styles of heterogeneity in dolomitized platform carbonate reservoirs: examples from the central...

Journal o f Petroleum Science and Engineering, 6 ( 1992 ) 301-339 301 Elsevier Science Publishers B.V., Amsterdam

Styles of heterogeneity in dolomitized platform carbonate reservoirs: Examples from the Central Basin Platform of the

Permian Basin, southwestern USA

Noel Tyler, R.P. Major, Don G. Bebout, Charles Kerans, F. Jerry Lucia, Stephen C. Ruppel and Mark H. Holtz

Bureau of Economic Geology, The University of Texas at Austin, Austin, TX 78713-7508, USA

(Received June 15, 1991 ; revised and accepted August 28, 1991 )

ABSTRACT

Tyler, N., Major, R.P., Bebout, N.G., Kerans, C., Lucia, F.J., Ruppel, S.C. and Holtz, M.H., 1992. Styles of heterogeneity in dolomitized platform carbonate reservoirs: Examples from the Central Basin Platform of the Permian Basin, southwestern USA. J. Pet. Sci. Eng., 6: 301-339.

The Permian Basin of West Texas and southeastern New Mexico, southwestern U.S.A., is the premiere oil basin of the United States. At discovery, reservoirs in this prolific province contained more than 100 billion barrels of oil, almost a quarter of all the oil discovered in the U.S. Almost half of this resource (43%) was contained in a single reservoir type, dolomitized platform carbonate. Dolomitized platform carbonates were deposited on shallow shelves fringing the basin and on a horst block, the Central Basin Platform, that divides the basin and separates sites of deep-water siliciclastic sedimentation in the adjacent subbasins. The Central Basin Platform hosts many large combined structural/stratigraphic trap reservoirs in dolomitized platform carbonates. These range in size up to four billion barrels of original oil in place.

Integrated geoscience and engineering characterization of four of these fields; Dune, Emma, Penwell, and Taylor-Link, allows comparison of the styles and scales of heterogeneities that influence recovery in this reservoir type. Facies compo- sition and architecture exert fundamental controls on paths of fluid movement during production. Principal facies are extensive, deep subtidal fusulinid wackestones that shoal upward into shallow-subtidal and intertidal packstones and grainstones in which the dominant productive facies are grainstone bars and shoals, and shorezone systems that are dis- sected by dip-oriented tidal systems. Rapid lateral facies changes together with highly cyclic shoaling sequences result in pronounced permeability variations both laterally and vertically in the section.

Superimposed on the depositional framework is a multi-event diagenetic overprint. These carbonate reservoirs are thoroughly dolomitized and partly cemented by sulfates. A post-burial leaching event increased permeability in some parts of these rocks. Karst processes have a large affect on reservoir quality in the southern part of the Central Basin Platform. Even though these carbonate reservoirs have undergone substantial diagenetic modification depositional facies still exert the primary control on remaining, and in particular, mobile, oil saturations.

Introduction

By any standard the Permian Basin of West Texas and southeastern New Mexico, U.S.A., is a petroleum province of world class propor- tion. It is estimated that at discovery reservoirs in the basin, which are of Paleozoic age, contained a total of 105.7 billion barrels of oil (Bbbl), a volume that amounts to 23% of the total U.S. domestic resource. Cumulative pro-

duction to 1985 totals 25.3 Bbbl and proved reserves are estimated to be 5.9 Bbbl (Fig. 1 ). Conventional recovery at current technological levels will be a little less than 30%, and a resource of 74.5 Bbbl of oil will remain in Permian Basin reservoirs at depletion (Tyler and Banta, 1989).

By far the greatest volume of oil discovered in the basin was contained in dolomitized platform carbonates, which are of Permian age.

0920-4105/92/$05.00 © 1992 Elsevier Science Publishers B.V. All rights reserved.

302 N. TYLER ET AL.

•

~ CUMULATIVE i~~/////~,/~ PRODUCTION

.o . , . . . ,o , , - ,n . , , . c .

UNRECOVERED " / ~ ,

xP .AT,O. - / ;

I ~ / / / / / H 1777";I Dolomitized platform carbonate ~ • v / / / , , ' / n ~ Open shelf carbonate ~ ~ / ~ /,/// i . w _ ~ / vJ-J77.7~ Submarine fan sandstone ~ "~ , ~

' ~ ~ Ramp/unconformity related ~ 3.5 "~, 7 ~ Platform margin carbonate

Fluvial/parallicsandstone C. Unte¢overed mobile oil D. Res idua l oi l ITITITfl Atoll/mound carbonate

QA15622c

Fig. 1. Oil resource distribution in the Permian Basin (center) and allocation of these resources to major producing depositional systems, which are all of Paleozoic age. Numbers are volumes of oil in billions of barrels (Bbbl).

These deposits contained at discovery 45 Bbbl of oil, or 43% of the oil discovered in the Permian Basin (Fig. 1 ). This volume ofoil was far greater than that contained in any other depositional system. Of the remaining systems, approximately equivalent amounts of oil were discovered in platform margin, ramp, and open shelf carbonates, and submarine fan sand- stones. With one major exception, low-recovery submarine fan reservoirs, projected conventional ultimate recovery at current technological levels reflects the distribution of original oil in place (Fig. 1 ).

The oil remaining in existing Permian Basin reservoirs in addition to reserves is composed of a combination of residual oil and unrecovered mobile oil. Residual oil is immobile at reservoir conditions and requires application

of enhanced oil recovery strategies to mobilize and produce the oil. Miscible flooding using carbon dioxide injection is the favored enhanced oil recovery strategy in the Permian Basin. Equally important, however, is the unrecovered mobile oil remaining in these reservoirs. In contrast to residual oil, mobile oil is that oil that is free to move in the reservoir and is producible through natural reservoir drive mechanisms aided by gas or water injection. Remaining (or unrecovered) mobile oil is the volume of mobile oil that remains in the reservoir after a history of conventional recovery. Prevented from migrating to the wellbore by geologic heterogeneity or permeability strati- fication, the remaining mobile oil is available for recovery by advanced conventional strategies and is calculated by subtracting produced

D O L O M I T I Z E D P L A T F O R M C A R B O N A T E RESERVOIRS 303

i Cochran

Yoakum " ' I

~ × t.LJ, [.LI

~-'-~UWard

Gaines

Andrews i

Cb

_ Emr.

. . . . . . __j :;8 j Wlnkter I r (

j U"I Oem,

~. Crane

Hocktey I Lubbock I

NORTHERN

4- Terry I Lynn

i i

! ' _ _ G I . . . . .

I Crosby

SHELF I i

Dawson ] Borden

i

I i

M(]rhn" 1 Howard

i -ctor Midland

d~

\ I Upton D u n e

Glasscock

Reagan

(::) - - "L ._~ ~ . i~,, ~

r" . \PecOP ~ ~L,-_ "- ' , -~v/ j - - ...... ~:.:

/ "~,o, ,,% " . . \ ~7,,. 9

" \ /4,

I. __ Crockett

,¢

I # / o 20 I I I

i . o 2'0

i

Gorza

-Scurry r

E A S T E R N I

i~ SHELF i _I ___J Sterhng ~-~

4'0

40 mi , ---~

60 km QA 12908

San Andres and Grayburg fields m Fields discussed in this paper

Fig. 2. Middle-Late Permian paleogeography of the Central Basin Platform with the locations of major San Andres/ Grayburg fields and the four fields that are the subject of this paper.

304 N. TYLER ETAL.

oil, reserves, and residual oil from the original oil in place.

The aggregate volume of unrecovered mobile oil in excess of proved reserves in the Permian Basin is estimated to be 30.0 Bbbl (Tyler and Banta, 1989). Graphs illustrating the distribution of the remaining mobile oil and residual oil resource basin (Fig. 1 ) em- phasize the importance of dolomitized platform carbonates as a primary focus for contin- ued conventional and tertiary recovery in the basin. Dolomitized platform carbonates, containing 44% of the unrecovered mobile oil resource and almost 40% of the residual oil resource, are the dominant components in the resource base. In total, dolomitized platform carbonate reservoirs still contain more than 30 Bbbl of unrecovered (mobile and residual) oil. This is a volume greater than the historical production from the basin and almost an order of magnitude greater than reserves remaining in the basin.

The overwhelming importance of dolomitized platform carbonate reservoirs to all as- pects of the resource base in the Permian Basin led to the initiation, by the Bureau of Eco- nomic Geology, The University of Texas at Austin, of an ongoing program of integrated geoscience and engineering characterization of this reservoir type. The objective of the research program was to identify scales and styles of heterogeneity and the controlling effects of these parameters on the recovery of hydrocarbons. Particular emphasis was placed on tar- geting locations of unrecovered mobile oil as a prelude to initiation of advanced secondary recovery. This paper contrasts styles of heterogeneity and production response in four such reservoirs (Fig. 2).

Geologic setting of dolomitized platform reservoirs in the Permian Basin

Permian Basin paleogeography was controlled by Pennsylvanian tectonism that de- formed the Precambrian basement and pre-

Pennsylvanian sedimentary rocks. During Permian t ime sedimentation in the region occurred in two deep-water basins, the Delaware Basin to the west and the Midland Basin to the east, separated by the south-southeast-trending Central Basin Platform (Fig. 2 ). The Cen- tral Basin Platform was the site of shallow- water ramp carbonate sedimentation, whereas the central parts of the Delaware and Midland Basins were the sites of deep-water siliciclastic deposition (Galley, 1958; Ward et al., 1986).

The Permian stratigraphic section on the Central Basin Platform is dominated by Wolf- campian, Leonardian, and Guadalupian shallow-water carbonate strata, many of these now thoroughly dolomitized, and by relatively mi- nor siliciclastic-rich carbonates. Guadalupian carbonates are in conformable and gradational contact with overlying Ochoan evapo- rites and siliciclastic red beds deposited dur-

SYSTEM S E R I E S STRATIGRAPHIC UNIT

Dewey Lake Ochoon Rustler

PERMIAN

PENNSYLVANIAN

MISSISSlPPIAN

Guadolupion

Leonardion

DEVONIAN

SILURIAN upper

ORDOVICIAN

CAMBRIAN

middle lower

Relative production • ~ •

Solado Castile

Capitan I Yates •__ I Seven Rivers ~"

Goat Seep I Queen

Son And

Clear = "~ Sproberry = Fork • ~ . Dean "

Wolf camp Clsco "~o t se

~ , ~ /shoe Atol Strown • Bend

Mississ~pp=a n

DcYOnign •

Fusselman

Montoya Simpson

Ellenburqer •

Fig. 3. Simplified stratigraphic section, Paleozoic Er- athem, Permian Basin, USA. Size of filled circles is proportional to the relative cumulative oil production from each unit. The San Andres and Grayburg Formations contain the most prolific oil reservoirs (modified from Galloway et al., 1983).

D O L O M I T 1 Z E D P L A T F O R M C A R B O N A T E R E S E R V O I R S 3 0 5

ing increasingly restricted marine conditions in the Permian Basin. The Permian carbonates of the Central Basin Platform contain numerous moderate to giant hydrocarbon reservoirs (Fig. 2), the largest of which contained up to 4.0 Bbbl of oil-in-place at the time of discovery. Most Central Basin Platform oil production has been from the Guadalupian San Andres and Grayburg Formations (Fig. 3), which contained an estimated 13.2 Bbbl of original oil in place and which now contain 3.3 Bbbl of remaining mobile oil in fields with cumulative production of more than 3.7 Bbbl (Finley et al., 1990).

Characterization of dolomitized platform carbonate reservoirs in the Permian Basin

Each of the reservoir studies discussed in this paper began with geologic analysis, which was subsequently incorporated into supporting petrophysical, well-log, and production engineering investigations (for more detailed re- ports, see Bebout et al., 1987; Ruppel and Cander, 1988a; Major et al., 1990; Fogg and Lucia, 1990). The fundamental objective of all reservoir studies was the quantification and geographic delineation of original and current oil saturations and the development of strategies for optimal recovery of the remaining oil.

Dune Grayburg reservoir

Dune field, in Crane County (Fig. 2), was discovered in January 1938, and since that time more than 1,200 wells have been drilled in the l l,640-ha (28,764-acre) field area, which is developed on an average well spacing of 9.7 ha (24 acres). The Guadalupian Gray- burg pay zone is approximately 24 m (80 ft) thick. Original oil-in-place across the entire field area is estimated to be 978 million barrels (MMbbl)and through 1987, 171 MMbbl of oil had been produced from the field.

