- Home
- Conferences
- Conference Proceedings
- Conferences
81st EAGE Conference and Exhibition 2019 Workshop Programme
- Conference date: June 3-6, 2019
- Location: London, UK
- Published: 03 June 2019
21 - 40 of 93 results
-
-
Porosity Prediction from Shallow Subsurface Seismic Investigation - A Rock Physical Model Approach
Authors Guillaume Sauvin, Maarten Vanneste, Park Joonsang and Madshus ChristianSummaryIn this paper, we present a rock physical model to estimate the porosity from the seismic velocities and compare it with existing rock physical models. A parameter sensitivity analysis is also conducted and the various models are validated with lab data. We also propose a workflow to predict the porosity from seismic velocities at field scale and apply it to a case study.
-
-
-
Time-Lapse Imaging of the Shallow Subsurface at Decimetre Scale Resolution
Authors M. Faggetter, M. Vardy, J. Dix, J. Bull and T. HenstockSummaryHigh-resolution seismic data provides information for many applications including offshore engineering work, where an accurate characterisation of the shallow marine subsurface is essential. However, single 3D volumes only provide a temporal snapshot and do not fully capture the highly dynamic nature of shallow water environments. Although changes at the seabed can be interpreted from repeat bathymetry, only very limited information about the substrate below. Here, we discuss the application of multiple, collocated, ultra-high-resolution (kHz-range) 3D seismic surveys as a tool to investigate changing processes in the marine subsurface. Examining data acquired with the 3D Chirp sub-bottom profiler, two case study examples will be presented. Results illustrate the capability for quantitative mapping of subsurface differences at decimetre-scale resolution using bin sizes of 0.25 cm and smaller.
-
-
-
High Resolution Imaging and Quantitative Analysis of HV Cable and Pipeline Trenching in the Marine Environment
Authors Justin Dix, Mark E Vardy, Michael Faggetter, David White and Peter AllenSummaryThe life time performance of both HV cables (ORE inter -array and export cables and cross-continental shelf interconnectors) and oil and gas pipelines are limited by the physical properties of the sediment in which the cable/pipeline is buried. In the case of HV cables the burial material and burial depth have implications for heat dissipation from the cable, which in turn plays a primary role in cable rating and its lifetime operation and maintenance. For a pipeline changes in the density and strength of the overburden material can impact on buckling potential once in operation. Our current understanding of the key physical parameters of the sediment (e.g. grain size, porosity, permeability, thermal conductivity, relative density and strength) are based on in situ measurements of the ambient condition and rarely take account of physical property changes during the trenching process. We provide initial acoustic inversion results from high resolution 3D Chirp volumes from both a prototype scale, CPT calibrated, tank experiment and in situ trenched cables in a range of substrates. We shall demonstrate the potential of acoustic inversion to non-destructively quantify trench disturbance in this critical engineering scenarios.
-
-
-
Sub-Surface Imaging at the Rockall Basin, Using Travel Time Tomography and Full Waveform Inversion
More LessSummaryThe geological structure beneath the Rockall basin and the nature of the crust are largely undefined because of the sill intrusion, lack of the seismic data coverage and deep well data penetration. The basaltic rocks prevent the seismic waves from travelling underneath them and make it hard to image below these high velocity layers. Here, we perform travel time tomography on long streamer data sets along a 80 km long profile to get a smooth P-wave velocity model using the first arrival travel time. The 2D seismic lines were acquired in 2013–14 using 10 km long streamers. We pick first arrival travel time from the shot gather after cleaning the data set. The velocity model obtained here, indicate the velocity from 1.6–4 km/s for the sediments and we also observe very high velocity ~ 6–7 km/s just 3 km below the sea-floor. This high velocity structure could be the lower crust pinching out at the Rockall high. This velocity model is used as a starting model for full waveform inversion (FWI) to get a higher resolution velocity structure.
-
-
-
Opportunities and Challenges in Multi-Component Processing
More LessSummaryThis talk gives an overview of the opportunities for a wider use of multi-component data and the requirement for the processing which needs to be met in order to ensure that the full value of the data are realized.
-
-
-
Multi-Component Processing on Land for Complimentary PP and PS Imaging and Characterisation
Authors R. Johnston and A. AyreSummaryThe use of multi-component seismic data has risen steadily since the ‘pure’ shear-wave acquisition activity onshore of the 1980s, through the converted-wave ‘revolution’ offshore in the 1990s and beyond. To benefit the most from multi-component seismic we hope for a situation where data quality of the entire elastic wavefield is similar which allows the complimentary information from compressional and shear waves to contribute together. Where either of these waves are compromised, the combined potential will necessarily be sub-optimal. To respond to Canada's Northern Alberta shallow unconventional reservoirs and provide high resolution imaging and characterisation, the industry has developed using point source and multi-component point receiver seismic. We describe the challenges addressed in processing for high resolution PP and PS imaging which deliver improved images and complimentary PP and PS attributes for characterisation. In this heavy oil unconventional reservoir setting, PS images are revealing new details about the geology not previously seen on P-wave data alone. The multi-component data have positively influenced an appraisal drilling program and subsequent pad layout, to take advantage of the geological variations and high grade the reservoir development plan.
