Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Oilfield Review Spring 2009 Electromagnetic Geophysics Deepwater Project Planning Volcanic Reservoirs 09-OR-0002 Marine CSEM: Evolution of a Technology Interest in marine controlled-source electromagnetics (mCSEM) emerged during the cold war era. A postulated high-resistivity layer within the oceanic lithosphere was proposed as a means for secure communication in the event of nuclear war. In a more practical sense, mCSEM became a reality because pioneering measurements in the 1970s by Charles Cox and coworkers at the Scripps Institution of Oceanography showed that the natural electric field noise level at frequencies around 1 Hz is extremely low (~1 pV/m). Cox realized that weak fields induced within the earth by a near-seafloor, horizontal electric dipole source could be detected many kilometers away. He also understood that mCSEM is preferentially sensitive to relatively resistive zones under the seafloor. In contrast, the widely used magnetotelluric method detects conductive material. Thus, mCSEM made it possible to study the earth in a new way. Subsequently, Cox and his colleagues designed and built apparatus to carry out mCSEM surveys in the deep ocean. First applied in 1979, it continued in use through the 1990s. Other mCSEM programs evolved at the University of Toronto, Canada, in the early 1980s and the University of Cambridge, UK, in the late 1980s. The former was primarily focused on shallow targets of geotechnical interest, while the Scripps and Cambridge groups used mCSEM for the academic study of mid-ocean ridges and related features. I joined Cox as a postdoctoral researcher in 1980. At that time, we recognized that mCSEM had potential industrial applications on continental shelves. In particular, like many others, we appreciated that petroleum-bearing formations are typically resistive compared with the substrate, giving mCSEM advantages over magnetotellurics for hydrocarbon surveys. Concomitantly, Len Srnka at Exxon investigated the method. After a brief flurry of interest, industrial mCSEM dwindled after the mid-1980s—the result of a combination of low oil prices, the emergence of 3D seismic reflection technology and a primary focus on drilling in water under 300 m deep, where mCSEM is problematic because of interference from a strong component propagating along the sea surface–air boundary. Interest re-emerged in the late 1990s, resulting in field demonstrations in 2000–2002 over producing fields off Angola and West Africa by Statoil and ExxonMobil, respectively. The evaluations showed the viability of mCSEM for direct detection of hydrocarbons through their low resistivity. Startup companies to provide mCSEM services to the petroleum industry were spawned, including EMGS, OHM and AOA Geomarine. The latter was acquired by Schlumberger in 2004. Unfortunately, the resurgence of interest also precipitated exaggerated claims for the effectiveness of mCSEM and wild predictions by stock analysts that the technology would consume 25% of the industry’s marine exploration budget by 2009. The reality is closer to 5%. Nevertheless, I predict a bright future for mCSEM as one of a range of tools that a marine petroleum explorationist will employ. Rather than being a panacea that yields clear discrimination of producing discoveries from dry holes, mCSEM is evolving into one component of an integrated exploration approach involving seismic reflection and other technologies. This conjoining of technologies will be especially important as exploration moves into deeper water where the cost of drilling is extreme. As a consequence, petroleum companies that do much of their data analysis and interpretation in-house, such as ExxonMobil, and integrated services companies, such as Schlumberger, will have a business advantage over specialist companies that provide only mCSEM. I think consolidation in the industry is likely. None of this would have been possible in the 1980s. Although the apparatus in current use by industry is essentially that of academia 20 years ago, incremental improvements have been incorporated as technology has evolved. The biggest difference is one of scale: The number of receivers that can be deployed has increased dramatically, and sources are substantially more powerful. However, the most important technical improvements have occurred in interpretation. In the 1980s, modeling was limited to 1D structures, and 2D analysis was at the cutting edge. Today, 3D modeling is becoming routine and is rapidly evolving in sophistication. Careful 3D modeling is essential for the interpretation of mCSEM in hydrocarbon applications. Coupled with other exploration measurements, increasing 3D capabilities will move mCSEM into the mainstream of hydrocarbon exploration. Alan D. Chave Woods Hole Oceanographic Institution Alan Chave is a Senior Scientist in the Deep Submergence Laboratory at the Woods Hole Oceanographic Institution. He holds a BS degree in physics from Harvey Mudd College, Claremont, California, USA, and a doctorate in oceanography from the MIT-WHOI Joint Program, Woods Hole, Massachusetts, USA. Alan was involved in early development of mCSEM and maintains an active experimental research group focused on marine electromagnetics, optics and ocean observatory technologies. 1 Schlumberger Oilfield Review Executive Editor Mark A. Andersen Advisory Editor Lisa Stewart Senior Editor Matt Varhaug Editors Rick von Flatern Vladislav Glyanchenko Tony Smithson Michael James Moody 4 Electromagnetic Sounding for Hydrocarbons Deep-reading electromagnetic surveys examine subsurface resistivity, providing information that is complementary to seismic data. This article introduces magnetotelluric and controlled-source electromagnetic technologies. Marine examples include case studies from the Gulf of Mexico, offshore Brazil and Greenland. Contributing Editors Rana Rottenberg Glenda de Luna Design/Production Herring Design Steve Freeman Illustration Tom McNeff Mike Messinger George Stewart Printing Wetmore Printing Company Curtis Weeks 20 Near-Surface Electromagnetic Surveying Electromagnetic surveys deliver information by probing the resistivity of the earth. One method employed in land surveys uses an inductive loop to investigate the near surface. Two case studies from the United Arab Emirates illustrate its use. One study mapped an aquifer for a water-storage project. The other identified the bottom of the low-velocity layer in an area of dunes, which helped determine how to apply static corrections for a seismic survey. On the cover: WesternGeco specialists install a navigation beacon onto the transmitter antenna cable during a controlled-source electromagnetic survey. The yellow floats provide buoyancy to keep the cable at a specified distance above the seabed. Toisa Valiant, the vessel in the inset, is equipped for performing these electromagnetic surveys. Useful links: Schlumberger www.slb.com Oilfield Review Archive www.slb.com/oilfieldreview Oilfield Glossary www.glossary.oilfield.slb.com 2 Address editorial correspondence to: Oilfield Review 5599 San Felipe Houston, Texas 77056 USA (1) 713-513-1194 Fax: (1) 713-513-2057 E-mail: [email protected] Address distribution inquiries to: Tony Smithson Oilfield Review 12149 Lakeview Manor Dr. Northport, Alabama 35475 USA (1) 832-886-5217 Fax: (1) 281-285-0065 E-mail: [email protected] Spring 2009 Volume 21 Number 1 26 A Plan for Success in Deep Water Advisory Panel Abdulla I. Al-Kubaisy Saudi Aramco Ras Tanura, Saudi Arabia Deepwater E&P operations present the upstream industry with unprecedented technological and economic challenges. Addressing them requires a fundamental change in the way the offshore oil and gas industry operates. The projects in these environs are best viewed as a single, integrated effort, from exploration to production and even beyond. Dilip M. Kale ONGC Energy Centre Delhi, India Roland Hamp Woodside Energy, Ltd. Perth, Australia George King Independent consultant Houston, Texas, USA Eteng A. Salam PERTAMINA Jakarta, Indonesia Jacques Braile Saliés Petrobras Houston, Texas 36 Evaluating Volcanic Reservoirs Richard Woodhouse Independent consultant Surrey, England Volcanic rock can contain oil and gas in commercial quantities, but evaluating these reservoirs is not straightforward. By extending techniques designed for evaluating sedimentary reservoirs, some companies find profitable opportunities in areas that others might deem unworthy of consideration. Examples from China and India demonstrate successful petrophysical evaluation of volcanic formations using neutroncapture spectroscopy, nuclear magnetic resonance, borehole resistivity images and conventional logging technology. 48 Contributors 51 New Books and Coming in Oilfield Review Oilfield Review subscriptions are available from: Oilfield Review Services Barbour Square, High Street Tattenhall, Chester CH3 9RF England (44) 1829-770569 Fax: (44) 1829-771354 E-mail: [email protected] www.oilfieldreview.com Annual subscriptions, including postage, are US $200, subject to exchangerate fluctuations. Oilfield Review is published quarterly by Schlumberger to communicate technical advances in finding and producing hydrocarbons to oilfield professionals. Oilfield Review is distributed by Schlumberger to its employees and clients. Oilfield Review is printed in the USA. Contributors listed with only geographic location are employees of Schlumberger or its affiliates. © 2009 Schlumberger. 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 publisher. 3 Electromagnetic Sounding for Hydrocarbons James Brady Tracy Campbell Alastair Fenwick Marcus Ganz Stewart K. Sandberg Houston, Texas, USA Marco Polo Pereira Buonora Luiz Felipe Rodrigues Petrobras E&P Rio de Janeiro, Brazil Chuck Campbell ACCEL Services Inc. Houston, Texas Leendert Combee Oslo, Norway Arnie Ferster Kenneth E. Umbach EnCana Corporation Calgary, Alberta, Canada Tiziano Labruzzo Andrea Zerilli Rio de Janeiro, Brazil Edward A. Nichols Clamart, France Steve Patmore Cairn Energy Plc Edinburgh, Scotland Jan Stilling Nunaoil A/S Nuuk, Greenland Oilfield Review Spring 2009: 21, no. 1. Copyright © 2009 Schlumberger. For help in preparation of this article, thanks to Graeme Cairns, George Jamieson, Jeff Mayville, Fred Snyder and Xianghong Wu, Houston. MMCI and Petrel are marks of Schlumberger. 4 Recent advancements in identifying subsurface features by resistivity contrasts have added a significant tool in the quest to locate hydrocarbon resources. The electromagnetic sounding technique comprises two related technologies, magnetotelluric and controlled-source electromagnetic surveys, that provide distinctly different insights into the subsurface. Their ability to clarify structures and to help identify possible hydrocarbon deposits before drilling is exciting explorationists. The Sun provides us with energy in many forms. A surprising connection between exploration for energy resources and the Sun is becoming increasingly significant for the E&P industry. Ions emitted by the Sun experience a complex interplay with the Earth’s magnetic field, generating propagating electromagnetic fields that penetrate the Earth and interact with its conductive layers. As the industry’s search for hydrocarbon resources intensifies, more geoscientists are relying on these electromagnetic fields to probe areas that are difficult to image with seismic methods. The study of electrical currents in the Earth, called tellurics, is not new. Conrad Schlumberger, one of the founders of Schlumberger, used the phenomenon in early surface studies that he directed in the 1920s, prior to his start in wireline logging.1 Louis Cagniard, a professor at the Sorbonne in Paris, first reported combining a measurement of electric and magnetic fields, termed magnetotellurics (MT), for exploration of the Earth’s subsurface in 1952.2 However, MT has become an important tool for explorationists in the E&P industry only within the past few years—thanks to advances in 3D modeling and inversion technology. Now, MT results can be combined more efficiently with seismic and gravity surveys, resulting in a more-calibrated model of the earth. Although Cagniard also discussed a method related to MT that uses an artificially imposed electromagnetic field, techniques for generating and detecting a signal strong enough for use in the E&P industry came decades later, in the 1960s on land and then in the 1980s in the marine environment. This method is now termed controlled-source electromagnetics (CSEM). The interaction of the earth with impinging electric and magnetic fields is complex. Two important factors in MT analysis are the frequency spectrum of the fields and the resistivity (or its inverse, the conductivity) of the particular medium through which the field waves propagate. Analyzing data from the frequency spectrum helps obtain an apparent resistivity as a function of frequency.3 This apparent resistivity can be related to the true resistivity of the formation at various depths. If the subsurface is homogeneous, the measured apparent resistivity is the same as the true resistivity, but if the resistivity changes with depth, apparent resistivity is a conflation of measurement effects and some average of the resistivities. Through data analysis, interpreters can determine the depths of bodies with contrasting resistivities, providing a result termed an MT sounding. Oilfield Review This article discusses the physics of these electromagnetic interactions and how they are interpreted to give information useful in basin and reservoir evaluation. It also describes the equipment used to detect and, in the case of CSEM, to generate relevant electromagnetic fields. Case studies from the Gulf of Mexico, Brazil and Greenland illustrate these technologies for offshore salt mapping and reservoir illumination. A companion article describes near-surface applications of CSEM on land (see “Near-Surface Electromagnetic Surveying,” page 20). The next section focuses on natural electromagnetic fields and their interactions with the Earth. Blowing in the Wind The solar wind is a stream of positive and negative ions emitted by the Sun. Wind intensity varies, increasing during periods of high sunspot activity. This ionic wind “blows” through space; auroras manifest its interaction with the Earth’s magnetic field in spectacularly colorful ways.4 Although most solar ions are deflected by the magnetic field in a region known as the magnetopause, which is several Earth radii out in space, some ions leak in. Those that reach the upper atmosphere can ionize particles in the ionosphere, which ranges from 75 to 550 km [50 to 340 mi] above the surface of the Earth. In the ionosphere, the particle velocities are high enough and the particle density low enough that charged ions do not immediately recombine into neutral atoms and molecules: They form a plasma of charged particles. This plasma makes the ionosphere a conducting layer, unlike the nonconducting layers of the lower atmosphere where the particle density is too high to maintain charged ions for a significant period of time. The motions of charges in the ionosphere are constrained by the Earth’s magnetic field, whose lines of force stretch from pole to pole. When solar ions enter the plasma within this magnetic field, they generate electromagnetic (EM) pulses 1. Leonardon EG: “Some Observations Upon Telluric Currents and Their Applications to Electrical Prospecting,” Terrestrial Magnetism and Atmospheric Electricity 33 (March–December 1928): 91–94. 2. Cagniard L: “Basic Theory of the Magneto-Telluric Method of Geophysical Prospecting,” Geophysics 18 (1953): 605–635. 3. Apparent resistivity is a volume average of the true resistivities of the media within the volume measured by a device, such as a resistivity or induction tool, or a magnetotelluric receiver. 4. For a recent discussion about the origin of the auroras: Brown D and Layton L: “NASA Satellites Discover What Powers Northern Lights,” NASA News & Features, http://www.nasa.gov/home/hqnews/2008/jul/HQ_08185_ THEMIS.html (accessed March 2, 2009). Spring 2009 5 10 Magnetic field spectral amplitude 1 0.1 0.01 0.001 0.0001 0.00001 0.000001 0.001 0.01 0.1 1 10 Frequency, Hz 100 1,000 10,000 > Typical magnetic-field amplitude spectrum from the atmosphere. The ionospheric signal originating from interactions of the Earth’s magnetic field decays rapidly with increasing electromagnetic frequency. Lightning generates signals in a region called the Schumann bands in the spectrum between about 7.8 and 60 Hz. that resonate in the ionosphere, traveling along the magnetic field lines. The result is analogous to plucking the string of a guitar; just as the string resonates at characteristic frequencies, so too does the ionosphere resonate electromagnetically. The complex interaction of magnetic field, atmospheric plasma and solar ions results in a broad spectrum of EM frequencies, including the visible-light phenomena of the aurora borealis and the aurora australis. The spectral range useful for E&P-related MT extends from frequencies of about 0.001 Hz to 10 kHz; for studies extending to the Earth’s mantle even lower frequencies are used (above). Frequencies above 1 Hz are severely attenuated through conductive seawater and thus create no subsea earth response, making this the effective upper-frequency limit for marine MT. The amplitude and frequency spectrum of the signal is highly variable.5 The fluctuations in the solar wind reflect the 11- to 14-year cycle of sunspot activity. The spectrum also depends on the season and time of day, since sunlight influences the degree of polarization in the ionosphere. Signal levels in equatorial regions are low, whereas they are high in polar regions. Geomagnetic index 16 12 8 Annual 2007 average EM_FIGURE 01 2005 2006 2008 4 0 1/1/08 4/1/08 7/1/08 10/1/08 12/31/08 Date > Electromagnetic activity. Planetary electromagnetic activity is estimated from measurements of a geomagnetic index taken by the US National Oceanic and Atmospheric Administration (see reference 5) at several locations. The activity fluctuates both annually and weekly, as shown for 2008 (black). The solar cycle is currently in a quiet period. 6 This stronger signal near the poles or near the peaks of the solar-activity cycle results in higher-quality MT data; conversely, obtaining data from deepwater equatorial areas, especially during low-activity periods, is more challenging (below left). A part of the frequency spectrum is influenced by lightning. A lightning discharge can generate current in the range of 20 to 50 kA, which initiates a strong interaction in the ionosphere. The charge pulse follows the magnetic field lines around the Earth, reflecting near the poles and playing its own resonance notes.6 The EM fields resulting from a lightning strike are global.7 The lower atmosphere is a poor electrical conductor, so the EM waves propagate with virtually no attenuation.8 This lack of attenuation allows radio broadcasts to be heard far from the source when atmospheric conditions are right for refracting them to listeners. In contrast, once the waves reach the surface layers of the Earth, they interact with seawater and formations that are electrically conductive to a greater or lesser extent. Conductive bodies attenuate EM waves. Most of a rock’s solid matrix conducts electricity poorly. However, various saturating fluids have differing conductivities. Brine conducts well, but oil and gas have high resistivities. Adjacent formations with a marked resistivity contrast—such as a hydrocarbon-bearing zone surrounded by brine-saturated strata—affect the propagating EM field in different and potentially measurable ways. The resistivity contrast is also high between brine-filled sedimentary layers and some specific lithologies, such as salt, basalt and resistive carbonates. The EM waves interact with conductive formations and induce a response wave that propagates back to the surface. Although the geometry of signal and response is sometimes depicted as analogous to that of a seismic reflection, the EM effect has a different physical origin and different behavior than a reflected seismic wave.9 The time-varying EM signal induces a current loop in the conducting layer. The properties of this induced eddy current depend on the resistivity of the conducting formation and the magnitude and time rate of change—or the frequency—of the source signal. The eddy current, in turn, induces a magnetic field, which propagates from the formation. Sensors on the surface measure this response field. Oilfield Review ω=2 ω=5 ω = 10 ω = 10 ° ~0 ° ~0 ° ~ 10 ° ~1 Skin depth Skin depth Skin depth Skin depth > Skin effect. A downward-moving electromagnetic field (blue curve) leaving a highly resistive medium, such as air, begins to decay when it enters a more-conductive medium, such as rock. Lower-frequency waves (left ) propagate farther than higher-frequency waves (center left and center right ), and waves propagate farther in less-conductive media (right ). The amplitude has an exponential decay (red) that is a function of the conductivity of the medium, σ, and the frequency of the wave, ω. The skin depth is the distance at which the amplitude has decayed to 1/e of the incident value. The wave in the conductive medium also experiences a gradual delay in the phase. Since the phase change is difficult to see in this example, one illustration (far left ) also shows an attenuated wave without the phase change (violet). Frequency and conductivity values are relative among these examples. The eddy current in the conducting formation opposes the change in the source field. The result of the eddy current and the transfer of energy to the response signal is attenuation of the incoming EM wave. Thus, as the wave passes successively deeper into the conductor, the eddy current becomes incrementally weaker, making the response field smaller also. As this process continues, the incident signal decays, while weaker response signals form at each successive increment of depth within the conducting formation. This decay is known as the skin effect (above). A characteristic distance for penetration of the signal into a conductor, termed the skin depth, is obtained by determining when the field amplitude drops by a factor of 1/e, the inverse of the exponential function. Attenuation is frequency dependent; high frequencies attenuate more rapidly than low frequencies. It is also a function of the formation conductivity: In more-conductive formations the impinging field induces greater current flow that partially cancels the source field. In a typical geological section, the natural frequencies used in MT have skin depths of a few tens to a few tens of thousands of meters. The high-frequency components useful for detecting thin, shallow formations are present only for land-based (or extremely shallow-water) surveys because of attenuation by conductive seawater. The deeper a target structure is buried, the larger it must be to enable detection through MT evaluation; this basic MT-resolution problem at depth is more severe than resolving small, deep features using seismic waves. The response signal contains information in the impedance value about the resistive properties of the formations. Impedance is a complex term—comprising real and imaginary parts— EM_FIGURE that designates the difficulty 04 of propagating the EM energy through a medium. It is determined from the amplitude and phase relationship that exists between the measured electric and magnetic fields.10 It is also a tensor quantity that can be related to the apparent resistivity of the formation. Impedance varies with the frequency of the incoming signal. Because the source is so distant, the MT fields impinging on an E&P survey area can be approximated over a wide bandwidth as vertically incident plane waves with the electric field horizontally polarized.11 MT fields are sensitive to large conductive features, making them useful in studies of large salt, basalt and carbonate bodies due to the contrast of these resistive features with the conductive surroundings. However, the attenuation of the MT fields with depth—the skin effect—makes them insensitive to resistivity contrasts of thin formations such as 5. Data are available from the US National Oceanic and Atmospheric Administration, http://www.swpc.noaa. gov/ftpmenu/indices/old_indices.html (accessed May 5, 2009). 6. This response to lightning is termed a Schumann resonance, after German physicist Winfried Otto Schumann, who predicted the resonances mathematically in 1952. 7. Active storms generating lightning seem to be linked: Synchronized lightning strikes from widely spaced geographic locations have been observed from the NASA Space Shuttle. For more on synchronized lightning strikes: Yair Y, Aviv R, Ravid G, Yaniv R, Ziv B and Price C: “Evidence for Synchronicity of Lightning Activity in Networks of Spatially Remote Thunderstorms,” Journal of Atmospheric and Solar-Terrestrial Physics 68, no. 12 (August 2006): 1401–1415. 8. Electromagnetic waves propagate through a vacuum with no attenuation. 9. EM energy in a conductive medium has a diffusive nature rather than a wave nature. 10.The phase of a wave describes where it is in its amplitude cycle of maximum to minimum and back to maximum as the phase angle goes from 0° to 360°. The electric and magnetic fields of a propagating wave are not necessarily at 0° phase at the same time, and the difference between the two is also referred to as the phase angle. 