Exxonmobil Upstream Research Company

Exxon Production Research Company

Christine E Krohn

John E Anderson

Ramesh Neelamani

Max Deffenbaugh

Jerome R Krebs

Partha Sarathi Routh

Charlie Jing

Thomas A Dickens

Marvin L Johnson

David L Hinkley

Peter Traynin

Jerry Krebs

Arthur H Thompson

Warren S Ross

James A Rice

Enru Liu

Mark A Miller

Louis M Houston

Julianna J Toms

Alan J Katz

Mamadou S Diallo

Nathan Downey

J Dave Fox

Method for selecting an acquisition geometry for a seismic survey based on ability to resolve an a priori velocity model. Two or more candidate acquisition geometries are selected (), differing in areal coverage and cost to perform. Then compute a synthetic seismic dataset for each geometry using a detailed geometrical reference model of the subsurface (). Invert the synthetic seismic datasets preferably using simultaneous source FWI, and preferably with Volume of Investigation constraints, to determine model updates (). Quantitatively assess the value of the additional traces in a fuller dataset relative to a subset (), using one or more statistics based on the accuracy of the updated models, such as improved match to the reference model, better fit of seismic data, or rate of change in improvement with iterations. Inversions may be cascaded for further efficiency ().

Seismic survey design using full wavefield inversion

Method for acquiring, at reduced acquisition cost, seismic data using simultaneous, field-encoded sources in the field (), and then constructing pseudo source-records () that better meet the requirements for using additional simultaneous computer-encoded sourcing for computer simulations or forward modeling () as part of () iterative FWI (Full Wavefield Inversion) or RTM (Reverse Time Migration), with additional reduction in computational costs. By better meeting the requirements of simultaneous sourcing for FWI or RTM (), artifacts and crosstalk are reduced in the output. The method can be used for marine streamer acquisition and other non-fixed spread geometries to acquire both positive and negative offsets and to mitigate the 'missing data' problem for simultaneous-source FWI. It can also be used for land data to overcome issues with moving spreads and long continuous records.

Simultaneous sourcing during both seismic acquisition and seismic inversion

Method for inverting seismic data to obtain reflection properties by estimating or measuring () the source signature for the seismic data, then inverting () simultaneously for subsurface formation properties or reflection properties and for the amplitude attenuation and velocity dispersion effects integrated over the raypath from the source to the reflectors and to the receiver.

Seismic inversion for formation properties and attenuation effects

Method for estimating fluid heterogeneity in a subsurface region from seismic wave attenuation or velocity dispersion that is either measured or extracted from geophysical data (). A rock physics model in the form of a mathematical relationship is selected that relates attenuation or velocity to frequency and to physical properties that are related to fluid heterogeneity (). The model, or asymptotes () representing the model's behavior at frequency extremes, is inverted () and that relationship is used to obtain () one or more of the physical properties related to fluid heterogeneity, such as characteristic length scale for fluid saturation heterogeneities () and relative volume fractions of the fluids saturating the pore space ().

Estimation of production sweep efficiency utilizing geophysical data

A seismic data processing method for predicting and removing unwanted near-surface noise, e.g. refracted waves and surface waves, from composite seismograms acquired with multiple, encoded seismic sources in which the time records from the different sources overlap or interfere. Surface-consistent properties are determined () for the near-surface noise and used to predict the noise waveforms () for the composite simultaneous-source seismograms, and the waveforms are then subtracted or adaptively subtracted () from the seismograms.

Near-surface noise prediction and removal for data recorded with simultaneous seismic sources

A method for transforming seismic surface waves into depth profiles of dynamic properties of the soil such as the soil shear modulus or wave velocity or wave attenuation. The invention uses seismic data recorded preferably with a plurality of sources and receivers with a two-stage inversion approach. First, variations in surface-wave waveforms are decomposed () into surface-consistent transfer functions preferably for each source, each receiver, and each small region () of the surface. Then the transfer functions for each region are inverted () to determine soil properties or near-surface properties (such as the shear modulus) as a function of depth. The method can solve for the complex multi-mode nature of the surface waves for a media with both vertical and lateral changes in properties. Errors and limits in resolution for traditional methods from misidentification of ground roll modes or from assuming laterally uniform soil properties are avoided.

Estimation of soil properties using waveforms of seismic surface waves

The invention is a method to predict surface-wave waveforms () and subtract them () from seismic data. Prediction is done by estimating a set of surface-consistent components (transfer functions in the frequency domain or impulse responses in time domain) that best represent changes in the waveforms for propagation along the surface from source to receiver (). The prediction uses a mathematical expression, or model, of the earth's filtering effects, both amplitude and phase, as a function of frequency. The desired surface-consistent components are model parameters, and model optimization is used to solve for the surface-consistent components. The surface-consistent components may include filter transfer functions for each source location, each receiver location, and for propagation () through each region () of the surface that exhibits lateral variation.