-
-
-
Extraction of Acquisition and Processing Attributes from a Time-Lapse Ocean-Bottom Seismic Elastic Finite-Difference Study
Authors R. Whitebread, P. Kristiansen and M. BranstonSummaryObserving changes of seismic amplitude and event timing over time within recorded seismic data are a well-established path to understanding the subtler changes to reservoir properties during injection and production phases of reservoir development. To understand the impact of acquisition tolerances and factors such as noise, ghost, and multiple energy on these reservoir changes, we must move away from real recorded seismic data and look at key issues in isolation with synthetic data.
We created a small four-component ocean-bottom seismic (OBS) data set using elastic finite-difference modelling where we have full control of the physical changes at the reservoir as well as free-surface effects and deviations from nominal acquisition geometry.
We chose elastic finite-difference modelling to include interface waves into PP and VZ data as well as to allow full modelling VX and VY components to infer 4D response from the subsequently derived radial component.
Our results confirm that geometry errors rapidly degrade the observed 4D response, but that the cumulative effect of geometry deviations, tidal statics, and water velocity corrections could also mask expected reservoir changes.
-
-
-
Multi-Component Seismic Data Processing of a 3D 3C Dataset for Sand Filled Channels and Rock Property Identification via PP/PS Joint Inversion : A Case History
More LessSummaryMulti-component seismic data processing of a 3D 3C dataset for sand filled channels and rock property identification via PP/PS joint inversion: A case history.
In this case history, we present improvements in sand filled channel identification and associated fracture detection, including fracture density and orientation. This was achieved by performing seismic data processing of vertical component (PP) and converted wave (PS) data for a 3D 3C dataset.
A superior quality image of the P-wave using 3D 3C data will be demonstrated by the robustness of the applied processing sequence and consistency of the utilized processing and Q.C. tools. In addition, we will demonstrate in detail the converted wave measured from the horizontal component processing.
Incorporating the interpreter's geological knowledge into the processing workflow is extremely crucial where intermingling of the data processing sequence, algorithms and geological knowledge plays a major role in obtaining successful final PS data results. Integrating PP / PS images by introducing both components in one processed solution should alleviate the interpretation challenges at the target reservoir interval, due to the added precision in elastic properties derivation through the PP/PS joint inversion process.
-
-
-
Multi-Component Seismic Data Processing of a 3D 4C OBC Dataset for Lithological Identification Through PP/PS Joint Inversion: A Case History
More LessSummaryIn this case history we present the advantages and improvements in lithological discrimination and identification by implementing a robust seismic data processing workflow of vertical component (PP) and converted wave (PS), for a 3D 4C dual sensor OBC dataset.
The image quality of the P-wave using dual sensor data will be demonstrated by the robustness of the applied processing sequence and quality control tools. We will demonstrate in details the converted wave measured from the horizontal component processing, where the interpreters geological knowledge, incorporated into the integration of processing applications, plays a major role in the success of the final PS data results.
Integrating PP / PS images and introducing PP, PS processed data in one solution should enhance the interpretation resolution at the target interval. Improved Vp/Vs for prominent markers and more accurate AI through PP/PS joint inversion is continuously evaluated throughout the processing workflow. Eventually, our main objective is the utilization of PP/PS processed data in joint inversion to produce accurate lithology discrimination maps.
-
-
-
Anisotropy processing of multi-component seismic data – Part 1: Theory and Implementation
More LessSummaryAn oil-gas reservoir can be treated as an orthorhombic medium which can be considered as a combination of VTI and HTI media. It means, for a given azimuthal direction, that an orthorhombic medium can be treated as a VTI medium. To process the seismic reflection data acquired from such a medium (the reservoir), the common practice is to separate the processing into two stages: anisotropy processing for a VTI medium and anisotropy processing for a HTI medium. In the first stage, we can ignore the HTI features and apply VTI anisotropy processing to this data, from which we can estimate the velocity and anisotropy parameters and obtain seismic images. Actually, the results are the average results for the whole dataset. Then we use the estimated velocities and anisotropy parameters as an initial model for HTI processing. In HTI processing, we need to separate the seismic data according to the azimuthal direction of the offset. Then we can carry on VTI processing for each set of azimuthal data. The results will be dependent on the azimuthal direction. The azimuthal dependence of velocity and anisotropy parameters can be used to estimate the fracture direction and strength of anisotropy of the reservoir.