11.The waves impinge vertically because air is not conductive. The uniformity of signal for MT surveys is based on the large distance to the ionosphere compared with the length of a survey line. However, if the signal comes from a lightning strike that is close to the survey area, the plane-wave assumption does not hold and the local geometry influences the interpretation. Spring 2009 7 Ex (t) Marine MT Hy (t) Ex = H V Marine CSEM Passive (atmospheric) source Active controlled source Plane waves, vertically incident Localized dipole source Basin scale Reservoir scale Detection of structure and lithology Detection of resistivity contrast, such as that caused by a resistive pore fluid against a conductive background Wave-field sensitive to conductors Wave-field sensitive to resistors > Comparison of marine MT and CSEM survey technologies. V L X L Ex Z= = iρ ° µ a Hy Y ρa = ° = µ= Z= E= H= t= V= L= i= formation resistivity frequency magnetic permeability formation impedance electric field magnetic field time voltage drop across dipole dipole length –1 > Sensing impedance. A vertically incident EM wave interacts with the Earth through the formation impedance, Z. The Z value can be determined by measuring the horizontal electric field, E, and the magnetic field, H, at the surface or on the seabed (tan). The apparent resistivity, ρa, is the aggregate resistivity of the formation layers beneath the electric dipole antenna and the magnetometer coil of a sensor (yellow). In the case shown, E and H are in phase; if the zero crossings of the two fields were out of synchronization, there would be a phase angle between the two fields. hydrocarbon-bearing sediments. Generally, to be resolved by MT, the layer’s thickness should be at least 5% of its burial depth, and the layer should be more conductive than its surroundings. These limitations led to the development of the CSEM EM_FIGURE 07 method (above right). The CSEM method imposes a powerful, artificially generated EM signal. The source is a localized electric dipole with a controlled signal that extends over a narrow bandwidth, often just a few fundamental frequencies and their harmonics. The EM fields generated by such a source are not plane waves. The composition and geometry of the signal are chosen to make it more sensitive for detection of a thin formation at a particular 8 hypothesized location and with a resistivity value that contrasts to that of surrounding formations. This difference between MT and CSEM source signals affects the method of processing the data and impacts the type of structures that can be detected by the two methods, as discussed in the next two sections. to obtain a model of the earth. The result is not unique, so the process iterates until the result is acceptable. Many algorithms are in use for converging the inversion on a particular model. A key step of preacquisition planning is to determine if different models will be distinguishable in the data. This is typically accomplished by first forward modeling the response of various Deep Vision with MT predicted scenarios, then possibly employing The atmospheric source for MT signals varies inversion on modeled synthetic data. To invesrandomly in time, but at any given time the ver- tigate whether the original model can be tically incident waves are uniform over a large recovered, the synthetic data include noise reparea. The wavefields are planar and vertically resenting expected background or measurement incident on the surface of the Earth; the electric noise. This step can help justify the usefulness of field has only horizontal components, as does the a proposed survey or, alternatively, advise against orthogonal magnetic field. As a matter of nomen- its application. Acquisition parameters such as clature, the portion of the electric field that can the location of instruments and how long they be resolved along the strike of a geologic feature must remain on the ground are also results of this is termed the transverse electric (TE) mode; the process. In CSEM surveys, the optimal frequenportion across the strike is the transverse mag- cies can also be determined through modeling. netic (TM) mode. Recent interest in MT measurements has Because of the vertical and planar geometry focused on evaluations in marine environments, of MT, the impedance of the subsurface can be driven by the increasing costs of drilling in deep obtained by taking the ratio of the horizontal water and the complexity of imaging below salt electric field in one direction to the horizon- and basalt. As a result, technologies that increase tal magnetic field in the orthogonal direction the chance of economic success after locating (above left).12 This calculation removes the tem- drilling targets have great value. As with seismic surveys, EM surveys require poral variability of the incident signal, leaving EM_FIGURE 05 deployment of equipment, either on land or at only the desired formation response. The complex impedance can be calculated to sea. Marine MT surveys are acquired using small obtain the apparent resistivity, ρa, of the underly- vessels and small crews. CSEM surveys need ing formations and the phase angle, φ, between larger vessels to handle the source equipment the electric and magnetic fields. Geoscientists and larger crews to operate and maintain that use these results to interpret the subsurface equipment. Typically, both MT and CSEM surstructure through forward modeling or through veys are targeted, examining specific ambiguous inversion.13 Forward modeling assumes a struc- structures or promising anomalies on a seismic ture and certain properties, such as layer depth section. Thus, the duration and areal scope of and resistivity, and predicts the earth’s elec- these studies are typically smaller than those for tromagnetic response to the assumed model. seismic surveys. Subsea EM measurements—both MT and Comparing or normalizing processed data against this model assesses its goodness of fit. Inversion CSEM—are similar to land measurements aside is the reverse of forward modeling, using the data from the vast difference in the resistivities of seato step backward through the physical process water and air. At the air/land interface there can Oilfield Review be no vertical electric current because the air is not conductive, but on the seabed a vertical electric current can exist in the conductive water. The consequence of this difference is subtle. On land, the electric field responds significantly to changes in resistivity in subsurface layers, but the magnetic field has much less variation. In contrast, in the marine environment it is the magnetic field rather than the electric field that displays the greater variation with change in subsurface structure, although both fields carry information on structure.14 Measuring the Signal The two basic devices for measuring EM fields are a pair of electrodes to sense an electric field potential difference and a magnetometer to sense magnetic field variations. The pair of electrodes forms an electric dipole, allowing measurement of the potential voltage difference between them. A magnetometer is a coil of conducting wire that generates a measurable current based on the changing magnetic flux through the coil. When only two sensors of one type are used, they are oriented to measure the orthogonal field components in the horizontal plane. The vertical component of the field is measured only if a third sensor is used. The primary recent interest in the E&P industry has been offshore, and considerable effort has been made over the past decade to develop a sensor for marine use. The Scripps Institution of Oceanography in La Jolla, California, USA, developed the basic electric field sensor that is used by WesternGeco today. The magnetometers were developed by Electromagnetic Instruments Inc., which was acquired by Schlumberger in 2001.15 In the WesternGeco device two horizontal electric dipoles are formed by silver–silver chloride electrodes placed at the ends of four long fiberglass tubes, extending from each of the four sides of the receiver frame (above right). The present configuration includes a vertical dipole with a length of 1.82 m [6 ft]. Its length is restricted by the need to maintain orthogonality and stability—a longer dipole is more susceptible to seabed currents that move the dipole antenna and introduce noise into the measurement within the frequency bandwidth of interest. The magnetometers, multiturn coils in a nonmetal housing, are used to sense the magnetic flux. The magnetometer tubes are secured horizontally into holes in the frame. The operating range is from 0.0001 to 100 Hz. Calibration of both types of sensor is critical. The WesternGeco sensors and amplifiers are individually calibrated far from electromagnetic Spring 2009 Instrumented strayline float Dipole for electric field Gas flotation Induction coil magnetometers Logger Acoustics Burnwire release mechanisms Concrete anchor > CSEM receiver. Orthogonal dipole antennas on the receiver measure Ex and Ey and two induction coil magnetometers measure Hx and Hy . Each tube containing an antenna is 3.6 m [12 ft] long; coupled with the dimension of the frame, the electric dipole length formed by a pair pointing in opposite directions is 10 m [32.8 ft]. A concrete anchor carries the receiver to the seafloor, where it remains throughout the test. The electronic logger records for a set time. At the conclusion of the test, an acoustic signal from the ship triggers a mechanism to burn through the wire holding the device to the anchor. Airfilled glass spheres raise the receiver to the surface, where it is retrieved and the data are captured. In some cases, the receiver includes a vertical dipole to measure the vertical electric field, Ez (not shown). (Image courtesy of Scripps Institution of Oceanography.) noise in a remote part of the Norwegian countryside. In addition, data quality requires strict adherence to deployment procedures on the survey ship. A concrete block attached to the bottom of the receiver frame provides weight to take it to the ocean floor. This concrete anchor also helps to stabilize the instrument against forces from sea currents; antenna rotation as tiny as 1 μrad can easily be detected by the magnetic induction coil moving in the Earth’s magnetic field. At the conclusion of the survey, an acoustic signal from surface triggers release from the block, and air-filled glass spheres lift the receiver to surface for retrieval. The cost and logistics of establishing electrical connections with multiple receivers placed on the seabed in deep water are prohibitive, so engineers designed the receiver to operate independently and to be retrieved at the end of the test. Each receiver carries a data logger that controls operation and records the signals on a compact flash card. High-resolution data from the dipoles and magnetometers come from 24-bit analog-to-digital converters, which accurately record time so that the signals can be synchro- nized later with the source record and with each other. The unit has several independent battery packs. One provides power to the data-logger electronics. A separate battery powers the anchorrelease devices, and another powers an acoustic positioning beacon that indicates the unit’s location on the seabed. The battery pack that powers the data logger lasts up to 40 days; the long battery life provides time to deploy the sensors and then acquire data. The anchor-release battery pack lasts more than a year, in case circumstances prevent immediate removal of the device after the survey. 12.Cagniard, reference 2. 13.For more on inversion: Barclay F, Bruun A, Rasmussen KB, Camara Alfaro J, Cooke A, Cooke D, Salter D, Godfrey R, Lowden D, McHugo S, Özdemir H, Pickering S, Gonzalez Pineda F, Herwanger J, Volterrrani S, Murineddu A, Rasmussen A and Roberts R: “Seismic Inversion: Reading Between the Lines,” Oilfield Review 20, no. 1 (Spring 2008): 42–63. 14.Constable SC, Orange AS, Hoversten GM and Morrison HF: “Marine Magnetotellurics for Petroleum Exploration, Part I: A Sea-Floor Equipment System,” Geophysics 63, no. 3 (May–June 1998): 816–825. 15.Webb SC, Constable SC, Cox CS and Deaton TK: “A Seafloor Electric Field Instrument,” Journal of Geomagnetism and Geoelectricity 37, no. 12 (1985): 1115–1129. Constable et al, reference 14. 9 The seabed orientation of the horizontal sensors is random. The measurement directions are resolved to a desired orientation during processing. The newest devices have a compass, but in the past the orientation for each receiver was obtained either by comparison with land-based sensors or by orientation based on the direction of a towed source in a CSEM survey. Towfish Cable to survey vessel Streamer Antenna Neutrally buoyant cable Tow-cable termination 300-m dipole Transponder A Electrode 1 B Electrode 2 2.5 m Strain relief Instrument suite Transponders 20 m A: Telemetry and signal conditioning B: Transmitter power section > CSEM transmitter. The transmitter comprises a towfish—the head section containing power and instrumentation—and a streamer antenna with dipole electrodes at the ends of two cables. The dipole is the source of the CSEM signal. The signal transmission and waveform parameters are set from the survey vessel during operations, and results are telemetered to the operators for real-time quality control of the signal. The photograph (top) shows a towfish being removed from the ocean, with the antenna trailing in the water. 1.5 ω0 Five-term sum 1.0 Amplitude 0.5 9ω 0 3ω 0 5ω 0 7ω 0 0 –0.5 –1.0 –1.5 0 1 2 09 EM_FIGURE 3 4 Time, s Square wave (ω 0) = 4 ° sin(ω 0t) + sin(3ω 0t) 3 + sin(5ω 0t) 5 + sin(7ω 0t) 7 + sin(9ω 0t) 9 + ... > Square-wave components. A square wave (magenta) can be broken into an infinite series of sine waves by using the Fourier transform (equation). The fundamental frequency, w 0 , has the greatest amplitude; each subsequent odd harmonic has a lower amplitude. Even-numbered harmonics are not included because of the symmetry of the square wave. 10 CSEM: Focusing on Hydrocarbon Detection MT measurements are not sensitive to thin resistive layers, so they are not well suited for evaluating potential hydrocarbon reservoirs. Over the course of a few decades starting in the 1980s, several research institutes and companies developed the equipment, modeling and interpretation tools that became the marine CSEM technique (see “Marine CSEM: Evolution of a Technology,” page 1).16 The systems are now widely available. Since the same receivers function for both CSEM and MT measurement, both responses can be recorded during a survey. The CSEM technique focuses on measuring and interpreting the response from the controlled source, while between those measurements, MT data are recorded. The processed and interpreted MT data establish a background model for the CSEM interpretation or inversion. The typical marine CSEM transmitter source is a long horizontal dipole (above left). The source comprises two neutrally buoyant antenna cables, each terminating in an electrode, thereby forming a dipole. The electrodes are pulled through the water behind a streamlined sensor platform, called a towfish, that is towed by the ship at a nominal speed of 2.8 to 3.7 km/h [1.7 to 2.3 mi/h or 1.5 to 2.0 knots] at an altitude of 50 to 100 m [160 to 330 ft] above the seabed. To provide accurate values for processing, the towfish measures seawater conductivity, local sound velocity and altitude above the seafloor. The strength of the dipole source is given by its dipole moment. This value is the product of the magnitude of the electric current flowing through the electrodes—given by the strength of the first harmonic of the output signal—and the distance between the electrodes. The power to generate a high-current, lowvoltage source signal and propagate it along 16.The first development was by Charles Cox of Scripps Institution of Oceanography: Cox CS: “On the Electrical Conductivity of the Oceanic Lithosphere,” Physics of the Earth and Planetary Interiors 25, no. 3 (May 1981): 196–201. For a recent overview of the history of CSEM: Constable S and Srnka LJ: “An Introduction to Marine ControlledSource Electromagnetic Methods for Hydrocarbon Exploration,” Geophysics 72, no. 2 (March–April 2007): WA3–WA12. Oilfield Review [6.2 mi] away, the electric-field magnitude is small, less than 1 nV/m. For the typical 10-m span of a seabed receiver dipole, the measured 10 nV is about 80 million times smaller than a AAA battery’s 1.2 V. The response magnetic-field magnitude at that distance from the source is about 0.0001 nT, which corresponds to about 2 parts in a billion of the Earth’s DC magnetic field. The controlled source typically generates square waves or sequences of square waves at user-defined fundamental frequencies. Fourier analysis resolves the square wave into sinusoidal waves of many frequencies (previous page, bottom left). The strongest components are the primary frequency w0 and the odd harmonics 3w0, 5w0 and 7w0, each with sequentially decreasing magnitudes. The combination of the skin-depth relationship to frequency and use of multiple frequencies means this process samples at several depths and with several resolutions. several kilometers of cable is typically provided by a 250-kV.A system on the ship. Transformers convert this to a low-current, high-voltage signal sent along the cable. In the towfish the signal is transformed back to the high-current, lowvoltage signal. The towfish generates a designed waveform based on commands from the ship. The actual current waveform transmitted by the source electrodes is measured and recorded by a data logger in the towfish and transmitted to the vessel in real time for quality control via high-speed telemetry. Because the waveform transmitted by the antenna is affected by antenna impedance and wear and by water salinity, accurate monitoring of the actual waveform is required to correctly resolve the survey data. Although the power emitted at the source is large—nominally 50 kW—the signal decays rapidly with distance. At a receiver placed 10 km The data from the receivers are collected as time-series data, but for the CSEM method, they must be synchronized to the source square-wave signal through accurate time measurement. Thus, in addition to the source GPS synchronization, each receiver has a high-precision clock that is GPS synchronized upon deployment and recovery. The instantaneous dipole-source position and orientation must also be captured for accurate inversion. Acoustic transponders in several locations along the antenna give this information by transmitting their positions at 1- to 4-s intervals. Accurate measurement of the feathering or tilt of the antenna is important for correct processing. The measurements of the fields are timedomain data, but these are typically converted to the frequency domain using a Fourier transform (below). The data are stacked by overlaying responses from multiple, sequential square-wave Ex Ey Hx Hy Time, min 5 10 15 20 25 30 35 40 –8 Frequency, Hz 0.0625 0.1875 0.25 0.315 0.4375 0.75 1.25 1.75 Scaled electric amplitude, V/(A.m2) –9 –10 –11 –12 –13 –14 –15 –16 –10 –9 –8 –7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10 Source-receiver offset, km > Converting time-domain measurements to amplitude versus offset. Each receiver records data for two horizontal electricand magnetic-field measurements (top). A Fourier transform converts these time-domain signals into the frequency domain. Fourier conversions of similar measurements at many receiver locations allow development of a frequency-dependent amplitude versus offset relationship (bottom). This can be developed for each measured component of the electric field (only one is shown) and the magnetic field. The resistivity of the subsurface affects the shape of these curves. Spring 2009 11 Air-wave signal Direct sig Water nal Conductor Geologic signal Resistor 10-3 10-4 Electric field, V/m 10-5 10-6 -7 10 10-8 Reservoir 10-9 Background (no resistive formation) 10-10 10-11 10-12 0 1 2 3 4 5 6 7 Source-receiver separation, km 8 9 10 > Paths from marine source to receivers. Signal energy from the marine source reaches the receivers by following three types of paths. A direct signal passes through the water to the receiver; this signal is strongest at the near-offset receivers. Signal energy that enters the subsurface interacts with layers of varying resistivity and generates a response signal containing geologic information that travels up to the receivers. Signal energy that reaches the air interface travels along the interface as an air wave, which also travels to receivers. In shallow water or at long source-receiver offsets in deep water, the air-wave signal is strongest. series, called a time gather, to improve the signal/ noise ratio. The window for the time gather must be short enough that the source movement does not significantly alter the sampled volume of the subsurface. Since the objective of E&P prospecting is to detect hydrocarbons, the signal from the CSEM source is optimized to find thin, nonconducting layers (possible hydrocarbon-bearing formations) in a conducting background (waterbearing formations). The discussion on skin depth pointed out that detecting thin formations requires higher-frequency components than available using MT. The typical frequency range of the CSEM signal is between 0.05 and 5 Hz; 1 Hz is the effective upper limit for marine MT studies. As a first-order approximation, the signal can take three general paths between the source and the receivers (above). When the sourcereceiver offset distance is short, the direct path through the water dominates the signal. The strength of the signal decreases rapidly with 12 distance because of its attenuation in conductive water. A second contribution comes from the air wave. The electromagnetic field travels to the water surface, where it encounters highly resistive air. The resistance contrast forces the wave propagation to follow the air/water interface. In deep water, the air-wave signal dominates only at long offsets, normally beyond 10 km, because, unlike the signals following the other two paths, the signal at the air/water interface has little attenuation. The third portion of the signal travels through the subsurface. Under the proper conditions of frequency, water depth and subsurface conductivity, there is aEM_FIGURE range of offsets 11 for which the third path dominates the signal. For this path, waves propagate into the subsurface, where they interact with resistive formations and generate a response field; some of that energy travels back to the seafloor receivers. This response signal appears at receivers at offset distances that are typically greater than the reservoir depth below the seabed, but at even greater offsets it attenu- ates so much that the air-wave signal overwhelms it. Since the waves propagate more easily though a resistive than a conductive formation, the presence of a reservoir enhances the received signal compared to a uniform subsurface lacking a resistive layer. Geoscientists can identify resistivity anomalies and therefore infer geologic information by analytic means through comparing the observed data with predictive models or by numeric means through inversion. At a certain offset distance, the natural noise limitation of the receiver exceeds the strength of the signal that originated at the source transmitter, presenting an effective limit on the depth of investigation in the subsurface. This limitation, or noise floor, varies with frequency and depends on the characteristics of the receiver and its environment—such as mechanical noise generated by water waves moving the antennas. The noise floor can be lowered through improved instrumentation, such as quieter electronics or more-stable mechanics, or through intelligent signal processing to remove motion noise or coherent noise across the survey. The source, receiver and environmental characteristics can be incorporated into a presurvey analysis to determine whether a resistive target at a certain depth can be detected (next page). Carbonates, which are resistive, present a problem: A trap with low oil saturation inside a resistive carbonate host may have insufficient detectable contrast. The receiver data can be presented as electricor magnetic-field amplitudes and phases that are functions of the offset distance between source and receiver. The effect of a resistive anomaly can be highlighted by several methods: analytic methods using only measured data, modelbased methods derived during survey planning, and inversion. Oilfield Review Spring 2009 0 1.5 1.4 Seawater 500 1.3 1.2 1.1 1,000 Frequency, Hz 1.0 1,500 2,000 2,500 0.9 0.8 0.7 0.6 Amplitude ratio 6 0.5 5 0.4 4 3 0.3 3,000 2 0.2 Depth, m 1 0 3,500 2,000 1.5 4,000 Basalt 4,000 6,000 8,000 Transmitter-receiver offset, m 10,000 12,000 1.4 1.3 4,500 1.2 1.1 5,000 1.0 Frequency, Hz One of the analytic methods normalizes the electric- and magnetic-field amplitudes versus offset response over the anomaly to the response of a distant receiver that does not sense the anomaly. A second analytic method compares the normalized response of the inline measurement with the crossline measurement, essentially comparing the two horizontal components of the electric field, Ex and Ey. The presence of an underlying resistive structure, such as a hydrocarbon-bearing formation, has greater effect on the inline response because of the polarization of the signal. A third analytic method converts the field data to apparent resistivity in a 2D or 3D pseudosection plotted as a function of source-receiver offset and signal frequency.17 When the dataset is normalized to a section space that contains no anomaly, the anomalous apparent resistivity values appear as deviations from unity. Alternatively, presurvey models can be built when seismic data or data from nearby wells are available. Typically, a WesternGeco survey includes at least two 3D models that are based on the target properties and survey geometry. One model incorporates a resistive body; the other uses a uniform earth without a resistive body. Response curves are extracted from the 3D models for each receiver-site and tow-line combination. Once data are acquired, the observations can be normalized to each of the models to determine which provides the best fit. Beyond these analytic and model-based methods, CSEM inversion is a powerful way to derive the earth’s resistivity profile from observed data. However, like most inversion methods, the solution is not unique. Forward-modeling codes are run iteratively with model parameters perturbed until the output result matches the data within an acceptable range. Jointly inverting as many significant channels and frequencies as possible constrains the possible solutions, but at a cost of longer processing time. Additional constraints—such as placement of known geologic structures—are sometimes introduced. Log and seismic data provide a starting model to help constrain the inversion. MT data also have limited resolution, so the modeling step benefits from information based on other types of measurement. Seismic interpretations often serve as constraints. Gravity surveys provide an independent constraint, as do well logs. The WesternGeco MMCI multi measurement-constrained imaging technique uses an iterative approach with gravity, MT and seismic data to improve inversion results, leading to a final, more-constrained depth image. 5,500 6,000 0.9 0.8 0.7 Phase difference, ° 0.6 40 30 20 10 0 –10 –20 –30 –40 0.5 6,500 0.4 0.3 0.2 7,000 1 10 Resistivity, ohm.m 100 0 2,000 4,000 6,000 8,000 Transmitter-receiver offset, m 10,000 12,000 > Presurvey modeling. To optimize CSEM acquisition parameters, the subsurface is modeled as a series of resistive layers (left ). Two models having identical geometries are compared. One model incorporates a layer of highly resistive basalt (yellow and brown); the other model assigns that layer a lower resistivity (yellow only). The two models have different phase and amplitude responses to a simulated CSEM pulse. The amplitude ratio between the models (top right ) is maximum (red) at an offset—distance from source to receiver—of about 7,000 m and at a frequency of about 0.7 Hz. The phase difference (bottom right ) has a maximum (red) at about 8,500 m and at a frequency less than 0.1 Hz, and another maximum (violet) at long offset and high frequency. Based on the information in both plots, geoscientists determined that the optimal offset to maximize the chance of detecting this anomaly is about 8,000 m, at frequencies of 0.5 and 0.125 Hz. The contour lines indicate various levels of receiver noise floors (labeled by the power of 10), which depend on the sensors, electronics and the environment. Although the noise floor in some environments may be as poor as 10-14, these plots extend to a noise floor of 10-15, which can usually be achieved. Although marine MT and CSEM receivers Mexico, offshore Louisiana, USA.18 Exploration have been used in studies since the 1990s, the companies have had an interest in evaluating industry’s interest has risen rapidly in the last hydrocarbon potential in subsalt formations few years, resulting in a rapid increase in the in this area. The seismic data available at the total number of sites evaluated. A large, multi time, a legacy survey called E-Cat, had been phase study recently performed in the Gulf of reprocessed recently over Garden Banks, but it Mexico included more marine MT receivers than had insufficient resolution to reliably determine EM_FIGURE 05a the base of a salt intrusion. The objective of the the total deployed worldwide to that date. new study was to integrate marine MT, fulltensor gravity and seismic measurements using Finding the Base of Salt In 2006, WesternGeco began a test of the MMCI an MMCI evaluation to improve the interpretaconcept in the Garden Banks area of the Gulf of tion of the base of salt. 17.A pseudosection uses approximate or pseudo spatial coordinates. It provides a semiquantitative way to look at spatial data. 18.Sandberg SK, Roper T and Campbell T: “Marine Magnetotelluric (MMT) Data Interpretation in the Gulf of Mexico for Subsalt Imaging,” paper OTC 19659, presented at the 2008 Offshore Technology Conference, Houston, May 5–8, 2008. 13 Line 5 Line 4 Line 3 Line 2 Line 1 N Seafloor depth, m 750 975 1,200 Tamara well 1,425 1,650 Line 6 The Garden Banks study included 171 seabed receivers, more than any previous marine MT survey, although surveys of this density are more common today. The survey utilized five parallel north-south lines of receivers, about 2.5 km [1.6 mi] apart, and an east-west cross line (right). Additional receivers placed between these lines provided denser coverage near the center of the survey area. Bathymetry data indicated seabed expressions of the underlying salt domes. During the course of the project, two events provided additional data for this investigation. During October and November of 2007, WesternGeco acquired a multiclient wideazimuth (WAZ) seismic survey over the area, which provided significantly better resolution for base of salt than did the previous E-Cat narrow-azimuth survey. However, even with WAZ illumination, the base of salt was poorly resolved in some areas.19 The second event occurred near the end of 2007, when BP released logging data from its Tamara well in Garden Banks Block 873. This well was drilled through the central portion of the survey area. The gamma ray log indicating the base of salt became available after most of the MT interpretation was completed, providing a base-truth point for comparison. An approach combining 1D models for each receiver station detected the salt body, but the details of the structure were incorrect because of its complex geometry. Several 2D approaches 14 1,875 2,100 0 0 km 10 mi 10 > MT survey in Garden Banks area. The MT receivers (inset) were placed in five north-south lines and one east-west crossline. Additional receivers were placed in the central area, near the Tamara well. The color-coding indicates seawater depth from bathymetry. 19.For more on WAZ surveys: Camara Alfaro J, Corcoran C, Davies K, Gonzalez Pineda F, Hampson G, Hill D, Howard M, Kapoor J, Moldoveanu N and Kragh E: “Reducing Exploration Risk,” Oilfield Review 19, no. 1 (Spring 2007): 26–43. 