Geology and reservoir description of 1Y1 reservoir

Geology and reservoir description of 1Y1 reservoir, Oso Field, Nigeria using FMS and dipmeter LAIRD

B. T H O M P S O N & J. W. S N E D D E N

Mobil Exploration and Producing Technical Center, 13777 Midway Road, Dallas, Texas 75244-4390 (e-mail: [email protected]) Abstract: Mobil Producing Nigeria's Oso Field produces 120 000 barrels of oil per day from late Tertiary sands in the Niger Delta offshore. The reservoir consists of anastomozing fluvial deltaic sands which are difficult to correlate from well to well using traditional log correlation techniques. Image log data from Formation MicroScanner (FMS) logs and dipmeter data tied directly to cores permit detailed analysis of channel thickness, orientation, and distribution. This geographic understanding of the sand bodies enables geoscientists to construct a detailed 'plumbing' diagram of this complex reservoir, thus allowing for better exploitation. Studies describing the sandstone reservoirs in Nigeria are relatively scarce. Key publications by Weber (1971) and Doust & Omatsoli (1990) describe the sandstones producing in the central Niger delta region. With the exception of a brief discussion in Odior (1992), little has been published on the eastern portion of the Niger delta offshore. One of the major constraints upon investigation of these sandstones has been a lack of full-diameter core and modern image logs. However, over 4000 ft of core in eight wells was taken in the Oso Field, and FMS or dipmeter logs were run in 14 wells for the purpose of reservoir characterization (Fig. 1). Analysis of the cores, combined with dipmeter and borehole image data, palaeontology, and well-log correlations from Oso Field, has led to development of a highly constrained depositional and sequence stratigraphic model for the major Oso Field pay zone (the 1Y1 reservoir). This model provides a foundation for reservoir evaluation, zonation, volumetric estimation, and description of fluid flow/gas injection potential in this new producing field.

General description of the Oso Field

Depositional setting The Oso Field consists of two reservoir horizons, the upper 2Y2 and lower 1Y1 zones. These sands occur in the Miocene Biafra Member of the Agbada Formation (Fig. 2). The 2Y2 zone consists of shallow marine sands which are continuous across the field and thin from 175 ft

in the north to about 50 ft in the south (Fig. 3). The 2Y2 is of minor importance to the Oso Field volumetrically and will not be described in detail. In contrast, the 1Y 1 reservoir interval (Fig. 4), with typical thicknesses of 500-800ft (TVT), formed in a deltaic system intermediate in terms of wave, tide, and fluvial processes. Characteristic depositional facies include distributary channelfill, tidal creek, tidal flat, lagoon, mouth bar, delta front, and prodelta. These facies were initially described from conventional cores and were integrated with both standard wireline data and more detailed wireline log data sets available from dipmeter and Formation MicroScanner logs. The facies are summarized on Table 1 a n d briefly described below. The highest reservoir quality is found within distributary channel sands (facies 1), with pebblymassive lithologies exhibiting 10-plus darcies of permeability; cross-bedded variants have somewhat lower values. Channel-fills are typically 30-90 ft thick successions with sharp, erosional bases and sharp to gradational tops. Other depositional facies include tidal creek, mouth bar, delta front, and tidal flats sands. These reservoirs exhibit a strong tidal signature, with rhythmic alternations of shale and sand reflecting daily variations, and obvious bundling of sets related to spring/neap variations. Porosities are slightly higher, but permeabilities are lower in these units than in facies 1. Non-reservoir lagoonal/interchannel shales often exhibit soft-sediment deformation, implying rapid deposition and topographic relief related to steep channel-margin slopes. While these shales often register horizontal permeabilities

From: LOVELL, M. A., WILLIAMSON,G. & HARVEY, P. K. (eds) 1999. Borehole hnaging: applications and case histories. Geological Society, London, Special Publications, 159, 239-257. 1-86239-043-6/99/$15.00. @~The Geological Society of London 1999.