Removal of surface-wave noise in seismic data

The invention discloses a way to recover separated seismograms with reduced interference noise by processing vibroseis data recorded (or computer simulated) with multiple vibrators shaking simultaneously or nearly simultaneously (). A preliminary estimate of the separated seismograms is used to obtain improved seismograms (). The preliminary estimate is convolved with the vibrator signature and then used to update the seismogram. Primary criteria for performing the update include fitting the field data and satisfying typical criteria of noise-free seismograms (). Alternative ways to update are disclosed, including signal extraction, modeled noise extraction, constrained optimization based separation, and penalized least-squares based separation. The method is particularly suited for removing noise caused by separating the combined record into separate records for each vibrator, and is advantageous where the number of sweeps is fewer than the number of vibrators ().

Separation and noise removal for multiple vibratory source seismic data

Method for reducing the time needed to perform geophysical inversion by using simultaneous encoded sources in the simulation steps of the inversion process. The geophysical survey data are prepared by encoding () a group of source gathers (), using for each gather a different encoding signature selected from a set () of non-equivalent encoding signatures. Then, the encoded gathers are summed () by summing all traces corresponding to the same receiver from each gather, resulting in a simultaneous encoded gather. (Alternatively, the geophysical data are acquired from simultaneously encoded sources.) The simulation steps needed for inversion are then calculated using a particular assumed velocity (or other physical property) model () and simultaneously activated encoded sources using the same encoding scheme used on the measured data. The result is an updated physical properties model () that may be further updated () by additional iterations.

Iterative inversion of data from simultaneous geophysical sources

An exemplary method for filtering multi-component seismic data is provided. One or more characteristics of the seismic data corresponding to a relative manifestation of surface wave noise on the different components, such as polarization attributes, are identified. A time-frequency boundary in the seismic data is also identified in the time-frequency transform domain, the time-frequency boundary delineating portions of the seismic data estimated to contain surface wave noise (). Finally, filtered seismic data are created () by removing portions of the seismic data that correspond to the identified characteristics of surface waves and that are within the time-frequency boundary ().

Method for performing constrained polarization filtering

A method for simultaneously operating multiple seismic vibrators using continuous sweeps (little or no 'listening' time between sweeps) for each vibrator, and recovering the separated seismic responses for each vibrator with the earth signature removed. Each vibrator is given a unique, continuous pilot signal. The earth response to the motion of each vibrator is measured or estimated. The vibrator motion records for each vibrator and the combined seismic data record for all the vibrators are parsed into separate shorter records. The shorter records are then used to form a system of simultaneous linear equations in the Fourier transform domain, following the HFVS method of Sallas and Allen. The equations are then solved for the separated earth responses.

Method for continuous sweeping and separation of multiple seismic vibrators

The present invention is a method of processing seismic data in which one or more seismic vibrators are activated with one or more pilot signals and vibrator motions are recorded along with seismic data. Vibrator signatures are computed from measured vibrator motions, such as the ground force signal. A desired impulse response is specified from either a measured vibrator motion or from test data or field data from a location near the location from which the seismic data was acquired. A deconvolution filter is computed from the impulse response and the vibrator signature. Alternatively, a single separation and deconvolution filter is derived from the impulse response and from vibrator signatures from multiple vibrators and sweeps. The deconvolution or deconvolution and separation filter is used to process the seismic data. The vibrators are then moved to a new location, and the activation is repeated.

Shaped high frequency vibratory source

A drill bit drilling a borehole is used as a strong high frequency, subsurface seismic source in combination with at least one receiver in a preexisting borehole to determine the continuity of a subsurface formation layer extending between the borehole being drilled by the drill bit and the borehole in which the receiver is located. The receiver(s) in the preexisting borehole may be any conventional seismic receiver or receivers, such as hydrophones or geophones. The receiver or receivers are positioned in the preexisting borehole at a depth adjacent the formation layer whose continuity is to be determined. The depth (position), velocity and thickness of the layer is determined from a conventional well log of the preexisting borehole. A frequency power spectrum for a preselected frequency range (the range is selected based on the logging data) may be computed from the signals recorded by the receiver(s). Several such spectra may be computed from signals recorded for several different time intervals and then summed to increase the signal-to-noise ratio. A large amplitude in this power spectrum (or the detected signals) is indicative of a low velocity layer that is continuous between the drill bit source and the borehole containing the receivers.

Growing community of inventors

Houston, TX, United States of America

Christine E Krohn