20.An autochthonous formation is one that was deposited in its current location. This salt would be the source of the shallower salt bodies that moved to their current positions because of density difference and salt plasticity. EM_FIGURE 13 Oilfield Review Spring 2009 Gravity survey response Density 2.1 kg/m3 2.7 kg/m3 Volume = 3,600 m3 3,600 m3 2,800 m3 > Nonuniqueness of gravity survey. A gravity survey responds to the mass of an anomaly. A solution can propose one object or many, or have different density and size, as long as the mass and the center-of-mass location for the anomaly are the same. In this example, all three readings measure the same mass. The success of this proof-of-concept study was the impetus for a large-scale multisurvey MT project that has been active since May 2007 in other key areas in the Gulf of Mexico. For example, in the Keathley Canyon area, deter- mining base of salt from seismic data alone was difficult. Gravity data provided improvement, but several alternative interpretations could not be distinguished. By adding MT data and combining all the information through the MMCI approach, Distance, km –80 –40 0 40 2,500 5,000 80 120 SE Resistivity, ohm.m NW Depth, m were also used, but the results of all the 2D inversions indicated thinner salt bodies than shown by data from the Tamara well. The threedimensional nature of the body dictated a 3D approach to modeling. The first 3D approach taken by the study team was to fit the MT data independent of seismic and gravity data. The model started with a homogeneous and isotropic resistivity below the seabed. During iterations, each cell resistivity was allowed to change to match the apparent resistivity and phase measurements. A smooth inversion algorithm ensured that the resistivity changed as smoothly as possible from cell to cell. Agreement above the main salt body was good among the WAZ seismic result, the MT fit and the gravity model. In addition, the interpreted base of salt is within a few hundred feet of the log-derived base of salt in the Tamara well—a good match. However, the gravity model required adjustments to fit the measured gravity data. Similar gravity measurement results can be obtained for different configurations (above right). In this case, salt could be added either to the salt layer within the model space or to the autochthonous salt—which was mostly below the maximum depth of the seismic-velocity volume—or the subsalt formation densities could be altered to match the result.20 A second approach used the interpreted seismic survey to provide a starting point for the shape of the salt body. Resistivity for this a priori model was initially set at 50 ohm.m inside the salt body and at 1.2 ohm.m in the surrounding sediments. The inversion changes the values of the resistivity in the grid blocks to fit the measurement data while preserving the initial model as much as possible. The best interpretation used MMCI procedures, incorporating all available information, including MT, gravity and WAZ seismic data. Porosities were computed from the WAZ velocity field using local knowledge of the sand/shale ratio of the sedimentary section, and densities were computed from the matrix densities of sand and shale, the density of seawater and the velocity-derived porosity. Density in the salt was assumed constant at 2.16 g/cm3. The 3D model result in the Garden Banks area had an improved interpretation of the base of salt compared with that based on seismic data alone (right). Resistivity data indicated that a large lobe suggested by the seismic interpretation is not a part of the salt, but belongs to an underlying formation. 10 1 EM_FIGURE 16 7,500 10,000 > Confirmation by drilling. MT measurements detected a high-resistivity salt intrusion (pink). The Tamara well, drilled near MT receiver Line 3, provides a point of reference for interpretations of base of salt. The base-salt interpretation (gray) of the best WAZ data available shows a lobe to the southeast that is not supported by the MT resistivity data; the 35- to 50-ohm.m area of resistivity (pink) excludes that lobe from the salt. The 3D MMCI interpretation of seismic, gravity and MT data indicates a base of salt (white) within a few hundred vertical feet of the base determined from the well gamma ray log (turquoise). At the base of salt, the well log resistivity (orange) decreases significantly. MT receiver locations are shown on the seabed (white squares). 15 the analysis team obtained a consistent interpretation of the structure, including the base of salt (below). In parts of the survey area, the difference in interpretation of the base of salt was almost 3,000 m [9,700 ft]. EM Studies Offshore Brazil Marine MT surveys have also improved depth imaging in other parts of the world. The Santos basin, offshore Brazil, contains recent subsalt discoveries made by Petrobras. High-resolution Distance, km 128 136 144 152 160 NW SE 2,500 Depth, m 5,000 7,500 10,000 12,500 Distance, km 128 136 144 NW 152 160 SE 2,500 Resistivity, ohm.m 10 Depth, m 5,000 1 7,500 10,000 MMCI base salt 12,500 Seismic base salt > Keathley Canyon interpretation. The base of salt is difficult to find in the WAZ seismic section (top). The best pick based on the seismic data had a thick section to the right of middle (white outline, bottom). MT resistivity data (colors) add significant new information. Combining seismic, MT and gravity data in the MMCI evaluation improves the previous interpretations of the base of salt and gives interpreters greater confidence in their result (yellow dashed line). 16 seismic imaging has mapped the stratigraphy of hydrocarbon-producing turbidite reservoirs and the geometries of salt structures, including a thick sedimentary sequence in a syn-rift structure beneath the salt.21 The lithology of this sequence was defined by the first discovery well of the Tupi area. An MT survey northwest of Tupi confirmed the complex structure and demonstrated the utility of marine MT surveys to Petrobras.22 To the east of the Santos basin MT survey just described, Petrobras and WesternGeco performed a marine CSEM survey in the Tambuatá block of the basin as part of a cooperative project (next page, top).23 The survey location was about 170 km [106 mi] south of Rio de Janeiro. The water depth was taken from bathymetry data, and processing also included the variation in seawater resistivity as a function of depth. The survey used 180 receivers spaced approximately 1 km [0.6 mi] apart and deployed on the seabed over known reservoirs. A vessel towed the source over the receiver lines. Data acquisition used 0.25- and 0.0625-Hz square-wave signals that are also rich in odd harmonics of these frequencies.24 Analysts processed multicomponent electricand magnetic-field responses for all frequencies in the survey using an advanced workflow based on instantaneous measures of dipole length, dipole moment, dipole altitude, feather angle and dip. The data interpretation proceeded in stages, starting with generating a background model to compare with the processed measurements. Borehole measurements provided information on background resistivities, but the log data have more detail than CSEM measurements can discriminate. Thus, analysts reduced the number of layers in the resistivity model to reflect the resolving power of CSEM, but they ensured the resampled well logs retained the same CSEM response as the detailed log-based layering would have. To determine where the boundaries had to be placed, both the cumulative resistance and cumulative conductance were calculated from the well logs and coupled with stratigraphy. This not only clarified the locations of the layer interfaces but also determined the resistivities of the layers and the anisotropy caused by interbedding low- and high-resistivity layers. Analysts conducted detailed 3D modeling based on the blocked well log resistivities and based on model geometries derived from seismic sections without incorporating any reservoirs. The resulting models generated reference background fields, which provided a basis to normalize processed multicomponent field data at each receiver location. Oilfield Review Risking Prospects in the Arctic Frontier As operators move into increasingly difficult environments, the Arctic beckons as one of the last mostly unexploited frontiers. In 2008, the US Geological Survey (USGS) estimated undiscovered resources north of the Arctic Circle at 14 billion m3 [90 billion bbl] of oil and 47.8 trillion m3 [1,669 Tcf] of gas—of that total the province west of Greenland and east of Canada had an estimated 1.1 billion m3 [7 billion bbl] of oil and 1.5 trillion m3 [52 Tcf] of gas.25 21.Syn-rift refers to events that occur at the same time as the process of rifting. A syn-rift basin is formed along with, and as a consequence of, the rifting process. In the Santos basin, the rifting refers to the early stages of the separation of the South American and African continents. 22.de Lugao PP, Fontes SL, La Terra EF, Zerilli A, Labruzzo T and Buonora MP: “First Application of Marine Magnetotellurics Improves Depth Imaging in the Santos Basin–Brazil,” paper P192, presented at the 70th EAGE Conference and Exhibition, Rome, June 9–12, 2008. 23.Buonora MP, Zerilli A, Labruzzo T and Rodrigues LF: “Advancing Marine Controlled Source Electromagnetics in the Santos Basin, Brazil,” paper G008, presented at the 70th EAGE Conference and Exhibition, Rome, June 9–12, 2008. 24.The strongest harmonics are 0.75, 1.25 and 1.75 Hz for the 0.25-Hz signal, and they are 0.1875, 0.3125 and 0.4375 Hz for the 0.0625-Hz signal. 25.Bird KJ, Charpentier RR, Gautier DL, Houseknecht DW, Klett TR, Pitman JK, Moore TE, Schenk CJ, Tennyson ME and Wandrey CJ: “Circum-Arctic Resource Appraisal: Estimates of Undiscovered Oil and Gas North of the Arctic Circle,” USGS Fact Sheet 2008-3049 (2008), http://pubs. usgs.gov/fs/2008/3049/ (accessed March 31, 2009). Spring 2009 –46° –48° –44° –42° Altitude, m 1,365 –22° N 662 Rio de Janeiro 0 MT São Paulo –135 –2,286 B Selected tow lines were interpreted using a 2.5D inversion. The 2.5D analysis incorporates a 2D geological model and solves for multiple transmitter positions simultaneously, but the sources and receivers are not confined to the plane of the geological model. Thus, realistic acquisition geometries can be simulated (bottom right). The known reservoir underlying the survey area appeared in the EM response as a zone of higher resistivity than the surrounding formations. As with the MT project farther west in the Santos basin, the CSEM project also shows promise for adding considerable value in upstream applications. Both projects underscore the need for advanced integrated interpretation to improve the result over individual seismic, well log and electromagnetic measurements. They also advance the case for the industry to include these novel integration paradigms in standard applications. Petrobras has a technical collaboration agreement with Schlumberger to develop technology that integrates marine EM into other technologies for enhanced depth imaging and reservoir characterization. a s rea ya rve u S –24° CSEM s nto Sa e ad ci –3,784 Ocean depth, m 0 km 0 100 miles –26° –48° Tupi area 100 –46° –44° –26° –42° > Marine MT and CSEM surveys, offshore Brazil. Three lines of receivers for the MT survey (red) extended offshore toward the southeast and into deeper water. The main line was about 148 km [93 mi] long, starting about 42 km [26 mi] offshore, and the two adjacent lines were each about 54 km [34 mi] long. The CSEM survey lines (white) to the east of the MT survey covered the Tambuatá block (red). The map shows the ground elevation and ocean depth. 0 LTAM1 N N 0 0 km 10 mi 10 Resistivity, ohm.m 40 EM_FIGURE 17 10 1 0.4 > Combined analysis for the Tambuatá block. Reservoirs (green and pink outlines, top) identified by seismic interpretation were the targets of a CSEM and MT study. Receivers (white triangles) were laid in orthogonal sets, and the CSEM source was towed along the same lines (black). A 2.5D MMCI inversion based on EM and seismic data resulted in a section color-coded for resistivity, with seismic data providing texture (bottom). Along tow line LTAM10 N, a 20-ohm.m resistive anomaly (red) is clearly distinguished from the more-conductive background of about 1.2 ohm.m (green). Seismic results constrained the anomaly shape—by defined control points (white circles, bottom)—for the data inversion. 17 Prospect without resistive anomalies Volcanic flows Volcanic flows Resistivity, ohm.m 20 18 16 14 12 10 8 6 4 2 Prospects with resistive anomalies N > Prospects with resistive anomalies. Several prospects in a block west of Greenland were interpreted from seismic data (green outlines). The survey design placed lines of CSEM receivers (white icons) along the source tow lines (white lines) above the seismically determined prospects. The CSEM study distinguished the structures with vertical resistive anomalies (oranges and yellows) from those with no anomaly (representative locations labeled). Volcanic flows above the target formation are also identified along the lines. In this view, resistivities less than 10 ohm.m are not shown. The contour lines indicate depth of the seismic horizon of the target; each contour line represents a 100-m [328-ft] depth difference (also represented as the background color sequence). EnCana Corporation and its joint-venture exploration targets. For more information on (JV) partners Nunaoil A/S and Cairn Energy volcanic formations, see “Evaluating Volcanic have exploration prospects in two blocks in the Reservoirs,” page 36. Before conducting the CSEM survey, frontier basin offshore Greenland, 120 to 200 km [75 to 124 mi] west of the capital city, Nuuk. WesternGeco performed extensive 3D resistivity The ocean depth over the prospects ranges from modeling over each prospect. This step confirmed 250 to 1,800 m [820 to 5,900 ft]. Geologists that the survey could help define the presence of believe this area’s rifting and sedimentary-fill hydrocarbon-bearing reservoirs at up to 3,000 m history is similar to that of the productive North below the seafloor. Synthetic data were used in Sea basins. However, the nearest well control is forward-modeling and inversion methods. Based more than 120 km away, and there are no proven on well log data from key distant offset wells, petroleum systems in the basins. The JV needed a simplified starting model was created that a way to lessen the risk of drilling dry holes, so a included a reasonably uniform, 1.5-ohm.m clastic sedimentary section from the seafloor to the CSEM survey was acquired to help identify potenEM_FIGURE 26 target depth, a deeper layer with 4-ohm.m resistial hydrocarbon-bearing features.26 Sedimentary filling of the basin following rift- tivity extending to the basement, and a 60-ohm.m ing created a fairly simple geology, with the major basement formation. As part of this presurvey analysis, geosciencomplication coming from Paleocene volcanic activity. The volcanic flows are easily identifiable tists optimized the design for target sensitivity, geologically, seismically and magnetically. These presence of volcanic cover, reservoir proximity to volcanic rocks are the only known resistive litho- basement and signal waveform, as a few examlogic units in the survey area above basement, ple parameters. This optimization helped the and they are well separated from the Cretaceous EnCana JV plan a survey covering the vast area in a cost-efficient way. 18 The survey layout based on this analysis comprised 24 transmitter lines and 182 receivers. The tow-line geometry generated data from multiple angles on the receivers. The resulting vertical resolution was designed to be 50 m [164 ft] for the Cretaceous targets at depths of 3,500 m [11,500 ft] below the seafloor. A high-quality CSEM dataset was obtained in the summer of 2008. Processing the electric- and magnetic-field measurements yielded amplitude and phase responses at each receiver. Starting with electric-field responses, geoscientists analyzed these data using a complex 3D anisotropic-resistivity model. The starting geometry used the JV’s seismic interpretation and well log resistivity information, but no potential reservoirs were included. The 3D inversions required considerable computation time and interpreter input.27 The results were numerically stable with electrical models that were geologically consistent. The inversion process identified resistive anomalies over 8 of 14 prospects. The team used Petrel seismic-to-simulation software to visualize the resistivity volume data for these eight anomalies with geologic, seismic, gravity, magnetic and marine MT data (left). The results were insensitive to reasonable variations in the starting model, with each variation converging to a similar resistivity solution. The known Paleocene volcanic rocks provided another indication that the inversions were robust and geologically meaningful. Although the isolated volcanic features were not included in the starting models for the inversions, the inversion procedure located them correctly. The EnCana JV’s objective for obtaining the CSEM study was to improve the assessment of the probability that the structures were charged with hydrocarbons. With firm data lacking prior to the study, the hydrocarbon-charge probability was indeterminate and the JV assigned it an initial value of 50% for each of the eight prospects. The team’s analysis increased the probability of hydrocarbon charging for several features and decreased it for others. The prospect with the greatest probability for hydrocarbon charging displays many of the characteristics the geoscientists looked for in the analysis. Its resistivity anomaly conforms well with the target interval. The CSEM inversion resistivity within the anomaly increases upward from 10 ohm.m at the base of the structure to 35 ohm.m at the crest. Finally, the anomaly base is flat, which could suggest a hydrocarbon/ water contact. Oilfield Review > Deployment of CSEM receiver. Each receiver is assembled on the deck using defined deployment protocols. Then the receiver is hoisted and dropped at a specified location. EnCana and its partners are now prioritizing their prospects to identify the most prospective drilling candidates based on the geology, the geophysical mapping and the CSEM 3D model inversion results. The risk for exploration in this frontier Arctic basin remains great, but CSEM technology offers promising potential to reduce dry holes. Sounding for the Next Generation Although MT and CSEM surveys have been performed for many years, commercial use of the marine technology in the E&P industry is relatively new. The industry is still in its infancy in interpreting this electromagnetic survey data and combining the information with that of seismic surveys. 26.Umbach KE, Ferster A, Lovatini A and Watts D: “Hydrocarbon Charge Risk Assessment Using 3D CSEM Inversion Derived Resistivity in a Frontier Basin, Offshore West Greenland,” CSPG CSEG CWLS Convention, Calgary, May 4–8, 2009. 27.Mackie R, Watts D and Rodi W: “Joint 3D Inversion of Marine CSEM and MT Data,” SEG Expanded Abstracts 26, no. 1 (2007): 574–578. 28.National Petroleum Council (ed): Hard Truths: Facing the Hard Truths about Energy. Washington, DC: National Petroleum Council, 2007. Also available online at http:// www.npchardtruthsreport.org/ (accessed May 5, 2009). 29.WesternGeco regularly performs 3D modeling studies and offers 3D CSEM inversion including the use of algorithms in which the MT data are jointly inverted to help constrain the CSEM inversion. Spring 2009 The seabed receivers used by WesternGeco follow the basic design developed by Scripps Institution of Oceanography, but the devices and methodologies are continually being upgraded to improve instrument efficiency and reliability. In addition to changes in materials used in the manufacture of the dipoles and magnetometers and their packaging, new equipment has been added to the receiver pack, such as a highprecision compass. The dipole source for CSEM is also under going improvement by the industry. Equipment vendors have worked to refine the timing synchronization of the source waveform and the precise positioning of the source antenna. Major obstacles to marine EM efficiency are the cost and time involved in data collection. Seismic measurements over large 3D areas are efficient because vessels tow multiple receiving streamers and source array guns. In contrast, CSEM surveys cover less area because either sources or receivers, or both, are deployed individually, with receivers remaining stationary during the survey and then recovered (above). The development of a purely surface-towed, deep-reading EM system is likely at the forefront of R&D activities at many geophysical companies. The problems are noise inherent in the motion of sensors through the water and signal attenuation in seawater, which dramatically reduce the coupling of the source with the seafloor and the amplitude of the response field. The dipole antennas are long, and even with the present seabed configuration, currents can move the antennas and impact data quality. The National Petroleum Council (NPC), an industry body that advises the US government, studied several advancements related to CSEM, rating them as highly significant for exploration activities.28 To secure energy resources for the future, this expert group identified two improvements in CSEM technologies needed over the short term. Development of fast CSEM 3D modeling and inversion could reduce the number of false positives, or resistive anomalies that currently may be misinterpreted as a commercial petroleum response. These include hydrates, salt bodies and volcanic lithologies. The second short-term goal is integration of CSEM with structural information from seismic surveys to improve the resolution of the EM data. As discussed in the case studies in this article, this work is currently underway through efforts such as the MMCI method.29 Over a longer term, the NPC experts also rated advancing the realm of CSEM studies into shallow water, onshore, and deeper formations as highly significant. The signals in shallow water and onshore are much noisier than in deep water because of the air wave. Signal strength now limits the depth of the CSEM surveys, but the NPC group saw that developments leading to evaluating deeper formations would extend the application to new basins. Alternative acquisition geometries might play a role in ultradeep reservoirs. The term “electromagnetic sounding” is not yet commonly heard in the E&P industry, but impressive results from this generation of tools and interpretation methods have already sent a clear message. With commercial success will come further advances in technology and a wider variety of applications. —MAA 19 Near-Surface Electromagnetic Surveying The E&P industry typically focuses on deep formations, but frequently the nearsurface layers also need to be evaluated. Land-based electromagnetic surveys provide insights into this often complex zone. The interpreted resistivities of these layers help map and define features for applications as diverse as seismic studies and aquifer delineation. Mohamed Dawoud Environment Agency–Abu Dhabi Abu Dhabi, UAE Stephen Hallinan Milan, Italy Rolf Herrmann Abu Dhabi Frank van Kleef Dubai Petroleum Establishment Dubai, UAE Oilfield Review Spring 2009: 21, no. 1. Copyright © 2009 Schlumberger. For help in preparation of this article, thanks to Marcus Ganz, Houston. 1. The principle describing the changing magnetic field is Faraday’s law of induction. Including the sign of the induced current is Lenz’s law. The converse principle involving a changing current or electric field is Ampère’s law. These laws are included in Maxwell’s equations. 20 The near surface of the Earth is a complex place, the result of dynamic action by wind, water and other forces of nature. In those first tens of meters below the surface, the jumbled detritus of weathering is gradually buried. The near-surface layers, like those below, vary in resistivity according to their mineral and fluid compositions. This property allows their investigation using electromagnetic (EM) surveys. Often, surveys are performed using an artificial source of EM radiation, rather than the magnetotelluric (MT) radiation resulting from the interaction of the solar wind with the Earth’s magnetosphere. On land, there are two general methods of controlled-source electromagnetic (CSEM) measurement for generating the signal and detecting the response. The grounded-source method requires burial of source and receiver electrodes in electrical contact with the earth. The inductive-source method uses a current loop on the surface to induce a variable magnetic field, and the same or another loop to detect the response signal. The grounded-source method is efficient and sensitive to horizontal resistive targets because the electric field has a vertical component. The grounded receivers measure the response electric field; response magnetic fields are also measured to provide control during modeling. On land, however, the surface conditions must be suitable for creating and maintaining electrical contact. This prerequisite excludes the practical application of this method in large, arid dunes—sand grains are nonconducting. But in some areas, contact can be improved by drilling patterns of shallow holes for source and receiver electrodes and wetting the soil as the hole is refilled. Deeper investigation into the Earth requires stronger current sources, among other factors, and the high contact resistances on land mean this requires high-voltage systems to drive that current. The inductive-source method does not require electrical contact, since the current loop generates a magnetic field through a time-varying signal. This field generates a response electric field, but because the electric field is largely horizontal, the process is not as efficient for imaging horizontal layers of resistive hydrocarbons as the direct injection of current using the grounded-source method. Again, both the electric and magnetic response fields can be measured using the inductive-source technique. Coils for the current loops are square and for nearsurface investigation range from about 10 to 300 m [30 to 1,000 ft] on a side. Far larger loops have been used for deeper, but low-resolution, investigation. A companion article (see “Electromagnetic Sounding for Hydrocarbons,” page 4) describes the basic physics of the EM interaction with the Earth and discusses marine EM studies. It also covers MT in detail, because the objectives of those studies are similar for land and marine environments. This article focuses on investigations Oilfield Review using the inductive-loop method for nearsurface imaging, illustrated by two WesternGeco cases from the United Arab Emirates. One study mapped an aquifer in Abu Dhabi for a waterstorage project. The second determined nearsurface resistivity variations in Dubai sand dunes, providing valuable input for making static corrections in a seismic survey of the area. Stacking Time Sequences Maxwell’s equations describe the basic physics behind the interplay of electric and magnetic fields in a time-varying current loop. A current loop generates a magnetic field. If the current changes, the induced field also changes. The opposite is also true: Changing the flux of a magnetic field within a conducting loop induces a changing current.1 A simple way to generate such a current is to move a magnet toward or away from a wire loop. The movement changes the flux through the loop, inducing a current. This current induces a response magnetic field oriented to oppose the change in flux through the loop caused by the moving magnet. No actual magnet is required for this effect to take place. One coil with an imposed timevarying current sets up a time-varying magnetic LAND EM_OPENER Spring 2009 21 > Basic induction loops. An alternating current passing through a set of coils (blue) induces a cyclic magnetic field. When this field passes through a second set of coils (red), it induces a cyclic current in that circuit. Thus, energy passes from one circuit to another without a direct electrical contact. This is the basis for a transformer, which is a device that converts from an input voltage to a different output voltage by having different numbers of loops in the two coils. Ramp time Batteries and signal electronics Transmitter current and primary magnetic field Ramp time Time on Time off Time on field in response. Current is induced in a second coil positioned close enough to experience the changing flux. This is the configuration of a transformer (left). Energy passes from one circuit to the other through the changing magnetic field. One method that uses inductive measurement for evaluating the near surface is a time-domain electromagnetic (TDEM) survey. A conducting loop of wire set out in a square on the surface of the Earth acts as the first coil, and the second loop forms in conducting formations of the Earth itself. The primary magnetic field from the transmitter loop generates horizontal currents, called eddy currents, immediately beneath the loop. These currents induce a response field that can be detected at a surface receiver loop, but this field also travels farther into the subsurface, generating progressively weaker eddy current loops with larger radii and smaller response fields (below).2 Source current loop Measurement period Induced electromotive force in nearby conducting layers Eddy currents induced by field change LAND EM_FIGURE 03 Depth Response magnetic field induced by eddy currents Receiver-coil voltage from response magnetic field Time Secondary magnetic field Eddy currents at later times > TDEM inductive method. A large square loop is placed on the surface at the sounding site (top right). Passing a current pulse through this loop generates the primary magnetic field. The field induces a secondary, eddy current loop in the ground, as described by Faraday’s law (middle right ). This secondary current induces a response magnetic field, which can be recorded by a receiver loop on the surface. In this case, the same loop is used as source and receiver. The primary field decays with depth into the ground, generating response fields from each subsequent depth (bottom right ). The timing of the signals and response fields is also shown (left). 