The Dune Grayburg reservoir study focused on the Mobil University Unit 15/16, Univer-

sity Block 30 (Fig. 4). Production in this unit was established in 1938. However, major drilling programs were not conducted until 1954 to 1957, when 8-ha (20-acre) well spacing was

' °

";.i!.~;;/~A

(b)

8 / 9 IO , ; / /

I o o o : o o o ,/ I. ..\° ,,io ". , °

Contour intervoI : I 0 0 Mbbl oil t

/ O Ijrni I ,

0 I k m

14

b

23

Fig. 4. (a) Location of the Dune field reservoir study area. The study area, which concentrated on Sections 15 and 16, is stippled. (b) Isoproduction contours of Sections 15 and 16 prior to waterflooding, illustrating a major difference in oil production between these two sections. The map is based on production information from wells drilled in the initial development program between 1954 and 1957, and only production data from these wells through December 1980 are included so that any production in response to waterflooding is excluded. However, because waterfiooding was initiated in areas bordering Section 15 between 1969 and 1971, some of the production before January 1981 may be in response to these bordering waterflooding operations. Cross section A - A ' is shown in Fig. 5, and cross section B - B ' is shown in Fig. 7.

306 N. TYLER ET AL.

completed in Section 15; between 1971 and 1974 the 8-ha (20-acre) program was completed in Section 16. Between 1978 and 1986 Section 15 was converted to 4-ha (10-acre) well spacing. Water injection began in 1976 in Section 16 and in 1980 in Section 15; there are now 39 injection wells in these two sections.

The availability of cores from several closely spaced wells and modern wireline logs and the production history for each well in the Mobil University Unit 15/16 made possible a detailed geologic and engineering study of this area. In addition, heterogeneity was well dis- played in the unit by significant production in- equalities between Sections 15 and 16 (Fig. 4). The cumulative production from Section 15 is about 10 MMbbl, whereas that from Section 16 is only 2 MMbbl. Furthermore, wells from the same reservoir within Section 15 have yielded widely varying amounts of total production.

Geological setting Dune field is located on the east side of the

Central Basin Platform, on the edge of the Midland Basin. The structure at this field and neighbouring McElroy field to the south is partly controlled by drape over fault blocks of a buried Late Pennsylvanian fault system. Platform subtidal and tidal-fiat carbonates and siltstone accumulated in the area west of the field, whereas slope and basinal carbonates formed to the east. The single thick, dominantly marine cycle of the Grayburg Forma- tion at Dune is equivalent to multiple cycles of subtidal to tidal-fiat sediments farther shelfward.

Facies distribution The San Andres and Grayburg Formations

comprise several upward-shallowing cycles with more open-marine facies at the base of the cycles and more restricted supratidal (pisolite) facies at the top (Fig. 5 ). The Grayburg represents the topmost cycle and overlies sev-

eral similar cycles of the older San Andres Formation.

The Grayburg Formation in the Mobil Uni- versity Unit 15/16 and the surrounding study area of Dune field has been subdivided into three units on the basis of available cores: ( 1 ) the lower unit extending from the base of the Grayburg Formation to the M gamma-ray marker, (2) the middle unit extending between the M marker and A siltstone marker, and (3) the upper unit extending from the A siltstone marker to the top of the Grayburg Formation (Fig. 5 ).

The lower unit comprises fusulinid wackestone in all wells studied. This fusulinid wackestone of the lower unit rests with sharp contact on the underlying San Andres Forma- tion, which at the top is composed of siltstone and pisolite beds in the western part of the area and marine brachiopod-dominated facies to the east. The fusulinid wackestone in the lower unit typically has very low matrix porosity (approximately 5%), and the fusulinids are preserved as open molds or molds filled with anhydrite and gypsum. The contact of the lower unit with the overlying middle unit is also sharp, suggesting a significant geological break. The gamma-ray curve shows a pronounced low-gamma shoulder at this contact, designated here as the M marker (Fig. 5 ), and provides ready correlation throughout the local study area.

The middle unit includes the section from the M marker up to the base of the A siltstone marker. Fusulinid wackestone composes the upper 6-8 m (20-25 ft) of the section and in all wells is in sharp contact with the underlying facies. Beneath the fusulinid wackestone, the crinoid packstone/grainstone facies extends northwestward across the eastern two-thirds of the area, and the vertical-structured facies is distributed across the western one-third. Car- bonate fabrics range from wackestone to grainstone within a single core of the crinoid packstone/grainstone facies; the wells in which this facies is dominantly grainstone are lo-

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS 307

f l 07

5O

IOC

A A' SW NE

1559 1560

1625 1628 pX 1535 1540

PISOLITE LxJ~.33° O ~ l ~ - - ~-" D

:3100 y 3100

L ~ L ~ 3 1 o c ~ . Z l=

m ~ ~;ILTSTONE ~ C . o o - . ,

H -15 320C = " ~ 340U - - ]l l l lr "~-aP. V~J'

Ei - 25 . _ _ _ _ ~ 2 0 0 ] • ~ F U S U L I N I D ~ W A C KESTONE . . . .

VERTICALLY STRUCTURED

FACIES

Porosity - - I f 330(

I k so/° LOg only

LI5% o o.3

r l l l 2 5 % 3300 I , I ~ ~ L

0 0,5 Ikm

I r a i I I

3700

CRINOID PACKSTONE/ GRAINSTONE

E

3800 I

Fig. 5. Facies dip section A - A ' across Dune field. The letters on the right margin of this figure indicate the location of gamma-ray markers. The location of the section is shown in Fig. 4.

cated along northwest-trending bands. Poros- ity is best developed along these grainstone trends.

The upper unit extends from the top of the A siltstone up to the top of the formation. This upper unit comprises fusulinid wackestone at the base, pellet and ooid grainstone near the top, and pisolite grainstone and anhydrite at the top. Siltstone beds are thicker and more closely spaced toward the top of the unit.

In general this facies pattern is that of pisolite facies in the west and ooid and pellet grainstone and fusulinid wackestone in the east. This general facies pattern shifts upward from west to east in the section. This shift represents the eastward progradation of the pisolite facies, having low porosity and permeability (less than 5% and 0.1 mD) , over the more po-

rous and permeable pellet grainstone ( 15% and 80 md), fusulinid wackestone (5% and 0.2 mD ), and crinoid packstone/grainstone ( 11% and 1.31 mD) facies.

Depositional environments Abundant fusulinids, burrows, and carbon-

ate mud indicate that the lower unit of the Grayburg Formation was deposited in normal-marine water below wave base, in low-energy conditions. However, extensive high-energy shoals and tidal fiats equivalent to most of this subtidal section are expected to occur to the west, toward the interior of the platform.

The vertical-structured facies of the middle unit is interpreted to represent a low-energy shallow-water bank composed largely of carbonate mud. Some vertical structures, how-

308 N. TYLERETAL.

(a)

~ Vertically structured facies

~ Crinoid packstone/grainstone oA 4e,5

~ _Mudstone and algal mudstone ~ ~ " / ~ Pisolite groinstone ~ J /

Ooid groinstone Pellet grainstone Fusulinid wockestone

QA 4844

Fig. 6. Dune field depositional models of (a) the MA zone of the middle unit and (b) the CZ zone in the upper unit.

ever, suggest oriented heads of calcareous sponges and blue-green algae. These structures are characterized by abrupt horizontal changes in carbonate textures across the core surfaces. The banks were probably oriented approximately perpendicular to the tidal energy, and they focused higher energy tidal currents between them. Crinoid packstone/grainstone accumulated in channels between the banks

and as tidal deltas adjacent to the bank. De- velopment of lower energy conditions basinward of the bank and tidal-delta trend is indi- cated by the muddier crinoid packstone, which may represent low-gradient slope deposits. Lo- cal low-energy grainstone bars developed on the slope parallel to the bank (Fig. 6a).

The upper unit contains an upward-shoaling succession that is interpreted to represent a


progradational sequence from shallow-water subtidal to arid tidal-fiat environments (Fig. 6b). Pisolites, sheet and shrinkage cracks, and tepee structures at the top of the sequence indicate an arid subaerial environment sub- jected to desiccation. Associated with the pisolite facies are laminated mudstones and algal-laminated mudstones that were probably deposited in restricted ponds on the tidal fiat and islands. Highest energy occurred along the edges of these islands, where crossbedded and laminated ooids accumulated as fringing bars and beaches. Basinward of the ooid facies, pellet grainstones represent a broad area of low- energy, burrowed stable grain fiat that formed generally below normal wave base. Farther offshore, the fusulinid wackestone facies represents the extensive shallow-water subtidal shelf.

Porosity Dune field produces from intergranular and

intercrystalline pore space and very little vuggy pore space. Intergranular pores in grainstones are located between peloids that average 180 #m in diameter. Intercrystalline pores are located between dolomite crystals that have per- vasively replaced wackestones and mud-dominated packstone. The dolomite crystals range from 30 to 80 #m in diameter and average about 50 #m. Most samples contain either intercrystalline or intergranular porosity, but some samples contain both types coexisting on a scale of centimeters. Samples with both types of porosity are referred to as grain-dominated packstones, which are grain-supported carbonate rocks in which the intergranular areas are partly filled with carbonate mud. There- fore, three "pore families" are recognized: dolomitized grainstone with intergranular pore space, wackestone with intercrystalline pore space between 30 and 80 #m dolomite crystals, and a grain-dominated packstone with both intergranular and intercrystalline pores.

The presence of as much as 55% gypsum in the Dune reservoir complicates porosity cal-

culations. Routine core analysis uses temper- atures higher than 60°C, and bound water from gypsum is released, resulting in erro- neously high porosity and permeability. Only cores analyzed using a special low-temperature technique were used in this study.

Gypsum has a large effect on neutron- and density-log responses and little effect on acoustic-log response. The neutron-log meas- ures the hydrogen ion content of the rock, and porosity is calculated from these measurements assuming that all hydrogen ions are in the fluids. Hydrogen ions in pore water and in bound water of gypsum are recorded as porosity on the neutron log and, when large volumes of gypsum are present, this results in a large error in porosity calculations and original-oil- in-place calculations.

The acoustic log is the porosity log least affected by the presence of gypsum and was, therefore, used as the porosity tool in this study. The acoustic log does a poor job of measuring vuggy porosity, but detailed studies of cores from this field have shown that very little vuggy porosity exists.

Permeability Lucia ( 1983 ) demonstrated that the perme-

ability of nonvuggy carbonates is related to particle size and interparticle porosity. A similar relationship among particle size, interparticle porosity, and permeability has been established in Dune field. Dune field samples segregate into separate porosity/permeability relationships for intergranular pore space between 180 #m particles and intergranular pore space between 20- to 100-#m particles. Sam- ples having a mixture of intergranular and intercrystalline pore space generally plot at values intermediate to these two relationships.

Permeability profiles were calculated for all wells in Section 15 having acoustic logs and la- terologs. Permeability cross sections illustrate considerable lateral and vertical variation (Fig. 7 ). Permeability changes of up to four orders of magnitude occur over a distance of 150 m

310 N. TYLER ET AL.

(500 ft), and vertical changes in permeability are also large. Where high-permeability beds are continuous between wells they may be thief zones, causing cycling of injected flood water. The large permeability changes over short dis- tances occur within units that otherwise would be considered continuous pay if only porosity were considered, and the permeability distribution is probably much more complicated than depicted on the cross section (Fig. 7 ).

Water saturation The three pore families have unique water

saturations. The intergranular pore family has the lowest water saturation (less than 20%), the intercrystalline pore family has the highest water saturation (greater than 25%), and the mixed intergranular-intercrystalline pore family has intermediate water-saturation values (20-25%).

The relationship between water saturation and pore family is interpreted to be due to different pore-size distributions characteristic of

each family. Thin-section examination determined that the intergranular pore family has the largest pore sizes, the intercrystalline pore family has the smallest pore sizes, and the mixed family has intermediate pore sizes. Therefore, connate-water saturation is highest in the intercrystalline pore family and lowest in the intergranular pore family.

Original oil-in-place Stock-tank original oil-in-place was calcu-

lated as the product of porosity, oil saturation, and thickness (SoOH). Data were derived only from wells drilled in Section 15 after 1978. Po- rosity values have not changed since initial development 40 years ago, and water saturation does not change significantly because Dune field produces by pressure depletion. Water encroachment has occurred, probably from offset waterflooding operations, but has been accounted for in the calculation of the water- saturation values. The decrease in pressure has

B B' Northwest 1560 Southeast 1552 1543

1545 1553 1556

PERMEAB~UTY ~ < 0 . 1 ~ 0 . 1 - 1 1-10 ~ ' 0 - ' 0 0 I ' ' 0 0 S"' ~ b e d s Z Morker bed (rod)

Fig. 7. Permeability cross section B-B' of Section 15, Dune field. Location of the section is shown in Fig. 4.

