«PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Technische Universiteit Delft, op gezag van de Rector Magniﬁcus Prof.ir. K.C.A.M ...»
True-Amplitude Seismic Imaging
Beneath Gas Clouds
True-Amplitude Seismic Imaging
Beneath Gas Clouds
ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van de Rector Magniﬁcus Prof.ir. K.C.A.M Luyben,
voorzitter van het College voor Promoties,
in het openbaar te verdedigen
op maandag 3 januari 2011 om 10:00 uur
Ahmad Riza GHAZALI
Bachelor of Science in Geophysics,
University of Leicester, United Kingdom geboren te Penang, Maleisi¨ e
Dit proefschrift is goedgekeurd door de promotor:
Prof.dr.ir. A. Gisolf
Dr.ir. D.J. Verschuur
Rector Magniﬁcus, voorzitter Prof.dr.ir. A. Gisolf, Technische Universiteit Delft, promotor Dr.ir. D.J. Verschuur, Technische Universiteit Delft, copromotor Prof.dr. H. Jakubowicz, Imperial College London, Prof.dr.ir. C.P.A Wapenaar, Technische Universiteit Delft, Prof.dr.ir. R.J. Arts, Technische Universiteit Delft, Dr. D.P. Ghosh, PETRONAS Research (Maleisi¨), e Dr.ir. W.E.A. Rietveld, BP Angola (UK), Prof.dr.ir. P.M. van den Berg, Technische Universiteit Delft, reservelid ISBN 978-90-8570-711-0 Copyright c 2011, by A.R. Ghazali, Laboratory of Acoustical Imaging and Sound Control, Faculty of Applied Sciences, Delft University of Technology, Delft, The Netherlands.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the author A.R. Ghazali, Faculty of Applied Sciences, Delft University of Technology, P.O. Box 5046, 2600 GA, Delft, The Netherlands.
SUPPORTThe research for this thesis was ﬁnancially supported by the Petroliam Nasional Berhad (PETRONAS) and DELPHI consortium.
Typesetting system: LTEX.
A Printed in The Netherlands by W¨hrmann Print Service.
Dedicated to my love:
Cik Junidza Mat Nayan & my beautiful family and for my future genetics...
Contents 1 Introduction 5
1.1 Wave physics: propagation through a gas cloud................. 5
1.2 Geology: uncertainties............................... 8
1.3 Problem description: kinematics and dynamics................. 9
1.4 Current approaches to gas-cl
1.1 Wave physics: propagation through a gas cloud In several areas around the world (e.g. offshore South-East Asia, North Sea, Black Sea) seismic exploration is complicated by the presence of so-called gas clouds. A gas cloud, as deﬁned by Sheriff (2001), is an overburden region of low-concentration gas, escaping and migrating upward from a gas accumulation. Generally it shows as a region of severely deteriorated seismic data quality associated with low velocity and with velocity sags (push down) underneath the gas cloud overburden. The reﬂected events in this region appear with lower amplitude and frequency content, which is often referred to as ‘Q-attenuation or Q-absorption’ or anelastic or intrinsic loss. The effects are probably caused by incoherent scattering, absorption and poor stacking because of nonhyperbolic normal moveout. Swaves appear to be little affected by gas chimneys.
Seismic imaging below gas cloud is complex and current solutions that are based on a linear imaging do not yield satisfactory results. The simpliﬁed Greens functions used in these imaging algorithms do not properly describe both the kinematic and dynamic behavior of the actual wave propagation, even when the best possible velocity model is used. In addition, ﬁnding such a good velocity model itself is a major problem, due to the localized nature of the anomaly.
Chapter 1. Introduction
Figure 1.1 An extracted inline section from a 3D seismic ﬁeld data (time-migrated cube) in the Malay basin over a gas cloud.
The red ellipse shows reﬂected events underneath the gas cloud that appear with lower amplitude and frequency content, often refers to as ‘Q-attenuation or Q-absorption’ or anelastic or intrinsic loss. The red arrow shows the traveltime delays or time sag often apparent on reﬂections within and below the gas cloud due to low velocity overburden. (Data by permission from PETRONAS) As an example, Figure 1.1 shows a strong imprint from transmission through a complex overburden containing a gas cloud.