22 The transmitter and receiver loops can be the same wire coil when an appropriate time sequence of current steps is applied, or coaxial but separate coils in a typical setup. In all inductive-loop TDEM surveys, the time sequence begins by turning the current on to a constant DC value. Sufficient time elapses for transient responses in the subsurface to decay. Then, electronics shut the current off in a rapid, controlled ramp, inducing a known electromotive force in the immediate subsurface. The transient electromotive force generates eddy currents, producing a secondary magnetic field that decays with time. The secondary field is detected by the receiver coil. After enough time has elapsed, the sequence repeats with opposite polarity. Stacking many repeated responses improves the signal/noise ratio. As the eddy currents move progressively deeper into the subsurface, the response field contains resistivity information about deeper layers. Fundamentally, the resistivity variation with depth determines the rate of decay of the transient; higher conductivity results in slower decay. Inversion of the stacked data from a TDEM sounding reveals the distribution of nearsurface resistivity. The measured signals are very small, so land surveys must consider and avoid, if possible, any sources of noise. Electric trains, power lines, electric fences, buried utility cables, pipelines and water pumps distort the local measurement; large temperature variations and wind impact stability; and variations in soil moisture and permeability affect uniformity.3 TDEM is not commonly used directly in oil and gas exploration, although it has utility in evaluating surface statics for seismic studies. The method is applied widely for exploration within the mining industry, employing both terrestrial and airborne sources. It is also a tool for environmental and water-resource management, as shown in the first of the following case studies from the Middle East. Sounding for Water Storage A recent land-based EM application helped to locate potential water-storage sites in the UAE.4 The Environment Agency–Abu Dhabi (EAD) is managing a study for the government to evaluate storage plans for 30 billion British imperial gallons [136 million m3, 36 billion galUS] of fresh water in the northeast area of the Emirate.5 The country wants a freshwater reserve for emergency periods and to meet peak Oilfield Review summer demands. The water for this aquifer storage and recovery (ASR) project will be transported via pipeline from a water-desalination plant in the Emirate of Fujairah. EAD retained Schlumberger to identify and test a potential ASR site. This involved defining the subsurface storage zone and surrounding formations, aquifer thickness and related hydraulic parameters. Schlumberger selected a preferred site and constructed three pilot wells, which were tested to determine the aquifer’s potential. Geologic studies of the area found that deep-seated faults had been reactivated in the Late Tertiary Period by the northeastward displacement of the Arabian Peninsula, resulting in a series of tightly folded faults. The overlying layer of Quaternary Period sediments, consisting of eolian sands and alluvium, were primarily deposited along the reactivated faults and in the synclines between them, giving the sediment thickness a directional bias. The directionality can influence groundwater flow, creating a preference for flow parallel to the structure. Schlumberger evaluated this geologic structure in 2006 during the drilling, logging and testing periods. Logs from the pilot wells indicated a resistivity contrast between the targeted sand-and-gravel aquifer and the underlying clayrich layer. Because TDEM data are useful for aquifer characterization, the evaluation over the ASR site included a survey to define the lateral extent of the aquifer—necessary to compute the potential water-storage volume. In a time-domain survey, the apparent resistivity of the underlying formations is determined from the time variation of the electric and magnetic response fields. For the ASR study, the same set of coils placed on the surface was used for both current and receiver loops. The survey covered a 6- by 7-km [3.7- by 4.4-mi] area. A 1D Occam inversion at each receiver yielded resistivity input for constructing a 3D model.6 The maximum depth obtained by the inversion was about 250 m [820 ft]. Resistivity logs from the three wells were compared with the inversion results at adjacent sounding locations (right). The top of the underlying clay unit is clear in the eastern part of the survey area but less obvious in the western part. 2. Nabighian MS: “Quasi-Static Transient Response of a Conducting Half-Space—An Approximate Representation,” Geophysics 44, no. 10 (October 1979): 1700–1705. 3. Constable SC, Orange AS, Hoversten GM and Morrison HF: “Marine Magnetotellurics for Petroleum Exploration, Part I: A Sea-Floor Equipment System,” Geophysics 63, no. 3 (May–June 1998): 816–825. Spring 2009 Sounding S049 Depth, m TDEM Resistivity ohm.m 100 1 Well SWS17 1 Log Resistivity ohm.m 100 Sounding S071 1 Well SWS15 Log Resistivity TDEM Resistivity ohm.m 100 ohm.m 100 1 Sounding S013 1 Well SWS16 TDEM Resistivity Log Resistivity ohm.m 100 1 ohm.m 100 Depth, m –280 –280 –270 –270 –260 –260 –250 –250 Water table –240 Water table Water table –240 Bottom of aquifer –230 –230 Bottom of aquifer –220 –220 –210 –210 Bottom of aquifer –200 –200 –190 –190 > Comparison of TDEM soundings with resistivity profiles from logged wells. The TDEM resistivity measurements from soundings correlate closely with the resistivity logs from adjacent wells. The soundings at S049 and S013 show reasonable correlation for the contact between the aquifer and the clay-rich unit below (violet), which is true for most other soundings in the eastern part of the investigated area. The contrast is not as clear at S071, where the more resistive lower unit does not provide a sufficient contrast for the TDEM measurement. This trend follows for most of the soundings in the western part of the survey. The top of the water table (blue dashed line) was determined from a map of the water depth by interpolating between wells in the area. 4. For more on water storage: Black B, Dawoud M, Herrmann R, Largeau D, Maliva R and Will B: “Managing a Precious Resource,” Oilfield Review 20, no. 2 (Summer 2008): 18–33. 5. A British imperial gallon is equivalent to 1 galUK. 6. An Occam inversion is a smooth inversion that does not predefine the number of layers. 23 2 km 80 m 2 km N > Resistivity discontinuity in the bottom of an aquifer. The discontinuity is a band with 15- to 20-ohm.m resistivity (yellow and green), which contrasts with the 1- to 10-ohm.m resistivity (blue and violet) elsewhere in the resistivity cube. The resistivity of the clay at the bottom of the aquifer (violet) to the east of the discontinuity is lower than that to the west. This discontinuity roughly aligns with a thrust fault (tan) that was identified at about 3,000 m [9,800 ft] by a seismic interpretation. Outcrop studies performed to the south of the survey area (not shown) support the surface expression of the fault being slightly west of the deeper seismic interpretation, and those observations are consistent with the location of the near-surface resistivity discontinuity. The syncline axis (blue) from the seismic interpretation, another thrust fault (purple) and some wells are also shown. (Seismic lines adapted from Woodward and Al-Jeelani, reference 7.) LAND EM_FIGURE 24 The TDEM data clearly show a discontinuity in the resistivity distribution in the clay unit (left). There is also a difference in resistivity between the eastern and western compartments; the western part exhibits significantly greater resistivity at a given depth. The anomaly aligns with a seismically mapped thrust fault.7 The seismic interpretation was based on a survey acquired in the early 1980s and reprocessed in 1992 to highlight the shallow structures. The shallow units above the anomaly are expected to show some structural complexity; they exhibit rapid variations in saturated thickness and possibly no saturated thickness in some areas. On the eastern side of the discontinuity, the horizontal extent of saturated thickness is suitable for an ASR unit. The western side of the discontinuity shows some potential, but has a larger risk: The interpretation in that area has a greater uncertainty because of the poor resistivity contrast between the saturated sands and the underlying clay. The discontinuity in the clay formation should not be seen as a complete hydraulic barrier in the shallower aquifer layer. Paleochannels or tear faults—those striking perpendicular to the overthrust fault—are expected to provide preferential flow paths from east to west across the line of the discontinuity. This TDEM study suggests that around 4 billion imperial gallons [18 million m3, 4.8 billion galUS] of water can be stored at this site, giving it the capacity for daily production of more than 20 million imperial gallons [91,000 m3, 24 million galUS] for 200 continuous days. Mapping the Dunes Within the same regional setting as the waterstorage site, a 2D seismic survey was conducted for Dubai Petroleum Establishment (DPE). The same clay layer forming the base of the storagesite aquifer is present in this area; it provides a marker for the base of the weathered surface layer. The depth of the clay varies across the survey area, and lines of dunes add local variation 7. Woodward DG and Al-Jeelani AH: “Application of Reprocessed Seismic Sections from Petroleum Exploration Surveys for Groundwater Studies, Eastern Abu Dhabi, UAE,” paper SPE 25538, presented at the SPE Middle East Oil Show, Bahrain, April 3–6, 1993. 8. An uphole is a shallow well used to determine nearsurface velocities for a seismic survey. 9. Colombo D, Cogan M, Hallinan S, Mantovani M, Vergilio M and Soyer W: “Near-Surface P-Velocity Modelling by Integrated Seismic, EM, and Gravity Data: Examples from the Middle East,” First Break 26 (October 2008): 91–102. > Desert terrain at the ASR site. The blue cable is part of a TDEM sounding loop. The building houses a submersible pump and water tank for the nearby wells (blue caps) of the ASR project. 24 Oilfield Review 240 North Resistivity, ohm.m South 150.0 200 92.8 UH-08 UH-09 Elevation, m 160 57.5 UH-06 35.6 22.0 120 13.6 8.4 80 5.2 40 3.2 2.0 0 4 5 6 7 8 Distance, km 9 10 11 > Soundings along a seismic line. Interpolation between sounding points yields a detailed 2D model of resistivity along a seismic line. The sharply layered model at each sounding point (filled black square) is shown as a narrow column (top). Uphole sites (UH-09, -06, -08) contain domains of constant velocity in the weathered layer (yellow stippling) of about 1,400 m/s [4,600 ft/s] and in the underlying clay and limestone (gray stippling) of greater than 2,000 m/s [6,560 ft/s]. The variation in properties both within the dunes and in the lower layer is evident along the entire seismic line (bottom). The higher resistivity of the bottom layer (south end of the seismic line) indicates a relatively clay-poor region. The survey comprised 505 sounding sites to the depth of the weathered layer. Sand dunes generally exhibit low seismic velocity, and defin- using square loops of 50 m [164 ft] on a side, ing the velocity variation and thickness of the except for a few sites that used 75-m [246-ft] surface layer is crucial for deriving a long-wave- square loops for deeper penetration.9 Spacing length static correction for the seismic data. between sounding points was generally about The seismic crew drilled several upholes to 1,000 m [3,280 ft]; GPS was used to position the log the velocity of the surface and underlying sites. The effective time for decay ranged from clay layer.8 The upholes typically were at seismic- 0.01 to 10 ms; the pulse repetition rate was 6.3 Hz. Given the subhorizontal nature of the zone line intersections and, for practical reasons, away from the higher dune crests. However, this of investigation and its shallow depth compared pattern often does not sample the near-surface with the TDEM station spacing, 1D resistivity variations found in sand dune areas, so finer sam- inversion modeling was selected for the analyTwo 1D resistivity inversion methods were pling was desired. DPE elected toLAND use aEM_FIGURE TDEM sis.MARGHAM resistivity survey to map the area because it applied. The first one incorporated about 15 laywould be more cost-effective than drilling more ers extending to a depth of about 200 m [650 ft]. upholes and would avoid additional drilling on Layer thickness increased logarithmically with depth. Resistivity was a free parameter, and the environmentally sensitive dunes. this inversion yielded a detailed, smooth variation of resistivity. Spring 2009 The detailed fit provided a starting point for the second inversion. Termed a layered fit, it used the minimum number of layers required to fit the data to less than 5% root-mean-square misfit. This was typically two to five layers. Analysts selected the starting definition of these layers from the detailed fit. The layered model generated stronger resistivity contrasts than the detailed one. Interpreters created a 2D model with grid blocks 200 m wide and 5 m [16 ft] deep along a seismic line. They used the sounding sites along the seismic line to evaluate the surface layering. Model resistivity values were obtained by interpolating between the 1D smooth inversions at those sounding sites (left). The result is a detailed description of the location of the bounding clay layer at the bottom of the lowvelocity zone. The resistivity data were not calibrated to seismic velocities. The seismic processing team used these maps during the estimation of the surface static corrections. The velocities for the surface zone were interpolated from velocity measurements taken at the upholes. The TDEM approach gave the seismic interpreters a geologically consistent way to remove the effects of the laterally varying sand velocities. The resistivity analysis also highlighted variations within and below the lowvelocity weathered layer. Sounding Deeper Both case studies used TDEM methods to examine near-surface features. However, because using inductive-source techniques is inefficient for defining deep targets, the method is not the exploration tool of choice for examining deeper structures. The industry is improving techniques for using the alternative, grounded-source method to inject current into the earth. The source for the grounded method must be able to inject a large current at a voltage that is sufficient to overcome contact resistance in areas where the soil is dry. This combination has been difficult to achieve. The method using direct injection of current is more sensitive to resistive targets, making it a more likely method than the inductive-loop option to provide a direct hydrocarbon indication. —MAA 25 A Plan for Success in Deep Water Deepwater oil and gas are conventional resources in an unconventional setting; operations are notable mainly for their high risk and high reward. Because of the scope and complexity of projects beyond the continental shelves, the difference between success and failure often hinges on good planning. Adwait Chawathe Umut Ozdogan Chevron Corporation Houston, Texas, USA Karen Sullivan Glaser Houston, Texas Younes Jalali Beijing, China Mark Riding Gatwick, England Oilfield Review Spring 2009: 21, no. 1. Copyright © 2009 Schlumberger. Petrel and SMC are marks of Schlumberger. For help in preparation of this article, thanks to Robert Clyde, Debra Grooms, Scott Scheid and Drew Wharton, Houston; Nils A. Solvik, Framo Engineering, Bergen, Norway; Merrick Walford, Pau, France; and Jeremy Walker, Rosharon, Texas. 1. Curole MA and Turley AJ Jr: “Mars Debottlenecking Project,” paper SPE 69199, presented at the SPE Annual Technical Conference and Exhibition, New Orleans, September 27–30, 1998. 2. Wetzel RJ Jr, Mathis S, Ratterman G and Cade R: “Completion Selection Methodology for Optimum Reservoir Performance and Project Economics in Deepwater Applications,” paper SPE 56716, presented at the SPE Annual Technical Conference and Exhibition, Houston, October 3–6, 1999. Amin A, Riding M, Shepler R, Smedstad E and Ratulowski J: “Subsea Development from Pore to Process,” Oilfield Review 17, no. 1 (Spring 2005): 4–17. In 2001, while constructing its massive deepwater Mars tension-leg platform, Shell concluded the plans for the facility required major adjustments. The changes were needed to take advantage of just-introduced advances in well completion technology that would boost production beyond the original design parameter of a maximum 1,750 m3/d [11,000 bbl/d] of oil per well. Fortunately, because the Mars team comprised experts from many project disciplines, it was aware of overall project parameters, and Shell was able to implement the necessary changes in the construction yard before the giant floating platform sailed off.1 This Shell experience clearly demonstrates the case for planning practices that consider the development as a whole—from subsurface modeling to completion strategies to first oil and beyond. By considering every aspect of development at the planning stage, operators are more likely to find they still have viable options before or during deployment and operational phases. Such flexibility may become critical as new information about a reservoir, the available Load and interpret LWD data during drilling. Plan a new well based on an up-to-date 3D reservoir model. Make live updates to the 3D model with the new data. Plan follow-up steps (completion, next well) with latest subsurface views at hand. > Iterative processes. Project plans are fine-tuned constantly as a field is developed. Beginning with a 3D reservoir model, drilling experts select drilling locations, target zones and trajectories. Model updates take place as wireline and LWD measurements are acquired, enabling changes that reflect the most recent information. This process repeats throughout the development drilling program. 26 Oilfield Review technology or any number of related parameters becomes apparent during project commissioning, drilling, completion or production. The penalty for inefficient or incomplete planning could be an inability to change designs or accept compromises in critical elements such as well location, completion type, well size or field configuration once work has begun. The result could be a less-than-optimal development, which almost always translates to negative outcomes such as reduced ultimate recovery, lower productivity rates and significantly higher capital and operating costs. Adoption of proper deepwater project planning practices will probably require more cultural change than technological innovation. This is because the upstream industry traditionally treats the various operations that make up field development as separate tasks performed in series by experts working independently. More importantly, oil industry operators, contractors and service companies have a long history of working from out-of-date plans or those too general to be of much use. This mode of operation forced them to deal with individual problems in a reactionary mode rather than planning in advance for potential problems and possible solutions. In deep water, where stakes are high and the time between concept and first oil can be as much as a decade, segregation of responsibility and use of static plans that cannot be adjusted to respond to changing circumstances are no longer options. It is, therefore, essential that experts of all disciplines take a longer, more integrated view. For example, it is usually desirable to begin a drilling program by considering the type of completion required to best exploit the reservoir. While this reservoir-driven approach is common, the ultimate goal of a thought-out plan is the profitability of the overall project. This being the case, the well’s productivity becomes only one factor in the selection of completion type.2 Other considerations include cost-sensitive factors and the risks associated with overall project expense, interventions, well longevity, sand production and flow assurance. In recent years, many drilling and completion engineers have made progress toward a more integrated approach. But deepwater project planning requires extension of that practice beyond well construction to connect the entire enterprise—from early exploration to final production—while using each step in between to refine the process. Spring 2009 O Therefore, a typical deepwater project plan not only includes each of the following elements, but also considers their influence on each other: • subsurface reservoir model • drainage strategy and bottomhole locations • field development plan • well design engineering and technology • intervention methodology • pipeline and platform design and installation. From a practical standpoint, planning begins at the exploratory stage. Once the reservoir has been characterized through seismic data interpretation, petrophysical information about the target formation is gathered during the drilling process using such tools as wireline logs, LWD operations and dynamic testing (previous page). The resulting combination of data about 27 > Deepwater drilling units. The complex, dynamically positioned units capable of drilling in extreme water depths are relatively rare. The cost to build and outfit them has been reported to be as high as US $750 million, and despite a recent spate of new construction, demand outstrips supply. The investment needed to drill in water depths greater than 1,800 m [6,000 ft] is reflected in the unit lease rate—often US $1 million per day. (Photograph courtesy of Transocean Ltd.) the reservoir matrix, fluid properties and producibility serves as a basis for the many decisions that will be made about the field throughout its life. One such technical decision is well trajectory within a reservoir. Since efficient reservoir drainage—using as few wells as possible to access and produce the maximum volume of oil and gas at the most advantageous rate—is key to profitability in deepwater project planning, well angle and reach are decided early in the process. However, in a holistic approach these calculations must include more than maximum reservoir exposure—the most common driver for the use of extended-reach wells. Completion designs for these deepwater wells must also consider the optimal flow rate for the long term, which requires a balance between maximizing ultimate recovery through prudent production practices and maximizing immediate returns through high flow rates. These decisions both drive and are driven by available drilling technologies and their associated completion configurations. Operators 3. Perforation density is the number of holes per linear foot of borehole, reported as shots per foot (spf). Perforation phasing refers to the angle at which the perforations are offset from the toolstring axis. Thus, in a 30° phasing, each perforation is separated by 30°. 4. Iledare OO: “Profitability of Deepwater Petroleum Leases: Empirical Evidence from the US Gulf of Mexico Offshore Region,” paper SPE 116602, presented at the SPE Annual Technical Conference and Exhibition, Denver, September 21–24, 2008. 5. Mouawad J and Fackler M: “Dearth of Ships Delays Drilling of Offshore Oil,” http://www.nytimes.com/ 2008/06/19/business/19drillship.html (accessed December 11, 2008). Ultradeep water depths are considered by industry to be those beyond about 1,800 m [6,000 ft]. 6. Lifting cost is the operator’s total financial outlay for bringing oil and gas to the surface and is generally calculated in US dollars per barrel of oil equivalent. 7. Cullick AS, Cude R and Tarman M: “Optimizing Field Development Concepts for Complex Offshore Production Systems,” paper SPE 108562, presented at the SPE Offshore Europe Oil & Gas Conference and Exhibition, Aberdeen, September 4–7, 2007. 8. For more on inversion: Barclay F, Bruun A, Rasmussen KB, Camara Alfaro J, Cooke A, Cooke D, Salter D, Godfrey R, Lowden D, McHugo S, Özdemir H, Pickering S, Gonzalez Pineda F, Herwanger J, Volterrani S, Murineddu A, Rasmussen A and Roberts R: “Seismic Inversion: Reading Between the Lines,” Oilfield Review 20, no. 1 (Spring 2008): 42–63. 28 may choose to develop their fields through a few extended-reach wells, numerous vertical wells, multilateral wells, intelligent wells or some combination of these and other scenarios. Throughout exploration, assessment and development drilling, virtually all development parameters—such as well location, completion type and flow rates—may be altered as the reservoir model is refined by information gathered from new wells. Real-time data and the actions taken in response to confirmation of, or changes to, assumptions about a reservoir are used throughout the life of the field. Improved knowledge about rock stresses, for instance, impacts such vital details as perforation density and phasing and choice of sand control system.3 Updated models of porosity, permeability and fluid characteristics do not just shape the drilling and completion program, they are also fundamental inputs to key decisions about flow assurance and facility design. Modeling fluid parameters over the life of a deepwater project is itself a far more complex undertaking than for onshore or shallow-water fields. In deep water, economics dictate that multiple reservoirs—often with different and changing characteristics—share facilities, pipelines and other infrastructure. Because of the complexity-driven risk and large reserves potential involved, deepwater developments are economically more sensitive than most other E&P endeavors. According to an analysis of Gulf of Mexico lease data conducted by the US Minerals Management Service, both risk and reward rise significantly with increased water depth.4 Given that relationship, it is clear that operations planned in waters deeper than 3,050 m [10,000 ft] have increased the stakes to such a level that even seemingly minor missteps may conspire to quickly overwhelm project economics. Current extreme costs in the deepwater play are rooted in two major outlays: the costs of facilities, pipelines and other infrastructure and high rental rates—dayrates—that contractors must charge to make a reasonable return on their investment in rigs. For a rig capable of operating in ultradeep water, that investment is about US $500 million in construction costs alone (above left).5 As a consequence, the spread rate—dayrate plus all other required equipment and services for any given operation—for such a drilling vessel is approximately US $1 million per day, or nearly US $42,000 per hour. In deep water, well construction typically accounts for about 50 to 60% of the total lifting costs, split evenly between drilling and completion.6 A field infrastructure often requires capital expenditures of more than US $1 billion.7 Oilfield Review Surface Seismic Data Optimization • Seismic cube, line • Seismic gathers • Seismic velocities Reservoir Characterization Well Services • Fracture design • Microseismic Petrophysicists Geologists Fracture Characterization Drilling Measurements • Seismic while drilling • Sonic Inversion specialists Geophysicists Rock physicists Reservoir engineers Logs • Dipole sonic • Density • Geology, reservoir • Borehole seismic Reservoir Operations Pore-Pressure Prediction Geohazards • Field development planning • Reservoir modeling • Mechanical earth modeling • Wellbore stability and planning • Production enhancement > Integrated reservoir characterization workflow. Schlumberger uses a multidisciplinary team with expertise in petrophysics, geology, geophysics, inversion and rock physics to evaluate new reservoirs. Through attribute analysis and inversion of seismic traces and log data, the team derives the acoustic and elastic parameters required to discriminate hydrocarbons, estimate rock properties and characterize fracture systems. Collaboration with geomechanics and field development planning experts results in a predrill seismic assessment of drilling hazards and computation of a pore-pressure cube by high-grading the surface seismic velocities. As well data are collected by logs, drilling measurements and well services, they are used to refine initial reservoir characteristics. While these absolute costs are significant, well construction costs typically include 24 to 27% nonproductive time (NPT)—a time loss that is aggravated by working in reactive mode. Subsea architecture and installation of production facilities routinely incur 30 to 35% NPT. It is clear, given the investments involved, that these percentages represent a significant amount of money and underscore why minimizing NPT is a key goal of operators. This article looks at the many obvious and notso-obvious parameters considered in deepwater project planning. In some instances, the project is the entire venture from seismic survey to abandonment. In others, the project is more specific—testing, cementing or some other major component of a larger operation. Deepwater case histories from the Gulf of Mexico, West Africa and the North Sea demonstrate why and how deepwater operators and service providers must take a long-term integrated view in this challenging environment. Spring 2009 From the Bottom Up Reservoir drainage strategy essentially drives deepwater projects. Engineers must have a thorough understanding of the reservoir before they can optimize well locations and make informed decisions about wellbore size, sand control, artificial lift, perforation and all other facets of a drilling, completion and production program. They must also make decisions as to wellhead type, pipeline and manifold configuration and the type of host platform to be used. As with all modeling systems, a wrong first step in deepwater project planning endangers all decisions that follow. In the case of deepwater field developments, the earliest planning steps occur at the seismic-to-reservoir simulation stage. Good reservoir simulators have been available for more than 20 years, but for much of that time, preparing input and