240

L. B. THOMPSON & J. W. SNEDDEN

Fig. 1. Structural contour map on top of the 1Y1 reservoir zone and well locations, Oso Field, Nigeria. Contour interval is 100 ft.

in the range of 10-100md, special core analyses indicate that these shales are effectively impermeable in a vertical sense. Log correlation indicates that these thick (5-50 ft) shales are generally laterally discontinuous, due to incision and cutouts by distributary channel-fills. This ancient deltaic system probably had some resemblance to the modern Niger delta, which is also a mixed tide-and wave-dominated system. However, the dominance by distributary channel-fills and coarse grain size of the channel bedload suggests that the ancient 1Y1 delta was more fluvially-dominated than its modern counterpart. This may be due to a higher sediment supply in the Late Miocene than at present. In fact, many of the eastern Niger delta distributaries today are inactive or less active than earlier in the Holocene and Pleistocene (see N E D E C O 1961 ; Oomkens 1974). In addition to the depositional complexity noted above, several intra-lY1 faults have been penetrated, and associated fractures and deformation bands described. These features do not

appear to be initially affecting producibility of the field but may become more important as the field reaches later stages of production.

Dipmeter and borehole image data sets As the Oso Field was drilled, the stratigraphic complexity was recognized after difficulty in correlating logs from the first few,wells. Once a commitment was made to understand this reservoir in detail, several wells cut conventional core, and logging with electrical imaging tools (notably, the Formation MicroScanner (FMS of Schlumberger)) was done to extrapolate environments away from cored intervals. As the reservoir model was developed, the later stage drilling did not need the extra detail of these dense data sets. More economical drilling was done with oil-based muds, and the Oil Based Dipmeter Tool (OBDT) of Schlumberger, was used to get directional data in order to define sand body geometries.

GEOLOGY AND RESERVOIR DESCRIPTION OF OSO FIELD

Age

Formation

Members

241

Pay

Recent j

_

Pleistoc.

Benin

Qua Iboe

ZON~ I

Ptiocene ............

~NE 4

Agbada Biafra

___/ ZOnE

Miocene

7-

Z0~[ 7

10~

Miocene

Akata

11

Oliqocene Fig. 2.

Summary stratigraphic column for the Oso Field and Nigeria offshore.

The 'standard' dipmeter tools used by the logging industry gather the same data as the electrical imaging tools; they just have fewer sensor buttons and so gather less data. Typical data collected for the Oso Field by these tools are shown on Fig. 5. The first four columns show a data set collected by the OBDT tool; only four curves, each orthogonal to its neighbors, are gathered. The fifth column shows the raw FMS data for one p a d - 16 button curves are displayed. The sixth column shows a false-color image of the FMS data. The right-hand column shows several dips which were hand picked from the image data and presented as a 'tadpole' plot. The 'body' of the dip 'tadpole' indicates the dip magnitude from 0-90 ~ and the 'tail' points to a compass direction indicating the azimuthal orientation of the dipping plane. Any OBDT dips picked from this interval would be generated by mathematical algorithms which look for planar surfaces intersecting the four button curves. Such algorithms work well in sediment where there are consistent resistive patterns such as the lower half of the display on Fig. 5. When the sediments consist of more

complex patterns of electrical conductivity, the computed dip results become less reliable. In the upper portion of Fig. 5 there are some intervals of apparent soft sediment deformation (for about 2 ft around 12 716 ft). Interpretations can be made from the image which are impossible from the OBDT curves. Oriented dense data sets from dipmeters and amaging tools can be used to investigate a broad range of geologic f e a t u r e s - structural as well as sedimentological. When used in combination with other well data such as cores and conventional logs, these data can be invaluable in defining complex depositional systems such as the Oso Field.

Dipmeter/FMS Interpretion Structural Information Regional structure dip, faults, and fractures are examples of the types of structural data which can be derived from dipmeter and image data sets. Figure 6 shows the lower part of an

242

L. B. T H O M P S O N & J. W. S N E D D E N

o 'm

o

..~

GEOLOGY AND RESERVOIR DESCRIPTION OF OSO FIELD

243

Fig. 4. 1Y1 reservoir depositional model.

Table 1. Oso 1 II1 jacies and characteristics Facies

Zone

Characteristics

Palaeoenvironment

1

1Y1

Course-medium-grained sandstone, locally conglomeratic. Cross-bedded variant known previously as Facies 2. High permeability and moderate porosity.

Distributary channel-fill

I Y1

Fine-to very-fine grained sandstone, often interbedded with shale in rhythmic fashion. Shales usually 20% of gross interval. Heterolithic, occasional high permeability.

Tidal creek channel (TC) And sandy tidal flat (TF)

4

1Y1 & 2Y2

Very fine-grained sandstone, interbedded with Shale (especially at base). Flat and low angle. Bedding dominant physical structure. Moderate porosity and permeability.

