4D SEISMIC HELPS TRACK DRAINAGE, PRESSURE COMPARTMENTALIZATION

Roger N. Anderson, Albert Boulanger, Wei He, Y.F. Sun, Liqing Xu
Lamont Doherty Earth Observatory
Palisades, N.Y.
David Sibley
Chevron U.S.A. Production Co.
Lafayette, La.
John Austin, Richard Woodhams
Pennzoil Exploration & Production Inc.
Houston
Richard Andre, Kent Rinehart
Texaco Exploration & Production Inc.
New Orleans

3D seismic technologies have progressed to the point that new reservoir production and engineering applications are being discovered yearly. In particular, 4D, or time-lapse, seismic monitoring offers the possibility that computer control of field development may someday be the industry norm.

We see a future in which continual refinements to the 4D seismic solution are strongly coupled to forward and inverse models of different attributes (reservoir, elastic, and acoustic), all of which are used to understand changes observed in pressures, temperatures, and fluid contents coming into control centers from the monitoring of all wells in a field in near real-time. Teams of production engineers, geologists, and geophysicists will evaluate these diverse datasets. These professionals, in turn, will control reservoir production through a "visualization nexus" that allows this interaction to occur.

As a beginning, we examined a partially depleted oil reservoir along the Eugene Island Block 330/338 boundary. Texaco and partners Chevron, Pennzoil, Exxon, Mobil, and Cockrell Oil Corp. had evaluated the production history of this LF sand reservoir, and planned a well into it. In spring 1994, a 1,200 ft horizontal well was drilled. Production exceeded 1,500 b/d in August 1994, and the well was still flowing over 900 b/d in March 1995.

Global Basins Research Network (GBRN) modeled this reservoir to test and calibrate its 4D seismic analysis technologies. These 4D technologies examine the changes in seismic amplitudes between repeated 3D surveys to track the movement of fluids caused by drainage within a given reservoir.

In the 4D analysis, two vintages of 3D seismic data over the LF reservoir were merged and spectrally filtered, and high amplitude regions (HARs) were "grown." The similarities, differences, and connectivity of HARs were then tracked in space and time. We identified amplitude "dim-outs," possibly caused by pressure drop or water invasion, amplitude "brightening" caused by GOR increases, and sustained high amplitude regions that might represent bypassed hydrocarbons. We have computed a 2D acoustic model of the observed seismic amplitude changes, varying reservoir pressure and oil/gas/water mix to account for the observed changes in reflectivity over time. In addition, a subtle fault boundary separating the two wells in the reservoir into pressure compartments was identified by both the company and GBRN 4D analyses.

EUGENE ISLAND 330 FIELD

Observational and modeling results suggest that migration of hydrocarbons has occurred very recently in EI 330 field off Louisiana.1 2

This natural fluid dynamics, combined with the large volumes of hydrocarbons produced since 1972 (500 million bbl of oil equivalent), led GBRN to test whether the effects of drainage could be quantified by analyzing changes in seismic amplitudes between 3D surveys conducted over the field a number of years apart.

In EI 330 field, four vintages of 3D seismic surveys have been recorded: in 1985, 1988, 1992, and 1994 (Fig. 1)(77822 bytes). We report here on the first of an ongoing set of time-lapse studies, that between the 1985 and 1988 seismic datasets. We concentrated on an isolated segment of the LF reservoir along the boundary of El blocks 330 and 338; a sand confined on all sides by splays of the "Red" and "Blue" fault zones bounding the Eugene Island minibasin.3 A re-evaluation of production histories and seismic interpretations has been under way for several years by Texaco and Chevron in EI 338/339 and Pennzoil, Exxon, Mobil, and Cockrell in EI 330. This reservoir study, aimed at identifying bypassed pay, provided the fluid withdrawal history required to compare against changes observed over time by GBRN in high amplitude seismic events within the reservoir.

THE LF RESERVOIR STUDY

The LF sand extends east-west over a 240 acre area along the property line between El blocks 330 and 338 at a depth of 6,900-7,000 ft (Fig. 2)(36244 bytes). A large, high amplitude seismic anomaly (bright spot) is imaged from 1.98 to 2.06 see of two-way travel time in both seismic surveys. Average porosity is 27%, water saturation is 35%, and permeability is 500 md. There are proven reserves of up to 3.3 million bbl of oil in 6,600 acre-ft, and probable reserves of 0.9 million bbl of oil in an additional 1,800 acre-ft.