analyzing the results of reservoir simulation were difficult tasks. A lack of integration between pre- and postprocessing tools and the need for many manual, time-consuming data transfers and dataformatting steps frequently caused operators to avoid what was often cumbersome simulation work even while making critical business decisions about their developments. Today’s software overcomes this hurdle to best practices by clarifying the intersections of seismic data and reservoir modeling. Geoscientists now interpret and quantify reservoir properties using processes that integrate seismic data with all available petrophysical data through inversion and reservoir modeling. An important prerequisite to this process is the conditioning of both seismic and reservoir data for inversion and integration into a reservoir characterization workflow. Well logs and vertical seismic profiles provide calibrated earth properties that enable well-tie processing and testing of inversion models.8 The objective of this exercise is to characterize the reservoir and obtain estimates of reservoir lithology and fluid distributions by estimating rock properties such as porosity, sand and shale volumes, density and water saturation (above). One operator drilled three Gulf of Mexico deepwater wells with mixed results 29 > Taking the proper angle. Seismic gathers that are flat out only to an angle of 30° (left ) are of poorer quality than the ones made after data optimization, which makes the gathers flat out to an angle of 54° (right ). Optimization resulted in a good density determination. before turning to Schlumberger to manage the risk and cost of drilling subsequent wells. While some wells were successful, others drilled in promising formations found only residual gas instead of commercially viable accumulations— a difficult distinction to make from seismic amplitude data alone. The solution was to calculate density through an amplitude variation with offset (AVO) process. To determine the percent gas in a reservoir, a geoscientist must first model the densities of both the rock and fluids. Doing so requires the ability to see the variation in the far angles of the seismic gathers. This information in turn allows interpreters to perform a three-term AVO inversion calculation that includes density. The outputs of the three-term inversion process are relative acoustic impedance, shear impedance and density volumes. Shared Earth Model Planning The lithology and fluid types, using properties derived from log measurements in nearby wells, along with varying percentages of gas are modeled to see the effects on the seismic AVO signature. The workflow process then compares the modeling results with the inverted seismic volumes—acoustic impedance, shear impedance, density volumes and Poisson’s ratio—and the results are the basis for generating lithofacies and fluid saturation prediction volumes. Using these probability volumes, with statistical uncertainties, geoscientists can give better predictions of reservoir quality and distribution. These calculations require data to a fairly high-angle range in the seismic gathers. In this instance, the Schlumberger team was able to extend the usable seismic angle from 30 to 54°, enabling accurate density determinations (above). Calibration of the seismically derived Execution, Real-Time Monitoring, Replanning Evaluation G&G, RE, PE Scope risk Offset wells Key historical information Detailed engineering Actual versus plan Drilling operations TD Final well report Lessons learned Replan > Refining the plan. A shared earth model is the basis for collaborative well planning. From this starting point, geoscientists (G&G), reservoir engineers (RE) and production engineers (PE) define subsurface targets. Using offset well correlations or simulations from reservoir analysis software, this model delivers profiles of pore pressure and rock fracture strength with depth. The drilling engineer uses these inputs plus well objectives for conceptual design or scoping. The output assists in detailed engineering decisions on rig selection, drilling engineering technical risk and probabilistic time and cost estimates. Operators undertake drilling operations and modify them based on actual drilling performance. As each well reaches total depth, the team incorporates lessons learned with other offset well data and uses the update to modify the conceptual model and plan the next well. This iteration process repeats throughout the development-well drilling phase. 30 reservoir properties of porosity and hydrocarbon saturation with measurements from an existing wellbore validated the seismic facies predictions. With this otherwise unattainable piece of information, the operator was able to determine, without first drilling the well, that the prospect would not produce. Further, the company is using the seismically derived reservoir measurements to evaluate other prospects in the field and to manage development by ranking drilling locations according to risk and the probability of commercial success. Once seismic data have been used to characterize the reservoir, reservoir modeling integrates geophysics, geology and reservoir engineering to build one earth model. Reservoir engineers use this model to predict drainage patterns and design injection strategies. Drilling and production engineers use it to plan well trajectories.9 The latest versions of these tools, such as Petrel seismic-to-simulation software, enable seismic- and geomechanics-centered disciplines to build shared earth models, rendering a more accurate subsurface picture than that created by one of them working independently. Changes in the seismic interpretation or geological model easily cascade through to the reservoir simulation model and back. The workflows upon which these software packages are constructed are increasing the role of seismic data in understanding dynamic reservoirs (below left). Surface Work The better the understanding of the target reservoir, the less potential there is for surprises such as noncommercial volumes of hydrocarbons, flow assurance issues, early water breakthrough, sand production and changes to fluid makeup. Similarly, a well-planned overall field development strategy—completion configurations, well locations, processing facility types and sizes, and intervention decisions—is key to efficient ultimate oil and gas recovery. This is because the consequences of poor planning are often not felt until near the end of the field’s projected life span. Significant revenue is lost when fields are abandoned prematurely because the costs of remediation or operating expenses are greater than the value of the remaining reserves. Avoiding these risks requires bypassing the pitfalls created by strict separation of engineering disciplines. Traditionally, reservoir engineers have focused on well count, well placement and recovery mechanisms; production Oilfield Review and completion engineers on well design; and facilities engineers on the subsea layout, facility size and topsides design. These seemingly disparate groups must instead be convinced to perform their tasks independently yet understand the connection imposed by production and therefore the economics of the project. In turn, these economics are dependent on the physical limitations of the overall system. To function properly, each discipline must be aware of how it impacts the work of others; each member or team involved in a development must work from a common forecasting system. One such system that comprises dynamically linked models of the field’s subsystems— reservoir, well and facility—is called an integrated production model or integrated asset model (IAM).10 During development planning stages and before operations are undertaken, asset teams use these integrated models to analyze the interaction of proposed subsystems within a project. IAMs represent a break with traditional field development practices that are more likely to be centered on capital expenditure and focused on implementing modifications that drive down costs. A common pitfall of the traditional approach is a failure to properly quantify the effects of changes on system deliverability that in turn may ultimately lead to suboptimal designs. In contrast, an IAM uses a reservoir simulation model to calculate fluid movement and pressure distribution. Then, at the subsurface coupling point—the well locations in the reservoir model—these factors are put into the well model, which establishes conditions at the sandface. The sandface condition is used as a boundary to compute the fluid rates or pressures at the surface coupling point—the wellhead—where the well model is linked to the surface facility.11 The interaction of well–surface boundary conditions makes it possible to calculate the backpressure of the production system for each well. This is then conveyed back through the system to the reservoir. The process iterates to balance the full network. The result is stabilized solutions for fluid flow from the reservoirs into 9. Hopkins C: “Go Beyond Reservoir Visualization,” E&P 80, no. 9 (September 2007): 13–17. 10. For more on integrated asset modeling: Bouleau C, Gehin H, Gutierrez H, Landgren K, Miller G, Peterson R, Sperandio U, Traboulay I and Bravo da Silva L: “The Big Picture: Integrated Asset Management,” Oilfield Review 19, no. 4 (Winter 2007/2008): 34–48. Spring 2009 Modeling Framing Production engineer Flow assurance engineer Static QC Facilities engineer Decision analyst Reservoir Middle engineer management Simulation modeler IPM Initialization Dynamic QC Material balance Simulation > Integrated production model (IPM). Chevron engineers planning the company’s ultradeepwater Jack field used a five-step workflow to build an integrated production model for field development. In the framing step, the entire team describes the project in terms of objectives, time frames, givens, assumptions and deliverables. Once all parties were aware of these inputs, subsurface, wellbore and network models were created during the modeling step. A static QC was conducted by comparing the reservoir-to-separator model inputs with available data from logs, cores, fluid samples and other measurements. The wells in the subsurface (material balance or simulation) model were linked to their pairs in the surface facility model during intialization. The linked system was then run to check if the model converged to a solution. Following this initialization, the whole system was run for the entire prediction period. At the end of the period dynamic pressure and temperature responses at separator, manifold, wellhead and bottomhole were plotted during the dynamic QC step to help the team understand the operating pressures and temperatures of seafloor boosters, wellheads and bottom hole. Field schedules and constraints including water-handling capacity, well maximum drawdown and minimum bottomhole pressure were then checked to see if they were honored by the model. the well and from the well into the surface system and then to sales points. In this way, the IAM technique considers the response of the surface system in fluid-flow calculations.12 Chevron engineers used integrated production management as a forecasting tool to couple models of the subsurface with a surface network via a wellbore model at their deepwater Gulf of Mexico Jack field. A steady-state model calculated temperature and pressure changes within the wellbore. The surface-network model included the subsea and surface elements such as manifolds, seafloor pumps, wellheads, risers, flowlines and separators. The surface and subsurface models were linked at a bottomhole node.13 The Chevron model was constructed using a five-step workflow (above). Those steps included • defining the problem in terms of objectives, time frames, givens, assumptions and deliverables • modeling • quality-checking of reservoir-to-separator model input data against available data under static conditions • linking the surface and subsurface models • quality-checking the full system for the entire prediction period under dynamic conditions. Once the workflow steps were completed, the integrated project model determined that seafloor boosting, coupled with downhole artificial lift, would best exploit reservoir deliverability. A Chow CV, Arnondin MC, Wolcott KD and Ballard ND: “Managing Risks Using Integrated Production Models: Applications,” Journal of Petroleum Technology 52, no. 4 (April 2000): 94–98. 11. Tesaker Ø, Øverland AM, Arnesen D, Zangl G, AlKinani A, Torrens R, Bailey W, Couët B, Pecher R and Rodriquez N: “Breaking the Barriers—The Integrated Asset Model,” paper SPE 112223, presented at the SPE Intelligent Energy Conference and Exhibition, Amsterdam, February 25–27, 2008. 12. Tesaker et al, reference 11. 13. Ozdogan U, Keating JF, Knobles M, Chawathe A and Seren D: “Recent Advances and Practical Applications of Integrated Production Modeling at Jack Asset in Deepwater Gulf of Mexico,” paper SPE 113904, presented at the SPE Europec/EAGE Annual Conference and Exhibition, Rome, June 9–12, 2008. 31 second study using experimental design then allowed the Chevron team to identify the key parameters of the artificial lift system.14 The operator considered time of installation, seafloor boosting inlet pressure, ESP horsepower and setting depth. The company concluded that the ESP horsepower was the most significant design parameter across various recovery mechanisms. The operator then switched from an integrated to a modular approach—using the wellbore model alone—that would support design of the Jack field water-injection facility and calculate the recovery trade-offs for various topsides pressures and injection rates. Topsides pressures were converted into bottomhole pressures (BHPs) using the wellbore model. From the resulting recovery contouring study the team concluded that maximum recovery would require a range of BHPs and injection rates. A BHP of 21,000 psi [144.7 MPa] was infeasible because the proposed high injection rates and topsides pressure requirements limited available pressure. Modeling, on the other hand, showed that lowering the injection rates would not result in significant production loss given the field recovery response.15 Further developments to the integrated model were used to optimize the number of flowlines, number of seafloor boosting pumps and platform location. Driven to Succeed Deposited on a seafloor of steep slopes, gullies and canyons, hydrocarbon-bearing formations beyond the continental shelves may be quite unlike those found in shallow water. So it follows that project plans are driven by considerations specific to deep water. Because they are submersed in nearfreezing water, subsea flowlines, mudline wellheads and manifolds pose flow assurance concerns. As a consequence, subsea surveillance instrumentation is installed throughout critical points along the well-to-surface flow path. The surveillance data are fed into fluid models that enable engineers to take preemptive steps to prevent hydrate or paraffin blockages (next page). Because of the ability of hydrate, paraffin and asphaltine deposition to impact project economics, flow assurance is a primary consideration in many deepwater project plans. For example, given the wide separation of the five fields and comparatively small reserves base that make up BP’s Western Area Development (WAD) in Angola’s deepwater Block 18, project planners paid special attention to flow assurance issues and system deliverability. 32 At the planning stage, engineers applied a numerical method coupled with engineering software to calculate multiphase thermalhydraulic behavior in an IAM.16 The goal was to avoid potential flow disruptions caused by the solidification of gas hydrates in the flowlines traversing the cold seafloor. In performing the thermal analyses of the subsea systems at the WAD, the operator considered conventional wet insulation, pipe-inpipe systems and flexible pipelines to determine the time required for the coldest location in the production flowline system to fall to temperatures at which hydrates form. Known as cooldown time, this parameter indicates how long the operator has before hydrate-prevention measures must be taken following an unplanned shut-in. Analysts also used IAM to perform deliverability calculations for numerous field architectures and to investigate the effects of tubing and pipeline sizes, looping pipelines and subsea multiphase boosting. Their findings enabled BP to combine representative production profiles with capital and operating expense models to derive a full economic evaluation and screening of the company’s options. Evolution or Revolution Industry response to the unique challenges of deep water has done more than spawn procedural changes; it has also generated a nearly overwhelming burst of innovative drilling and production hardware in a relatively short period of time. As demonstrated by the Shell Mars platform experience, this onslaught of new technology has at times threatened to outpace engineers’ efforts to keep abreast of it. When coupled with the economic requirement that deepwater fields be developed with as few wells as possible, this flood of new tools and procedures has made it particularly difficult for completion engineers to be certain of delivering optimal solutions throughout the project. The effort to use the most-effective equipment is also thwarted by the time elapsing between project commissioning and well completion installation—often more than five years—during which tool design and availability may change radically. To deal with these issues, experts have designed modeling techniques to facilitate quantitative analysis of wellbore-reservoir interactions. These methods allow engineers to plan individual wells based on surface and subsurface characteristics as well as the current state of technology. All this is done while taking into account the constraints and characteristics imposed on and by the associated disciplines of geology, drilling, reservoir evaluation and production operations. One such method treats well design in much the same way the processing industry handles engineering-design problems. Process plants are designed with feed and effluent flow streams, using a flow diagram to capture the process. This diagram then becomes the basis for detailed design, component specifications and other considerations.17 Using this method, engineers divide the well design into conceptual and detail phases with iterations between the two. The conceptual design phase includes simple diagrams that highlight the impact of critical surface and subsurface attributes on well architecture. These attributes are examined by the interdisciplinary team so that alternatives may be considered for the key components of well design—well trajectory, formation completion and wellbore components. Choice of well trajectory is a function of local geology, reservoir properties and drilling capabilities. Formation completion refers to the interface between the wellbore and the reservoir; its configuration is determined by such factors as rock lithology, mechanical properties, grainsize distribution and operational constraints in the process. Wellbore components are important elements in the architecture of all wells. But they are especially critical in deep water, where modification after the fact can be costly and technically challenging and where conventional technologies are sometimes insufficient. One way to choose the appropriate technology for a particular completion is to sort available hardware options into four categories by function—packers, valves, pumps and sensors—and then use simple block diagrams to discern the optimal type of each. 14. Experimental design is a branch of statistics that outlines the way in which experiments should be carried out so that statistical information can be maximized with the minimum number of trials. 15. Ozdogan et al, reference 13. 16. Watson MJ, Hawkes NJ, Pickering PF, Elliot J and Studd LW: “Integrated Flow Assurance Modeling of Angola Block 18 Western Area Development,” paper SPE 101826, presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, September 24–27, 2006. 17. Sinha S, Yan M and Jalali Y: “Completion Architecture: A Methodology for Integrated Well Planning,” paper SPE 85315, presented at the SPE/IADC Middle East Drilling Technology Conference and Exhibition, Abu Dhabi, UAE, October 20–22, 2003. Oilfield Review Sensor Systems Acquisition Systems Fluid-Property Models Process Models Operations Changing parameters Facilities simulator Multiphase flowmeters Dynamic dataacquisition system Distributed temperature sensors Thermodynamic models Pressure and temperature gauges Multiphaseflow models Electrical submersible pump monitors Static-data storage system Flowline simulator Monitoring Wellbore simulator Optimization Deposition models Model conditioning > Prevention not remediation. Flow assurance strategies are an integral part of production operations in deep water. Surveillance instrumentation such as multiphase meters, distributed temperature sensors, pressure and temperature gauges, an SMC subsea surveillance system and ESP monitors installed in the flow path (bottom) delivers data in real or near-real time. This data stream feeds predictive fluid-property models for solids deposition, corrosion, rheology and thermodynamics. The planning team uses these models to create process models that include facilities, flowline and wellbore simulators. Over time, continuous monitoring of changing parameters closes the loop and includes fluids data, flow models and real-time measurements (top). This loop drives the optimization of mitigation strategies and should be included in the production system design as early as possible. In this way, overtreating by a chemicals-delivery system set to worst-case scenarios can be minimized. Monitoring and modeling also provide the information for decisions on the most appropriate proactive, preventive treatments or remedial techniques—thermal, chemical or mechanical—to be put into place to prevent plugging and other impediments to flow, thus avoiding expensive rig-based interventions. Spring 2009 33 Reservoir Geology and Geometry Reservoir Property Yes Surface setting: subsea Large Yes Property: kv /kh > 0.25 Geometry: flat areal structure No Small Horizontal, highly deviated wells No Poor mobility Vertical, low-deviation wells Vertical, low-deviation wells Vertical, low-deviation wells > Well completion technology. In complex deepwater developments, where most wells are subsea completions, each well is a critical contributor to overall production. In the simplified decision tree shown, basic geological and reservoir attributes are used to determine, generally, the appropriate wellbore angle—low to vertical or high to horizontal. Knowledge of necessary well angle and reach informs operator decisions about requirements for drilling rig capabilities and directional drilling. In a small areal structure, low-angle or vertical wells are usually sufficient to drain the reservoir. This may allow the use of lessrobust and therefore less-costly equipment for development drilling. In larger structures, wellbore deviation severity is subject to secondary parameters of formation permeability and fluid mobility. The selection process begins with an evaluation of basic geological and reservoir attributes that constrain options for wellbore trajectories (above). Similar exercises can be used to decide the advisability and type of subsea wellbore configurations such as multilateral completions, stimulations, sand control and artificial lift. The resulting conceptual well design can be used for more detailed analysis that is then added to a project workflow. This workflow considers a quantitative assessment of well performance, preliminary economic analysis and detailed design. Planning for the Unexpected As subsea wellheads increasingly become the completion design of choice in deep water, operators strive to contain costs by minimizing interventions. The high cost and technical uncertainty associated with entering wellheads located on the ocean floor beneath thousands of feet of water were, in fact, early motivations for developing intelligent completions in the 1990s. Although remote downhole monitoring and operations, along with more-robust completions, have done much to reduce their frequency, rigbased deepwater well interventions are unlikely to be eliminated entirely (next page). 34 The links between intervention, production and costs have been demonstrated clearly and often. Frequent, immediate remedial action to repair failed components yields higher production rates but increases operating expense. A policy of fewer or delayed maintenance operations is less costly but results in lower production volumes and revenues. The challenge then is to develop a plan that strikes a balance of the two options, which is commonly done by dividing interventions into immediate, performance-based or campaign (numerous interventions performed within a field in sequence) repair strategies. To make such a policy workable through the life of the project requires including proactive maintenance during the front-end engineering and design (FEED) stage, rather than relegating it the traditional “as needed” role. Norske Shell used a modified version of the three-repair strategy to help mitigate intervention costs at its Ormen Lange project—Norway’s first deepwater subsea development 120 km [74 mi] northwest of Kristiansund. During the FEED stage planners instituted a simulation approach for estimating future intervention, maintenance and repair costs of various strategies. Risk expenditures included the direct cost of the intervention to repair a specific component plus the lost revenue incurred as a result of the failure. This method allowed the operator to assess the impact of different intervention strategies throughout the life of the project.18 The overall result of this type of modeling is to direct intervention, maintenance and repair efforts toward areas where they yield the highest value. It also focuses attention on critical equipment packages within the project that contribute the most to risk expenditures—the sum of the revenue lost to producing well failures and the cost of intervention. This information may be used to improve designs, increase equipment reliability and initiate smarter, lessexpensive ways to fix failed units and thus reduce risk expenditures. Finally, such a strategy often allows the operator to continually update forecasting information within the model as the project progresses into the operational phase. In the case of Ormen Lange, for example, the model was initially built on rough estimates of the 18. Eriksen R, Gustavsson F and Anthonsen H: “Developing an Intervention, Maintenance and Repair Strategy for Ormen Lange,” paper SPE 96751, presented at the SPE Offshore Europe Oil & Gas Conference and Exhibition, Aberdeen, September 6–9, 2005. 19. Eriksen et al, reference 18. 20. For more on the Jack field: Aghar H, Carie M, Elshahawi H, Gomez JR, Saeedi J, Young C, Pinguet B, Swainson K, Takla E and Theuveny B: “The Expanding Scope of Well Testing,” Oilfield Review 19, no. 1 (Spring 2007): 44–59. Oilfield Review Similarly, while a great deal of progress has been made in the past decade, the deepwater environment still holds technological challenges for the industry. Just as operators once purchased leases in water depths greater than 2,300 m [7,500 ft] knowing they could not economically produce them, today they are asking service companies and drilling contractors to customize tools to extend depth, pressure and temperature barriers to 3,050 m and beyond. For example, Schlumberger engineers were recently asked to develop a perforating system specifically for Chevron’s Jack field test well.20 The solution required re-engineering existing tools based on casing size and the unique bottomhole pressure conditions expected in this ultradeepwater well. In the end Schlumberger delivered a unique combination of tools that together operated as a 172-MPa [25,000-psi] perforating system. Such requests will continue to be made as operators bring their ultradeepwater prospects beyond the exploratory stage to testing and development. But it is important to consider that it took nine months for Schlumberger to develop, qualify, quality-test and deliver the necessary system working from a similar tool developed for Chevron’s deepwater Gulf of Mexico Tahiti field. Given the task, nine months was a remarkably short time in which to deliver. But it clearly demonstrates the necessity of maintaining a broad view if costly delays caused by technological shortcomings are to be avoided in the complex and unforgiving deepwater operating environment. Such foresight demands a culture of planning that simultaneously encompasses a long-term vision of integrated tasks and short-term detailoriented designs. —RvF > Light well interventions. Rigless subsea well interventions from multiservice vessels are considerably less expensive than those performed from deepwater drilling rigs. Light interventions commonly support remotely operated vehicles used to replace or reset controls or sensors on subsea trees, pipelines or manifolds. They may also facilitate rigless downhole intervention using wireline or coiled tubing, as shown here, to perform such functions as safety valve maintenance, perforation or coiled tubing wellbore cleanout. Currently, these types of interventions are restricted to water depths of about 500 m or about 1,500 ft. field layout. As the project moved forward, the Norske Shell team continually refined the model to assist in equipment configuration and to predict performance, which is used in economic evaluations.19 The Big Picture More than any previous shift in its environment, operating in deep water has forced the upstream industry to change how it conducts business. This cultural shift is due to the outsized rewards and Spring 2009 risks attached to deepwater operations, but perhaps even more so it is the consequence of the unprecedented length of time between the decision to exploit a prospect and first oil. It is impossible to predict oil and gas prices or the state of the global economy over such a time span. Operators must make critical investment decisions without the benefit of traditional economic fundamentals that come with more immediate returns on investment. 