Method
Two approaches were used to calculate reservoir quality for the Miocene sandstones of the Southern Red Sea. First, the limited poroperm data that did exist were analyzed in terms of the controls on porosity and permeability, using the methodology of Cade et al.
(1994). The results from this analysis were compared with the qualitative petrographic descriptions.
Second, reservoir quality data were evaluated using the broad geological data available for the area. The methods for porosity and permeability synthesis are given below.
Correct prediction of porosity requires that the vol-ume losses due to compaction and cementation are quantified. Permeability prediction further requires knowledge of the grain size and sorting characteris-tics, cement types, and their distribution.
Porosity loss due to compaction was calculated using the methodology of Gluyas and Cade (this volume).
Cement types and volumes were calculated on the basis
of a BP Exploration in-house regional diagenesis study (Primmer, 1993; Primmer et al., this volume), in which the links between sand mineralogy at deposition, depo-sitional environment, burial, and thermal history were quantified.
Grain size and sorting data were adopted from the existing well information. Permeabilities were calcu-lated from the cement data and estimates of grain size and sorting, using the sphere pack modeling approach of Cade et al. (1994).
Data Analysis
The reservoir intervals of the two Al Meethag wells contain quartz-poor, feldspar- and volcaniclastic-rich, fine- to medium-grained sandstones. Their diagenetic history is complex, with calcite, dolomite, chlorite, smectite, zeolite, quartz, illite, and halite cements.
Given that the sediments are at most 15 m.y. old and even now buried only to 1.5–1.7 km, all of these processes must have occurred in a short geological time and at shallow depth.
An attempt to construct an empirical porosity depth plot proved futile. The problems encountered are illus-trated in Figures 3 and 4. In short, there are too few data from which to draw any valid conclusions as to how, or if, porosity varies with depth in this basin.
Porosity and permeability data from conventional core analysis for the two Al Meethag wells are plotted in Figure 5. Plotted on the same graph are modeled curves for the porosity-to-permeability relationship in similar grain size (fine- to medium-grained) clean, compacted, and/or quartz cemented sandstones (Evans et al., this volume). Most of the data from the two wells describe two distinctly different prolate clusters of reasonably similar permeability range but significantly different porosity range. The outliers to these two trends are medium/coarse grained sand-stones, carbonate cemented sandsand-stones, and, in one instance, a fractured core analysis plug.
The poroperm data for both wells lie well below the modeled clean sand lines. The steep porosity-to-permeability gradient is indicative of a sand with a large proportion of poorly interconnected porosity:
either intragranular secondary porosity or micropo-rosity trapped between clay fibers and plates. The sim-ilarity of poroperm gradients in the two wells was taken to indicate that the process controlling permeability evolution in both was similar. This relationship did not hold for porosity. The inferred importance of clay in controlling the permeability of these sandstones is fully supported by the petrographic descriptions.
In order to explain the porosity difference between the wells, a process is needed to reduce porosity with only a minor (relative to the clays) effect on permeabil-ity. At the high porosities seen in these cores, two processes could have been responsible: compaction and/or syntaxial quartz precipitation (Cade et al., 1994).
There is insufficient difference in burial depth of the two sandstones to account for the porosity difference in terms of compaction alone, even when the 13 MPa overpressure in well W1 is taken into account (Robin-son and Gluyas, 1992). It is possible that quartz
cementation may account for much of the difference.
This suggestion is supported by the qualitative descriptions of the petrography of the sandstones from the two wells. Quartz cement was described from W2 but not from W1. Modeling data (see the following section) also lend some support to this suggestion.
No equivalent quantitative reservoir quality data were available for the sandstones derived from the Pre-Cambrian acid igneous and gneiss terrains.
Data Synthesis—Modeled Poroperm Evolution The following criteria were used to construct a semiquantitative diagenetic history for the Miocene sandstones (Figures 6, 7).
Volcaniclastic Sandstones
• Volcaniclastic sandstones are likely to react in situ at temperatures below 25°C to produce aluminium and iron smectites, zeolites (clinoptilolite), and chlorite. By 75°C, the same assemblage can further react to produce higher temperature zeolites at the expense of aluminum smectite. At 100°C, lau-monite is likely to be the stable zeolite alongside albite and quartz and the persistent chlorite (Bloch and Helmold, 1995; Primmer et al., this volume).
