primary reflections models with the SWD prediction operator to create a hybrid prediction operator comprises:
overlaying Green's function on top of the SWD prediction operator.

8. A method for predicting a shallow water multiples model in seismic receiver data comprising:
generating a series of underwater seismic wavelets by one or more sources;
acquiring seismic data from a plurality of receivers, and storing the seismic data;
estimating a multi-channel prediction operator F by shallow water demultiple (SWD) technique;
generating a model of the Green's functions of the water layer primary reflections;
merging the generated water layer primary reflections model with the multi-channel prediction operator F to create a hybrid multi-channel prediction operator F H ; and
convolving the hybrid multi-channel prediction operator F H with receiver data to determine the shallow water multiples model.

9. The method according to claim 8 , wherein the step of estimating a multi-channel prediction operator F by shallow water demultiple (SWD) technique comprises:
using a model of water layer related multiples, M with respect to said acquired receiver seismic data, P; and
determining the multi-channel prediction operator F according to
M=ΔPw −1 P and
M=FP,

wherein the multi-channel prediction operator F is equivalent to ΔP, a primary response, and w −1 is a source wavelet.

10. The method according to claim 8 , wherein the step of generating a model of the Green's functions of the water layer primary reflections comprises:
estimating a travel time of the transmitted seismic wavelets from one or more sources to each of a plurality of receivers; and
generating the model of the Green's functions of the water layer primary reflections using the estimated travel time and Green's function.

11. The method according to claim 10 , wherein Green's function is evaluated according to the expression
G ( {right arrow over (s)},{right arrow over (r)} ;ω)=∫ G ( {right arrow over (s)},{right arrow over (x)} ;ω) F ′( {right arrow over (s)},{right arrow over (x)},r ) G ( {right arrow over (x)}, r ;ω) d {right arrow over ( x )}

where {right arrow over (s)} and {right arrow over (r)} are source and receiver locations, respectively, F′ is an auto-picked water bottom event from the multichannel prediction operator F at location {right arrow over (x)}, and ω is frequency.

12. The method according to claim 11 , wherein auto-picking is performed using one or more the following criteria: estimation of the time of the water bottom reflection event, amplitude, amplitude ratio, and neighbouring trace information.

13. The method according to claim 10 , wherein the step of estimating a travel time of the transmitted seismic wavelets from the one or more sources to each of the plurality of receivers comprises:
evaluating a second Green's function according to the following expression

Assignees

• Cgg Services Sa

Inventors

• Yang Kunlun
• Hung Barry

Classifications

• G01V1/362 Primary

  Effecting static or dynamic corrections; Stacking · CPC title

• G01V1/28 Primary

  Processing seismic data, e.g. for interpretation or for event detection (G01V1/48 takes precedence) · CPC title