In general, there are two major theories to explain how waves can lose their energy in
complex media, such as gas clouds:
1. Anelastic attenuation in terms of relaxation mechanism i.e. intrinsic attenuation.
The usual explanation is squirt-ﬂow which dissipates the wave energy into ﬂuid turbulence and ultimately converts it into heat. Strong intrinsic attenuation is also widely observed in soft sediments and partially saturated rocks, from laboratory to seismic scales. Anelastic energy loss is due to friction between grains or fracture surfaces when the propagating wave passes through the sediments, such that part of the propagation energy is transferred into heat (White, 1975). Attenuation and dispersion effects have been modeled using a complex-valued velocity (Aki and Richards, 1980; Dvorkin and Mavko, 2006). Thomsen et al. (1997) proposed that the presence of anisotropy, implied by a shallow gas, can result in a mismatch between seismic depth and well data.
Batzle et al. (2005) demonstrated via laboratory experiments on well cores that the seismic energy loss is frequency-dependent. Furthermore, they showed that bulk ﬂuid
1.1. Wave physics: propagation through a gas cloud motion is the primary loss mechanism, which can be further affected by the interaction with the water bound in the shale. They also concluded that observations of 1/Q made at seismic frequencies will not normally agree with sonic-log measurements. Vogelaar and Smeulders (2007) found that a uniform porous medium with gas/water layering leads to strong attenuation and dispersion in the typical seismic frequency band.
2. Scattering attenuation in terms of physical causality.
Waves also attenuate due to internal scattering in strongly heterogeneous media. In ﬁnely-layered media this is due to intrabed multiples and mode conversions. O’Doherty and Anstey (1971) were amongst the ﬁrst to discover that multiple-scattering of a waveﬁeld due to cyclic bounces in ﬁnely-layered media produce a so-called coda (see Figure 1.2). It also produces dispersion-like effects. This phenomenon is also called ‘stratigraphic ﬁltering’. The effect of multiple-scattering in a ﬁnely-layered medium depends on both the ratio of wavelength to the layer thickness and the strength of the reﬂection coefﬁcients. For the low frequencies, layered structures behave as effective medium and for higher frequencies whereby the medium can be described by the time average velocity (Stovas and Arntsen, 2006; Stovas and Ursin, 2007). Muller and Shapiro (2004) tried to quantify the combined effects of scattering attenuation due to thin layering, random diffractions and refractions, in a lossless medium in which no occurrence of an intrinsic wave attenuation.
Chapter 1. Introduction In ﬁne layering, multiple interference of the back-scattered waveﬁeld is the main source of scattering attenuation.
3D effects such as focusing and defocusing due to diffractions and refractions can also signiﬁcantly increase the transmission loss. Campman et al. (2005) have shown in laboratory experiments, that near-surface scattering due to heterogeneities can produce the same effect as Q-attenuation. Van der Baan et al. (2007) suggest that using a lower-frequency source wavelet will reduce the multiple-scattering problem in strongly heterogeneous media, at the cost of losing temporal resolution.
1.2 Geology: uncertainties Today, there is still debate about what is the physical cause or nature of shallow gas, gas clouds and gas chimneys. They are usually described as a vertical disturbed zone, which is associated with poor image quality, caused by gas accumulations from leakage through sediments. It is generally believed that faults or fractures are the main pathway for the migration of gas towards shallower unconsolidated clastic sediments. Sometimes the gas is trapped in a shallower reservoir sand, which typically contains channels a few hundred metres below the seabottom, sealed by mud, clay, carbonate-cemented sediments or shale. Biogenic or thermogenic processes can also produce locally charged shallow gas (Schroot and Schuttenhelm, 2003). Heggland (1997) and Schroot and Schuttenhelm (2003) have also described the interpretation and the geological surface expression of shallow gas, together with its occurrence and origin.
The shallow gas accumulations commonly found in a channel complex, are recognized as strong reﬂection amplitude anomalies, phase variations along seismic reﬂections and areas of ‘acoustic blanking’ where no reﬂectors can be seen below the gas. Judd and Hovland (1992) described that the ‘acoustic turbidity’ which appears as chaotic reﬂections may result from the scattering of the acoustic energy though the presence of only 1%-5% of gas. The reﬂections underneath the gas exhibit a ‘push-down’ effect due to the decrease in the acoustic velocity (vp ) in the gas-bearing zone, and as in the example in Figure 1.1, are also weak.