• At temperatures >70°C, burial rates exceeding
~100 m.y.–1(meters per million years), and heat-ing rates exceedheat-ing 2°C m.y.–1, quartz is likely to be an important cement phase during open sys-tem diagenesis (Gluyas et al., 1993).
Al Meethag 1 Al Meethag 1
overpressure corrected
Al Meethag 2
-2000 -1800 -1600 -1400 -1200 -1000 -800 -600 -400 -200 0
0 5 10 15 20 25 30 35 40
Porosity (%)
Depth (m, subsea)
Figure 3. Porosity depth plot for cored intervals from Antufash License. Al Meethag 1 is plotted twice, at its current burial depth and at its hydrostatic equivalent burial depth. The Miocene sands in Al Meethag 1 are overpressured by 9 MPa (1300 psi); 1 MPa is ~80 m of burial in a hydrostatic system at these burial depths (Gluyas and Cade, this volume). Data are averages for wells; individual plug data are plotted in Figure 5.
Circle = Al Meethag 1; square = Al Meethag 2; diamond = Al Meethag 2 overpressure corrected.
Figure 4. Porosity–depth plot for log data from intervals in Al Meethag 1. Porosity and shale percentages were calculated from a combination of neutron density and resistivity logs.
porosity (%)
burial depth (m)
-1800 -1700 -1600 -1500 -1400 -1300 -1200 -1100 -1000 -900 -800
0 10 20 30 40
0-10% shale 11-20% shale 21-50% shale 50-75% shale
>75% shale core
• There are sufficient components for illite to form, although significant quantities are unlikely to exist at temperatures below about 100°C (Small et al., 1992).
• In a depositional system containing some marine influence, a little early diagenetic carbonate is to be expected (Bjørkum and Walderhaug, 1990); some of this cement is likely to have been dissolved and reprecipitated during the later stages of diagene-sis. Some decarboxylation carbonate may have also been added (Gluyas and Coleman, 1992).
• In sequences interbedded with evaporites, there is a possibility that any residual porosity will have been filled by halite and other evaporite minerals.
This point is speculative. We do not yet have information that would allow us to describe the process or timing of such cementation.
• Finally, we made the assumption that of the compo-nents required for silicate mineral cementation, only silica is likely to have been imported to the sands in quantities large enough to appreciably affect poros-ity (Gluyas and Coleman, 1992). This point could be considered controversial given the current debate in the literature with respect to the sources of silica for quartz cementation. However, we imply no scale of transport here; import could mean derivation from local silica sources, such as nearby pressure dissolu-tion seams, or more distant sourcing from unspeci-fied sources. Other elements such as potassium and
0.01 0.1 1 10 100 1000 10000 100000
0 5 10 15 20 25 30 35 40
Porosity (%)
Permeability (md)
Figure 5. Porosity and permeability data from the cored intervals of Antufash License wells. Al Meethag 1 (blue) high porosity; Al Meethag 2 (brown) lower porosity. The 900-md outlier is from a fractured plug; the remaining outliers are from thin, medium-grained sandstones.
Figure 6. Synthesized diagenetic history for volcani-clastic sandstones in the Antufash License.
Figure 7. Synthesized diagenetic history for arkosic sandstones in the Antufash License.
15 10 5 0
Deposition Carbonate precipitation
Chlorite precipitation Compaction Smectite &
zeolite ppt.
Quartz precipitation Illite precipitation
Oil migration
Porosity &
permeability evolution
high
low
porosity
permeability
Ma
Deposition Carbonate precipitation Compaction Kaolinite precipitation
Quartz precipitation Illite precipitation
Oil migration
Porosity &
permeability evolution
high
low
porosity permeability
Ma
and / or
15 10 5 0
aluminum are likely to have been supplied inter-nally (Gluyas and Leonard, 1995).