Two wells have produced a cumulative total of 1.2 MMBOE from the LF reservoir between 1972 and 1988: the easternmost A5 well 450,000 bbl of an average 2,000 GOR oil and the westernmost All well 370,000 bbl of 5,000 GOR oil (Fig. 3)(68487 bytes). The A5 well was producing at a formation pressure of 2,143 psi just before it sanded out in late 1987, whereas the pressure in the All had fallen from 2,575 psi in 1987 to 1,123 psi when it was shut-in in 1990. Initial reservoir pressure in both was approximately 5,100 psi in 1972.

SUBTLE BOUNDARY

Though production began in 1974 in both the A5 and All wells, significant seismic amplitudes remained in both the 1985 and 1988 seismic surveys to the east of the A5 well. The pressure anomaly between the A5 and All wells was of considerable concern, however, to the partners involved in planning a drilling program targeting bypassed pay in the reservoir. The risk was that pressure depletion might be found to the east of the A5 well if the All pressure drop turned out to be representative of remaining condition in the entire LF reservoir. Rock property research and modeling convinced the partners that the reservoir was exhibiting 4D effects. That is, the A5 and All wells were no longer in pressure communication. This would explain the dimming-out of the All compartment before either 3D survey was acquired. It was possible that a sealing fault not resolvable in the seismic data (now labeled "A" in Fig. 2)(36244 bytes)f existed to separate the two wells into pressure compartments within the LF reservoir.

A8ST HORIZONTAL WELL

The A8ST was drilled in spring 1994 into the eastern compartment of the LF reservoir to the east of the A5 well by Texaco, as operator of EI 338 for the partners (Fig. 4)(76481 bytes). Pressures were found to be 2,500 psi, somewhat higher than the last bottomhole pressure in the A5 well from 1987. A 1,200 ft horizontal borehole was then sidetracked right along the property line between EI 330/338 and the well completed. Flow rates reached a maximum of more than 1,500 b/d of low GOR oil during production tests in August (Fig. 5)(29465 bytes). By February 1995, the well had stabilized at approximately 900 b/d from a 16/64 in. choke, and cumulative production has reached more than 250,000 bbl of oil. No water has been produced.

PREDICTING DRAINAGE EFFECTS

For several years, laboratory measurements of the changes in acoustic reflection coefficients caused by changes in oil, gas, and water mixes, as well as effective pressure decreases, have been predicting that seismic monitoring of enhanced recovery processes such as water and steam floods should produce noticeable acoustic differences over time.4 Successful monitoring of oil/water boundary movements during waterfloods and steamfloods has been carried out in several fields worldwide.5 6

In order to quantify the seismic effects expected from the production of hydrocarbons in the LF reservoir, we conducted a seismic amplitude modeling study by varying the acoustic properties across a hypothetical, constant thickness sandstone reservoir undergoing pressure depletion (Fig. 6)(87916 bytes). Tuning was avoided in this modeling study, although the thickness of the LF reservoir is indeed near tuning for a 30 Hz acoustic source (a tuning thickness model is presented below). Gas/oil ratio (GOR), oil/water contact (O/W) , and gas/water (G/W) changes, as well as the increase in differential pressure caused by depletion, are all shown to affect seismic amplitudes through changes in acoustic impedance across the reservoir boundaries.

Differential pressure (lithostatic minus reservoir fluid) increases as hydrocarbons are drained from a reservoir, reducing seismic amplitudes in a gas sand7 (Fig. 6)(87916 bytes). If, however, a GOR increase occurs during production, as with the formation of a secondary gas cap, seismic amplitudes are predicted to increase, or "brighten" over time, because the acoustic affects of the fluid change dominate over the pressure depletion affects. Similarly, seismic amplitudes can dramatically decrease, or "dim-out," if in addition to pressure depletion, the O/W contact migrates in a reservoir. Here, the amplitude decrease is caused by pressure depletion and is further enhanced by the drop in impedance caused by the replacement of low velocity oil and/or gas by relatively high velocity water.

In addition, bypassed hydrocarbons can be identified by "near-zero" changes in high seismic amplitude regions over time, if there has been little increase in differential pressure. Also, areas of sustained high seismic amplitudes can be the highest permeability, drainage pathways, through which hydrocarbons are moving to get to wellbores. Thus, the detection of the oil, water, and gas volumes remaining in a reservoir might be derived from observing changes in seismic amplitudes over time, after carefully calibrating these observations to changes measured within wells.

Next: 4D analysis technique