Delta front/shoreface

5

IY1

20-50% shale. Highly laminated in rhythmic fashion. Few burrows, abundant slumping and soft-sediment deformation. Low net to gross. Non-reservoir.

Lagoon/bay

6

1Y1

Shale, with thin siltstones. Upward coarsening pattern. Non-reservoir.

Prodelta

7

1Y1

Medium-to fine-grained sandstone, well-sorted, compound unidirectional cross-bedding, climbing ripples. High porosity and moderate permeability.

Mouth bar

amalgamated channel sequence (from 1085510875ft) underlain by thin-bedded silts and shales (10 875 to bottom of image). These thinbedded, originally horizontal sediments are usually best for using indetermining regional structural dip. In the illustrated case, however, these dips should not be used as indicators of structural dip, since they occur directly under a

channel sand. The deposition of the sand was rapid and may have differentially compacted the underlying silts and mud, thus giving an incorrect orientation for structural dip. The preferred location for picking an unbiased regional dip is in thin-bedded material which is deposited on top of the main deltaic section and is thus unaffected by the channel sands.

244

L. B. T H O M P S O N

& J. W. S N E D D E N

d o

"-a

..a r.,t3

t'-I r

9

E 9

9

o

c,..)

G E O L O G Y A N D R E S E R V O I R D E S C R I P T I O N OF OSO FIELD

245

, ,..,q

~

246


2

9

z 9

O

+";'

r.~

+..a ,~"O

tt3

9

8.-~ o~

.-~ .~ ~-.~

9

0

G E O L O G Y A N D R E S E R V O I R D E S C R I P T I O N OF OSO F I E L D Table 2. Structural dip for 3 0 s o wells Well number

Structural dip (above 2Y2 zone)

Structural dip (below 2Y2 zone)

24C

3 degrees at azimuth 15


28E



31D

1 degree at azimuth 300


247

the fault are generally dipping to the southeast, whereas the sediments above the fault are dipping to the northeast. While image data allow for direct observation of a fault, such a feature often affects sediments for several tens to hundreds of feet a r o u n d a n d thus is frequently found by a dipmeter.

Faults are often seen by the F M S (Fig. 7). The faulted surface is highlighted by a c o n t i n u o u s line on the image, and its orientation is represented on the tadpole plot by the red features at a b o u t 10404ft which indicate an orientation of 70 ~ dip to the N N W . The sediments below

Fig. 9. Fault indicated by OBDT on Fig. 8 seen in core. Note that adjacent sediments have been disturbed by fault movement.

Fig. 8. Fault at about 11 040 ft in the Oso 24C well. Note the increasing dip magnitude above the fault highlighted by the large arrow. Low angle bedding above is flat, but below the fault is dipping northward (below 11 085 ft).

Fig. 10. A fault with associated cement-filled fractures, Oso 24C well. Dark sand on the right side of the core is oil soaked. The sand in the upper left portion of the core contains no oil.

248


OBDT data from three wells in the Oso Field show faults (Table 2). In the 24C well, two faults can be documented with the OBDT. The lower fault is also present in the conventional core (Figs 8 & 9). Note that the dipmeter fault indicators are necessarily more indirect than the image indicators; one needs to observe a change in structural dip or a steepening of bed dips adjacent to the fault (note arrow on Fig. 8) to make such an interpretation with dipmeter data. Figure 10 shows a fault and associated cement-filled fractures in the core from the Oso 24C well. The fault is a thin fracture zone in which the individual sand grains are crushed as a result of the shear stress. This type of fracture can occur in a course-grained sand and results in a local permeability barrier. Note that the sand is oil-soaked to the right of the deformation band and shows no oil to the left. This type of feature should show up on an FMS electrical image, since the microresistivity of the sand

should be quite different across the deformation band. The OBDT data consist of too few microresistivity curves to provide an interpretable pattern. As a result, the OBD would not show these fractures.

Sedimentological information Depositional environments Depositional facies are often seen easily with the FMS images but must be carefully inferred from the dipmeter and are best corroborated with conventional core. An interval in the 18 well (Fig. 6) shows an FMS image containing higher values of microresistivity above 10 875 ft and lower values below that point. The gamma ray curve seen in the rightmost column shows a blocky sand signature consistent with a channel deposit. Conventional core shows that the sandy

Fig. 11. Coarse grained, pebbly sand in upper channel of the 1Y1 zone, Oso 8 well. The coarse 'graininess' of the FMS image reflects the sediment texture.