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS 31 1

liberated dissolved gas, calculated at 9% gas saturation in 1978.

The original oil-in-place for each well in Section 15 was calculated as the product of porosity times oil saturation times thickness (So~)H). Intervals having less than 6% porosity or greater than 1,000 g2 resistivity were considered to be 100% water saturated and were omitted from the calculations. The SoOH values were contoured using depositional models as guides (Fig. 8). The total original oil-in-place for Section 15 is calculated to have been 30.90 MMbbl; more than half of this, 58%, resided in the MA zone, and 25% in the BC zone. These zones are vertically separated by the AB zone, which has low original oil-in- place.

Cumulative production Cumulative production of oil and water are

among the more reliable data available from old fields. Although there are usually insuffi- cient pressure data to estimate fluid migration between wells, isoproduction contours can provide patterns of the areal distribution of production capacity. The cumulative production from Section 15 (Fig. 4) increases from northwest to southeast. Depositional facies maps of the MA and the CZ zones show trends of grainstones similar to those of the isoproduction contour map.

Cumulative-production figures show regional depletion of the field, but they provide little insight into pattern (areal recovery) or conformance (vertical recovery) efficiency of the recovery process. Since most wells are completed in multiple zones and have been pumping most of the time, the stratigraphic distribution of production from each well is unavailable from production statistics alone. However, using permeability data calculated from logs, production was allocated to individual zones, resulting in three-dimensional geographic displays of the remaining mobile oil in the field.

Distribution of remaining mobile oil As of January 1988, 11 MMbbl of oil had

been produced from Section 15. Using a residual-oil value of one-third the entire oil in place, about 10 MMbbl of mobile oil (inclusive of reserves) remains in the reservoir in Section 15.

Remaining mobile oil for each reservoir zone in Section 15 (Fig. 8 ) was calculated by subtracting produced oil and residual oil from original oil-in-place. Thirty-eight percent of the total oil produced from Section 15 came from the CZ zone, even though that zone contained only 8% of the original oil-in-place. This high productivity is due to the high average permeability for the CZ zone. Only 200 thousand barrels (Mbbl) of mobile oil remain in the CZ zone in Section 15. Forty-two percent of the total oil produced came from the MA zone, which originally contained 58% of the original oil-in-place. As a result, the MA zone contains by far the largest volume of remaining mobile oil: 7.3 MMbbl, or 73%. The BC zone contains 2.5 MMbbl of remaining mobile oil, or 25%. Only a small amount ofoil has been recovered from the AB zone, resulting in 886 Mbbl of remaining mobile oil, which is not considered an immediate target because the zone has low permeability.

Strategies for recovery of remaining mobile oil in Dune field

Ten million barrels of remaining mobile oil still reside in the Grayburg Formation in Sec- tion 15 of Dune field. This oil is located in northwest-trending carbonate sand bars developed in three major geological zones: MA (7.3 MMbbl), BC (2.5 MMbbl), and CZ (0.2 MMbbl). Because of the complexity and dis- continuity of the permeability within these sand bars, geologically targeted infill wells are required to improve recovery.

Production rate increased as a result of the infill-drilling program initiated in 1978. How- ever, the second infill-drilling and waterflood- optimization phase, using advanced geological

312 N. TYLER ET AL.

CZ

~::::

ac I!!iii:, !iiiiiiii. liiiiiiiiii: iiiiiiiiii~

MA

~ I : : ~ : :: "" '7~:::!:~:::::." • • ~ ~ :".;?~"~ .!::'i~ ::::::::::::::::i :::::::::::::::::::::::::: I~i!i::::ii ~ i i i : : : : ~ I ~ I I :::::::::::::::::::::::::::::::::::::::::POiiOt" "~'~::" "~9.~ " •

;;s:;ili::~i. :;h~iiiii::~:.:~!,:7!~r:~ ~'!i !

l~:i:i:!GrOi nsi6ne-:~'~ Peckstone ~!::: dom n ~ - ~ . . ~ . ~ ~ ~ ~ii:i:i:~ i i i~'!i i~',:~~'~~:~:!':':"""'"'"' 2::::::::: ~ i i l iii!ii:i!i!!:~!i~ ~ ' ' : : i : : : ::::::::::::::::::::::::::::: Oo

:::::::::::::::::::::::::::::::::: %

~ i i ! ! i~:::::::::::::::::::::: :'.::: ..

Facies Original oil-in-place Remaining mobile oil Contour interval: Thickness (ft) Contour interval: Net SoPhi H Contour interval: Mbbl/acre

Fig. 8. Facies, original-oil-in-place, and remaining-mobile-oil maps of the MA, BC, and CZ zones at Dune field.

models and engineering information, increased product ion by 60%. Recovery efficiency as of November 1987 in the MA zone of Section 15 is est imated at 45% of mobile oil originally in place, or 26% of original oil-in-

place. Well spacing in Section 15 of Dune field is already small enough to tap large scale heterogeneities produced by depositional facies. However, smaller scale, interwell heterogeneities that cause compartmental izat ion and by-


passing of oil continue to hamper oil recovery efficiency. Several simulation experiments were conducted using two-dimensional cross sections of the oil-rich MA zone in Section 15. Results show the effects of interwell heterogeneity on fluid flow and oil recovery efficiency.

Both deterministic and stochastic simula- tions of permeability distribution were conducted to determine why recovery efficiency is low and how it might be improved with infill drilling. The deterministic interpretation in- volved correlating permeability values from well to well and assuming gradational changes in permeability where lateral discontinuities occurred. This technique produces relatively high pay continuity. The stochastic technique known as conditional simulation was used to generate numerous permeability patterns that are thought to be more realistic. The permeability patterns range from high to low continuity.

The results of black-oil simulation experiments indicate that targeted infill drilling would significantly increase mobile oil recovery efficiency. Low-continuity models produced oil recovery and water/oil ratio values that closely resemble field recovery and thus support the assumption that continuity between wells is low. With the current well spacing of 4 ha ( 10 acres), mobile oil recovery efficiency is 45 to 50%. In the model, the addition of two infill wells reduced well spacing to 1 ha (2.5 acres) and increased mobile oil recovery efficiency by 27 to 32% points. Adding four infill wells, 0.7-ha ( 1.7-acre) well spacing, increased mobile oil recovery efficiency to 84-92% (Fogg and Lucia, 1990).

The Dune field study indicates 17.8 MMbbl of original oil-in-place in the MA zone of Sec- tion 15, of which 12.4 MMbbl is mobile oil. Infill drilling from 4- to 2-ha (10- to 5-acre) well space would increase recovery by about 15% points, which is equal to an additional recovery of 1.86 MMbbl ofoil from the MA zone. Remaining mobile oil in the MA zone is concentrated in the 65 ha ( 160 acres) that encom-

pass the grainstone trend. Targeted infill drilling of this 65 ha (160 acres) to 2-ha (5-acre) spacing would require 16 wells for a per-well recovery of about 110 Mbbl.

Emma San Andres reservoir

Emma San Andres field is located in south- central Andrews County (Fig. 2). After discovery in 1937, early wells produced at initial rates as high as 1,600 barrels per day from the San Andres Formation. Water injection began in 1965, by which time cumulative production had reached about 11 MMbbl. As of 1987, the reservoir had produced nearly 100% of the projected ultimate recovery (approximately 20 MMbbl ).

Geologic setting Emma field is one of several fields that have

been developed along the eastern margin of the Central Basin Platform. Like many fields containing San Andres reservoirs on the Central Basin Platform, Emma field is developed on an asymmetrical, northwest-trending anticline (Fig. 9 ) that is subparallel to the eastern margin of the Central Basin Platform. Until recently, hydrocarbon production has been largely restricted to the axis of the anticline. In the later 1970s, however, significant new production was established farther downdip, on the southwest limb of the fold.

Oil production in the field is confined to about a 75-m (250-ft) interval in the upper 100 m (350 ft) of the San Andres Formation (Fig. 10 ), which is composed of dolostone and relatively small amounts of nodular and poi- kilotopic anhydrite. Thin beds of terrigenous siliciclastics are present in the uppermost part of the formation above the producing interval (Fig. 10 ). These beds are persistent in the area and form readily traceable markers.

Facies and depositional environments The upper San Andres Formation in Emma

field comprises nine intergradational but dis-

314 N. TYLERETAL

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :1::: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : ;: : : : : : : : : : : : :7::: : : : : :;

2 4 : : : : : : : : ; : : ; : ; : : : / 3 0 :::::::::::::::::::::::::::::

"" \ i i i i ! i i i i t l ~"~c, ,i . . . . . .

. . . . . . . . ~ U N I V I ' I ' ~ : ~ I I "f L / - ~ I ~ I U : ~ I ~ L U U I ~ ~::::::~:~:~:~:~:~:~:~:?:~:;:?:i:!:i:i:i:i:i:?:!:!:i:i:i:i:i:i::~:~1~i:!;?~i~i;i~!~r~ ...................................... iii~;~ii;~?i!!?1.3 . ....%::..

:::::::::::::::::::::::::: ::: :::::: ::::::::::::::::::::: ::::::::::::::::::::: !:: :::: ti:: :::::: l~ i ! i i i ' : : : ; . ( o ~ o ~ il :::: ::i::!: :: i ~ ;.i::: ::.::::i:::: ~::i~i!i::!i::!i i-~:i!: [,~ ::iiiiiii~ i ! i ! i i : : ! ! ! i i !~ :~ i i ' i i i i i i i::r% ::::!i!i!!'~i~,~]iii:ibi~fvi~RSiTV~i::i:::::.:! .. Iiiii:.:. : : : . : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :: ::

! ! ! ! ! t i ! i : i i i i i i i i i i i i i ~ : . ..... ~ i ! i i : . ! ! i ~ : : ~ d i i ~ i l ~ i~ i ! i i~ [::i~::i i : i ] : :: : B L O C K 9: :: :: :: :: :: : : i : : : : i i~i i i l l ~i i i i i i i i ~: : i i ; i i i i i i i i i i i l ) i l l ? i i i i i i i i i : i : i : : i i i i i l :: ?: ill/i::ii;'~'~ i::ii::ii::ii::ii::ii::i!:ii::ii::ii::ii::ii::i!::ii: ! ! i !~i i i i i i i i i i i i i ioi .i::::!i[.~;,~:iiiiiiiiii:iiiii??ii ii:!i:iiiii iill iiii::::': :::::::::::::::::::::::::::::::::::::::::::::::::::::

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : i~;~i;~i :. :::::: :: ; i :! : : ~ ":: :: :: :: ii!i :: :: : , ; : :: :: !:: !::i!: : : : : : : : : : : : : : : : : : : : : : : : : : =============================== : : : : : : : : : : : : : : : : : : : : : : : : : " , . . . . . i::~ . . . . . . . . 1:..1 ..::..1:..:

iii!iii', :iiii ,ii',ii',i!:,:,iill: ', ============================================================

:: • ~ =========================

" " " ~ i i i i i i i i i i i i i i i i i i i i i i i

= = = = = = = = = = = = : = = : = = : : : = = = : =

• • :::: ° o o • .:q~ ..,~:..::..::..: ===================== ::..::..::..: • J , : - . - . . . : . . . : .. • . . . . . . : . . . . . . . .

2000 400Oft 0 , ~ , ,' E X P L A N A T I O N 0 460 860 1200m • Well • Cored .e l l ~i~:~ Unit boundory

C o n t o u r intervol 2Oft D o t u m m e o n s e o l e v e l ~ S l r u c t u r ( ] l axis

Fig. 9. Structure map contoured on top of the San Andres Formation, E m m a field area. Cross section C-C' shown in Fig. 10; cross section D-D' shown in Fig. 12.

D O L O M I T I Z E D P L A T F O R M C A R B O N A T E R E S E R V O I R S 3 1 5

C Southwest

,q[

,.~E o',

TOP OF SAN ANORES FM, ~.