35 Evaluating Volcanic Reservoirs M.Y. Farooqui Gujarat State Petroleum Corporation (GSPC) Gandhinagar, Gujarat, India Huijun Hou Dhahran, Saudi Arabia Guoxin Li PetroChina Exploration and Production Company Limited Beijing, China Nigel Machin Saudi Aramco Dhahran, Saudi Arabia Tom Neville Cambridge, Massachusetts, USA Aditi Pal Jakarta, Indonesia Chandramani Shrivastva Mumbai, India Yuhua Wang Fengping Yang Changhai Yin Jie Zhao PetroChina Daqing Oilfield Company Daqing, China Xingwang Yang Tokyo, Japan Oilfield Review Spring 2009: 21, no. 1. Copyright © 2009 Schlumberger. DMR, ECS and FMI are marks of Schlumberger. For help in preparation of this article, thanks to Martin Isaacs, Sugar Land, Texas, USA; Shumao Jin, Brett Rimmer and Michael Yang, Beijing; Charles E. Jones, University of Pittsburgh, Pennsylvania, USA; Andreas Laake, Cairo; and Hetu C. Sheth, Indian Institute of Technology, Mumbai. 36 Hydrocarbons can be found in volcanic rock—sometimes in significant quantities. Petrophysical methods originally developed for sedimentary accumulations are being used to evaluate these unusual reservoirs. In the early days of petroleum exploration, the discovery of hydrocarbons in anything other than sedimentary rock was largely accidental, and such accumulations were considered flukes. Serendipity is still part of exploration, but geologists now know that the presence of oil and gas in such rock is certainly no coincidence. Igneous rock—created by the solidification of magma—hosts petroleum reservoirs in many major hydrocarbon provinces, sometimes predominating them. In general, igneous rocks have been ignored and even avoided by the E&P industry. They have been ignored because of a perceived lack of reservoir quality. However, there are many ways in which igneous rocks can develop porosity and permeability.1 Far from inconsequential, igneous activity can influence every aspect of a petroleum system, providing source rock, affecting fluid maturation and creating migration pathways, traps, reservoirs and seals.2 Igneous rocks have been avoided for other reasons. They tend to be extremely hard, although improvements in bit technology are helping drillers cope with these tough lithologies.3 Because they typically prevent deep penetration of seismic energy, igneous layers are considered an impediment to evaluation of underlying sediments as well. New seismic methods are advancing solutions to this problem, but with their strong refractive qualities, igneous reservoirs remain difficult to characterize.4 Once hydrocarbons are found in igneous reservoirs, assessing hydrocarbon volumes and productivity presents several challenges. Log interpretation in igneous reservoirs often requires adapting techniques designed for other environments. Logging tools and interpretation methods that succeed in sedimentary rock can give meaningful answers in igneous rock, but they often require artful application. Furthermore, because mineralogy varies greatly in these formations, methods that work in one volcanic province may fail in another. Usually, a combination of methods is required. This article describes the complexity of volcanic reservoirs and presents technologies that have proved successful in characterizing them. The discussion begins with a review of igneous rock types and follows with an examination of the effects of igneous processes on petroleum 1. Srugoa P and Rubinstein P: “Processes Controlling Porosity and Permeability in Volcanic Reservoirs from the Austral and Neuquén Basins, Argentina,” AAPG Bulletin 91, no. 1 (January 2007): 115–129. 2. Schutter SR: “Hydrocarbon Occurrence and Exploration in and Around Igneous Rocks,” in Petford N and McCaffrey KJW (eds): Hydrocarbons in Crystalline Rocks, Geological Society Special Publication 214. London: Geological Society (2003): 7–33. 3. Close F, Conroy D, Greig A, Morin A, Flint G and Seale R: “Successful Drilling of Basalt in a West of Shetland Deepwater Discovery,” paper SPE 96575, presented at the SPE Offshore Europe Oil and Gas Conference and Exhibition, Aberdeen, September 6–9, 2005. Salleh S and Eckstrom D: “Reducing Well Costs by Optimizing Drilling Including Hard/Abrasive Igneous Rock Section Offshore Vietnam,” paper SPE 62777, presented at the IADC/SPE Asia Pacific Drilling Technology Conference, Kuala Lumpur, September 11–13, 2000. 4. Hill D, Combee L and Bacon J: “Over/Under Acquisition and Data Processing: The Next Quantum Leap in Seismic Technology?” First Break 24, no. 6 (June 2006): 81–95. White RS, Smallwood JR, Fliedner MM, Boslaugh B, Maresh J and Fruehn J: “Imaging and Regional Distribution of Basalt Flows in the Faeroe-Shetland Basin,” Geophysical Prospecting 51, no. 3 (May 2003): 215–231. Oilfield Review Oilfield Review Winter 09 Volcanic Fig. Opener ORWINT09-VOL Fig. Opener Spring 2009 37 systems. Two field examples highlight formation evaluation in volcanic rocks. A case study from a gas-rich reservoir in China presents a technique that combines conventional logging measurements and image logs with neutron-capture spectroscopy and nuclear magnetic resonance. Plume Ash-cloud surge Pyroclastic flow Traps Eruption column Volcaniclastic rocks Laccolith exposed by erosion Dikes Volcano Granite wash Lava flow Dike Plutonic rock Laccolith Sill Country rock Pluton Basement > Emplacement of igneous rocks. Plutonic rocks, formed by cooling of magma within the Earth, display well-developed crystals with little porosity. Plutons and laccoliths—bulging igneous injections into sedimentary layers—are examples of plutonic rock. Volcanic rocks, formed when magma extrudes onto the surface and cools rapidly, show very fine crystalline or even glassy textures. Buildup of pressures within the Earth can cause explosive eruptions; these result in the accumulation of fragments of volcanic material in pyroclastic deposits. Rock containing clastic fragments of volcanic origin is termed volcaniclastic. Complex porosities and permeabilities can develop as a result of these different processes. Structures Textures Flow—Flows form when the fabric of lava aligns in parallel rows or ropy waves. Brecciated—Most angular particles exceeding 2 mm in diameter are volcanic breccia. Typically, particles form from the movement of partially solidified rock, not from the ejection of fragments. Pillow—Lava that erupts under water and quickly develops a cool skin around a molten core forms pillow structures, which are bulbous piles of rock. Pillow lava often incorporates seafloor sediments. Glassy—Lava that cools rapidly forms volcanic glass such as obsidian, tonalite and pitchstone, which differ mainly in their alkali feldspar content. Porphyry—One of the most common porphyritic Oilfield ReviewTuffaceous—Consolidated pyroclastic material less structures is phenocrysts, 1- to 2-mm [0.04- to 0.08-in.] crystals embedded in a fine-grained, often glassy matrix. 09 than 2 mm [about 0.08 in.] in diameter is tuff. Winter Andesite and basalt often have olivine and pyroxene tuff is ash. Both can be deposited far Volcanic Fig. 1Unconsolidated phenocrysts. from their source. A common epiclastic, or weathered ORWINT09-VOL Fig. 1reservoir rock is tuffaceous sand, in which volcanic, Pyroclast—Pyroclasts are sharp, chiseled rock reworked tuff accounts for less than half the volume of fragments created during a volcanic explosion. Glass rock. When tuff makes up more than half the rock, shards are often a key component. Sharp shards indicate the deposit is called sandy tuff. rapid burial or minimal postdepositional reworking. Vesicular—Gas expanding in cooling lava creates pores called vesicles. Often unconnected, they are the reason very porous volcanic rock, such as pumice, can > Structures and textures in volcanic rocks. float but has negligible permeability. Vesicles often fill with secondary minerals, usually hydrated silicates Variations in structure and texture give rise to called zeolites. These filled vesicles, called amygdules, the wide range of porosity and permeability reduce intergranular porosity in the same manner as observed in crystalline and pyroclastic rock. clay in sandstone. 38 An example from India demonstrates the importance of incorporating borehole resistivity images in the evaluation of oil-bearing volcanic rock. About Igneous Rocks Igneous rock is formed through the solidification of magma—a mixture of water, dissolved gases and molten to partially molten rock. Igneous rocks vary from one reservoir to another because their constituents have diverse chemistries, originating from magma that mixes material from the Earth’s mantle, crust and surface— typically oxides of silicon, iron, magnesium, sodium, calcium and potassium. They also have diverse structures and textures—leading to complex porosities and permeabilities—depending on how they were emplaced. Emplacement mechanisms include sudden explosive eruptions, syrupy viscous flows and slow, deep subsurface intrusions. Subsequent weathering and fracturing can further complicate rock properties. Igneous rocks form under a wide range of conditions, and therefore display a variety of properties (left). Molten rock that cools deep beneath the surface forms intrusive, or plutonic, rocks. Slow cooling of deep magmas forms large crystals, resulting in coarse-grained rock. These formations typically have low intergranular porosity and insignificant permeability, making them of little interest to the oil industry. The one exception is fractured granites, which can produce hydrocarbons.5 Magmas that approach the surface tend to cool more rapidly. This allows less time for the formation of crystals, which therefore tend to be smaller, resulting in finegrained crystalline rock. Extrusive, or volcanic, rocks are created when magma erupts through the Earth’s surface. Magma may extrude in flows of molten lava that, when cooled, form fine- to very finegrained crystalline volcanic rock. Sometimes, cooling occurs so quickly that crystals cannot form, resulting in volcanic glass, such as obsidian. When magmas contain large amounts of water and dissolved gases, buildup of excessive pressure under the ground can cause explosive eruptions of volcanic material. Ejected fragments, or pyroclasts, can range in size from fine volcanic ash to “bombs” tens of centimeters in diameter. Once they have been ejected, individual fragments accumulate to form pyroclastic rock. Lava flows and pyroclastic deposits may be a few centimeters to a few hundred meters thick, covering thousands of square kilometers. These deposits can have sufficient porosity and permeability to make them viable hydrocarbon reservoirs. Oilfield Review Spring 2009 Coarse Grained Peridotite Basalt Andesite Dacite Rhyolite Gabbro Diorite Granodiorite Granite 100 Calcium-rich plagioclase feldspars 80 Quartz Potassium feldspar 60 Sodium-rich plagioclase feldspars Olivine 40 Pyroxene 20 Biotite Amphibole 0 45% Increasing silica content 75% Increasing calcium, magnesium and iron content Increasing potassium, sodium and aluminum content 1,200°C [2,200°F] Increasing temperature of crystallization 700°C [1,300°F] > Classifying igneous rocks by mineral composition. Fine-grained and coarse-grained rocks of similar composition have different names. For example, a magma containing quartz, potassium feldspar, sodium-rich plagioclase and biotite may cool slowly and form coarse-grained granite. If the same magma is extruded, it will form fine-grained rhyolite. Olivine-rich magmas do not commonly extrude, but crystallize at depth, and so form only coarse-grained rocks. Clast or Crystal Size, mm Sedimentary Clasts Boulders 256 Cobbles 16 Pyroclastic Fragments Crystalline Rocks: Igneous, Metamorphic or Sedimentary Blocks and bombs Very coarse grained Very coarse crystalline Gravel 64 Pebbles Lapilli Coarse grained 2 1 0.5 0.25 0.125 Coarse crystalline Granules Very coarse sand Medium grained Coarse sand Medium sand Fine sand Very fine sand Oilfield Review Coarse ash Winter 09grains Volcanic Fig. 2 ORWINT09-VOL Fig. 2 Sand 4 Medium crystalline Fine grained Fine crystalline 0.032 Silt 0.004 Clay Very fine grained Mud 5. For example, recoverable oil reserves in the fractured granite of the Cuu Long basin offshore Vietnam are estimated at 2 billion bbl [320 million m3] or more. For more: Du Hung N and Van Le H: “Petroleum Geology of Cuu Long Basin—Offshore Vietnam,” Search and Discovery Article #10062, http://www.searchanddiscovery. net/documents/2004/hung/images/hung.pdf (accessed April 6, 2009). The giant Suban gas field in southern Sumatra contains estimated reserves of 5 Tcf [140 billion m3] in fractured granites. For more: Koning T: “Oil and Gas Production from Basement Reservoirs: Examples from Indonesia, USA and Venezuela,” in Petford N and McCaffrey KJW (eds): Hydrocarbons in Crystalline Rocks, Geological Society Special Publication 214. London: Geological Society (2003): 83–92. Landes KK, Amoruso JJ, Charlesworth LJ Jr, Heany F and Lesperance PJ: “Petroleum Resources in Basement Rocks,” Bulletin of the AAPG 44, no. 10 (October 1960): 1682–1691. 6. An acidic rock contains proportionately more nonmetallic oxides than a basic rock and forms an acid when dissolved in water. A basic rock contains proportionately more metallic oxides than an acidic rock and forms a base when dissolved in water. 7. The term “mafic” is derived from the words magnesium and ferric, whereas “felsic” is a combination of feldspar and silica. Hyndman DW: Petrology of Igneous and Metamorphic Rocks, 2nd ed. New York City: McGraw-Hill Higher Education, 1985. Fine Grained Mineral composition, volume percent The different modes of formation of igneous rocks—cooling of lavas, either under the ground or at the surface, and agglomeration of fragments ejected during explosive eruptions—allow a subdivision of igneous rocks into two groups: crystalline igneous rocks and fragmental igneous, or pyroclastic, rocks. A simple and common compositional classification of crystalline igneous rocks is based on silica [SiO2] weight percentage. Rocks low in SiO2 (less than 52%) are classed as basic, rocks high in SiO2 (more than 66%) are acidic and those with SiO2 between 52 and 66% are intermediate.6 A parallel classification system groups rocks by weight percent of dark-colored minerals. Rocks rich (more than 70%) in dark minerals, such as olivine and pyroxene, are mafic; those containing few dark minerals (less than 40%), and therefore more light minerals, such as quartz and feldspar, are silicic, sometimes called felsic.7 Mafic rocks, such as basalt, tend to be basic; silicic rocks, such as granite, tend to be acidic. A different classification encompasses emplacement mechanism, crystal size and mineralogy, dividing crystalline volcanic rocks into four main types (above right). The trend from basalt to andesite, dacite and rhyolite forms a continuum of mineralogy. Pyroclastic rocks, on the other hand, are typically classified by grain size, as are clastic sedimentary rocks. Relative proportions of three grain-size classes—blocks and bombs, lapilli and ash—are used to classify a pyroclastic rock (right). Pyroclastic and crystalline rock types exhibit differences in texture and structure that lead to differences in porosity and permeability (previous page, bottom). Fine ash grains Very fine crystalline Cryptocrystalline > Classifying pyroclastic rocks by grain size. Pyroclastic rocks are identified based on grain size, in a similar fashion to clastic sedimentary rocks. 39 ARGENTINA Chaitén CHILE Plume Ash cover ATLANTIC OCEAN 0 km 100 0 miles 100 > Image of the Chaitén volcano, southern Chile, from the NASA Terra satellite. The volcano, thought to be dormant before its May 2, 2008, eruption, sent a plume of ash and steam 10.7 to 16.8 km [35,000 to 55,000 ft] into the atmosphere. This image, acquired three days after the eruption, shows the plume extending eastward more than 1,000 km across Argentina and into the Atlantic Ocean. The volcanic plume (white) is distinguishable from the clouds (turquoise). The land surface is dusted with tan-gray ash. [From “Chile’s Chaiten Volcano Erupts,” http://earthobservatory.nasa.gov/IOTD/view.php?id=8725 (accessed April 6, 2009)]. Volumes of Volcanics Petrologists have calculated that the shallow part of the Earth’s crust contains a volume of volcanic rock—formed by the ejection of lava at the surface—of 3.4 to 9 x 109 km3, an order of magnitude greater than the volume of sedimentary rock. This estimate includes extrusions at seafloor rift zones, where oceanic plates are pulling apart and new crust is created by volcanic activity. The presence of volcanic rocks in hydrocarbon provinces is common because volcanic activity has taken place in or near many sedimentary basins at one time or another. Volcanism can also affect distant basins—large volcanoes can push pyroclastic flows up to 1,000 km [about 600 mi] from their origin and wind can carry ash thousands of kilometers (left). Consequently, blankets of ash and tuffs, or consolidated ash, may be found far from their source. Hydrocarbon-producing igneous rocks occur the world over (below). The earliest documented oil discovery in volcanic rock may be the Hara oil field of Japan, which began producing in 1900.8 The field produced oil from three tuffaceous layers. Other early production was recorded in Texas, in 1915, along a trend of seafloor volcanoes that erupted during deposition of the Austin Chalk.9 The buried volcanic formations produced 54 million bbl [8.6 million m3] of oil from 90 fields in more than 200 igneous bodies. Oilfield Review Winter 09 Volcanic Fig. 5 ORWINT09-VOL Fig. 5 Hydrocarbons associated with igneous rocks or igneous activity > Distribution of hydrocarbon-bearing igneous rocks. Gold dots represent locations of hydrocarbon seeps, shows and reservoirs in igneous rocks. (Adapted from Schutter, reference 36). 40 Oilfield Review Volcanic reservoirs may contain significant accumulations. As of 1996, cumulative produc tion from the volcanic tuff and associated layers of the Jatibarang field, West Java, was 1.2 billion bbl [190 million m3] of oil and 2.7 Tcf [76 billion m3] of gas. Speculated reserves are 4 billion bbl [635 million m3] of oil and 3 Tcf [85 billion m3] of gas.10 Reservoir analysis yields porosity values of 16 to 25% and permeability up to 10 darcies. In this reservoir, the volcanic rocks are also source rocks.11 Christmas Tree Laccolith Punched Laccolith Petroleum Systems Volcanism can affect all aspects of a petroleum system, producing distinctive source rocks, accelerating fluid maturation, facilitating fluid migration, and creating traps, reservoirs and seals. Source Rock—Although most hydrocarbons found in volcanic rocks come from sedimen tary source rock, some volcanic rocks are also source rocks. Vegetation entrained in ash flows may contain enough water to protect it from the heat of emplacement. Subaerial volcanism may create lakes and swamps with kerogen-rich sediments, and the volcanically warmed water in these basins encourages nutrient growth, further enhancing the production of organic material. Maturation—By adding heat, igneous bod ies can accelerate hydrocarbon maturation. Large intrusive bodies, such as thick dikes and sills, cool slowly and may affect great volumes of surrounding rock, causing overmaturation.12 Volcanic flows cool relatively quickly, so they usu ally have less impact on maturation. The impact of igneous activity on fluid maturation can be assessed by petroleum systems modeling.13 In addition to direct heat, the circulation of hydrothermal fluids in the heated zone also may affect maturation. For example, scientists working in the Guaymas basin of the Gulf of California have reported that hydrothermal fluids heated to 400°C [752°F] are responsible Traps—Igneous intrusions into surrounding for alteration of organic matter and the creation of petroleum.14 The process is rapid, taking hundreds sedimentary layers, called country rock, often to thousands of years rather than the millions of result in closed structures within the intruded formations. The Omaha Dome field in the Illinois years typically needed to generate oil.15 Migration—There are several ways for hydro basin, USA, was formed by this type of trap. The carbons that originated elsewhere to become trapping structure is a Christmas tree laccolith produced by an ultramafic intrusion (above).16 trapped in volcanic rocks: •Hydrocarbons can pass vertically or later The field was discovered in 1940 and produced ally from sedimentary rocks into structurally about 6.5 million bbl [1 million m3] of oil from Oilfield Review higher volcanic rocks. sandstones that are in contact with the intrusion. Winter 09 •Compaction of sedimentary rocks can force Reservoirs—Igneous rocks share another Volcanic Fig.characteristic 7 hydrocarbons downward into volcanic rocks. with sedimentary reservoir rocks; ORWINT09-VOL Fig. 7 •Hydrothermal fluids are capable of dissolving they can have primary porosity and sometimes hydrocarbons and depositing them in igne develop secondary porosity. But unlike sedimen ous rocks. tary rocks, igneous rocks lose their porosity quite •If the vapor pressure in volcanic rocks becomes slowly with compaction. Primary porosity may low enough during cooling, hydrocarbons may be intergranular or vesicular—a type of poros be drawn into the pore spaces. ity resulting from the presence of vesicles, or gas 8.Mining in Japan, Past and Present. The Bureau of Mines, Department of Agriculture and Commerce of Japan, 1909. 9.Ewing TE and Caran SC: “Late Cretaceous Volcanism in South and Central Texas—Stratigraphic, Structural, and Seismic Models,” Transactions, Gulf Coast Association of Geological Societies 32 (1982): 137–145. 10.Kartanegara AL, Baik RN and Ibrahim MA: “Volcanics Oil Bearing in Indonesia,” AAPG Bulletin 80, no. 13 (1996): A73. 11.Bishop MG: “Petroleum Systems of the Northwest Java Province, Java and Offshore Southeast Sumatra, Indonesia,” USGS Open-File Report 99–50R (2000), http://pubs.usgs.gov/of/1999/ofr-99-0050/OF99-50R/ ardj_occr.html (accessed April 7, 2009). 12.Schutter, reference 2. 13.Yurewicz DA, Bohacs KM, Kendall J, Klimentidis RE, Kronmueller K, Meurer ME, Ryan TC and Yeakel JD: “Controls on Gas and Water Distribution, Mesaverde Basin-Centered Gas Play, Piceance Basin, Colorado,” in Cumella SP, Shanley KW and Camp WK (eds): Understanding, Exploring and Developing Tight-Gas Sands: 2005 Vail Hedberg Conference, AAPG Hedberg Series, no. 3 (2008): 105–136. 14.Simoneit BRT: “Organic Matter Alteration and Fluid Migration in Hydrothermal Systems,” in Parnell J (ed): Geofluids: Origin, Migration and Evolution of Fluids in Sedimentary Basins, Geological Society Special Publication 78. London: Geological Society (1994): 261–274. Spring 2009 > Traps caused by laccolith intrusion. The trap of the Omaha Dome field in Illinois was caused by a Christmas tree laccolith (left ) of mica-peridotite intruding into limestones and sandstones. Traps (green) can also be caused by punched laccoliths (right), which lift overlying layers along bounding faults. 15.Kvenvolden KA and Simoneit BRT: “Hydrothermally Derived Petroleum: Examples from Guaymas Basin, Gulf of California, and Escanaba Trough, Northeast Pacific Ocean,” AAPG Bulletin 74, no. 3 (March 1990): 223–237. 16.English RM and Grogan RM: “Omaha Pool and Mica-Peridotite Intrusives, Gallatin County, Illinois,” in Howell JV (ed): Structure of Typical American Oil Fields, Special Publication 14, vol. 3. Tulsa: American Association of Petroleum Geologists (1948): 189–212. 41 Fresh basalt Weathered basalt Nonbasalt rocks Fresh basalt Payun Payun Weathered basalt Basalt with sparse vegetation Nonvolcanic sediments 0 0 km 20 mi 20 Vegetation > Remote sensing in volcanic provinces. Satellite data from visible, near-infrared, infrared and thermal bands help geophysicists assess topography and ground surface character before planning seismic survey acquisition. In this example from Argentina, satellite data (bottom) from several spectral bands are combined and color-coded to distinguish different surface characteristics. Recently erupted basalt flows are highlighted as dark red in both satellite images. Acquisition crews use the information to determine whether the terrain is accessible to vibrator trucks and other equipment (top). The photograph of the survey vehicles shows the Payun volcano seen from the south. bubbles, in igneous rock. Porosities in vesicular surface mapping of elevated structures has basalts and andesites may reach 50%.17 Secondary revealed volcanic deposits. For example, in porosity is important for many volcanic reservoirs Japan, rhyolitic volcanic rocks containing large and is sometimes the only porosity present. It may hydrocarbon accumulations have been discovresult from hydrothermal alteration, fracturing ered by mapping structural highs.18 Another traand late-stage metamorphism—metamorphism ditional method, the recognition of hydrocarbon Oilfield Review during the late stages of igneous activity that seeps at the surface, is used to find deeper reserWinter 09 alters the minerals formed earlier. Sills and lac- voirs. Oil and gas sometimes rise to the surface Volcanic Fig. 8 coliths may become reservoirs, especially when alongFig. contacts between igneous and sedimentary ORWINT09-VOL 8 they intrude into source rocks. They may fracture rocks. Seeps in the Golden Lane area of eastern upon cooling, providing porosity, permeability Mexico have been associated with steeply dipand migration pathways. ping igneous rocks that have penetrated thick Seals—Igneous rocks can provide seals. After oil-rich carbonate layers.19 alteration to clay, extrusive layers may act as Advanced techniques are also used. Satellite tight seals. Impermeable intruded rocks, such as imagery has been applied to evaluate the basaltlaccoliths that form traps, also may seal hydro- covered Columbia basin in Washington and carbons in formations beneath them. Oregon, USA.20 Geochemical analysis of groundwater in the same region has detected significant Exploration in Volcanic Provinces levels of methane over a large area, indicating Hydrocarbon exploration in and around igneous potentially commercial quantities of natural gas rocks may involve a variety of geological, geo in Columbia River basalts.21 physical and geochemical techniques. Traditional 42 Depending on the properties of the volcanic rocks, gravity and magnetic techniques may be useful. These were among the earliest geophysical approaches applied, and they contributed to the successful exploitation of the 1915 Texas volcanic play mentioned previously. Mafic igneous rocks—richer in dense and magnetic minerals than felsic igneous rocks—offer better contrast with regional sediments, so they may show up distinctly on gravity and magnetic surveys. Aeromagnetic surveys have been effective in identifying prospects in mafic flood basalts in the Otway basin, southeastern Australia.22 Magnetotelluric (MT) methods have also been used, usually in conjunction with other techniques, to investigate high-resistivity volcanic rocks as potential reservoirs (for more on MT, see “Electromagnetic Sounding for Hydrocarbons,” page 4). For example, MT surveys in the Yurihara oil and gas field in Japan are aiding exploration of areas surrounding producing reservoirs.23 On some MT lines, resistive uplifted volcanic layers have been identified as possible prospects. Integration of MT surveys with surface seismic information was valuable in characterizing the internal structure of an oil- and gas-producing basalt layer. Seismic methods, while extremely useful for detecting sedimentary structures, have had mixed success in volcanic provinces. Massive basalts without internal layering have high effective seismic quality, meaning they are not highly absorptive, so seismic waves pass through them with little attenuation. Seismic surveys are relatively successful in delineating the tops and bottoms of such layers. However, layered basalts, especially those with interspersed weathered surfaces, tend to scatter seismic energy and may yield poor data.24 To improve the quality of seismic data in volcanic provinces, survey planners use satellite sensing to determine lithology and topography, and are incorporating the results in assessments of survey logistics, acquisition parameters and processing requirements (above left).25 In areas with highly attenuating volcanic layers, borehole seismic surveys have shown some promise in improving seismic image resolution. Such was the case with an offset vertical seismic profile (VSP) acquired in a 4,750-m [15,600-ft] exploratory well in the Neuquén basin, Argentina.26 At the well location, the surface was covered by approximately 150 m [490 ft] of basalt that strongly attenuated surface seismic energy. The VSP produced an image with higher resolution than the surface seismic results and illuminated other igneous bodies in the subsurface. Oilfield Review 17.Chen Z, Yan H, Li J, Zhang G, Zhang Z and Liu B: “Relationship Between Tertiary Volcanic Rocks and Hydrocarbons in the Liaohe Basin, People’s Republic of China,” AAPG Bulletin 83, no. 6 (June 1999): 1004–1014. 