Arkosic Sandstones
The diagenesis of arkosic sandstones is likely to have been very different (Figure 7). The most common low-temperature product is likely to have been kaolin-ite, precipitation of which could have accompanied ingress of undersaturated water of near-surface, mete-oric, or connate origin (Gluyas, 1985; Bjørkum et al., 1993). In an open system, quartz is likely to have pre-cipitated once the sandstones exceeded 70°C. By 100°C, illite will have been the most likely clay phase to precipitate (Small et al., 1992).
The presence of carbonate and evaporite cement is likely to be common to both the volcanic and arkosic sourced sandstones, since both cements would have been supplied from largely outwith the sandstone.
Estimating Porosity and Permeability—Antufash-1 The location for Antufash-1 is shown in Figure 1. The area was a poorly explored anticline/diapir fairway comprising upper-middle Miocene reservoirs. The prospect lay above a well-defined NNW-SSE–trending salt-cored anticline with four-way dip closure through-out the Pliocene and Miocene sections. Multiple reser-voirs were expected to be present in transgressive sands
of the upper-middle Miocene sections. Four such reser-voirs were included in the volumetric calculations. The seal was expected to be salt. Depth to crest of the upper-most prospective horizon was estimated at 850 m, and gross reservoir thickness was calculated at 450 m. The surface temperature is 25°C, and the present thermal gradient is 60°C km–1. The reservoir was expected to be normally pressured. The sands were estimated to be fine grained and well sorted. The sediment source area was thought to be along Wadi Mawr (Figure 1), which drains dominantly granitic terrain and is likely to have yielded arkosic sands.
Modeled Porosity
Using the above criteria, the effects of compaction and quartz cementation were modeled. The likely effect of carbonate cement on bulk porosity was assumed to be small, by analogy with the Al Meethag wells, while the potential for evaporite plugging of porosity was estimated to be large.
The modeled porosity–depth curve for either arkosic or volcaniclastic sands in the Antufash acreage is shown in Figure 8. In order to generate such a porosity–depth curve, several simplifying assump-tions were made. Those that might have introduced a systematic error in the porosity estimate are:
• Formation of overpressuring during burial, lead-ing to a low porosity estimate
• Conversion of labile volcaniclastic grains to duc-tile “clay clasts,” which are more susceptible to compaction than rigid grains, leading to an over-estimation of porosity
The sensitivity of the porosity estimate at 1 km bur-ial, errors in pressure, or ductile grain content are examined in Table 2.
The porosity gradient associated with cementation, 16 ± 5% km–1, is based on the empirical observation that subregional porosity gradients resulting from quartz cementation covary with thermal gradients (Rønnevik et al., 1983) (Table 3; Figure 9).
Modeled Permeability
In addition to the data required for porosity calcula-tion, data on grain size, sorting, and cement mineral-ogy were required for the permeability estimate. Grain size data were taken from the Al Meethag wells and, for want of hard data, sorting was assumed to be mod-erate. Two cases were run for the cement mineralogy.
For the volcaniclastic sands, a case based on pervasive, pore-lining smectite and/or zeolites was calculated, while the arkosic sand calculation was based on a pore structure comprising clean, “grain-lined” pores with randomly scattered kaolinite-filled pores.
Modeled curves of the porosity to permeability relationship for both the arkosic and volcaniclastic cases are shown in Figures 10 and 11.
Using the porosity/permeability relationships and porosity-to-depth relationships, it is possible to determine depth equivalencies for the permeability cutoffs (100 md, 10 md) (Table 4).
Figure 8. Synthesized and simplified porosity–depth relationship for all types of Miocene sandstones of the Antufash License. Curve C-C is a pure com-paction curve for a rigid grain hydrostatically pres-sured sandstone; curve Q-Q is the expected porosity decline of 16% km–1hung from a depth equivalence of ~70°C (800 m) at the time of silicate cementation (quartz, clays, and/or zeolites). Q+ and Q– are the potential ±5% km–1variance on the expected value.
Porosity (%)
Q-Prospect-Specific Estimates of Porosity and Permeability
The estimated depth to top reservoir was 850 m. At this shallow depth, the risk on reservoir effectiveness was low for arkosic and volcaniclastic sandstones.
Both types of sandstones were likely to have porosities
~36% and permeability >2 darcys.