, , * - = - - ° . - ' a - ~ b' "- - o ' . " g'~" " ' ~ ' . ' ; , . . ; - c ' ° . . . . . .~,---~;-'.-:...: ........ .~.~---. c TM |

........ !'~'~."!"" ...... - ~,~.":~ ~i~-'.~. "

C /

Northeast

I

- o o • o o" FUSULINID PACKSTONE/WACKESTONE o • o o o o o = % o O a ° o' 0 " = o = o o " = o o ' o = • o o e o o o ° o o ° o o ~ o = ° o ~ ° a o l o o = ° . % ° o o o = o

o o . a = °e = = o a • o o o o o ' I ° o ° D ° = o o o ~ o o o o = o= o c o = o ~ e = ° o *o 0 0 ' * o = o . % ~ o o = o== = • 0 o = = = o o o = o ~ = =

• o o = ° = O o o . = e = o ^ o . " • % ° = e ' * = = "

I P o r o s i t y ~ C o r e

E X P L A N A T I O N

GR G a m m a - r o y log D a t u m B o s e of A m a r k e r

Z M a r k e r b e d

I Q

e '

f? m

IO0130 20

I0

0 I 0

IO00ft I L I t

I I I I I 100 200 500m

Fig. 10. Cross section C-C' through Emma field, depicting vertical and lateral facies relations. The letters on the right margin of this figure indicate the location of gamma-ray markers. Line of section is shown in Fig. 9.

tinct facies (Fig. 10) that represent four major depositional environments: open platform, shoal, restricted inner platform, and supratidal (Fig. 11 ). Development of reservoir-quality porosity and permeability, however, is restricted to the Shoal and Open Platform facies.

Open platform. Open platform deposits comprise three distinct facies that collectively form the lower porosity zone in the reservoir: fusulinid packstone/wackestone, fusulinid/crinoid packstone, and burrowed wackestone (Fig. 10). Throughout most of the areas, the base of the reservoir section is formed by a

thick, continuous blanket of fusulinid packstone/wackestone (Fig. 12). These rocks are characterized by abundant anhydrite nodules. Where not filled with anhydrite, fusulinid molds account for much of the observed reservoir porosity. Fusulinid/crinoid packstone is present in the lowermost part of the section in the eastern part of the area only (Fig. 10).

The presence of fusulinids and crinoids in the lower part of the open platform sequence indicates that these deposits accumulated in a normal-marine setting.

Shoal. Shoal deposits consist of thin, 3-6 m

316 N. TYLERETAL.

SUPRATIDAL \

- ~ ; :\< ~ ? : ~ . ~ ~ . ~ . . . . . ~ . ~ - . : . ~ . , 3~ : ........... :

~ o o ° Oo o ~. - -~ " ~ o I o Oo o o

" ° o o 0 a • o l • o o o o o(~ o co©

~ ° O O o o o ° D o o o o * ° o o o oo~ EXPLANATION ~o o o

~ ] Field outline • Cored w e l l ~

RESTRICTED INNER PLATFORM

Cryptolgol mudstone

~ Pisolile groinstone

Burrowed mudstone

" ' i. MIGRATING ~ / /.: i: i!:!":" i..i;.~. !i SAND SHOALS

//

:i:..::::! -. / . ::!/;"':: '4 ..... ;!. ...!::. ~ ....... ./ ,

.... ::::.:. :.i!.i. ... , ~ / ~ .:'.::,.. , / / / ,=:~:, " " "<i , / / ...... : -~ ' : : . : : . . / o/"

: !i: ,':'. / / ~... ....:::: ~ ~ ~ LATFORM/// r c~, ~ :'" *o'" // o 0 o I ! / /

/ //

Skeletol grainstone

Burrowed wockestone

-~ Fusulinid pocksione/wockestone

Fusulinid / crinoid pockstone

Fig. 11. Paleoenvironmental reconstruction of the Emma field area during development of(A) the lower reservoir interval and (B) the upper reservoir interval.

( 10-20 ft), laterally discontinuous intervals of skeletal grainstone in the upper part of the San Andres section between the X and Z markers (Fig. 10). These rocks, which directly overlie open platform rocks throughout most of Emma field, contain abundant clasts of calcareous algae and fusulinids. Thickest accumulations of grainstone define northwest-trending axes of sand bodies (Fig. 13). The axis of greatest thickness in part coincides with the present structural axis in the field; however, significant thick areas are also present off structure to the north and south (compare Figs. 9 and 13). Intervals of grainstone exhibit distinct lateral and vertical discontinuities throughout the area because Shoal grainstone and pack-

stone is interbedded with muddier restricted inner platform deposits (Fig. 10). Shoal grainstone constitutes the upper porosity zone in the reservoir.

Skeletal grainstone and associated packstone are interpreted to represent deposition in a migrating complex of skeletal sand shoals (Fig. 11 ). Variations in mud content and lateral and vertical continuity (Figs. 10 and 12 ) probably reflect lateral migration of shoals and deposition in slack-water areas developed on and around the shoal complex. The orientation of thickness trends, oblique to subperpen- dicular to regional depositional strike, suggests that accumulation of these deposits may have been controlled by current-modified tidal


D Southwest

{ P~

D I

Northeast

GR ~_Son GR

I. [b ~ UPPER POROSIT ZONE I0

i1 , ,ooo : Poro~slty I ,t Fusulinid pockstone/wockestone 0 ' 200 ' 400 6()Ore

Fig. 12. Cross section D-D' i l lustrat ing the d is t r ibu t ion o f the two reservoir pay zones in the Emma San Andres reservoir. Note lateral and vert ical var iat ions in the thickness and extent o f the upper (shoal grainstone reservoir) porosi ty interval. Line of section is shown in Fig. 9.

or storm-related processes. Similarly trending grainstone accumulations have been reported from the Grayburg Formation in Dune field, as was discussed previously, suggesting that controls on their accumulation may have been widespread on the Central Basin Platform.

The shoal grainstone facies is overlain by peritidal and supratidal rocks of the supratidal facies. These rocks, which are interbedded with laterally continuous siliciclastic beds, are rich in carbonate mud and anhydrite and form the top seal of the reservoir.

Paleogeography and depositional history From the vertical sequence of facies docu-

mented above it is apparent that the upper San Andres Formation in Emma field consists of an upward-shallowing sequence of shallow subtidal to peritidal and supratidal deposits that accumulated on a shallow-water carbonate ramp (Fig. 11 ). Open platform packstone and wackestone represent deposition in a moderately low energy (near effective fair

weather wave base), shallow-water subtidal setting that apparently became somewhat more restricted (lower energy) through time. Shoal grainstone accumulated in shallower water, high-energy conditions. These shoal deposits sharply overlie open platform rocks and suggest a major shift in paleoenvironments at this time, perhaps caused by sea-level fall and subsequent rise.

Diagenesis San Andres rocks in Emma field have been

substantially modified since deposition by a complex series of diagenetic events. From the standpoint of porosity evolution, three main stages in the diagenetic history are significant: dolomitization, sulfate emplacement, and sulfate dissolution. Although pervasive, dolomitization of the San Andres in the Emma field area was primarily, if not exclusively, repla- cive. Because of this, primary porosity was not markedly altered. Processes of sulfate emplacement and subsequent dissolution, how-

318 N. TYLER ET AL.

24 25

__/9 .....

50

0 2000 4 0 0 0 ft I , ' , I , ~ , ,~ E X P L A N A T I O N

0 4 0 0 8 0 0 1 2 0 O r e • Well m C o r e d wel l i i i i i i iEmma f ie ld b o u n d a r y Contour interval ;)Oft

~Structural axis ~ O-20ft ~ >2Oft

Fig. 13. Thickness and d is t r ibu t ion of porous skeletal grainstone (shoal facies) at E m m a field. Note tha t this map is also a net-pay map for the upper reservoir interval.


ever, have had a much more significant effect on porosity development and distribution.

Following dolomitization, sulfate (either in the form of gypsum or anhydrite) filled existing void space and replaced dolomite, both grains and matrix. Although difficult to quan- tify, porosity reduction during sulfate emplacement was substantial. The Emma San Andres reservoir locally contains as much as 20-30% anhydrite.

Subsequent dissolution of void-filling sulfate has exhumed some previously occluded porosity. Sulfate dissolution has also, at least locally, created new porosity as a result of sulfate replacement of dolomite. Subsequent dissolution of this anhydrite has actually increased porosity. Much of the porosity in the highly porous and permeable skeletal grainstone interval was created in this way.

Porosity and permeability Significant porosity development is re-

stricted to two major zones (Fig. 12): an upper zone of thin beds (shoal grainstone) and a lower, thicker zone (open platform packstone and wackestone).

Lower porosity interval (open platform deposits). Porosity in open platform fusulinid wackestone/packstone deposits ranges from about 4 to 15% and extends well below the reservoir interval, a thickness of at least 60 m (200 ft) locally. Permeability in this lower porous interval averages less than 2 mD but reaches 25 mD in some thin zones.

Both moldic and intercrystalline pore space is common in open platform rocks. Open fusulinid molds (average, 1-2 mm wide) are locally abundant and contribute to porosities of as much as 18% in fusulinid packstone and wackestone. Such zones however, are rare and thin, usually less than 0.3 m ( 1 ft) thick.

Intercrystalline porosity is locally abundant in open platform rocks. Visual estimates from thin sections suggest that intercrystalline pore volume locally ranges as high as 10%. In most

cases, intercrystalline pores range in size from a few to a few hundred (very rare) microme- ters. Intercrystalline porosity in open platform rocks is noticeably higher in irregular, generally lighter colored "recrystallized" patches. These features are present in several other San Andres/Grayburg reservoirs on the Central Basin Platform, including all the other reservoirs discussed in this report. Data from Pen- well field, in particular, indicate that these zones are much more permeable than surrounding unaltered dolomite. These patches may play a major role in the development of porosity and permeability in Open Platform rocks in Emma field and elsewhere.

Upper interval (shoal facies). The upper porous interval within the Emma San Andres reservoir comprises skeletal, shoal grainstone, packstone, and wackestone. Although porosity (average 8%) and permeability (average 3.5 mD) in these deposits is about the same as in the lower porous interval, porosities of 10-15% and permeabilities of 50-100 mD are common in mud-free grainstone intervals.

Shoal grainstone contains interparticle, intercrystalline, moldic, and intraparticle pore space. Interparticle pore size typically ranges from 200 to 400/~m; intraparticle pores vary in size to 700/tm in diameter. In extensively leached zones, intercrystalline porosity is high where not filled with anhydrite or calcite.

Distribution of porous facies. The two major intervals of porosity in the Emma reservoir exhibit significantly different distributions across the field area. Open platform rocks, which constitute the lower interval, extend as a blanket deposit across the area (Figs. 10 and 12 ). Although porosity varies locally on a small scale, porosity development in the lower reservoir interval is widespread across the area. The upper, shoal grainstone porosity interval, in contrast, is much more restricted in its distribution and contains distinct local variations in thickness. Porous grainstone intervals vary

320 N. TYLER ET AL.

in number and thickness across the field (Fig. 12).

The distribution of net pay in the lower porosity interval reflects the influence of structure because it is limited by the field oil-water contact (Fig. 12 ). Because of its stratigraphi- cally higher position, however, the upper porosity interval is almost entirely above the oil- water contact in the area (see Fig. 12). Thus, the distribution of net pay in the upper interval is not primarily controlled by structure. Al- though maximum net-pay thickness trends correspond to the structural axis, there is significant net pay off structure to the southwest and north (Fig. 13 ). The distribution of these rocks is a function of original deposition.

Production characteristics Cumulative production of oil in the Emma

San Andres reservoir totals approximately 19.5 MMbbl. Based on conventional estimates this represents more than 95% of the projected ultimate recovery. In recent years annual production from the reservoir has dropped markedly. Only 0.156 MMbbl were produced in 1986; most of this came from the southwest flank of the field.

Cumulative oil production prior to unitiza- tion and waterflooding in May 1965 totaled about 11 MMbbl. Although production during that time generally came from areas along the field structure, production patterns correlate more closely with the distribution of porous grainstone in the upper porosity interval than with structure. Particularly obvious in this re- gard is the volume of production obtained from areas off structure in the northern part of the reservoir unit on University Lands Blocks 9 and 10 (Fig. 14).

Production trends since the onset of waterflooding are generally similar to those observed prior to waterflooding, with one nota- ble exception. In the late 1970s production was established on the southwest flank of the field structure well downdip from previous producing wells. This new production has accounted

for about 1.4 MMbbl of the total recovery from the reservoir. As is the case in the rest of the field, production patterns in this area show a close correlation to the distribution of skeletal grainstone of the upper porosity interval (Fig. 13).

Permeability calculations suggest that most of the production in the Emma San Andres reservoir came from the upper part of the reservoir. This interpretation is supported by the similarity between the distribution of skeletal grainstone and production patterns (compare Figs. 13 and 14). Completion history data are consistent with this conclusion.