G E O L O G Y A N D R E S E R V O I R D E S C R I P T I O N O F OSO F I E L D

249

-A

9 9

,.a

be4

250


..-.Z

t----

ct~

q=

...a

,~o

xz~ ...o .,..~

e,,.)

o~ o.~

GEOLOGY AND RESERVOIR DESCRIPTION OF OSO FIELD interval does consist of two amalgamated channel sands underlain by inter-distributary silts and shale. When the channels are composed of very coarse-grained sand, even the grainy texture is imaged by the F M S (Fig. 11). By comparison, small-scale tidal bundles (with bedforms as small as 2-3 cm) are clearly seen by the F M S (Fig. 12). Finally, shales may be contorted by soft sediment deformation, burrowed, or left intact as thinly-bedded hemipelagic sediments. All these details can be seen with the FMS but must be inferred at best by the OBDT. When making depositional facies interpretations from dipmeter data, calibration to con-

251

ventional core is often a necessity, since the four densely sampled microresistivity curves may not provide sufficient details of the sediments. The Oso 18 well provides an example of this limitation (Fig. 13). The image data indicate channel foreset beds (10800ft and below) and lateral accretion channel fill features present from 10 800-10 790 ft. The gamma ray curve (the sparsely-sampled 'lazy' curve in the central column) shows only the presence of the channel sand and overlying shale. The dipmeter data (the four 'active' curves in the middle column) show the sand and shale, but miss the subtleties necessary to distinguish the channel forest beds from

Fig. 14. (A) Oso 3 ID well, 1Yl sand, zone 6. Foreset beds show channel flow to the southwest. (B) Oso 31D well, 1Y1 sand, zone 2. Channel fill pattern between 11 320 and 11335 ft shows channel flow direction of southwest. Note that 8~ of structural dip have been removed.

252


the channel fill features. Thus, while the dipmeter microresistivity curve data often yield a more detailed picture of the sediments than gamma ray data, this is not universally true.

Sand body geometries Besides helping to recognize depositional facies, dipmeter and image data also allow one to make palaeocurrent observations and thus predict sand body geometry. The most commonly preserved palaeocurrent indicators are the channel trough crossbeds, which show downstream direction, and the lateral accretion bedding associated with channel abandonment, which show a cross-stream direction. FMS images show both

current indicators (Figs 6 & 13). The lower portion of the channel sands contains crossbed foresets seen as brighter (less-silty) intervals with higher angle bed dips (represented by higher amplitude sinusoids on the images and also seen in the dip tadpoles). These beds show the direction of flow in the channel, while the upper portion of each channel section shows a cross-channel pattern of fining-upward sediments (increasingly silty images with lower angle sinusoids and dip tadpoles which decrease in amplitude upward and are orthogonal to the foresets). Dipmeter microresistivity curves undersample both channel foreset and channel features, but may gather enough data to allow for an accurate interpretation fill (Fig. 14). There are

Fig. 15. 'Classic' dipmeter interpretation in the Oso 8 well. A structural dip of 8~ magnitude, azimuth 250~ is removed from the rotated Schmidt vector plot. Tidal creek channels (1) above the distributary channels are shown oriented in an ESE direction. The upper channel (2) flowed SSW and shows a good channel fill pattern at the top. The lower channel (3) was flowing to the SSE.

GEOLOGY AND

R E S E R V O I R D E S C R I P T I O N O F OSO F I E L D

253

~A

o

O

o

3

Z ,d

254

L. B. T H O M P S O N

& J. W. S N E D D E N

r.,r

,..a

8

,....,

8

r.~ --1

Q

8

~5 ~Z

o,

GEOLOGY AND

R E S E R V O I R D E S C R I P T I O N O F OSO F I E L D

255

o

.o +.a

9

9 tg-)

9

7~

9

256


other depositional features which can produce the same dip patterns, so detailed interpretations such as these should be done with great care when only presented with dipmeter data. In the Oso Field, conventional core was available, and specific correlation of dips to core sediments in the Oso delta also allowed for orienting depositional d i r e c t i o n - the tidal creeks with their thin-bedded tidal sediments (Fig. 12). Figure 15 shows a 'classic' dipmeter interpretation of river channel and tidal creek dip signatures. Note that the tidal creek dip signature consists of good quality dips all with the same orientation reflecting the thin-bedded, consistent nature of the sediments. The tidal channels are often roughly orthogonal to the main distributary channel sands.