18.Komatsu N, Fujita Y and Sato O: “Cenozoic Volcanic Rocks as Potential Hydrocarbon Reservoirs,” presented at the 11th World Petroleum Congress, London, August 28–September 2, 1983. 19.Link WK: “Significance of Oil and Gas Seeps in World Oil Exploration,” Bulletin of the AAPG 36, no. 8 (August 1952): 1505–1540. 20.Fritts SG and Fisk LH: “Structural Evolution of South Margin—Relation to Hydrocarbon Generation,” Oil & Gas Journal 83, no. 34 (August 26, 1985): 84–86. Fritts SG and Fisk LH: “Tectonic Model for Formation of Columbia Basin: Implications for Oil, Gas Potential of North Central Oregon,” Oil & Gas Journal 83, no. 35 (September 2, 1985): 85–89. 21.Johnson VG, Graham DL and Reidel SP: “Methane in Columbia River Basalt Aquifers: Isotopic and Geohydrologic Evidence for a Deep Coal-Bed Gas Spring 2009 XS8 X,200 XS401 XS4 XS602 XS6 XS601 X,400 X,600 X,800 Y,000 Y,200 Y,400 0 km 0 Conglomerate Shale Upper volcanic Sedimentary Lower volcanic Basalt 2 mi 2 Y,600 Y,800 R U S S I A Daqing N MONGOLIA P A N. KOREA A Beijing C 0 km 400 0 mi H I N A S. KOREA J Gas-Bearing Volcanic Formations in China The giant Daqing field, discovered in 1959, is the largest oil field in China and one of the largest in the world. The field has produced more than 10 billion bbl [1.6 billion m3] from sedimentary layers 700 to 1,200 m [2,300 to 3,900 ft] deep. Stratigraphic wells—drilled to understand the basin-scale relationships between the reservoirs and the surrounding strata—encountered gas in volcanic layers at depths between 3,000 and 6,000 m [10,000 and 20,000 ft]. Because of the difficult environment and challenging reservoir rocks, these reserves were not immediately targeted for development. In 2004, PetroChina initiated a nine-well appraisal program and entered into a joint project with Schlumberger to better understand these deep volcanic reservoirs. The study area covered 930 km2 [360 mi2] and incorporated 3D seismic data along with wireline logs, borehole images and core analyses from 15 wells. To support development decisions, analysts constructed a workflow to evaluate these complex reservoirs and estimate the amount of gas in place.27 The initial step in the workflow involved building a structural model from seismic data. The top of the Yingcheng volcanic group is a significant seismic reflector, and interpretation of this horizon supplied the major structural control for the model. In addition to the top of the group, seismic interpreters distinguished three main volcanic sequences, with interbedded and bounding sedimentary sequences (above right). Within the structural model, each sequence was divided X,000 Depth, m Once a hydrocarbon-bearing volcanic deposit is discovered, evaluating the reservoir can be a challenge. Methods for assessing porosity, permeability and saturation in sedimentary rocks must be modified to work in volcanic provinces. Case studies from China and India demonstrate such techniques. 400 > Structure of the Yingcheng volcanic group beneath the Daqing field. Interpretation of seismic data determined the top of the volcanic group, and integration of seismic and log data allowed delineation of the upper volcanic, lower volcanic and predominantly basaltic sequences. into smaller cells that were later populated with physical properties. The reservoir consists mainly of interlayered crystalline rhyolites and rhyolitic pyroclastics, but a full spectrum of volcanics was encountered, ranging from basaltic to rhyolitic in composition and from crystalline igneous to pyroclastic in texture. Identifying rock types within the sequences and correlating them between wells were difficult tasks. Lithology classification for most types of rocks relies on mineralogy, which cannot be determined easily for the very fine-grained or glassy textures common in volcanic rocks. This led scientists studying volcanic rocks to focus on chemical composition as the key factor in classification schemes. With elemental concentrations from an ECS elemental capture spectroscopy tool, interpreters used these chemistry-based classification schemes to provide a continuous lithology description.28 However, chemical 26.Rodríguez Arias L, Galaguza M and Sanchez A: “Look Source in the Columbia Basin, Washington,” AAPG Ahead VSP, Inversion, and Imaging from ZVSP and Bulletin 77, no. 7 (July 1993): 1192–1207. OVSP in a Surface Basalt Environment: Neuquen Basin, 22.Gunn P: “Aeromagnetics Locates Prospective Argentina,” paper SPE 107944, presented at the SPE Areas and Prospects,” The Leading Edge 17, no. 1 Latin American and Caribbean Petroleum Engineering (January 1998): 67–69. Conference, Buenos Aires, April 15–18, 2007. 23.Mitsuhata Y, Matsuo K and Minegishi M: 27.Li G, Wang YH, Yang FP, Zhao J, Meisenhelder J, “Magnetotelluric Survey for Exploration of a VolcanicOilfield Rock Reservoir in the Yurihara Oil and Gas Field, Japan,”ReviewNeville TJ, Farag S, Yang XW, Zhu YQ, Luthi S, Hou HJ, Zhang SP, Wu C, Wu JH and Conefrey M: “Computing Geophysical Prospecting 47, no. 2 (March 1999): 195–218.09 Winter Gas in Place in a Complex Volcanic Reservoir in China,” 24.Rohrman M: “Prospectivity of Volcanic Basins: Trap Volcanic Fig. 9 paper SPE 103790, presented at the SPE International Delineation and Acreage De-Risking,” AAPG ORWINT09-VOL Bulletin 91, OilFig. and 9 Gas Conference and Exhibition in China, Beijing, no. 6 (June 2007): 915–939. December 5–7, 2006. 25.Laake A: “Remote Sensing Application for Vibroseis Data 28.Barson D, Christensen R, Decoster E, Grau J, Herron M, Quality Estimation in the Neuquen Basin, Argentina,” Herron S, Guru UK, Jordán M, Maher TM, Rylander E paper presented at the IAPG VI Congreso de Exploración and White J: “Spectroscopy: The Key to Rapid, Reliable y Desarrollo de Hidrocarburos, Mar del Plata, Argentina, Petrophysical Answers,” Oilfield Review 17, no. 2 November 15–19, 2005. (Summer 2005): 14–33. Coulson S, Gråbak O, Cutts A, Sweeney D, Hinsch R, Schachinger M, Laake A, Monk DJ and Towart J: “Satellite Sensing: Risk Mapping for Seismic Surveys,” Oilfield Review 20, no. 4 (Winter 2008/2009): 40–51. 43 Common depth point number 600 650 700 750 800 850 900 Pyroclastic flow Lava flow Pyroclastic fall Extrusive FMI Image 50 Porosity % 0 Facies Lava flow Tuff Pyroclastic flow Water laid Outer dome-building volcanic Pyroclastic flow Middle dome-building volcanic Pyroclastic fall Inner dome-building volcanic Surge flow Intrusive Upper lava flow Middle lava flow Surge flow Lower lava flow Pyroclastic fall > Correlation of igneous rock types with seismic data. Rock types were identified using FMI images, NMR T2 distributions and ECS elemental concentrations. Rock types were classified into seven crystalline lithologies (greens, pinks and purples) and four pyroclastic lithologies (orange and yellows). A sample correlation (bottom) shows an FMI image acquired through an interval of predominantly pyroclastic layers. A seismic section (top) through the central well is used to extend rock types across the field. The rock types observed in the central well are displayed at the well location using the color codes for volcaniclastic and crystalline lithologies. Rock types extrapolated away from the central well are displayed as semitransparent colors on the seismic section. 32.Kumar R: Fundamentals of Historical Geology and 29.Li GX, Wang YH, Zhao J, Yang FP, Yin CH, Neville TJ, Farag S, Yang XW and Zhu YQ: “Petrophysical Oilfield Review Stratigraphy of India. New Delhi: New Age International Publishers Limited, 2001. Characterization of a Complex Volcanic Reservoir,” Winter 09 Transactions of the SPWLA 48th Annual Logging 33.Negi Volcanic Fig. 10 AS, Sahu SK, Thomas PD, Raju DSAN, Chand R and Symposium, Austin, Texas, June 3–6, 2007, paper E. Ram J: “Fusing Geologic Knowledge and Seismic in ORWINT09-VOL Fig. 10 for Subtle Hydrocarbon Traps in India’s Cambay Searching 30.Freedman R, Cao Minh C, Gubelin G, Freeman JJ, Basin,” The Leading Edge 25, no. 7 (July 2006): 872–880. McGinness T, Terry B and Rawlence D: “Combining NMR and Density Logs for Petrophysical Analysis in 34.Pal A, Machin N, Sinha S and Shrivastva C: “Application Gas-Bearing Formations,” Transactions of the SPWLA of Borehole Images for the Evaluation of Volcanic 39th Annual Logging Symposium, Keystone, Colorado, Reservoirs: A Case Study from the Deccan Volcanics, USA, May 26–29, 1998, paper II. Cambay Basin, India,” presented at the AAPG Annual Convention and Exhibition, Long Beach, California, USA, 31.Short NM Sr and Blair RW Jr (eds): Geomorphology April 1–4, 2007. from Space. NASA (1986), http://disc.gsfc.nasa.gov/ geomorphology/ (accessed March 3, 2009). 44 composition is not the whole story; for example, if a particular rock has a rhyolitic composition, chemistry alone cannot distinguish between a crystalline rhyolite and a pyroclastic rhyolite tuff. Textural information from borehole images obtained by the FMI fullbore formation microimager provided the basis for distinguishing these rock types and tying together log data from all the wells. Magnetic resonance T2 distributions provided additional information to complete the lithology classification. By combining all available information, geologists were able to identify 11 igneous rock types in each well and then correlate them across the field using seismic data and conceptual geological models from other volcanic environments (left). Evaluating the petrophysical properties of each rock type was particularly challenging.29 Compared with the clastic and carbonate rocks that form conventional hydrocarbon reservoirs, these volcanic rocks exhibit the most problematic features of both; the complex mineralogy, including the presence of conductive minerals such as clays and zeolites, parallels that of the most challenging clastic rocks, and their texture and pore structure mimic those of the most complex carbonate rocks. This combination of features presents difficulties for the evaluation of porosity, permeability and fluid saturations. A robust scheme for lithology-independent evaluation of porosity in low-porosity, gas-bearing formations is the DMR density–magnetic resonance interpretation method, which combines bulk density and magnetic resonance porosity measurements.30 A relationship between matrix density and elemental concentrations derived from core analysis was applied to the ECS results to produce a continuous log of matrix density. The matrix density provided input to the DMR process for calculating high-quality estimates of porosity and indications of gas saturation in each well. To extrapolate porosity information to areas away from the wells, interpreters developed probability distributions of porosity for each rock type and used them to populate the model. Estimating gas saturation was a challenge because the complex rock texture prevented development of a suitable Archie-type saturation equation, so a capillary pressure–based approach was used to estimate saturation. Pseudocapillarypressure curves were derived from well-log magnetic resonance T2 distributions and calibrated to mercury-injection capillary-pressure Oilfield Review measurements performed on cores. Saturation values computed in this way showed a strong dependence on pore network geometry. For example, the core measurements showed the air-fall tuffs—volumetrically the most significant reservoir rock type—to be microporous, or having pore throats less than 0.5 μm in radius. Saturation profiles across these formations exhibited long transition zones extending hundreds of meters and covering most of the reservoir. The saturation results, validated with gas indications from the DMR method, downhole fluid analysis measurements and production data, were consistent with the assumption that the reservoir was a single-pressure system with one free-water level. The capillary pressure–based approach was subsequently used to populate the model with saturation values. Gas in place for the reservoir was calculated by summing the gas contained in each model cell. However, reservoir rock quality in this field is extremely heterogeneous. In addition, well control was limited, and the seismic data were imperfect in guiding the distribution of petrophysical properties. To cope with these difficulties, engineers employed a stochastic method to populate cells with porosity and gas saturation. Nearly 60 realizations were performed to evaluate the potential quantities of gas in place for the study area, providing an understanding of the range of uncertainty associated with field volumetrics. The results of the overall study supported the decision to develop the field. Oil in India’s Deccan Traps The Deccan Traps were formed by Late Cretaceous extrusion of flood basalts that today cover more than 500,000 km2 [190,000 mi2] of central western India. They are called traps, from the German word treppen for step, because they give rise to topography characterized by stepped terraces of resistant basalt layers (above right).31 The episode of volcanism was synchronous with the rifting of the Indian continent from southern Africa. Although the genesis and the mechanism of emplacement of these basalts are still debated, the general consensus is that they erupted under water.32 More than 40 such basalt layers have been identified, many of them interbedded with fluvial and estuarine limestones, shales and sandstones. In some places, total thickness of the traps exceeds 3,000 m. During the last 40 years, Cambay basin, one of the oldest hydrocarbon plays of western India, has produced hydrocarbons from sediments overlying the Deccan basalts.33 Until recently, Spring 2009 PA S KI TA N C H I N A NEP AL Cambay basin BANGLADESH Deccan Traps I N D I A Mahabaleshwar 0 0 km 500 miles 500 SRI LANKA > The Deccan Traps of India. The Deccan Traps are a sequence of approximately 40 basalt layers covering portions of central western India. Differences between the basalts, which are competent, and interlayered sands, shales and limestones, which are more easily eroded, give rise to the rough terrain (right ). This photograph was taken at the Mahabaleshwar escarpment in the Western Ghats. The Cambay basin (left) is a downdropped graben with oil-bearing sediments overlying the basalts. Basalt outcrops are shown in orange. (Photograph courtesy of Dr. Hetu C. Sheth, Department of Earth Sciences, Indian Institute of Technology, Mumbai.) the top of the volcanic deposits was considered by Well PK-2 was laterally extensive. Based on economic basement, below which commercial this model, Well PK-6 was drilled in 2005 just hydrocarbon reservoirs were not expected to be 600 m [1,970 ft] to the southwest of PK-2, but Oilfield Review found. However, in the past few years, oil has unfortunately it did not flow any hydrocarbon. Winter been discovered in these deeper volcanic rocks.09 This unexpected result encouraged GSPC to Volcanic Fig. 11 In 2003, Gujarat State Petroleum Corporation update ORWINT09-VOL Fig. the 11 reservoir model through further data (GSPC) initiated a six-well campaign in analysis, specifically considering the rock facies Block CB-ONN-2000/1. The first three wells and fractures and their interplay with faults exhibited oil shows in the volcanic layers. In within the volcanic layers.34 2004, the fourth well, PK-2, proved to be a signifiAs a first step, geologists developed a textural cant oil discovery, testing at 64 m3/d [400 bbl/d]. classification of the volcanic layers. Three main For planning the next well, a simplistic reservoir facies—vesicular basalt, nonvesicular basalt model was constructed that assumed the hydro- and volcaniclastic units—were identified using carbon-bearing topmost basalt layer penetrated borehole image logs, petrography from Well PK-1 45 Vesicular Basalt Nonvesicular Basalt Well PK-2 Volcaniclastic Rock Well PK-6 Top Basalt A 1,775 1,775 Top Basalt B 1,800 1,800 1,825 1,825 3 cm Top Basalt C 1,850 1,850 1,875 1,900 1,900 Depth, m 1,875 > Textural classification of Deccan basalt facies. Images from the FMI borehole resistivity imaging tool helped geologists identify three main rock types. Vesicular basalts (left) exhibited vesicles in image (top), in hand specimen sample (bottom) and also in sidewall cores from a neighboring well. Nonvesicular basalts (center ) showed no such gas bubbles in borehole images or in sidewall cores. Images of volcaniclastic basalts (right) showed fine-scale layering of angular particles. (Basalt photograph courtesy of Charles E. Jones, University of Pittsburgh, Pennsylvania.) and hand specimens of basalt (above). Next, the aluminum, iron and titanium for Basalts A, B and facies were correlated from well to well—an C showed that Basalt A, the top unit, is composiexercise that was far from straightforward. Lava tionally different in the key wells, while Basalts B flows can commingle, and after solidification and C are compositionally similar (next page). other changes can occur, such as hydrother- This suggests that the top basalt layer is disconmal alteration, weathering, cementation and tinuous laterally between the two wells, contrary structural deformation. These changes can be to the assumption in the original model. Following the facies analysis, the next phase identified in outcrop, but tracking them in the subsurface is not easy. Based on image facies and of the study involved characterizing natural fracOilfield Review log signatures, three main basalt layers, A, B and tures, which are abundant within the volcanic Winter 09 C, could be correlated between key wells PK-2 Fig.layers. Volcanic 12 In the discovery Well PK-2, the top basalt that flowed and PK-6 (above right). ORWINT09-VOL Fig. 12 hydrocarbon is thick, comprising a In outcrop studies, volcanic rocks can be cor- nonvesicular basalt layer overlying a vesicular related using geochemical analysis of major and basalt section with a number of fractures that minor elemental composition. In the subsurface, appear conductive on borehole images.35 The similar data can be acquired using the ECS tool. presence of open fractures and vesicles creates Crossplots of elemental silicon versus calcium, a good-quality reservoir with a dual-porosity system, and the fracture network enhances per35.In the absence of acoustic or testing data, conductive frac meability. In contrast, in Well PK-6, the top basalt tures on borehole images are considered open to flow. 36.Schutter SR: “Occurrences of Hydrocarbons in and layer, which is thinner, essentially nonvesicular Around Igneous Rocks,” in Petford N and McCaffrey KJW and less fractured, is not a good reservoir. (eds): Hydrocarbons in Crystalline Rocks, Geological Society Special Publication 214. London: Geological In addition to facies type and the presence of Society (2003): 35–68. fractures, the geometrical relationship between 46 1,925 Volcaniclastics Nonvesicular basalt Vesicular basalt Brecciated zone in nonweathered basalt > Initial well-to-well facies correlation. Texturebased facies classification allowed correlation of three basalt layers between Well PK-2 and Well PK-6. Basalt A (blue) is the producing zone in Well PK-2, but not in PK-6. Basalts B and C are nonproductive. fractures and faults also seems to play a crucial role in localizing hydrocarbon accumulations. In Well PK-2, the open fractures occur at high angles to a seismic-scale fault, while fractures in Well PK-6 are aligned approximately parallel to the fault. Interpreters developed a conceptual model in which the seismic-scale fault facilitates fluid communication, allowing the open fractures that intersect it to conduct hydrocarbons to producOilfieldaligned Reviewwith the fault are less ing wells. Fractures Winter likely to intersect it, 09 and therefore are unlikely to Volcanic Fig. 13 conduct hydrocarbons. This Fig. concept ORWINT09-VOL 13 was validated in a new well, PK-2A1, which contained conductive fractures oriented perpendicular to seismic-scale faults and also produced oil. Future Volcanic Activity Evaluation of hydrocarbons in volcanic rock presents many challenges, but creative application of techniques designed for sedimentary reservoirs is helping oil and gas companies characterize and exploit these complex accumulations. The Oilfield Review Depth, m Well PK-2 Gamma Ray Lithology Image Logs Elemental Concentrations, kg/kg 1,760 1,770 Ca/Si Fe/Si 0.20 0.15 Basalt A 1,780 0.15 0.10 1,790 0.10 1,800 1,810 0.05 Basalt B 0.05 1,820 0 0.10 1,830 0 0.15 1,840 1,850 0.25 0.30 0.35 Al/Si 0.14 Basalt C 0.20 0.05 0.15 0.35 Ti/Si 0.06 0.12 0.25 0.05 0.10 0.04 0.08 0.03 0.06 Depth, m Well PK-6 Lithology Gamma Ray 0.02 0.04 0.01 0.02 0 0 1,760 1,770 0 0.10 0.20 0 0.30 0.10 0.20 0.30 Basalt A 1,780 1,790 Ca/Si 0.15 Fe/Si 0.20 1,800 0.15 1,810 1,820 0.10 Basalt B 1,830 0.10 0.05 0.05 1,840 1,850 0 0.10 0 0.15 0.20 0.25 0.30 0 0.35 0.10 0.20 0.30 1,860 1,870 Al/Si 0.14 1,880 1,890 1,900 Ti/Si 0.06 0.12 0.10 Basalt C 1,910 0.06 1,920 0.04 1,930 0.04 0.08 0.02 0.02 0 1,940 0 0 0.10 0.20 0.30 0 0.10 0.20 0.30 > Comparison of basalts in two wells. Elemental concentrations (right) from the ECS tool are expressed as ratios of calcium, iron, aluminum and titanium to silicon (Ca/Si, Fe/Si, Al/Si and Ti/Si). Ratios are plotted for Basalts A (blue oval), B (green oval) and C (red oval). In each of the ratio plots, the red and green ovals have approximately the same relationship to each other, but not to the blue ovals. For example, in the Ca/Si plot for Well PK-2 (top), the red and green ovals are next to each other, and the blue oval is inside the red oval. However, in the Ca/Si plot for Well PK-6, the red and green ovals are still next to each other, but the blue oval is inside the green oval. This arrangement indicates that Basalts B and C correlate from one well to the other, but Basalt A does not. combination of borehole resistivity images with neutron-capture spectroscopy and magnetic resonance logs is becoming the new standard data suite for evaluation of volcanic reservoirs. With increased understanding of the capacity of volcanic rocks to contain oil and gas, other companies may consider reassessing volcanic formations they have bypassed, with a view to reevaluating their potential. Spring 2009 Unlike their sedimentary counterparts, volcanic rock reservoirs have not been studied systematically. In addition to the few examples Oilfield Review described inWinter this article, 09 hydrocarbons occur in or around igneous rocks Volcanic Fig.in14more than 100 counORWINT09-VOL Fig. 14 tries.36 In many instances, only oil shows and seeps have been documented, but further exploration may uncover significant reserves. The presence of volcanic rocks in a basin may not ever become a basis for exploration, but the possibility of such basins sustaining a viable petroleum system should be included within an array of options. While some operators might stop drilling after encountering “basement,” those with a better understanding of the potential of volcanic rocks may treat them like any other prospective reservoir rock. —LS 47 Contributors James Brady, Product Development Manager for WesternGeco Electromagnetics (EM) in Houston, oversees technology development for the company’s EM products. He joined Schlumberger in 1988 as an engineer at the Schlumberger Wireline Systems Development Center in Austin, Texas, USA. He subsequently worked on land seismic product development in Hannover, Germany. He has also served as geological product development manager, product planning and marketing manager, Petrel* integration manager and as research and innovation program manager. James earned an MS degree in electrical engineering at University of Texas at Austin, and a BS degree in electrical engineering and a BA degree in economics at University of California, Santa Barbara, USA. Marco Polo Pereira Buonora is Potential Method Geophysical Manager at Petrobras in Rio de Janeiro, where he is responsible for all nonseismic data acquisition and interpretation, such as gravity, magnetic and EM methods, particularly marine controlledsource electromagnetics (mCSEM). After earning a BS degree in geology at the Federal University of Pernambuco in Recife, Brazil, he began his career as a geologist at the Brazilian Ministry of Mines and Energy, working on gravity and magnetic data acquisition and interpretation. Later, he worked as a geophysicist on data acquisition, processing and interpretation of magnetics and gamma spectrometry data. In 1974, he was granted a Fulbright scholarship and attended St. Louis University, Missouri, USA, where he earned Master in Professional Geophysics (MprGph) and PhD degrees in geophysics, with concentrations in gravity and magnetics. After returning to Brazil in 1980, he joined Petrobras as a potential method scientist and has been active in gravity and magnetic interpretation of several onshore and offshore sedimentary basins in Brazil. In the last four years, he has been involved with the data acquisition and interpretation of mCSEM. He is also an associate professor at the Fluminense Federal University in Niterói, Rio de Janeiro, where he teaches applied gravity, magnetics and digital signal analysis. He is a member of the SEG and EAGE and of the Brazilian Geophysical Society, serving as its president from 1989 to 1991. Chuck Campbell is President and Senior Geoscientist at ACCEL Services Inc. in Houston. There, he provides interpretation services in gravity, magnetics and electrical methods. He began his career with Unocal in 1979, moving to Sohio/BP four years later. He has been with ACCEL Services since 1992. Chuck holds a BS degree in geology and geophysics from Western Illinois University in Macomb, USA. Tracy Campbell, WesternGeco Business Development Manager in the International Multiclient group, works to build the company’s multiclient portfolio. He joined Schlumberger in 2004 with the acquisition of AOA Geomarine Operations (AGO). Since then, he has been WesternGeco EM data processing manager in Austin, WesternGeco EM North America sales manager in Houston, and WesternGeco EM global projects manager in Houston. Tracy received a BS degree in physics from University of Alberta, Edmonton, Canada. 