Volumetrics The volume of original oil-in-place has been

calculated to be 48.4 MMbbl for the Emma San Andres reservoir. Individual determinations of original oil-in-place for the upper and lower reservoir intervals give values of 34.3 and 14.1 MMbbl, respectively. Cumulative production is about 19.5 MMbbl. These data indicate a recovery efficiency of more than 40%, which is exceptionally high for a San Andres reservoir (Galloway et al., 1983 ).

Despite the apparently efficient recovery in the Emma reservoir, calculations indicate approximately 15.0 MMbbl, or 43% of the original mobile oil, remains in the reservoir. Of this total original mobile oil, approximately 7.8 MMbbl, or 52% of the oil, remains in the upper porosity interval; about 7.2 MMbl lies in the lower porosity interval.

Strategies for recovery of remaining mobile oil

Consideration of permeability data and production history data suggests that much of the produced oil (as much as 85% ) has come from the upper skeletal grainstone porosity interval. The high apparent recovery efficiency calculated for the Emma San Andres may be the result of the fact that most oil has been produced from this interval, which contains relatively high and uniform permeability. Despite the

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS

24

-UNIVERSITY LANDS BLOCK 35

25 130

31

UNIVERSITY LANDS

BLOCK 9

321

I02 W E COWDEN [~!:"":~"*":~"~~~ ~~ : ' ] . 4 . . 2~ . :~ i ! . ' ; : i ! i ' i ' ~ i !i::!~ "SURVEY ~ i ~ : ~ ~ ~ ~ ! : i ! i . ! ~ ! i i ! i ! ' i ; i i ii'i!i!-~

~...: .: ' :. ~ . : . ' . . ' . - . " . . : " . ' - .1 . - .~, '~ , . . / : . ; .~ ...:....,-.!--.. :..:.:?..:.:..:.:-i ~y..-!::::.. ~; . , i ... .. ' ... ' .--...:.@... '..:. ' .::.:..: ' ;.:..-o~,=:.... ":.-::!i .. ..i;i!! i.!:i i~i; '~':::~:':'.. .:':.'. .. "':'" "".'.'.'".'.'::.'::::'~ iC:':.".':: ::~.::.:.Y::~':':~!

RAILROAD ~ ! : ~ : ; Z . ! ~ : ! ~ \ ~ " ] BLOCK 44 / ~ ~ ~ i ! ! ~ ~; ~ BLOCK 44

' . :;,f. i:::~ " ; : : " : ' : . ' , ~._~1"~ ";!:::;;~ ~:; '.::!

~iii:: '~:"!!i~iiil :~I:;: ~ ~ ~ ~ : ! i ~ i ! ~ 0.-'. .:..', :-:. " ":i;:-~:iii~:' ~i]i~

~:~!::~!~.~..:.:~.~.~!~̀~.~:~:'~:;:i~!!:~i;~.i.!i~i~.;i:~i:.~i~;i~i~i!~:~:~:.~:~;~:.;~::~! i i.!~i.!:.:i~i:i:~;:::!:,i~::~-:~ili;.i~i.:!i!~i!!~.~:i :::~.~:~;;.-,:,.:: :;:-:;!~:~':i~!i! ~ :iii;:~,~. . . . . . ~i~ ', 7!:~.~:...~:~?;~!~k~:~?~.~;~:~;i~!~;~7~4~!!~ ~ ~.:.;.:~:'~:/:;:~:~i~-.:.?~.~7".i:77::';"!: ~?17::-?!.'.,~',~:;~!::~7.:.=.::i;?;~:::...":~.~.~!~" ~'. ;.7%.:~;'..':::" ':lr~*~l :: ~.

] :.~,'.'.-:."~:?"..-'"- ?" '~?,.':.?..::::::.:~::'::::.':~,:-?... ~ : . ~ . / G ~ : : . . ~ : : : c : " ' " . ' ~ % ~ . . . : . ' . ' . ~ , ~ .

...-:::~.~':~.~:~5...~7~/.?~7~:~:~:~:~:~i~::~! ;::::!~..::!..:::.:i~.::'.:,lI::7.::...?. ~::::'~-?-:':::2".?.'..: '7::::;!f'::';.':~::,..:.':f..::?!. .';'.:~','.'.Z....:.;:::::~ ":.:';~ ~'~

0 2 0 0 0 400Oft I ' / , .j E X P L A N A T I O N 0 4()0 800 1200m e Well =Cored well ii~i~i~iiEmma field boundary

Contour interval I00,000 bbl I I Structural axis ~ 0-200~000 bbl

~ Infill tor(:Jet area ~ >200,O00bbl

Fig. 14. lsoproduction map for the Emma reservoir as of January 1986. Comparison with Fig. 13 illustrates several areas in the nor thern par t of the field where drainage from this, the major reservoir zone in the field, is incomplete.

322 N. TYLER ET AL.

apparently high recovery efficiency, conserva- tive calculations indicate that a significant amount of mobile oil, as much as 8 MMbbl, still resides in the upper skeletal grainstone reservoir interval.

Because of its favorable reservoir characteristics (pore types and distribution and facies geometry) the upper skeletal grainstone reservoir interval must be considered the primary target for future infill drilling and recompletion development. Comparison of isoproduction maps with the net-pay map for this reservoir interval (Figs. 13 and 14) indicates several areas in the field that are potential sites for in fill because of poor recovery relative to the thickness of skeletal grainstone. Especially prominent among these is an area in the southeastern corner of Section 35 and southwestern corner of Section 36, University Lands Block 10, and immediately adjacent parts of Section 102, W.F. Cowden Survey. This area contains up to 18 m (60 ft) of porous skeletal grainstone (Fig. 13 ) and is high on the field structure (Fig. 9). However, wells in this area have produced relatively small volumes of oil (Fig. 14). Effective exploitation of the skeletal grainstone reservoir zone must consider the lateral and vertical variations in the thickness and distribution of these skeletal grainstones (Fig. 13 ) in recompletions and new drilling.

The lower reservoir interval is considered a secondary target for recovery of remaining oil. Calculations suggest that this zone may contain as much as 7 MMbbl of remaining oil. However, the difficulty of mapping porosity and permeability distributions in this zone will make effective exploitation of this interval of longer term objective.

East Penwell San Andres Unit

The East Penwell San Andres Unit (EP- SAU ) is located approximately 24 km ( 15 mi) west of the eastern margin of the Central Basin Platform in University Block 35, Ector County (Fig. 2 ). Penwell is the northernmost field in

a five-field complex (Major et al., 1988) that produces oil from a combined structural and stratigraphic trap on the east flank of a broad, asymmetric anticline (Fig. 15 ). Production is from the Permian (Guadalupian) San Andres Formation reservoir at a depth of approximately 1,070 m (3,500 ft). The field was discovered in 1927 and has been on waterflood since 1970. The waterflood is a modified five spot, and current producing spacing is approximately 8 ha (20 acres) per well.

The main reservoir zone, which is the subject of this report, has been produced since discovery in 1927. A lower San Andres reservoir zone was not penetrated until 1985, and very few data are currently available from this new zone. The Unit has produced 43 MMbbl of primary and waterflood oil from an estimated 164 MMbbl of original oil-in-place. Ac- cording to our calculations from data in Rail- road Commission of Texas files, approximately 91 MMbbl of remaining oil is immobile. There is thus approximately 30 MMbbl of remaining mobile oil in the Unit.

Depositional facies The main San Andres reservoir at the EP-

SAU is composed of an upward-shoaling sequence of shallow-water ramp facies. The reservoir rocks are primarily porous open-marine grainstone/packstone overlain by generally nonporous tidal-flat mudstone and pisolite packstone. The following facies descriptions are based on examination of 13 cores from the unit.

Pellet grainstone/packstone. The volumetri- cally dominant open-marine facies in the upper San Andres Formation at Penwell field is composed of thoroughly dolomitized grainstone/packstone composed of spherical to ovoid fecal pellets. Common accessory skeletal grains are fusulinids and mollusks, which are rarely preserved and most commonly evident as molds. Where fusulinids or mollusks compose 10% or more of grains, this facies is


\

EXPLANATION

o Producing well

• Inactive well

& Inject ion

• Solid symbols • ~, indicate cased

borehole

Unit boundary . . . . University Lands

boundary

N

0 4 0 0 O f t

0 IdOOm Contour interval 15 f t Datum mean sea level

Fig. 15. Structure map contoured on top of San Andres Format ion in the East Penwell San Andres Unit . The trap is on the flank of an anticline; an updip stratigraphic seal lies west of the field unit boundary.

described as pellet-fusulinid grainstone/packstone of pellet-mollusk grainstone/packstone. Burrow structures are rare, but a complete lack of bedding suggests that this sediment was thoroughly bioturbated. This thorough biotur- bation and the presence of abundant normal- marine fossils indicate deposition in an open-

marine setting similar to Holocene open-marine pelleted mud seaward of the tidal flats in the Bahamas (Shinn, 1983).

Pellet grainstone/packstone is the primary San Andres reservoir rock at Penwell field. In- terparticle porosity is commonly well preserved and results in a relatively high permea-

324 N. TYLER ET AL.

bility rock. Fusulinid and mollusk molds contribute somewhat to reservoir porosity but have little effect on permeability.

Algal grainstone. Algal grainstone, with both micritized and well-preserved dasycladacean algae grains, occurs in thin and discontinuous beds. Some algal grainstones are bedded. This facies commonly occurs interbedded with pellet grainstone/packstone at or near the boundary of pellet grainstone/packstone and super- jacent pisolite packstone or mudstone. This stratigraphic position and the suggestion of local relatively high original depositional energy, in contrast to that of adjacent pelleted rocks, as evidenced by crossbedding and the small amount of mud matrix, suggest that algal grainstone was deposited in tidal channels similar to those that cross the Holocene tidal fiats of the Bahamas (Shinn, 1983) or to the Holocene tidal channels that transport relatively coarse sediment across the muddy open- marine sediments of Florida Bay (Jindrich, 1969). The algal grainstone facies is interpreted as having formed as tidal-channel deposits in a relatively high energy ramp-interior setting. Where not thoroughly cemented by sulfates, this facies has high effective interparticle porosity and high permeability.

Sponge-algal boundstone. Thin zones of sponge-algal boundstone occur near the bottom of cores interbedded with pellet grainstone/packstone and generally 60 m (200 ft) or more below pisolite packstone and mudstone. These bioherms lack any evidence of subaerial exposure or mechanical abrasion due to wave action and are apparently discontinuous. These zones are only 0.3-(k6 m ( 1-2 ft) thick and, although they contain some interparticle porosity, are not of sufficient volume to be considered a significant portion of the reservoir. The association with the strati- graphically deeper portion of the pellet grainstone/packstone facies and the lack of evidence of high-energy conditions suggest that

these rocks were formed as isolated ramp-margin reef mounds. The occurrence of these reef mounds is apparently restricted to the downdip (east) side of the anticlinal structure at Penwell field.

Crinoid grainstone. Crinoid fragments occur as rare accessory grains in pellet grainstone/ packstone, but are observed in sufficient quantities to constitute a separate facies in only one core. Crinoid grainstone is poorly sorted, thoroughly cemented by dolomite, and is not part of the reservoir. The presence of this facies as thin beds interbedded with the strati- graphically deeper parts of pellet grainstone/ packstone suggests that these rocks formed as crinoid meadows in the deeper portion of the open-marine ramp margin.

Mudstone. Much of the reservoir seal in the San Andres at Penwell field is dolomitic mudstone. These rocks are the lithified equivalents of carbonate mud; in some cases they are finely laminated and generally not pelleted, presum- ably because high environmental stress isolated these sediments from the organisms that produce pellets and bioturbate the sediment in deeper water, open-marine environments. The mudstone facies is generally cream colored and barren of fossils, although algal laminites and rare fusulinids and mollusks do occur in these rocks. This facies is commonly interbedded with pisolite packstone and occurs strati- graphically above the open-marine pellet grainstone/packstone. This stratigraphic position and association with pisolite rocks that contain evidence of subaerial exposure (see pisolite-packstone section below) suggest that the mudstone facies was deposited in hypersaline ponds on a tidal fiat landward of the open-marine facies. The rare fusulinids and mollusks were probably transported by storms from deeper water, open-marine environments.

Pisolite packstone. The pisolite packstone fa-


cies in the San Andres section exhibits evidence of syndepositional subaerial exposure. These rocks are composed of poorly sorted and fitted-fabric pisolites and are characterized by sheet cracks, fenestrae, and desiccation cracks. Pisolite packstone is commonly interbedded with mudstone and is characteristically the same cream color. This facies is also generally barren of skeletal grains. Intergranular pores in the pisolite packstone facies are generally thoroughly filled by anhydrite. Rare, thin, par- tially cemented zones only 0.3-0.6 m ( 1-2 ft) thick may be high-permeability floodwater thief zones. Minor karst dissolution is indi- cated locally by severe brecciation and in filling by greenish-gray siltstone. The abundant evidence of syndepositional desiccation, association with fossil-barren mudstone, and presence of minor karst dissolution indicate that the pisolite packstone facies formed in a tidal- flat environment that was frequently subaeri- ally exposed.

