Oilfield Technology - May/June 2020

May/June 2020

Oilfield Technology

the salt. These issues are difficult to resolve using existing streamer data

and conventional imaging technology (Figure 1).

However, the use of an appropriate data-drivenmodel building

algorithm, such as TL-FWI, with the right OBNdata can unlock the details

of the salt complexity. As seen fromthemodel comparison in Figures 2a

and 2b, the technologywas able to correct the initial saltmisinterpretation

and add precious velocity details that would be impossible to achieve

using conventional tomographic ormanual interpretationmethods.

The significant contrast between shallow, slowsediment and the deep

super-fast Cretaceous is accuratelymodelled and the salt can nowbe

easily interpreted. The image on the right-hand side of Figure 2b, produced

using this detailed velocitymodel, has capitalised on the benefits of

TL-FWI. In terms of the value to interpretation, there is significant uplift

for a number of important features. There is no longer distortion of the

Oligocene layer (6 to 7 km) and the truncations in the steeply-dipping

Cretaceous layers against the salt are greatly improved (below8 km).

Data-driven imaging for complexgeology

Structural imaging of theHerschel field is not the end of the story. The

RTM images still suffer from irregular illumination andmigration artifacts

originating fromboth the salt canopy and the small areal extent of the

survey, causing edge effects. An imaging process using a newLSRTM

approach

was implemented to better handle these challenges. This

approach estimates the inverse of the forwardmodelling operator

(i.e. the imprint of themigration process on the imperfect dataset),

then applies this for different structural dips, frequencies and offsets to

compensate for frequency-, angle- and offset-dependent illumination

effects. Compared to the RTM, LSRTM further improves the image

withmore balanced amplitudes, improved resolution, and an

enhanced signal-to-noise ratio (Figure 3). Figure 4, representing

the extracted amplitude along an interpreted horizon at a depth

of approximately 5 km, shows the extent towhich the loss of

illumination around the salt body observed in the RTM image (4a)

is compensated for in the LSRTM (4b).

Large-scaleOBNsurveys inCentralNorthSea

The latest generation of broadband towed-streamer surveys

have revealed additional insight into the remaining potential

of the Central North Sea (CNS). CGGhas regularly drawn on

its experience in this region to apply its latest acquisition and

processing technologies in order to improve the imaging of its

challenging geology. Currently, a contiguous data volume of

over 35 000 km

of recent streamer data is being imaged using

aQ-Tomo, Q-FWI, andQ-RTM. This compensates for absorption

effects (Q) and is designed to providemore reliable images and

attributes for quantitative interpretation (QI). Regional pore

pressure prediction and quick-look AVOattributes have been

derived fromthis data to provide insight for reservoir development

and near-field exploration. The images of the deeper high pressure

high temperature (HPHT) reservoir targets have been enhanced

with 8600 km

of dual-azimuth coverage data obtainedwith

orthogonal passes of towed streamers.

Therightdata

Despite the uplift in subsurface images brought about by

dual- andmulti-azimuth streamer surveys and the latest

imaging technology, there are specific scenarioswhere the

technical limitations of what can be achievedwithin the

geometrical constraints of towed-streamer data have likely been

reached. Overcoming the toughest challenges of the CNS, such

as complex salt diapirismor deep reservoir characterisation in

structures beneath the Base Cretaceous Unconformity (BCU),

requires the next step in seismic acquisition. If designed appropriately,

OBNacquisition is now the right technology tomake that next step. The

goal is to target the remaining reservoir opportunities below the BCUas

well as any potential upside hydrocarbon volumes in existing Tertiary

discoveries around salt diapirs (Figure 5).

OBN represents the best way forward to sample the subsurface

with a uniformdistribution of offset-azimuths and a rich bandwidth,

especiallywith noise-free low frequencies. Proven data-drivenmodel

building technologies, such as TL-FWI, will then reap the full benefit of

the recordedwavefield that has densely sampled every corner of the

subsurface. The technologywill properly addressmisinterpretation

of the strong velocity contrasts of Quaternary channels in the very

shallowsubsurface aswell as contourites present in the shallow

Tertiary section, and possibly highlight some inter-bedded sand injectites.

By resolving any existingmisinterpretation of strong velocity contrasts

and velocity details in the overburden, imaging through LSRTMwill then

increase the resolution and recover the amplitude of target horizons

impacted by various artefacts originating fromthe rugose top chalk

andBCU surfaces, or fromsand injectites and shallowpolygonal faults.

In addition, the increased fold gainedwithOBNwill provide amuch

cleaner image.

Reasonablywell sampledOBNdatasets result in better removal

of strong propagation effects. Advanced demultiple and deghosting

algorithms benefit fromreceiver positioning on the seafloor. In addition,

the anisotropy of the subsurface, which often skews and slightly distorts

images, will bemore accurately taken into account due to the full-azimuth

sampling of thewavefield in all directions. The tailored processing of

OBN surveys in the CNSwill generate data that is intrinsically suitable for

Figure 4.

OBNamplitude image alongan interpreted horizon (5 kmdepth) for a)

conventional RTMandb) LSRTM. White arrow indicates the amplitude recovery near

the salt.

Figure 3.

OBN images generatedby a) conventional RTMandb) LSRTM.