48 Alan D. Chave is a Senior Scientist in the Deep Submergence Laboratory at the Woods Hole Oceanographic Institution (WHOI) in Woods Hole, Massachusetts, USA. He holds a BS degree in physics from Harvey Mudd College, Claremont, California, and a doctorate in oceanography from the MIT-WHOI Joint Program in marine science. Alan was involved in the early development of mCSEM and maintains an active experimental research group focused on marine electromagnetics, optics and ocean observatory technologies. Adwait Chawathé is Subsurface Team Leader (Jack project) for the Chevron North America Exploration E&P Deepwater Business Unit. Based in Houston, he manages a subsurface team that evaluates the Jack and St. Malo developments. He began his career in 1995 as a senior research associate at the Petroleum Recovery Research Center in Socorro, New Mexico, USA. Two years later, he joined Chevron’s Simulation Consulting Team. Before assuming his current position in 2007, he spent more than three years in Kuwait as Chevron Ratawi asset team leader. Adwait earned his PhD degree in petroleum and natural gas engineering from The Pennsylvania State University, College Park, USA. Leendert Combee is Principal Geophysicist, Marine Seismic and EM Acquisition, at the WesternGeco Oslo Technology Center (OTC) in Norway. There, he is involved with the development of electromagnetic acquisition systems. In addition, he oversees the geophysical direction of marine seismic acquisition projects. He joined Geco-Prakla in Delft, The Netherlands, in 1992, as a research scientist studying the geophysical specifications for the Q-Land* system. He continued his research work, extending it to the Q-Marine* system, at Schlumberger Cambridge Research in England. Before joining the EM system development group in 2005, he was geophysical advisor for GecoPrakla receiver systems in Asker, Norway, responsible for all geophysical development of marine seismic systems including Q-Marine, Q-Seabed,* Q-Fin* and 4Dsteering. Leendert obtained MS and PhD degrees in electrical engineering and electromagnetic surveying from Delft University of Technology. Mohamed Dawoud is Manager of the Natural Resources Policy Department at the Environment Agency–Abu Dhabi, UAE. He is also Associate Professor (on leave) at the Research Institute for Groundwater, National Water Research Center in Egypt. Since 1991, he has maintained research, teaching and consulting activities in Egypt, Nigeria, Saudi Arabia and the UAE. His current research includes analysis of water supply-and-demand issues; development of a geographic information systems database; numerical modeling for groundwater flow and management; water management; and the role that improvements in water management can play in reducing poverty, improving environmental quality and enhancing food security. Dr. Dawoud has a BS degree with honors in civil engineering from Menoufia University, Egypt; and MS and PhD degrees from Ain Shams University, Cairo, through a joint program with Colorado State University, Fort Collins, USA. M.Y. Farooqui is General Manager (Planning and Development) with Gujarat State Petroleum Corporation (GSPC). He began his career with Oil and Natural Gas Corporation (ONGC) and supervised operations as wellsite geologist for more than 75 exploration or development wells. Since joining GSPC as senior geologist in 1994, he has played key roles in all aspects of the business. He represents the company on various operating and management committees and has helped secure domestic and international exploration acreage for GSPC. He has written several technical papers and is an active member of the SPE. He holds MSc and MPhil degrees in geology from Aligarh Muslim University, Uttar Pradesh, India, with a specialization in micropaleontology. Alastair Fenwick is Acquisition Development Manager for WesternGeco in Houston. He has also served as global sales and marketing manager for WesternGeco EM and as senior account manager for North American marine projects. After earning a BSc degree in oceanography and sedimentology from the University of East London in 1982, he joined OHP Ltd. as a hydrographic surveyor in Aberdeen. He moved to Geco-Prakla in 1990 as a navigation specialist, soon becoming a navigation processing supervisor. Alastair has had varied assignments with WesternGeco in marine sales and exploration services management in many locations around the world, particularly in Southeast Asia. Arnie Ferster, Exploration Manager, Greenland, for EnCana Corporation in Calgary, manages the company’s exploration program offshore southwest Greenland. He joined EnCana in 2000 as a staff geophysicist and worked on new ventures in Libya. He then worked on new ventures in West Africa before serving as team leader for projects in Ghana and later Oman. Arnie holds a BS degree in physics from the University of Victoria, British Columbia, Canada, and is an active member of APPEGA and AAPG. Marcus Ganz has been Marketing and Sales Manager for WesternGeco Electromagnetics in Houston since 2008. His main responsibility is the commercialization of the electromagnetics product group. He joined the company in 1981 to work on seismic field crews. He has also served as Schlumberger Oilfield Services manager in Argentina and Chile; WesternGeco region manager for South America, based in Rio de Janeiro; and WesternGeco land general manager, based in Gatwick. Marcus has a BS degree (Hons) in physics from University of Southampton, England. Karen Sullivan Glaser, who is based in Houston, is Manager of the Reservoir Consulting group, a part of Schlumberger Data & Consulting Services (DCS). She oversees a large group of geologists, geophysicists, petrophysicists and engineers who assist clients in improving characterization of their reservoirs. She joined Schlumberger GeoQuest in 1995 and subsequently worked for WesternGeco, Integrated Project Management and DCS in various technical, marketing and management roles. Before joining Schlumberger, she worked for Exxon Production Research as a research geologist focusing on sequence stratigraphy. Oilfield Review She has also worked for Amoco Production Company in the Permian basin. Karen has a BA degree in geology from Colgate University, Hamilton, New York, USA; an MS degree in petroleum geochemistry from University of Oklahoma in Norman, USA; and a PhD degree in geology from Rice University in Houston. Stephen Hallinan, WesternGeco Land EM Manager, is based in Milan, Italy, where he is involved in project sales, operations and interpretation of magnetotellurics, controlled-source electromagnetics (grounded dipole and inductive loop) and gravity surveys. He has more than 15 years of experience as an EM project geophysicist in various locations around the world. Stephen received a BA degree in geology from Trinity College, Dublin, Ireland, and a PhD degree from The Open University, Milton Keynes, England. Rolf Herrmann is a Technical Manager for Schlumberger Water Services in Abu Dhabi, UAE. As principal hydrogeologist, he is involved in subsurface exploration and evaluation of aquifers and reservoirs. He has carried out numerous projects in the assessment of hydrogeological systems and analysis of dynamic conditions of groundwater aquifers and reservoirs. He also provides expertise in the development of conceptual models and numerical simulations. Rolf has served as a project manager for the exploration of carbonate aquifer structures for underground gas storage, including planning and design, supervision and evaluation of geological well information and 2D and 3D seismic information. His specialties include aquifer and reservoir characterization, dynamic simulation and evaluation of geophysical logs and all aspects of aquifer storage and recovery systems. He has an MS degree in geology from The State University of New York and a BS degree in earth sciences from the University of Würzburg in Germany. Huijun Hou is a Senior Geologist for Schlumberger, providing LWD support in Dhahran, Saudi Arabia. He joined Schlumberger in 2000 as a geologist in Beijing, working on interpretation of borehole image logs. While there, he was part of the team that performed an integrated characterization of a volcanic gas reservoir in the Songliao basin. Before assuming his current position, he was operations manager in Sudan and geology domain champion in Bangkok, Thailand. Before joining Schlumberger, Huijun worked for the China National Logging Company. He earned a BS degree in geology from Jianghan Petroleum Institute in Wuhan, Hubei province, China, and an MS degree in geosciences from the University of Petroleum in Beijing. Younes Jalali is a Schlumberger Reservoir Engineering Advisor at the Beijing Geoscience Center. He has been with Schlumberger since 1990, with assignments in North Africa, Europe, the USA and now Asia. Prior to joining Schlumberger, he was a member of the Petroleum Engineering Faculty at Stanford University, California. He has BS and MS degrees in petroleum engineering from the University of Tulsa and a PhD degree in petroleum engineering from the University of Southern California. He holds a number of patents related to reservoir and well evaluation and is a regular contributor to SPE literature. Spring 2009 Tiziano Labruzzo, who is based in Rio de Janeiro, has been Senior Geophysical Programmer for WesternGeco Electromagnetics since 2007. He is responsible for development of new technologies for marine EM processing and for interpretation of CSEM and marine magnetotelluric data. He began his career in 1993 as a software engineer for the National Museum of Science and Technology in Milan, Italy. He also worked for Coas srl and Asforil srl, both in Milan, as a software engineer and consultant, respectively. Before becoming an engineering intern at Schlumberger EMI Technology Center in Richmond, California, in 2005, he was a consultant on EM processing and interpretation software for Geoinvest srl, Piacenza, Italy. Tiziano is a graduate of Università di Milano, with a degree in computer science. Guoxin Li is a Senior Petrophysicist and Director of the Engineering Technology and Supervision Department of PetroChina Exploration and Production Company, where he is responsible for drilling, petrophysics and mud logging. He has a bachelor’s degree in petrophysics and a master’s degree from the China University of Petroleum. Nigel Machin is a Senior Reservoir Geologist for the Central Arabia Group with Saudi Aramco in Dhahran. He began his career in 1993 with Enterprise Oil in London, specializing in the application of borehole images to reservoir evaluation. He has 16 years of reservoir evaluation experience and worked as a consultant geologist for several operators in Indonesia including Core Laboratories and Halliburton. He joined Schlumberger as geology domain champion for India in 2004 and moved to Saudi Aramco in 2007. His areas of interest are the evaluation of deepwater and fluvio-eolian depositional systems and fracture evaluation through borehole imaging techniques. Nigel is a member of the International Association of Sedimentologists (IAS) and Society for Sedimentary Geology (SEPM) and holds a BSc degree in mining and mineral exploration (applied geology) from the University of Leicester, England. Tom Neville is a Petrophysics Advisor and Acting Research Director at Schlumberger-Doll Research in Cambridge, Massachusetts. His research focuses on formation evaluation of unconventional resources. Before this, he led formation evaluation at Schlumberger Data & Consulting Services (DCS) in China. In his 13 years with Schlumberger, Tom has held a variety of field positions, in engineering, in research and at headquarters. Before joining Schlumberger, he had six years of experience as an exploration and development geologist with independent oil companies in Australia. Tom earned a BS degree in geology from the University of Queensland, Brisbane, Australia. Edward A. Nichols is an EM Specialist at Schlumberger Riboud Product Center in Clamart, France. Previously, he was EM discipline manager, EMI Technology Center, in Richmond, California, where he was responsible for land and marine geophysical instrumentation products and for borehole physics. He has also been project manager for technology development and adviser on EM instrumentation. He began his career in 1977 as a geologist-geophysicist in eastern Canada with Amax Minerals Exploration. In 1982, he became president of a Canadian consulting group, Capital Resources. From 1985 to 2004, he worked for Electromagnetic Instruments Inc. as vice president R&D, president, operations manager and consultant geophysicist. The author of numerous publications and holder of many patents, Edward holds a BS degree in mathematics (Hons) from Mount Allison University, Sackville, New Brunswick, Canada, and an MS degree in geophysics from McGill University in Montreal, Quebec, Canada. Umut Ozdogan, Lead Reservoir Engineer in the Chevron Angola Reservoir Management Group, heads reservoir simulation and engineering activities in the Takula asset, offshore Angola. He joined Chevron Corporation in 2003 as a reservoir engineer for the Tahiti field in the deepwater Gulf of Mexico. Since then he has led and participated in multiple reservoir engineering and simulation studies in the deepwater Gulf of Mexico, West Africa and Eurasia. Umut received a BS degree in petroleum and natural gas engineering from Middle East Technical University, Ankara, Turkey, and an MS degree in petroleum engineering from Stanford University, California. Aditi Pal, Schlumberger Borehole Geology Team Leader, is based in Jakarta. She joined Schlumberger in 2002 in Mumbai, where she worked on interpretation of borehole image data in both clastic and carbonate reservoirs. In 2005, she participated in a multiwell study involving facies and fracture analysis in a basalt reservoir in India. She assumed her current position in 2007. Aditi has a bachelor’s degree in geology from the University of Calcutta and MS degrees in applied geology and geo-exploration from the Indian Institute of Technology, Mumbai. Steve Patmore is Principal Explorer, Greenland, for Cairn Energy Plc in Edinburgh, Scotland. Last year he joined the new Greenland team with responsibilities for geophysical operations, interpretation and regional consulting. He began his career with Conoco in 1974 as a geophysicist and went on to work in Chad, Norway, China, North America, Egypt and West Africa. Since 1992, he has worked the UK Continental Shelf including the central Graben, southern North Sea and West of Shetlands areas. After joining Cairn in 2004 as principal geophysicist working in Indian assets, he then became asset manager for Nepal and northern India until taking his current post last year. Steve earned a BS degree (Hons) in geology and oceanography at the University of Wales in Swansea. Mark Riding, Schlumberger Deepwater Theme Director, is responsible for deepwater corporate strategic planning, sales and technology development worldwide. He began his 27-year career with Schlumberger as a field engineer for Flopetrol well testing and interpretation services. He transferred to Wireline openhole services in 1990 and has subsequently worked in various field, sales and management capacities, including district management for Wireline operations, Trinidad; business manager for Testing services, Asia; general manager for Testing services worldwide; and VP and general manager for subsea services worldwide. Recent corporate roles have enhanced his expertise in mergers and acquisitions, HPHT operations and knowledge management. Mark holds a BS degree in mining and chemical engineering from the University of Birmingham, England. 49 Luiz Felipe Rodrigues earned a BS degree in geology from the Federal University of Rio de Janeiro in 1996. Since then he has worked in acquisition, processing and interpretation of high-resolution seismic data for engineering and exploration projects. He joined Petrobras in 2000, where he specialized in geophysics, with emphasis on seismic methods, then acted as seismic interpreter in the Santos basin. His responsibilities include several exploration projects. Luiz is a member of the Brazilian Society of Geophysicists and the SEG. Stewart K. Sandberg, Senior Geophysicist at WesternGeco Electromagnetics in Houston, interprets electromagnetic data acquired offshore to map geology and to evaluate potential hydrocarbon horizons. His more than 30 years of industry experience have included work as supervising geophysicist, project manager, assistant professor and private consultant performing geophysical fieldwork, data processing and interpretation for geological, hydrogeological and environmental assessments. From 1996 to 2005, he was president of Geophysical Solutions, supplying geophysical contracting and consulting services in environmental, engineering and mining exploration applications. The author of many technical papers, Stewart received BS degrees in mathematics and in physics and an MS degree in geophysics, all from the University of Utah in Salt Lake City, USA. He also earned a PhD degree in geological sciences from Rutgers University, New Brunswick, New Jersey, USA. Chandramani Shrivastva is Schlumberger Geology Domain Champion, based in Mumbai. He began his Schlumberger career as a data management geoscientist in New Delhi, India, in 2002. In 2003, he transferred to Mumbai to work on interpretation of borehole image data. As a borehole geologist, he was involved in a multiwell project integrating borehole images, openhole log data and seismic data to characterize fractures and facies in a volcanic reservoir. He was promoted to his current position in 2007. Chandramani has a BS degree in geology from Patna University, Bihar, India, and an MS degree in geological engineering from the Indian Institute of Technology in Roorkee, Uttarakhand. Jan Stilling is Chief Geologist for Nunaoil A/S, Nuuk, Greenland. In 1996, after beginning as a geologist with PC-Laboratoriet in Fjerritslev, Denmark, Jan moved to Statoil ASA in Harstad, Norway, to work as a geologist and then senior geologist. He participated in evaluation of exploration opportunities offshore Norway and in the Svale development projects before joining Nunaoil in 2001. Jan obtained an MS degree in geology from Aarhus University, Denmark. Kenneth E. Umbach, a Geophysicist with EnCana Corporation in Calgary, has been responsible for geophysical interpretation and operations in Greenland and the Middle East since 2004. He began his career in 1978 as a geophysical interpreter with Amoco Canada and subsequently served in Houston, Jakarta and Calgary. In 1992, he joined PanCanadian to work on new ventures in North Africa, Europe, the Far East and Australia. He also worked on geophysical interpretation and operations in Australia and the Middle East for AEC, the predecessor of EnCana. 50 Frank van Kleef has been Lead Geophysicist for the Dubai Petroleum Establishment in the UAE since 2007. Before that he was a senior geophysicist for Gaz de France and worked on assets in Algeria and offshore Netherlands. Frank received an MSc degree in geology and geophysics from the University of Utrecht, The Netherlands. Yuhua Wang is Deputy General Manager of PetroChina Daqing Oilfield Company. He has more than 20 years of experience as an exploration geologist and earned a doctoral degree from the Chinese Academy of Sciences. Xingwang Yang, a Senior Petrophysicist and DCS Manager with Schlumberger in Tokyo, manages business for Japan, Korea and Taiwan. He joined Schlumberger in 1999 as a log analyst. Since then, he has worked in China and Saudi Arabia on various petrophysics projects. From 2004 to 2007, he was involved in evaluating deep gas reservoirs in volcanic rocks in several Chinese fields. Before joining Schlumberger, he worked for PetroChina as a wireline field engineer for two years. Xingwang has a BS degree in petrophysics from the University of Petroleum in Dongying, Shandong, China. Changhai Yin is Director of the Natural Gas Department in the Exploration and Development Research Institute of PetroChina Daqing Oilfield Company. He has more than 20 years of experience as a petrophysicist and holds a doctoral degree from the China University of Geosciences. Andrea Zerilli is a Research Scientist with WesternGeco Electromagnetic Services in Rio de Janeiro. With more than 30 years in the oil industry and worldwide experience in R&D, his current interests include emerging, deep-reading EM technologies, new high-resolution marine EM technologies, development of integrated solutions and management of multidisciplinary R&E and multiproduct projects. Before he joined Schlumberger in 2003, he worked for ENI as research project leader, for KMS Technologies as director of integrated geophysics, for the USGS as a visiting scientist and for the Colorado School of Mines as research associate. Invited speaker and organizer at many technical society meetings, Andrea holds a DSc degree in earth sciences from Parma University in Italy. Jie Zhao is a Vice Chief Engineer in the Exploration Company of PetroChina Daqing Oilfield Company. He has more than 20 years of experience as a petrophysicist and obtained a doctoral degree from the China University of Mining and Technology. An asterisk (*) is used to denote a mark of Schlumberger. An asterisk (*) is used to denote a mark of Schlumberger. Casing Drilling® is a registered trademark of Tesco Corporation. Oilfield Review Coming in Oilfield Review NEW BOOKS Guesstimation: Solving the World’s Problems on the Back of a Cocktail Napkin Lawrence Weinstein and John A. Adam Princeton University Press 41 William Street Princeton, New Jersey 08540 USA 2008. 320 pages. US $19.95 (paperback) ISBN 978-0-691-12949-5 Today, the ability to estimate is a crucial skill. This book is a collection of problems from daily life that allows those with basic mathematics and science skills to quickly estimate almost anything using plausible assumptions and basic arithmetic. Contents: • How to Solve Problems • Dealing with Large Numbers • General Questions • Animals and People • Transportation • Energy and Work • Hydrocarbons and Carbohydrates • The Earth, the Moon, and Lots of Gerbils • Energy and the Environment • The Atmosphere • Risk • Unanswered Questions • Appendixes, Bibliography, Index One excellent way to start honing such skills is with a few so-called Fermi problems, named for Enrico Fermi, the physicist who delighted in tossing out the little mental teasers to his colleagues whenever they needed a break from building the atomic bomb. … Dr. Adam and his colleague Lawrence Weinstein, a professor of physics, offer a wide and often amusing assortment of Fermi flexes in a book that just caught my eye, [Guesstimation]. Angier N: “The Biggest of Puzzles Brought Down to Size,” New York Times (March 30, 2009), http://www.nytimes.com/2009/03/31/science/ 31angi.html?ref=science (accessed April 22, 2009). Spring 2009 Flash of Genius: And Other True Stories of Invention John Seabrook St. Martin’s Press 175 Fifth Avenue New York, New York 10010 USA 2008. 384 pages. US $14.95 ISBN 0-312-53572-4 This book of essays is a collection of true stories about great ideas. New Yorker author Seabrook explores the moment when inspiration strikes in an otherwise average life, and what happens when that idea takes on a life, and commercial possibilities, of its own. Contents: • The Flash of Genius • The Fruit Detective • Game Master • Child’s Play • Sowing for Apocalypse • The Tree of Me • Fragmentary Knowledge • Invisible Gold • Selling the Weather • The Slow Lane • The Tower Builder • American Scrap • It Came from Hollywood • Tremors in the Hothouse • The Spinach King … characters, not events are at the heart of Seabrook’s excellent writing. Even the technical details of the ideas and inventions come second, almost incidental to his first-person explorations of the people imbued with or affected by the inspiration. … Seabrook fits in a surprising amount of technical detail. You’ll learn, for instance, how to recover scrap metal or design skyscrapers that won’t topple over. But… the real lesson of this book is a human one: Flashes of genius, no matter how small, can come from anywhere and perhaps anyone. Simonite T: New Scientist 199, no. 2675 (September 27, 2008): 46. When Science Goes Wrong: Twelve Tales from the Dark Side of Discovery Simon LeVay Penguin Group 375 Hudson Street New York, New York 10014 USA 2008. 304 pages. US $15.00 (paperback) ISBN 978-0-452-28932-1 Neuroscientist LeVay understands the high potential cost of erroneous theories and bad information. This book presents 12 stories of catastrophic blunders in a wide variety of scientific disciplines, from engineering geology and volcanology to microbiology and nuclear physics. Contents: • Neuroscience: The Runner’s Brain • Meteorology: All Quiet on the Western Front • Volcanology: The Crater of Doom • Neuroscience: The Ecstasy and the Agony • Engineering Geology: The Night the Dam Broke • Gene Therapy: The Genes of Death • Nuclear Physics: Meltdown • Microbiology: Gone with the Wind • Forensic Science: The Wrong Man • Space Science: Off Target • Speech Pathology: The Monster Study • Nuclear Chemistry: The Magic Island • Epilogue, Sources Petroleum System Modeling. Petroleum system modeling, sometimes called charge modeling, uses seismic interpretations, well logs, laboratory data and geological knowledge to model the evolution of a sedimentary basin to determine if a reservoir has been filled, or charged, with hydrocarbons. This article describes the steps in the process and explains how this modeling helps to establish fluid presence and type with confidence and to assess exploration risk before drilling. Coalbed Methane—A Global Resource. Commercial coalbed methane (CBM) production was originally a North American phenomenon. Techniques used to tap this unconventional resource, many introduced in the 1980s, have changed considerably as CBM development has increasingly become a global venture. This article presents the expanding geographical scope of CBM production and describes recently introduced methods for evaluating, drilling, completing and producing natural gas from coal. Crosswell Electromagnetic Surveys. To better manage producing fields, operators must understand and predict fluid movements between wells. A recently developed crosswell electromagnetic induction system can image resistivity distribution between wells. Formation resistivity is, in turn, a function of porosity and fluid saturations. This new crosswell system illuminates the interwell reservoir area using a transmitter in one well and a string of receivers in another, and it can propagate signals up to 1 km [0.6 mi] through a typical oilfield section. Adding to the drama of each story are the scientists’ own interpretation of events. … LeVay has tried, where possible, to interview all major players involved. Reading the characters’ own reflections and opinions in their own words and then comparing that to the “facts” makes for an absorbing read. [The book] is written for both the scientist and the layperson to enjoy. Wayman E: Geotimes 53, no. 7 (July 2008): 43. 51 • Lithology and Porosity Estimation • Saturation and Permeability Estimation • Index Well Logging for Earth Scientists Darwin V. Ellis and Julian M. Singer Springer P.O. Box 17 3300 AA Dordrecht, The Netherlands 2008. 692 pages. US $99.00 ISBN 978-1-4030-3738-2 This revised, expanded edition of the classic 1987 text provides detail on a variety of specialized well logging instruments used to obtain measurements from the borehole during and after the drilling process. The book contains information about the physical basis of borehole geophysical measurements, as well as an introduction to practical petrophysics—extracting desired properties from well log measurements. Contents: • An Overview of Well Logging • Introduction to Well Log Interpretation: Finding the Hydrocarbon • Basic Resistivity and Spontaneous Potential • Empiricism: The Cornerstone of Interpretation • Resistivity: Electrode Devices and How They Evolved • Other Electrodes and Toroid Devices • Resistivity: Induction Devices • Multi-Array and Triaxial Induction Devices • Propagation Measurements • Basic Nuclear Physics for Logging Applications: Gamma Rays • Gamma Ray Devices • Gamma Ray Scattering and Adsorption Measurements • Basic Neutron Physics for Logging Applications • Neutron Porosity Devices • Pulsed Neutron Devices and Spectroscopy • Nuclear Magnetic Logging • Introduction to Acoustic Logging • Acoustic Waves in Porous Rocks and Boreholes • Acoustic Logging Methods • High Angle and Horizontal Wells • Clay Quantification 52 The collaboration has resulted in a book that is both authoritative and lucid and a suitable text for university curricula. However, it is also an important reference book for industry users, describing both the fundamental physics of well logging and a historical development of tool design. I recommend this book highly. The authors have worked hard and meticulously on the text as a labor of love, as is obvious on every page. This book should be on the shelf of everyone who works with logs or aspires to, as an introduction, a refresher, and a well logging reference source. Doveton JH: AAPG Bulletin 93, no. 2 (February 2009): 293–294. • Psychrometry, Evaporative Cooling, and Solids Drying • Distillation • Equipment for Distillation, Gas Absorption, Phase Dispersion, and Phase Separation • Liquid-Liquid Extraction and Other Liquid-Liquid Operations and Equipment • Adsorption and Ion Exchange • Gas-Solid Operations and Equipment • Liquid-Solid Operations and Equipment • Reactors • Alternative Separation Processes • Solid-Solid Operations and Processing • Waste Management • Process Safety • Energy Resources, Conversion, and Utilization • Materials of Construction • Index physics. He divides impossibility into three classes: Class 1 for concepts beyond current technology, but which do not violate any known physical laws; Class II for concepts beyond present technology that also challenge interpretation of those laws; and Class III for concepts that defy known physical laws and would demand huge changes in our understanding of how the universe works. … [the book] remains the definitive resource for process engineering, and is a must have for university libraries and practicing chemical engineers everywhere. … it will enable a trained engineer to handle any process design contingency with confidence.… Highly recommended. The study of the impossible has opened up entirely new vistas for science, Kaku rightly points out. It is here that the book’s strength lies: the impossible is a gateway for discussing what we still do not understand, those gray areas that are surely the most fascinating part of physics. King MR: Choice 45, no. 8 (April 2008): 1366. … there is a surprising amount of heavyweight, cutting-edge science woven into the fabric of this book. … [The book] is, in fact, an easy-toread physics primer in disguise. Kaku has a huge reach as a writer and speaker. Perry’s Chemical Engineers’ Handbook, 8th ed. Don W. Green and Robert H. Perry (eds) McGraw-Hill Professional Two Penn Plaza, 23rd Floor New York, New York 10121 USA 2008. 2,400 pages. US $199.00 Contents: • Class I Impossibilities: Force Fields; Invisibility; Phasers and Death Stars; Teleportation; Telepathy; Psychokinesis; Robots; Extraterrestrials and UFOs; Starships; Antimatter and Anti-Universes • Class II Impossibilities: Faster Than Light; Time Travel; Parallel Universes • Class III Impossibilities: Perpetual Motion Machines; Precognition • Epilogue: The Future of the Impossible • Notes, Bibliography, Index Brooks M: New Scientist 197, no. 2645 (March 2008): 52. ISBN 0-07-142294-3 First published in 1934, this book has long been regarded as an expert source of chemical engineering information. This updated classic text covers every aspect of chemical engineering, from fundamental principles to chemical processes and equipment to new computer applications. Contents: • Conversion Factors and Mathematical Symbols • Physical and Chemical Data • Mathematics • Thermodynamics • Heat and Mass Transfer • Fluid and Plastic Dynamics • Reaction Kinetics • Process Control • Process Economics • Transport and Storage of Fluids • Heat-Transfer Equipment Physics of the Impossible: A Scientific Exploration into the World of Phasers, Force Fields, Teleportation, and Time Travel Michio Kaku Doubleday, Division of Random House, Inc. 1745 Broadway New York, New York 10019 USA 2008. 329 pages. US $26.95 ISBN 0-385-52069-7 In this book, noted physicist Michio Kaku explores the extent to which the technologies and devices of science fiction, deemed impossible today, might become commonplace in the future. From antimatter to time travel, the author explores the basics and the limits of the current known laws of Oilfield Review