Siltstone. Siliciclastic siltstone beds occur interbedded with the mudstone and pisolite packstone (tidal fiat) portion of the upper San Andres at Penwell field. Some of these siltstones are finely laminated, but most are massive. These rocks are often carbonaceous and in transitional contact with tidal-fiat mudstone and pisolite packstone/grainstone. The presence of this facies interbedded with rocks containing evidence of subaerial exposure and the lack of any regional sources for siliciclastic detritus suggest that these sediments were transported to the tidal-flat environment by eolian processes. Some reworking in shallow water subsequent to elolian transport is sug- gested by the laminations.

Depositional model The succession of facies in the upper San

Andres Formation at PenweU field comprises rocks formed from an upward-shoaling sequence of open-marine to tidal-flat sediments. The interpreted depositional environments are

illustrated schematically in Fig. 16. The open- marine section was characterized by pelleted mud and open-marine fauna, mostly fusulinids and mollusks, and sparse sponges, algae, and crinoids. The open-marine section contained rare, isolated sponge-algal bioherms. The shoreward tidal-flat environment was characterized by tranquil, high-salinity waters in which environmental stress excluded most fauna, resulting in deposition of barren carbonate mud. Tidal fiats were sites of pisolite formation and desiccation features such as sheet cracks and fenestrae. Lack of a continuous shelf-margin facies, such as a barrier reef or continuous grainstone shoal, and the lateral, sheetlike geometry of the pellet grainstone/packstone facies suggest that these rocks were deposited in a carbonate ramp setting. The tidal-fiat and open-marine portions of the ramp were locally cut by dip-oriented, relatively high energy tidal channels. These deposits were characterized by skeletal grainstone in which the grains are dominantly dasycladacean algae.

Diagenesis Induration of soft pelleted mud began early

in the diagenetic history of the San Andres Formation at Penwell field. Where induration resulted in pellet preservation, interparticle porosity is now preserved. Where pellets were compacted, most of the porosity is now de- stroyed. Thus, this early diagenetic event in- fluenced the formation of lateral porosity heterogeneities in the pellet grainstone facies, which in turn control the heterogeneous distribution of remaining mobile oil.

The entire reservoir section has been perva- sively dolomitized, and dolomitization of the original carbonate sediment was the major diagenetic event. Oxygen and carbon isotope data from Penwell field and from other San Andres reservoirs (Ruppel and Cander, 1988b; Leary and Vogt, 1990) suggest dolomitization by hypersaline brines that originated through evaporation of seawater (also see Chilingar et

326 N. TYLER ETAL.

Hypersaline pond

pellet

Tidal :channels

mounds

Fig. 16. Schematic reconstruction of depositional environments in the East Penwell San Andres Unit.

al., 1979). These San Andres carbonates, therefore, probably were dolomitized by hypersaline brines that originated o n arid tidal fiats and percolated through the shallow sub- surface during the Guadalupian. This hypersaline brine was also probably the source of the anhydrite and gypsum common in the San Andres Formation.

A diagenetic event that has major conse- quences for reservoir quality is a post-burial leaching that affected the pellet grainstone/ packstone in parts of the EPSAU reservoir zone. The pellet grainstone/packstone facies is locally mottled by light-colored, leached areas that in some cases has a geometry suggesting it was affected by burrows and in other cases a geometry suggesting the fluids responsible for this diagenetic alteration flowed preferentially along stylolites. Importantly, altered and unaltered rock are closely associated such that mottling occurs on a centimeter scale. This heterogeneity is too small to be measured by sampling techniques normally used for con-

ventional plug or whole-core permeability analyses. Minipermeameter data (Fig. 17), however, indicate the diagenetically altered (leached) rock has permeability values that cluster about a value of 10 mD, which contrasts with diagenetically unaltered rocks having permeabilities of approximately 1 mD (for a description of the minipermeameter instru- ment, see Chandler et al., 1988 ).

Petrographic evidence, such as anhydrite nodules with an outer rim of gypsum and rem- nant anhydrite within gypsum crystals, suggests that sulfates were probably entirely anhydrite at some time during the diagenetic sequence and are now partly hydrated to gypsum. Presence of gypsum in the formation is especially noteworthy because the bound water in this mineral affects interpretation of core- analysis data and wireline logs, as discussed previously in the Dune field section of this paper. Rare oil staining postdates sulfate leaching, suggesting that leaching may be associated with hydrocarbon migration.

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS 3 2 7

40

[ ] Unaltered (n ^~'

• A l tered(n=

3O

~ 20 g

u _

I I Irl

I I I

0.5 1.0 5.0 10 50 Permeability (md)

Fig. 17. Histogram of minipermeameter data collected from samples of diagenetically unaltered and altered (leached) dolomite. Diagenetically altered dolomite has permeabilities approximately one order of magnitude higher than those of unaltered dolomite.

Facies distribution Thirteen cores are available from the upper

San Andres Formation, and these data points may be used to construct facies isopach maps. Cross plots of wireline-log data were used in an at tempt to identify facies from log data so that wells without cores could be used as data points for mapping. However, no distinguish- able patterns were observed in log cross plots, and maps can be made only with data from cored wells.

The isopach map of net-tidal-fiat facies (mudstone and pisolite packstone) indicates that these facies are generally thicker in the western portion of the EPSAU and thin downdip to the east. Modifying this general pattern is an area of relatively thick tidal-fiat sediments in the southern third of the unit. The net-algal-grainstone data are sufficiently sparse that they could be contoured in more than one manner with equal degrees of confidence. The

net-algal-grainstone isopach map is contoured with a generally east-west grain, approximately parallel to structural and depositional dip, consistent with the petrologic interpretation of algal grainstone as tidal-channel deposits.

Production patterns Production maps are useful tools for evalu-

ating reservoir heterogeneity. In the case of old fields such as Penwell, however, production data are generally unavailable on a per-well basis. In the EPSAU, production data are available from the operator on a per-well basis only for the years postdating initiation of the waterflood in 1970. Production data before 1970 are available in the files of the Railroad Commission of Texas on a per-lease basis. These per-lease data may be apportioned to wells using the results of periodic tests. Thus, by combining records available from the operator and in the files of the Railroad Commis- sion of Texas, the per-well production history of the unit can be reconstructed.

The long production history of old fields such as Penwell results in a mixture of wells that have been on production for decades with wells that have been on production for only a few years. Moreover, the well spacing in this unit is uneven. These factors introduce "cultural effects" in production maps that obscure production patterns controlled by reservoir heterogeneity. To minimize these cultural effects, the production for each well was divided by the number of years that well had been producing, yielding an average production value. Next, average production for each well was apportioned within a 16-ha (40-acre) grid such that a single data point, expressed as Mbbl / year/acre, was assigned to each cell in the grid. The resulting map (Fig. 18) removes cultural effects and illustrates production anomalies resulting from reservoir heterogeneity. The low average cumulative production values in the southern part of the unit represent the relatively thick, nonreservoir tidal-fiat sediments

3 2 8 N. TYLER ET AL.

in this area. The elongated east-west zones of high production in the northern part of the unit represent inferred tidal channels.

Calibration of well logs to cores Porosity logs must be calibrated with cores

to provide porosity data in wells for which cores are unavailable. A major consideration in evaluating the log and core data in the EP- SAU is that this reservoir contains gypsum, as

has been discussed earlier. A calibration of acoustic transit time from wireline logs with core porosity collected using low-temperature analytic techniques yielded an excellent correlation (r= 0.90, n = 298 ), and this relationship allows calculation of porosities in wells for which low-temperature core analysis data are unavailable.

Oil saturations were calculated from wireline logs with the Archie equation (Archie,

0 I

, , ~ . . . . • • . . . . • . . . ' : ' .

300

~ 400

• . . . - - . . . . . - .

. . . . . 1 : : [ : . . 1 : : : [ . . . . . . .

J O

!ii! ii~ii??iiiiiiiiiiii~iii

iQ i i i i i ! i i i i i i i i : : : i i : : : : : :O : : : : : :X : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : ioiiiiiio!!iiii i!iiiiiii!iiiiiioii\

==========================================

:iiiiiiiiii! iiiii!o!ii ::: . . . . : iiiii

0 0 0

0 0 \o o

0 ©

o

EXPLANATION bbl/acre/yr

I I IOO -200

200-300

300-400

400-50O

l >500

0 Data point

0 I 0

N

400Oft J

I O 0 0 m

Contour interval I00 bbl/ocre/yr

Fig. 18. Average annual oil production from the East Penwell San Andres Unit normalized to a 40-acre grid (one data point to every 40 acres, or 16 ha). This mapping technique identifies area of highest reservoir quality by removing production-strategy effects such as the timing of developmental well drilling.


1942 ) using the acoustic log calibration based on low-temperature core analysis and a ce- mentat ion factor derived from point-count data collected from more than 50 thin sections. The Archie equation also requires a water resistivity value. Inasmuch as Penwell field has been waterflooded for many years and all resistivity logs in the unit were run after the initiation of the waterflood, the values of water resistivity vary across the field. Water analysis data provided by the operator were used to calculate water resistivity values, which were used to calculate oil saturations in wells for which both an acoustic log and a resistivity log are available.

Location of original and remaining mobile oil

Original mobile oil was located using the saturations calculated by the methods out- lined earlier. Whereas the top of the reservoir is clearly defined by the top of the San Andres Formation, the bot tom is not as well defined. No clear free-water level is present in the original producing zones, and recent wells encoun- tered deeper zones capable of producing oil. The bot tom of the reservoir, for the purpose of locating original mobile oil, was taken to be 40 m ( 130 ft) below an arbitrarily chosen gamma- ray marker, the approximate depth to which most wells were drilled. The combined effects of most wells not reaching the base of the reservoir and many wells not having a complete log suite resulted in an original mobile oil map that covers only 70% of the unit area, although the most prospective updip and northern parts of the unit have adequate data for evaluation.

The cumulative production data were sub- tracted from the original mobile oil in place data to yield a map of remaining mob i l e oil (Fig. 19). Note that there are no data points posted on the map illustrated in Fig. 19 because it is the difference of two contour maps constructed from different data points. A key feature illustrated in Fig. 19 is the concentration of remaining mobile oil in the northwest-

ern part of the unit. This concentration of remaining mobile oil corresponds to the tidal- channel trends inferred from Fig. 18, and is associated with high primary porosity preserved in pellet grainstone adjacent to porous algal grainstone.


The remaining mobile oil map indicates oil is concentrated in pellet grainstone adjacent to tidal-channel deposits in the updip portion of the unit. This geologically located concentration of remaining mobile oil is targeted for infill development drilling. It is emphasized that our calculation of remaining mobile oil is for the main reservoir at Penwell and that the newly drilled deeper zone will probably substantially increase reserves.

Taylor-Link West reservoir

The Taylor-Link West San Andres reservoir (Fig. 2) is karsted. It differs from most other San Andres reservoirs, which lack the significant impact of karsting on reservoir performance. These effects are particularly evident in the waterflood performance, which is discussed below.

Taylor-Link field was discovered in 1928 and covers approximately 800 ha (2,000 acres) on University Blocks 16 and 18, Pecos County, Texas. The reservoir zone is in the San Andres Formation, and the siltstones of the basal Grayburg Formation form the seal (Fig. 20). The trap is structural, being defined by a nearly symmetrical northeast-trending elon- gate dome. The crest of the structure is at 300 m (980 ft) subsea, and the oil-water contact is at approximately 265 m ( 875 ft) subsea.

Since discovery, the reservoir has produced about 10 MMbbl of the approximately 48 MMbbl of original oil in place. The initial development phase was from 1930 through 1945. Field production peaked in 1941 and has generally declined since that time. The field has

330 N. TYLER ET AL.

EXPLANATION Remaining mobile oil

Mbbl /acre

<o

0 5

5-10

IO-t5

>15

Rw = variable

N

0 400oft I i 1

0 IO00m

Contour interval 5 Mbbl/acre

Fig. 19. Contour map of remaining mobile oil in the East Penwell San Andres Unit calculated using Rw values obtained from produced water data.

produced very large volumes of water, some of which is from the overlying Cretaceous Trin- ity Sandstone aquifer. Water from this aquifer has been flowing down well bores for a number of years, producing an uncontrolled dump flood. A centered five-spot waterflood was initiated in 1985 following infill and relocation drilling of 114 new wells. High volumes of water were initially injected into the reservoir with limited results. Oil/water ratios of 0.01 or less were common and injection rates ranged

from 450 Mbbl to 1,400 Mbbl of water per month.

