Wei He, Roger N. Anderson, Liqing Xu, Albert Boulanger, Billy Meadow, Randall Neal
Lamont-Doherty Earth Observatory, Columbia University, and 4D Technology, Inc.
Even though more and more of the Earth's limited number of petroleum basins have been explored, the increase of newly discovered world oil reserves is barely keeping pace with burgeoning demand as developing countries are being industrialized. The crossover between supply and demand may be fast approaching. Under the circumstances, the petroleum industry is about to face what might be its greatest challenge ever to provide the world with sufficient supplies into the early part of the 21st century.
The yearly increase in 3D seismic data collection has accelerated to the point that this wonderful new technology has significantly dropped the price-per-barrel to find oil. It has also resulted in surprising increases in the production of more oil from existing fields in the world. Perhaps, the solution to the impending push to get more oil to market, faster, will come not from increased exploration success, but from new uses of 3D seismic data that maximize recovery efficiencies from existing oil and gas fields.
4D, or time-dependent, seismic reservoir monitoring, is an emerging technology that holds great hope that it will develop into a better oil production management system. Similar computer simulation environments have revolutionized other technologically-driven producers such as the automotive, aerospace, chemical and military industries.
Seismic technologies have progressed to the point that new reservoir production and engineering applications are being discovered yearly. Analyst James Wicklund of Rauscher, Pierce, Refsnes recently estimated that the 4D seismic market will grow to $1-3 billion annually within 5 years (Mack, 1995). The resulting revelation has been that we are only just beginning to visualize the real changes that occur within reservoirs as oil and gas are drained over time, and the changes that we see are surprising us. For example, gravity is often not nearly as efficient at sweep as we thought it was for all these years.
The physics and chemistry of one of the industry's most fundamental fluid-flow processes must be better understood before added economic benefits can be realized. In particular, 4D seismic monitoring offers the possibility that computer control of field development may someday be the industry norm. For example, 4D will unquestionably improve the performance of reservoir simulators.
However, continued refinements to the 4D seismic solution must be strongly coupled to forward and inverse models of seismic attribute changes; which are, in turn, needed to understand changes in pressure, temperature, and fluid content coming from the monitoring of wells. These 4D monitoring solutions will be driven by teams of production and reservoir engineers, geologists and geophysicists.
The average oil field yields only about 35% of the oil in place. Much that might have been recovered is bypassed, even with the best of current 3D technologies. There is general consensus arising in the industry that 3D technology has established a proven improvement factor in the 25-30% range in return on investment (ROI). The value driver for this improvement in production efficiency is increased "hit rate", or the finding of more oil with fewer wells, coupled with the elimination of many dry holes.
4D technology requires an additional investment in software and processing over 3D, but its unique time-dependent information promises to yield seismic differences that, when calibrated with well logs and production histories, identifies drainage patterns and locations of bypassed pay. The "right sizing" of high volume/high rate, deepwater production systems, improved location and timing of developmental drilling programs, the early verification of the reservoir simulation models, improved in-fill drilling and injector placements are all value drivers to widespread acceptance of 4D technologies.
4D is expected to extract up to an added 30% return over and above 3D's improvement. For example, in the biggest 4D development in the world so far, BP/Shell's Foenhaven Field off the western Shetlands, BP reports an expected increase in recovery from 40-50% of oil-in-place with conventional 3D to 65-75% with 4D (Kott, 1996). A permanent bottom cable seismic array has been installed there, and the industry is anxiously awaiting the first production effects to show up from repeated re surveying.
4D not only promises to produce new revenue for oil companies, but service companies will benefit as well. They sit on a "treasure trove" of legacy 3D seismic surveys. The "data exploitation" of these old datasets currently in company vaults will produce new reprocessing jobs and new reasons for reshooting over old survey locations.
4D seismic monitoring of hydrocarbon drainage is an integrated exploration and production technology. It requires not only the static description of reservoir geometry. but also the dynamic description of the fluid and pressure changes in the reservoir that occur during hydrocarbon production. To delineate these changes, the geological, geophysical and petroleum engineering data obtained during both exploration and production phases are required: thus the integration. In fact, the integration of high resolution measurements in the vertical direction from well log data with the high spatial and temporal resolution of time-lapse seismic datasets is the key to the success of the 4D seismic monitoring efforts.
In 4D seismic monitoring, we begin with a reservoir characterization model, containing lithology, porosity, and pore fluid pressure distributions. This provides a static description of the reservoir. To build time-dependence into such a description, new geological, geophysical, and engineering data must be recorded over and over in a field. Based upon changes in that data over time, a quantitative reservoir simulation is constructed using both inverse and forward seismic models of 4D seismic differences that can then be iteratively recomputed and compared with the reservoir simulations. Using such a seismic/reservoir simulation technique, the dynamic changes in the hydrocarbon reservoir can be monitored and simulated efficiently, and the results can then be used to understand and predict drainage occurring during production. New wells can then be placed to maximize the lives of oil and gas fields in order to achieve the highest hydrocarbon recovery rate possible.