The shallow gas itself shows strong reﬂected energy due to the large acoustic-impedance contrasts between the gas-ﬁlled ﬁnely-layered porous silt-sand-rich or clay-sand-rich clastic sediments. Typically, a high portion of the reﬂected acoustic energy is reﬂected to the surface, leaving a smaller amount of transmitted energy. An important observation here is that this reﬂection response contains information on the properties of the shallow gas body.
Below is a list of some of the physical geological interpretations of gas clouds (Schroot and Schuttenhelm, 2003; Judd and Hovland, 1992; Vogelaar and Smeulders, 2007; Arntsen et
• The geochemical signatures such as biogenic or thermogenic processes produce local gas e.g. methane, ethane, CO2 and hydrogen sulphide. These are often associated with hydrocarbon-related diagenetic zones. They create features such as pockmarks
1.3. Problem description: kinematics and dynamics and mounding if gas escapes to the surface.
• From the geochemical process, escaping ﬂuids (liquids and gasses including methane) produce macro-seepages/bubbles, which due to buoyancy, will migrate upwards through a leaking fault-system. These seepages are commonly large enough to be visible in seismic data, and are seen as vertical disturbances due to the upward movement of ﬂuid or free-gas.
• The migrated gas travels to a shallower depth and saturates the shallow reservoirs e.g. high-porosity channel sands or ﬁnely-layered ﬂuid-saturated porous sediments (poroelastic media), thus becoming a shallow gas-sand complex.
• Because of complicated faulting, the waveﬁelds experience multiple-scattering and become very complex.
1.3 Problem description: kinematics and dynamics Why are we interested in true-amplitude imaging beneath the gas clouds? The reason lies in the fact that there are many major hydrocarbon ﬁelds in the world e.g. Valhall ﬁeld, Sleipner ﬁeld, Haltenbanken ﬁeld, Tommeliten ﬁeld, Irong Barat ﬁeld, Kikeh ﬁeld and ﬁelds in Malaysia, Trinidad, Azerbaijan and Indonesia which are located underneath gas clouds (Heggland, 1997; O’Brien et al., 1999; MacLeod et al., 1999; Granli et al., 1999; Johnston et al., 2002; Khan and Klein-Helmkamp, 2005; Tanis et al., 2006a; Ghosh et al., 2010). A similar situation to the gas cloud problem is imaging underneath basalt, where often thin-basalt layers also produce complex wave propagation effects due to multiple-scattering. Note, however, that basalt has a high velocity as compared to the low velocity in gas (Li et al., 1998;
Hanssen et al., 2003; Van der Baan et al., 2007). Due to poor seismic imaging, reservoir management becomes more difﬁcult, as does quantifying the amount of hydrocarbons in the subsurface and conﬁrming compartmentalization of the reservoirs.
The reﬂectors underneath the gas cloud, as shown in the seismic stack section in Figure 1.3, are characterized by time delays (sagging), frequency and amplitude loss, and phase distortion. Due to strong acoustic impedance contrasts and the low velocity of the gas layers, the prestack gathers show strong internal short-period multiples and nonhyperbolic moveout (see Figure 1.4). The heterogeneous nature of shallow gas-ﬁlled areas results in scattering, dispersion, internal multiples, mode conversion during wave propagation and, possibly, anelastic losses. Acoustic processing, including the use of prestack depth migration (PreSDM), under these circumstances, has been fraught with difﬁculties. Figure 1.5 shows a comparison of a Kirchhoff prestack Time Migration (PreSTM) image and a PreSDM image of the data from Figure 1.3. It demonstrates that by better honouring the propagation effects in PreSDM an improved image is obtained. However, very heterogeneous media, such as ones containing gas clouds, give problems in migration velocity analysis. The localized nature of the velocity anomaly also poses problems in the parameterization of the velocity model.
Chapter 1. Introduction Figure 1.
3 Stack section in time from one of the 3D data acquired from a ﬁeld in Malaysia. The shallow gas accumulation (shallow overburden) is mainly located between 0.65 ∼ 0.8s in the middle of the section, as indicated by the red ellipse. (Data by permission from PETRONAS) Moreover, even if a proper velocity model is obtained, conventional imaging cannot correct for the complex transmission effects.
There have been several previous studies on complex wave propagation through gas clouds.
O’Brien et al. (1999) built two ﬁnite-difference elastic models, which created similar gas clouds effects to those observed on the data obtained from the Valhall ﬁeld. Youn et al.