Quantitative reservoir analysis The dipmeter and F M S / F M I tools collect relative microresistivity data, not true resistivity. These data are not consistent from well to well, and are sometimes not consistent within one well or even between pads on a single tool. If manipulated carefully, however, these data can be calibrated, resampled, or both, in order to generate quantitative information for reservoir analysis. Figure 16 shows a portion of the Oso 18 well. An apparent channel sand is shown on the gamma ray curve (on the left of the figure) from 10 785-10 880 ft. The sand present in the 10 90510 915 interval is not as clean as the channel, but a 10ft thickness is indicated. A more detailed image of that sand is seen on the F M S data. By setting cutoffs on the image data to model sand (white), silt (gray), and shale (black; see insert at upper right), the sandy interval is documented to contain only 49% net sand over the 10ft. This technique is best done by calibrating the FMS image to conventional core to get the most accurate cutoffs, then extrapolating away from the cored intervals. Another technique to quantify physical properties using image logs is to use satellite false color technology (Fig. 17). Every wireline lot curve consists of a set of values (e.g. 0-150api units for a gamma ray log). Any set of log values can be used to create a false color band (say, of red color). By convolving three such data sets (for example, the microresistivity data of an FMS with gamma ray, density, or sonic log) into one image, one can create a false color display called an RGB (for Red, Green, and Blue) image or an HLS image (for Hue, Lightness, and Saturation). Such images can also be created

from derived curves such as water saturation or porosity. By creating this presentation with the image, a more accurate picture and measure of the small-scale distribution of the attribute can be measured.

Summary Analysis of depositional environments coupled with palaeocurrent information to constrain sand body geometry allows one to construct a detailed stratigraphic model of the Oso delta. Maps of each stratigraphic layer are constructed using the image and dipmeter data, conventional core data, and wireline data available at the wells (Fig. 18). These layer maps show a set of facies 'polygons' existing in a distribution which is similar to the modern Niger deltaic system (cf. N E D E C O 1961). Channels border tidal flats and tidal creeks which split and subdivide downstream. In addition, measurement of the channel-fill dimensions demonstrates that they have widths and thicknesses falling in the range expected for straight to moderately sinuous channel systems of the modern and Miocene Niger delta (cf. Weber 1971). Most 1Y1 channel-fills have thicknesses of 30-90ft (TVT), which approximates the depth of modern distributary channels'of the Niger delta ( N E D E C O 1959; 1961). Mapped (non-amalgamated) sand body widths typically range from 1000-10 000 ft. This detailed depositional model has been coupled with petrophysical and engineering data for each facies to quantitatively map volumetrics, to predict flow connectivity within the reservoir, and to guide producibility by locating gas injectors at optimum positions (Fig. 1). Thus, the F M S / F M I image data serve as the connecting link between core and wireline curve data to allow for analysis of disparate data sets on a common basis. The image data bridge a broad range of scales for reservoir a n a l y s i s from thin-section to basin in scope. Emerging techniques of using the images quantitatively further their utility in reservoir characterization. The authors would like to acknowledge the aid and insights of several individuals: A. K. Bhatia, E. B. Ogiamien, and R. M. Vaught, all of MPN; T. W. Cooley and B. E. Welton (MEPTEC). The previous work of R. D. Kreisa and others of MEPTEC provided a firm basis for this reservoir characterization effort. The support of G. K. Baker and Mike Croft (MEPTEC) and V. K. Oyofo, D. O. Lambert-Aikhionbare, and J. Y. K. Blevins of MPN is also gratefully acknowledged.

G E O L O G Y A N D R E S E R V O I R D E S C R I P T I O N OF OSO F I E L D

References DOUST, H. & OMATSOLA, E. 1990. Niger Delta. In: EDWARDS, B. D. & SANTOGROSSI, W. (eds) American Association of Petroleum Geologists Memoir, 48, 201-329. NEDECO (Netherlands Engineering Consultants). 1959. River studies and recommendations on improvement of Nigeria and Benue. North Holland Publishing, Amsterdam.

257

1961. The waters of the Niger Delta, Report on an Investigation. Unpublished company report, 316. ODIOR, G. E. 1992. Ubit Field. Nigerian Association of Petroleum Explorationists, Transactions. OOMKENS, E. 1974. Lithofacies relations in the Late Quaternary Niger delta complex. Sedimentology, 21, 195 222. WEBER, K. J. 1971. Sedimentologic aspects of oilfields in the Niger Delta. Geologie en M~inbow, 50(3). 559-576.