Seismic inversion is becoming more and more important in the petroleum industry because the outcome, i.e., acoustic impedance, is more easily related to petrophysical properties such as lithology, porosity and oil saturations, than are seismic waveforms themselves. In this sense, seismic inversion is the bridge that connects seismic data to reservoir petrophysical properties. Thus, the advantages of using acoustic impedance data over seismic waveforms are profound in 4D seismic monitoring investigations.
There are numerous seismic inversion algorithms developed since the early 1980's, but these algorithms have yet to be applied to real 4D seismic data. To date, the stability and consistency of seismic inversions are the major concerns. New nonlinear, seismic waveform inversion techniques have improved the stability and the consistency to the point that useful interpretations of the meaning of seismic amplitude changes over time are emerging (He, 1996).
We have been conducting a 4D field monitoring experiment in the South Eugene Island (EI) Block 330 field of offshore Louisiana for several years now (Fig. 1), and observational and modeling results from one reservoir in that field, the LF, suggest that the effects of drainage can be quantified by analyzing 4D changes in seismic impedance between 3D surveys conducted over the field a number of years apart.
The locations of bypassed hydrocarbons in the LF reservoir in the EI 330 Field at the beginning of our study were thought to be gravitationally controlled (Fig. 2). The sweep from the beginning of production in 1972 to the beginning of the 4D study in 1992 was updip and systematic. It accounted for the watering-out of the downdip wells in a fashion consistent with the migration of a clear oil/water contact. Only the most updip wells were producing oil and gas by 1992 (Fig. 3).
The 4D seismic inversion of seismic surveys conducted in 1985 and 1992 (Figure 4) suggests that there was a much more complicated drainage history occurring in the LF reservoir, with water fingering controlled perhaps by sand-quality variations (Fig.4). While oil/water contacts were observed in both surveys, inefficient gravitational sweep left many remaining high amplitude, low impedance zones downdip (Fig. 5). The impedance-difference analysis further shows sustained low impedance zones (green in Fig. 6) that could yield bypassed pay.
The watered out wells seem to have drained in a non-radial and decidedly 3D sense (Fig. 4). For example, EI 330 well A-7 watered out in 1980, whereas well A-6, which is along the same structural contour, watered out in 1987, but the A-8 well, 100 ft downdip, did not water out until 1988 (Fig. 3). Both the A-12 and B-4 wells were still producing when the second 3D seismic survey was recorded in 1992. These observations are consistent with the irregular drainage indicated by the 4D impedance analysis (Figs. 4 and 6).
The seismic inversion, coupled with a geostatistical link between the impedance changes and well logs, produces an enhanced image of the state of the reservoir in 1992. The impedance inversion allows the extraction of not only lithology and porosity changes over time in the reservoir, also "effective oil saturation" (Fig. 7). Consistency is found between the effective oil saturation map and the bypassed hydrocarbon locations obtained from the 4D impedance analysis (Figs. 4 and 6). This suggests that new wells can be placed to further drain bypassed oil from the LF reservoir (Fig. 8).
In the EI 330 Field, one horizontal well has been drilled into bypassed pay identified by 4D and detailed reservoir characterization studies in the spring of 1994 (Anderson, et al, 1995). It has produced more than one million additional barrels, so far, from another fault block of the LF that had been in production since 1972. The cumulative prior production from this sand had totaled 1.2 million bbl. The well is still producing approximately 1000 bbl/d two years later.
Such results lead us closer to a future in which 4D technologies
off a "total quality control" system of reservoir management
and development. In a more fundamental sense, 4D technologies
result in an improved understanding of the hydrodynamic processes
that really occur inside reservoirs during drainage-- and that
has to help the business.
References:
Anderson, R.N., et al, 4D Seismic Helps Track Drainage, Oil and
Gas Journal, March 27-April 3, 1995.
He, Wei, 4D Seismic Monitoring, Hydrocarbon Drainage, and Geopressure
Prediction in the Offshore Louisiana Gulf of Mexico Basin, Doctoral
Dissertation, Columbia University, 1996.
Kott, Gary, Is This Rocket Finally Off the Pad?, Petroleum Engineer
International, January, 1996, p. 33.
Mack, Toni, The Fourth Dimension, Forbes, November 6, 1995, p.
350.
Authors' bionotes:
Wei He recently completed his Ph.D. in 4D Seismic Inversion
at Columbia University, and is now Chief Geoscientist of the Lamont-Doherty
Earth Observatory 4D Technologies research group. Roger Anderson
is Director of Applied Earth Sciences at Lamont, and Head of the
4D Technologies Group. Liqing Xu is Chief Computer Geoscientist
with the Lamont 4D Technologies Group, and previously worked for
the Global Software Corporation of China. Albert Boulanger
is Head of Technical Development for the Lamont 4D Technologies
Group, and before that, he worked in expert systems, artificial
intelligence and human interface design at BBN, Inc. Randall
Neal is CEO and President of 4D Technologies, Inc., a company
created by Columbia University to commercialize the products from
the Lamont 4D Technologies Group. He is past-President of DPC&A,
providers of economic and risk modeling software for the oil industry.
Billy Meadow is Vice President of Business Development
for 4D Technology, Inc., having recently served in a similar capacity
with BBN, Inc.