-
-
-
Anisotropy Processing of Multi-Component Seismic Data – Part 2: Processing Demonstration Based on CXtools
More LessSummaryIn Part 1, I introducted the theory for anisotropic processing in VTI and HTI media and CXtools, a processing package which implements the theory. In this part, I will show how to use CXtools to carry out anisotropic processing based on a 3D multicomponent seismic dataset. During the demonstration, I also show how to adjust the velocity and anisotropy parameters to improve the results. This live demonstration is separated into three sessions:
- How to run the CXtools
- Anisotropic processing for VTI media
- Anisotropic processing for HTI media
-
-
-
Anisotropy Processing of Multi-Component Seismic Data – Part 3: Pre-Stack Time Migration
More LessSummaryPrestack time migration is an efficient imaging method for processing seismic data due to its input/output flexibility and target-orientation, and has recently become a routine step in the seismic data processing flow. Prestack Kirchhoff time migration can produce high-quality migrated images from real seismic data with the correct velocity models. Important issues in applying prestack time migration are how to obtain the correct velocity model, and selecting suitable travel time formulae for different wave types (PP and PS) in different media. In this part, I will introduce the theory for prestack time migration in a VTI medium, and a method for obtaining the optimal velocity model for prestack time migration from seismic data.
-
-
-
How High Can We Go? Pushing Limits of Lateral and Vertical Resolution in Deepwater Seismic
Authors P. Hatchell, P. Dutta, S. Bakku and Z. YangSummaryWe acquired low-cost high-resolution 3D (HR3D) streamer seismic surveys using the P-Cable system with a small air-gun source array (300 in3) over deepwater fields with water depths ranging from 900m to 3000 m. The P-Cable HR3D streamer system employs multiple short streamers (100 m) connected to a cross-cable. In our surveys we deployed 18 streamer cables with each of the 100 m streamer cables having 16 hydrophone groups spaced at 6.25 m intervals. With shot intervals spaced every 12.5 m, the nominal bin size of this configuration is 6.25 m × 3.125 m and the fold is four. In the Shallow sections, we achieved migrated images with frequencies up to 200 Hz. The lateral resolution of the images is found to be superior to that from high-resolution processing of conventional data (streamer and OBN). In the deeper section, the frequencies dropped to much lower values indicating that the earth has more control on the high frequency end than the geophysicist.
-
-
-
Imaging with Near-Field Hydrophones
Authors Pete Nevill, Kevin Davies, Shanice Mohammed, Krzysztof Ubik, Richard Jupp, Ed Kragh and Philip ChristieSummaryNear Field Hydrophone (NFH) data are routinely collected during marine acquisition and historically these data were used to QC air-gun timings and/or other air-gun-related issues. More recently (last 10–15 years), these data were recorded and employed during processing to aid both 1D and 2D source-signature deconvolution. The results presented here demonstrate it is possible to obtain a high-resolution image (greater than 100 Hz) of the near subsurface (0–1 s) using passive NFH array data. Similar results are also achieved with the active NFH array although noise handling in processing is more difficult. NFH data for imaging, can be a natural by-product of conventional marine seismic acquisition, with minimal additional processing cost. Potential uses could include: replacing requirements for a high-resolution 2D acquisition for site surveys, utilizing a source-only vessel for localized overburden and reservoir 4D monitoring, as well as improved deghosting and integration with distributed acoustic sensor vertical seismic profile data.
-
-
-
Broadband De-Signature for Air-Gun Arrays
Authors R. Telling and S. GrionSummaryDe-signature processing of broadband seismic data demands reliable signature estimation in the band 2–200 Hz. Here we discuss estimation of signatures via inversion of near-field hydrophone data. This uses a model for the propagation of energy from each source point to each hydrophone in the array, incorporating bubble motion and ghosting at the sea-surface. In the standard approach we solve for a set of notional sources, assuming a simple model for the ghost, with rough sea effects treated statistically using a frequency and angle varying reflection coefficient corresponding to the observed sea-state.
We discuss the successes of this standard approach and observed problems, specifically with predicted ghost amplitudes, that in practice leads us to parametrize the model using effective sea state parameters often larger than observer logs suggest. The physical reasons for this are linked to onset of cavitation in the water column. We then present results for an alternative approach, employing additional hydrophones, that solves directly for the down-going part of the signature without need for ghost model parametrization. We assess the quality of signatures estimated via this approach, their application to de-signature processing and examine sensitivity of this inversion to noise compared to the standard parametrized approach.