Geological setting The Taylor-Link West San Andres Unit lies

along the southern margin of the Central Basin Platform (Fig. 2 ) along the Sheffield Channel. The field is located approximately 8 km ( 5 mi) landward of the platform margin in a position comparable to the most interior portions of the sand-shoal complexes rimming the margins of the Bahama Platform (Hine, 1977 ).

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS

-1400

- 1 7 0 0

GR CNL Interval Facies Depositional environment

API

Cored interval ,~ >

"-'1"---1 l [ T 125 30 0

Neutron porosity (percent)

Siliciclastic-dominated intercalated

siltstone/dolomudstone

Carbonate-dominated intercalated

siltstone/dolomudstone

Basal siltstone

conglomerate

Grainstone

Fusulinid wackestone

Crinoid-brachiopod wackestone

Caliche

Ooid-peloid grainstone

Bioclastic-peloid grainstone

Fusulinid-peloid packstone

Sponge floatstone

Supratidal/ deflation flat inner ramp

Supratidal- intertidal

inner ramp

/ Shallow subtidal ~ inner ramp /

Shallow subtidal

Transgressive shoreline

Subaerial

High energy subtidal ramp cresl

Low to moderate energy subtidal

outer ramp

Low-energy subtidal

outer ramp

331

Fig. 20. Representative gamma-ray/neutron log for the upper San Andres Formation at Taylor-Link West field showing characteristic lithologies of stratigraphic intervals and their interpreted depositional environments.

Depositional facies The San Andres reservoir section in the

Taylor-Link West field comprises an upward- shallowing succession of (1) bryozoan-crinoid-fusulinid packstone/grainstone, (2) crinoid-brachiopod wackestone, (3) mudstone, (4) fusulinid wackestone, and (5) ooid-fusulinid grainstone/wackestone (Fig. 20). Pro- duction is from the grainstone facies that cap the sequence.

The grainstone interval makes up more than 80% of the San Andres reservoir. Thus, knowl-

edge of the geometry and internal heterogeneity of this interval is essential for understanding reservoir performance. The grainstone interval contains four facies that can be recognized in core and thin section, but not from log signatures. These facies are ( 1 ) ooid-peloid grainstone, (2) fine-grained bioclastic-peloid grainstone, (3) fusulinid-peloid wackestone/packstone, and (4) sponge floatstone.

Ooid-peloid grainstone comprises 60% of the facies and consists of 200-/tm-diameter peloids and poorly preserved ooids; variable

332 N. TYLER ET AL.

amounts of primary interparticle porosity occur in the grainstone facies. Thin, 0.3- to 0.9- m ( 1 - to 2-ft) -thick, beds of fine-grained bioclastic-peloid grainstone are locally interbedded with the ooid-peloid grainstone facies. The texture is characterized by 50- to 100-#m-size particles and abundant small separate vugs oc- curring as moldic pores after leached bioclasts, probably brachiopods and/or mollusks. The grainstone interval comprises 0.3- to 0.9-m- thick (1- to 3-ft) beds of fusulinid-peloid wackestone/packstone facies, containing between l0 and 20% fusulinid molds.

The sponge-floatstone facies of the grainstone interval is characterized by poorly preserved molds of unidentified (probably calcareous) sponges in a dense, commonly microfractured, light-tan micritic matrix. Po-

rosity is moldic and fracture related. This facies occurs in cores from three wells and averages 33 m ( 10 ft) thick.

The distribution of the grainstone facies was mapped using core descriptions and log-facies mapping techniques. The distinctive low gamma-ray signature of the ooid-peloid grainstone and bioclastic-peloid grainstone was used to generate an isopach map of these facies for the grainstone interval (Fig. 21). The out- standing feature of the grainstone-facies isopach is a northeast-trending belt of thick grainstone extending from the southeastern corner of Section 14 and adjacent southwestern corner of Section 13, through the north- eastern part of Section 13 and the southeastern quarter of Section 12. This northeast- trending belt is characterized by massive clean

E X P L A N A T I O N N /

• Producer ~ Ooid- peloid [~ o Injector groinstone facies o 2ooo ff

- - I I I I A I I r, Disposal Other facies o 6oo m [] Cored

Lease not in unit Contour interval 5 ft

Fig. 21. Isopach map of the San Andres grainstone interval at Taylor-Link West field showing feet of gamma-ray response greater than 50 API units. Isopach values greater than 17 m (55 ft) are interpreted to be areas where the ooid-peloid grainstone face is dominant.

DOLOMITIZED PLATFORM CARBONATE RESERVOIRS 3 3 3

gamma signatures indicating clean grainstone facies of between 17 and 23 m (55 and 75 ft) in thickness. Core examination reveals thick areas of ooid-peloid grainstone in this belt. Cores from the northwest-trending belts within the northeast trend show a high percentage of fusulinid-peloid packstone and sponge floatstone relative to ooid-peloid grainstone.

Cores from the east side of the ooid-grainstone belt show a vertical transition from more abundant fusulinid-peloid packstone facies, at the base of the grainstone interval, to clean ooid-peloid grainstone facies, toward the top. On the western margin of the reservoir, core descriptions show the grainstone interval to consist of a combination of dark-gray, fine- grained, bioclastic-peloid packstone, fusulinid-peloid packstone-grainstone, and lesser ooid-peloid grainstone facies. An area of mainly clean ooid-peloid grainstone with thin mud-rich interbeds occur between the ooid- grainstone belt and the more muddy facies on the western margin of the reservoir.

Fractures, microbreccias, and large vugs The response of the Taylor-Link reservoir

performance to the recently introduced waterflood shows significant deviations from the anticipated response, with an average oil/water ratio to date of 1-2%. Explanations for the high water and low oil volumes are found in the complex diagenetic history of the Taylor-Link carbonates that produced a system of interconnected large vugs, microbreccias, and fractures (referred to as a touching-vug system later in this report).

The origin of the fracture, large-vug, and breccia system found throughout the reservoir is important because these features have a significant effect on reservoir performance, particularly of the waterflood. Fractures described from the Taylor-Link West cores are grouped into simple and wide-aperture fractures and microbreccias (dense fracture net- works). Simple fractures are those having lit-

tie or no visible aperture, a straight, near- vertical orientation, and no lining or filling ce- ments. Wide-aperture fractures range from less than 1 mm to 4-5 mm in aperture width, as measured on core slabs, but are typically short (4-10 cm) and display a random orientation in individual core sections. Large vugs are 1- to 10-cm, oval-shaped voids commonly lined with scalenohedral calcite crystals that occur in the muddier sediments. Microbreccias occur in equidimensional areas several centimeters across that contain a dense network of randomly oriented, interconnected fractures outlining breccia fragments.

Fractures, microbreccias, and large vugs are far more common in cores taken from the low- energy lagoonal facies tract along the western portion of the reservoir than in the cores from the grainstone bar area. A strong vertical sep- aration in fracture and breccia density occurs in cores from the central grainstone bar complex. Fracturing and brecciation in the ooid- peloid grainstone facies are minor compared with that in the underlying fusulinid-wackestone and crinoid-brachiopod-wackestone facies. The distribution of large vugs also shows a similar correlation with depositional facies, marked by a downward increase in abundance of vugs in the San Andres section.

The origin of the pore system of fractures, microbreccias, and large vugs is found in the diagenetic history. Abundant evidence found in the Taylor-Link West cores indicates three major stages in the early diagenetic history of the San Andres Formation:

( 1 ) penecontemporaneous hypersaline-re- flux dolomitization; (2) emplacement of re- placive and cementing calcium sulfates; and (3) subaerial exposure accompanied by karsting and pervasive sulfate dissolution, with only minor (up to 10%) replacement of calcite by quartz, and calcification. The nearly complete dissolution of sulfate has been largely the cause of the extensive system of interconnected vugs, microbreccias, and fracture porosity.

334 N. TYLER ET AL.

Petrophysics The conversion of the geologic model to an

engineering model was accomplished by the use of data from the 12 cores from wells drilled during the 1983-85 redevelopment program. Core data were used exclusively because wireline logs from the original wells cannot be calibrated and dump flooding has altered the water resistivities sufficiently to make saturation calculations from modern logs question- able. Therefore, core descriptions, core analyses, and capillary pressure measurements from cores were used to reconstruct the original oil saturation and to define flow characteristics.

Volumetrics Capillary-pressure relationships were used to

calculate original water saturations because of the lack of reliable wireline-log saturation calculations. Average original water saturations in the producing interval were determined from 15 brine capillary pressure curves and 7 mer- cury curves, assuming an average height of 45 m (150 ft) above the free water level. Water saturations corresponding to 45 m (150 ft) above the free water level were read from the capillary pressure curves and plotted against porosity (Fig. 22a). Because water saturation is partly controlled by particle size (Lucia, 1983), the capillary pressure curves were divided into two particle size fabrics, a grain- dominated fabric with 200-#m average grain size, and a mud-dominated fabric with 15-#m average dolomite crystal size (Fig. 22b). The grain-dominated fabric corresponds primarily to the ooid-peloid facies of the grainstone interval, whereas the mud-dominated fabric characterized most of the other facies.

Stock tank original oil-in-place values were calculated for the grainstone and fusulinid wackestone intervals as well as for various facies within the grainstone interval. The results show a total of 48.2 MMbbl of original oil-in- place, which is in good agreement with volumetric calculations made using average data

(a) 1.o_

t -

O

o~

0.1 0.05

[] i ~ i i i i

0.10 0.20 0.25 Interparticle porosity

Swf = (~-2.o521 x 10 . 2 . 3039

Sw L = (~ -2.oooo x 10 - 2 . 6575

• Ooid grainstone (2001.tm grain size) [] Fusulinid wackestone

(15~m crystal size) QA12958c

(b)

"0

E v

.Q

¢0 q)

E

12.

1000 -

1 O0

10

0.1

/

,'///;." i 1 1 i i

4 10 20 40 Interpart ic le poros i ty

K L = (501.19 x 106"°°) X (~p8.OO)

K = (0.25097 X 106"°°) x (~)p6.3237)

• Ooid grainstone (200p.m grain size) [] Fusulinid wackestone

(15t.tm crystal size)

Fig. 22. Petrophysical/rock fabric relationships for Tay- lor-Link West field showing (a) the relationship between particle size, interparticle porosity, and water saturation and (b) the relationship between particle size, interparticle porosity, and permeability.

D O L O M I T I Z E D P L A T F O R M C A R B O N A T E R E S E R V O I R S 335

obtained from the Railroad Commission of Texas. The highest concentration was in the ooid-peloid grainstone facies, which contained 16.7 MMbbl.

Remaining mobile oil is calculated by subtracting produced oil and residual oil to waterflood from original oil-in-place. A residual oil saturation of 0.28 was used (Galloway et al., 1983 ). The volume of remaining mobile oil is calculated as approximately 20 MMbbl. The highest concentration is found in the ooid- peloid grainstone facies (Figs. 21 and 23), which has 7.1 MMbbl of remaining mobile oil, or 7 Mbbl per hectare ( 18 Mbbl per acre).

Flow model Volumetric calculations indicate 48 MMbbl

of original oil-in-place, of which 10.6 MMbbl has been produced, for a recovery efficiency of

22%. Engineering analysis of Taylor-Link production by the field operator indicates a proved reserve of 1.5 MMbbl of oil. This leaves 36 MMbbl remaining in the reservoir; 20 MMbbl is mobile oil recoverable by conventional methods, and 16 MMbbl is residual oil that will require advanced extraction techniques for recovery. Geologic and engineering characterization of the reservoir shows that the highest concentration of remaining mobile oil, 7 Mbbl/ha (18 MMbbl/acre), is in the ooid- peloid-grainstone facies of the central productive area.

Geologic observations show the reservoir to be composed of two pore-type groups: a matrix group and a touching-vug (fracture) group (sensu Lucia, 1983 ). Characterization of fluid flow can best be accomplished by separating touching-vug from matrix permeability. Ma-

J o . • I

30

o / o I 0 •

• \ o I • ~_ o ~ ; I ° • • •

o B • • I o 31

el I

o • "/ " 7

o o o ¢ ,~ o/ ~O -4 [

. . . . . .

° o o o o

/0

23 2 4 N 6

EXPLANATION 1 P r o d u c e r ~ D i s p o s a l o ~ooo ft I n j ec to r 13 Cored I i ~ i ?