-
-
-
New Wave Seismics
More LessSummaryThis talk will focus on the field of seismic oceanography. Seismic oceanography exploits acoustic energy reflected from temperature and salinity boundaries in the water column to map oceanic structure at unprecedented horizontal resolutions. New insights into four-dimensional ocean dynamics that marine seismic data has to date provided, along side potential future applications, will be reviewed. In addition, the challenges and opportunities presented by processing water column seismic reflection data, as opposed to sub-surface datasets, will be discussed. Advances include: estimating temporal changes in the water column during seismic data collection; quantifying noise, water turbulence and wave energy from seismic data; and inversion techniques to extract the high resolution temperature and salinity structure with precision uncertainty. Such non-conventional approaches to seismic data processing, alongside better quantifying the influence of the water column on the quality of marine seismic data, will be of interest to this community. Finally, the integration of autonomous systems into the acoustic mapping of oceanic thermohaline structure will be considered.
-
-
-
Finite Difference Modelling Including Dynamic Speed of Sound in Water
Authors K. Davies, J. Stefani, L. Zhuo and T. JohnsenSummaryIdentifying and implementing fit for purpose Earth complexity in synthetic modelling is an important part of testing the ability to recover broadband signal. Effects of dynamic speed of sound in water Vw(x,y,z,t) have widely been recognized in 3D marine processing, although 4D often highlights the complexity and requirement for adequate measurement and correction. Corrections are applicable to conventional 3D and VSP marine applications, even if the effect is not immediately obvious. In 2013 Chevron in conjunction with WesternGeco investigated, quantified and ranked the effects of Vw(x,y,z,t) in streamer 4D using finite difference modelling. However, modelling with Ocean Bottom Node data presents additional challenges. In 2015, Chevron built a complex 4D dynamic water column model closely based on Vw(x,y,z,t) observation. The model was specifically designed for evaluation of 4D OBN data. Processing in conjunction with CGG demonstrated the significance of no-correction, after typical correction and with alternative methods, including Up-Down deconvolution, the latter demonstrating a remarkably good efficient solution. Furthermore, Up-Down deconvolution enables efficient source signature deconvolution for enhanced bandwidth. Including the complexity of Vw(x,y,z,t) in modelling has enabled estimation of 4D signal to noise and design of mitigation measures in both acquisition and processing.
-
-
-
Why 4D Broadband is the Next Standard for Reservoir Monitoring Studies?
Authors D. Lecerf, B. Caselizt and C. ReiserSummaryFor the last ten years, the seismic industry has been offering broadband 3D seismic. Broadband acquisition and processing technologies are appealing for 4D time-lapse surveys. As for any 4D requirement, they must deliver seismic signal repeatability for an extended frequency range to be qualified as a 4D broadband solution. Multisensor streamer systems offer an optimum platform for acquiring 4D broadband data in an efficient way. Deeper tow depths give a better signal-to-noise ratio and allow an improved acquisition weather window. In addition, multi-sensor recording provides receiver ghost-free data insensitive to the sea state.
We will discuss the different options for introducing a new broadband dataset into the 4D reservoir cycles. In a genuine 4D broadband acquisition context, we will describe how the repeatability of broadband streamer data can compete with the repeatability of 4D OBN and why an extended bandwidth 4D signal allows to better characterize the seismic image variations due to reservoir production. Finally, we conclude that 4D broadband acquisition, processing and interpretation can complete the challenge of quality and efficiency improvement.
-
-
-
Multi-Parameter Waveform Inversion Using Broadband Data from the North West Shelf of Western Australia
Authors C. Manuel, J. Washbourne, D. Sibley, L. Duranti, M. Merino and B. BoulahanisSummaryWe present a multi-disciplinary approach to determine overburden anisotropy on the North West Shelf of Australia. Beginning with a synthetic dataset we demonstrate the significance of incorrect overburden anisotropy in seismic imaging and develop a workflow for solving the problem. This is followed by application of our workflow to a field case study involving the use of geological inference, log-based rock property relationships, and surface seismic data. Comparisons are made between applying the workflow using conventional and broadband streamer datasets acquired over the same area.
-
-
-
High-Frequency Full-Waveform Inversion: Just How High Should You Go?
Authors A. Ratcliffe, S. Bretherton, B. Xiao and R. HaackeSummaryWe pose the question, “how high a frequency should you go to in FWI?” The answer depends on your objective: the traditional processes of imaging, reservoir characterization, and interpretation, or as a potential complete replacement for these. In this paper we discuss and demonstrate the impact of the maximum frequency in the FWI velocity model on these processes, using data sets from the North and Norwegian Seas.
-