. . . . Lease not in unit o 600 rn Contour interval 5 Mbbl/ocre

O O

Fig. 23. Isopach map of the remaining mobile oil at Taylor-Link West field in thousands of barrels per acre ( 1,000 bbl/ acre is approximately 405 bbl /ha) . Cross section E-E' shown in Fig. 24.

336 N. TYLERETAL.

trix permeability was estimated using relationships between grain size, interparticle porosity, and permeability developed for Taylor- Link West field (Fig. 22b). Fracture permeability was estimated by subtracting matrix permeability from laboratory whole-core permeability measurements.

Within the ooid-peloid grainstone facies of the grainstone interval, most of the permeability can be accounted for by matrix permeability. In the other facies of the grainstone interval, and in the fusulinid wackestone interval, the permeability is primarily due to the touching-vug pore system.

Reservoir model Fluid flow in the Taylor-Link West reservoir

can be characterized by two flow units: ( 1 ) the matrix permeability flow unit, which is dominated by interparticle permeability and contains high oil saturations, and (2) the touching-vug permeability flow unit, which is dominated by fracture, microbreccia, and large-vug permeability and contains low oil saturations (Fig. 24). Permeability values are similar in the two flow units but are controlled by different fabrics. Barriers to vertical flow are difficult to define because of the fracturing. However, the preferential concentration of calcite cementation of fractures, large vugs, and microbreccia in the fusulinid wackestone interval and in the lower part of the grainstone interval may result in local vertical permeability barriers.

The geologic/engineering model described above suggests that the high water volumes and low oil cut in this field result from injected water flowing through the touching-vug permeability flow unit, which has little oil saturation. This suggestion is supported by an injection test that demonstrated that at a rate of 600 barrels (bbl) of water per day, 52% of the fluid is entered the touching-vug permeability flow unit. When the rate was increased to 3,000 bbl water per day, 80% of the injected fluid entered that flow unit. Thus, as the injection rate

increases, less water enters the oil-saturated ooid-grainstone facies, resulting in lower oil production and lower oil cut. Production history shows that the oil cut and oil production rate are inversely proportional to the water injection rate.

A bottom-hole pressure map (unpublished data of T. Scott Hickman and Associates, Inc., 1988) demonstrates a low-pressure area corresponding to the structural high, despite the fact that the largest volume of water has been injected and produced in this area (Taylor- Link Corporation, pers. commun., 1989). The low-pressure area approximately corresponds to the area where the fractured wackestone interval is above the free water level. This coin- cidence of low pressure and the structurally high, fractured wackestone interval suggests that injection water is cycling through the fractures in the wackestone and not flooding the oil-saturated ooid-grainstone facies.

Pressure data in injection wells 102 and 104 suggest the presence of a permeability barrier between the grainstone and wackestone intervals. Before the wackestone interval in these two wells was sealed by cement, the bottom- hole pressures were 271 and 335 psi, respectively. After the wackestone was sealed with cement, the pressures in the grainstone interval were measured at 201 and 265 psi, respectively (pressures were normalized for depth). These data suggest that water injected into these wells entered the touching-vug system in the wackestone interval and that a horizontal flow barrier kept water from crossing upward into the grainstone interval.


Approximately 20 MMbbl of mobile oil will remain in the Taylor-Link field unless current production practices are improved. The reservoir characterization study shows that most of the remaining oil saturation in the Taylor-Link West reservoir is in the ooid-peloid grainstone facies, which has high matrix permeability.


E E' Well No. 59 Well No. 83 Well No. 67 Well No. 151

Permeability Vertical scale GRO.I tO I000 ,LI (m) (ft) Permeability t Permeability 15-1"- 50

i o ° ' : . . . . . ' . ,o , o o o . o- o

i ". . ~ ° ;*~*-°* : " : . . No ~ ~ .. ° °0 _.~,~.-~" ~ ~ ~ horizontal

i i i i i iiii:i: t. ) ~ ~ii~iiii:ii~ i ~::~ ,.L.~,.:,:~:,_.',~~ . . . . . . . i'i'i'i'i'~' ~ ~ ~ - . . . . . . . ~1:~--~* """ '""' 'F-. interval ' " ' " " ' " ~':'::::::::.Grainstone :::::.:. " Oil-water contact (+810ft)

"-FF. " '.

~ Oil-saturated reservoir rock with matrix permeability interval (ooid-peloid grainstone facies)

~ Touching-rug permeability with little matrix oil saturation (fusulinid, brachiopod, floatstone, packstone/wackestone facies)

~ Touching-vug permeability

- - Total permeability (core analysis) ...... Calculated matrix permeability

Fig. 24. Structural cross section showing the reservoir flow model for Taylor-Link West field. The right side of the depth logs shows matrix permeability in white and fracture permeability stippled. The gamma-ray log is shown on the left side. Line of section shown in Fig. 23.

Therefore, the recovery problem is to concentrate the waterflood in the ooid-peloid grainstone facies. This suggests three strategies for accomplishing this task: ( 1 ) infill drilling at closer spacing and penetrating only the grainstone interval, (2) sealing the fusulinid wackestone interval with cement in existing wells, and (3) using polymers to concentrate water injection into the ooid-peloid grainstone facies.

Conclusion: controls on reservoir heterogeneity in dolomitized platform carbonate reservoirs

The preceding four examples of San Andres and Grayburg reservoirs on the Central Basin Platform of the Permian Basin illustrate the wide variety of heterogeneity styles that control the distribution of porosity and saturation in dolomitized platform carbonate reservoirs.

Depositional facies patterns result in lateral and vertical discontinuities in porosity and permeability, which can control the distribution of original oil-in-place and the loci of highest primary and secondary oil production. Postdepositional diagenetic alteration over- prints depositional heterogeneity, resulting in additional, often overriding, lateral and vertical heterogeneities. These heterogeneities, when viewed on a field wide scale, constitute the internal reservoir architecture.

An understanding of internal reservoir architecture is critical for efficient production. Early in the life of a reservoir, knowledge of reservoir architecture is important, both for understanding the distribution of original oil in place and for understanding which parts of the reservoir will be the locus of the earliest production. The more completely the reservoir architecture is understood, the more effi- ciently wells can be located and perforation in-

338 N. TYLER ETAL.

tervals chosen. Later in the life of a reservoir, an understanding of reservoir architecture can help identify zones bypassed by waterflood or out of contact with wellbores. Advanced secondary recovery utilizing selective, geologically targeted, infill drilling and recompletion, together with waterflood optimization, can focus the waterflood on parts of the reservoir that have been bypassed by the flood front and, therefore, contain high concentrations of remaining mobile oil. Finally, for very mature reservoirs, knowledge of reservoir architecture is important for predicting the paths of injected carbon dioxide or surfactants, and for estimating the anticipated ultimate production of the field.

Acknowledgments

This research was funded by The University of Texas System as part of a larger study of hydrocarbon reservoirs on University Lands. We are grateful for access to data through the courtesy of the University Lands Office, the Railroad Commission of Texas, and numerous operators of oil fields on University Lands. In the course of this work we have benefited from assistance of and critical discussion with several colleagues, notably G.E. Fogg, C.M. Garrett, Jr., and C.R. Hocott. Publication is with the permission of the Director, Bureau of Economic Geology, The University of Texas at Austin.

References

Archie, G.E., 1942. The electrical resistivity log as an aid in determining some reservoir characteristics. Pet. Technol., 5: 54-62.

Bebout, D.G., Lucia, F.J., Hocott, C.R., Fogg, G.E. and Vander Stoep, G.W., 1987. Characterization of the Grayburg reservoir, University Lands, Dune field, Crane County, Texas. Univ. Texas, Austin, Bur. Econ. Geol. Rep. Invest., 168, 98 pp.

Chandler, M.A., Googin, D.J. and Lake, L.W., 1988. A mechanical permeameter for making rapid, non-de- structive permeability measurements. J. Sediment. Petrol., 59:613-615.

Chilingar, G.V., Zenger, D.H., Bissell, H.J. and Wolf, K.H., 1979. Dolomites and dolomitization. In: G. Larsen and G.V. Chilingar (Editors), Diagenesis in Sediments and Sedimentary Rocks. Developments in Sedimentology, 25A. Elsevier, Amsterdam, pp. 423- 536.

Finley, R.J., Laubach, S.E., Tyler, N. and Holtz, M.H., 1990. Opportunities for horizontal drilling in Texas. Univ. Texas, Austin, Bur. Econ. Geol., Geol. Circ., 90- 2, 32 pp.

Fogg, G.E. and Lucia, F.J., 1990. Reservoir modeling of restricted platform carbonates: geologic/geostatistical characterization of interwell-scale reservoir heterogeneity, Dune Field, Crane County, Texas, Univ. Texas, Austin, Bur. Econ. Geol. Rep. Invest., 190, 66 pp.

Galley, J.D., 1958. Oil and gas geology in the Permian Basin in Texas and New Mexico. In: L.G. Weeks (Ed- itor), Habitat of Oil-a Symposium. Am. Assoc. Pet. Geol., Tulsa, Okla., pp. 395-446.

Galloway, W.E., Ewing, T.E., Garrett, C.M., Tyler, N. and Bebout, D.G., 1983. Atlas of major Texas oil reservoirs. Univ. Texas, Austin, Bur. Econ. Geol. Spec. Publ., 139 pp.

Hine, A.C., 1977. Lily Bank, Bahamas, history of an ac- tive oolite sand shoal. J. Sediment. Petrol., 47:1554- 1581.

Jindrich, V., 1969. Recent carbonate sedimentation by tidal channels in the lower Florida Keys. J. Sediment. Petrol., 39: 531-553.

Leary, D.A. and Vogt, J.N., 1990. Diagenesis of the San Andres Formation (Guadalupian), Central Basin Platform, Permian Basin. In: D.G. Bebout and P.M. Harris (Editors), Geologic and Engineering Ap- proaches in Evaluation of San Andres/Grayburg Hy- drocarbon Reservoirs-Permian Basin. Univ. Texas, Austin, Bur. Econ. Geol., pp. 21-48.

Lucia, F.J., 1983. Petrophysical parameters estimated from visual description of carbonate rocks: a field classification of carbonate pore space. J. Pet. Technol., 35: 629-637.

Major, R.P., Bebout, D.G. and Lucia, F.J., 1988. Depo- sitional facies and porosity distribution, Permian (Guadalupian) San Andres and Grayburg Forma- tions, P.J.W.D.M. field complex, Central Basin Plat- form, West Texas. In: A.J. Lomando and P.M. Harris (Editors), Giant Oil and Gas Fields, a Core Work- shop. Soc. Econ. Paleontol. Mineral. Core Workshop, 12: 615-648.

Major, R.P., Vander Stoep, G.W. and Holtz, M.H., 1990. Delineation of unrecovered mobile oil in a mature dolomite reservoir: East Penwell San Andres Unit, Uni- versity Lands, West Texas. Univ. Texas, Austin, Bur. Econ. Geol. Rep. Invest., 194, 53 pp.

Ruppel, S.C. and Cander, H.S., 1988a. Effects of facies and diagenesis on reservoir heterogeneity: Emma San Andres field, West Texas. Univ. Texas, Austin, Bur. Econ. Geol. Rep. Invest., 178, 67 pp.

DOLOM ITIZED PLATFORM CARBONATE RESERVOIRS 3 3 9

Ruppel, S.C. and Cander, H.S., 1988b. Dolomitization of shallow-water platform carbonates by sea water and seawater-derived brines: San Andres Formation (Guadalupian), West Texas. In: Vijai Shukla and P.A. Baker (Editors), Sedimentology and Geochemistry of Dolostones. Soc. Econ. Paleontol. Mineral. Spec. Publ. No. 43: 245-262.

Shinn, E.A., 1983. Tidal flat environment. In: P.A. Scholle, D.G. Bebout and C.H. Moore (Editors), Carbonate Depositional Environments. Am. Assoc. Pet. Geol. Mem., 33: 171-210.

Tyler, N. and Banta, N.J., 1989. Oil and gas resources remaining in the Permian Basin: Targets for additional hydrocarbon recovery. Univ. Texas, Austin, Bur. Econ. Geol., Geol. Circ., 89-4, 20 pp.

Ward, R.F., Kendall, C.G.St.C. and Harris, P.M., 1986. Upper Permian (Gaudalupian) facies and their association with hydrocarbons-Permian Basin, West Texas and New Mexico. Am. Assoc. Pet. Geol. Bull., 70: 239- 262.

styles of heterogeneity in dolomitized platform carbonate reservoirs: examples from the central...

Documents