The Effect of Grain-Coating Microquartz on Preservation of ...

ABSTRACT

Clay coatings have been widely accepted bymany workers as an explanation for preservinghigh porosity in deeply buried sandstones, but fewworkers have realized that similar effects can beproduced by microcrystalline quartz coatings. Thisphenomenon can be expected only under specialcircumstances, but in such cases it can have pro-found consequences for exploration.

In the Central Graben area of the southern NorthSea, unusually high porosity (20–27%) and perme-ability (100–1000 md) are found in certain zones inUpper Jurassic sandstones at depths of 3.4–4.4 km.The porosity in these zones is 5–15% higher thanexpected based on average porosity–depth trendsfrom Brent and Haltenbanken sandstones. We pro-pose that the high porosity is due to continuousgrain coatings of euhedral microcrystalline quartzcrystals that are 0.1–2 µm thick. The distribution ofmicrocrystalline quartz coatings is controlled bythe presence of siliceous sponge spicules (Rhaxella),which implies a sedimentological control on thereservoir quality. We present a thermodynamicmodel showing how continuous microcrystallinequartz coatings inhibit development of normalmacrocrystalline quartz overgrowths sourced main-ly from stylolites. High porosities in parts of variousUpper Jurassic oil fields (Ula and Gyda) have previ-ously been explained by inhibition of quartzcementation by early hydrocarbon charge. We sug-gest that the microcrystalline quartz coatings pro-vide a more plausible explanation.

INTRODUCTION

In Upper Jurassic sandstones in the CentralGraben area of the North Sea, anomalously highporosity (>20%) and permeability can be found atdepths greater than 4 km. The porosity and perme-ability are significantly higher than expected basedon average porosity–depth trends from the MiddleJurassic Brent Sandstones in the northern NorthSea or Haltenbanken Sandstones (Bjørlykke et al.,1992; Giles et al., 1992). Diagenetic quartz cemen-tation is commonly regarded as the most importantporosity-reducing process at depths greater than2500 m (Bjørlykke et al., 1992; Giles et al., 1992),and anomalously high porosity and permeabilityvalues in deeply buried sandstones commonly havebeen explained by early hydrocarbon trapping inthe reservoir (e.g., Selley, 1978; Sommer, 1978;Gluyas, 1985; Gluyas et al., 1990, 1993; Robinsonand Gluyas, 1992b; Burley, 1993; Rothwell et al.,1993) or highly overpressured sandstones(Bjørlykke et al., 1989, 1992; Ramm and Forsberg,1991; Ramm, 1992; Ramm and Bjørlykke, 1994).High-quality sandstones in the Gyda and Ula fieldscommonly have been explained by early hydrocar-bon trapping (Bjørnseth and Gluyas, 1991;Rothwell et al., 1993); however, Giles et al. (1992),Ramm (1994), and Walderhaug (1994a) concludedthat there was no relationship between the pres-ence of hydrocarbons and high porosities.Walderhaug (1994b) also concluded that there wasno relationship between overpressure and theamount of quartz cement, and Bjørkum (1984,1994, 1996) showed that the dissolution of quartzis not a pressure-driven process.

Grain coatings, such as early diagenetic chlo-rite, have generally been observed to inhibitquartz cementation (e.g., Heald and Larese, 1974;Ehrenberg, 1993). Anomalous reservoir porosityat depth has been reported for phosphatic sand-stones, which also contain diminuative euhedral”diagenetic quartz (Cazier et al., 1995). To whatextent porosity has been preserved by phosphat-ic or microquartz coatings is unclear, however.The purpose of this study is to evaluate the

1654 AAPG Bulletin, V. 80, No. 10 (October 1996), P. 1654–1673.

©Copyright 1996. The American Association of Petroleum Geologists. Allrights reserved.

1Manuscript received May 26, 1995; revised manuscript received January26, 1996; final acceptance May 15, 1996.

2Statoil, 4035 Stavanger, Norway.We thank BP and the license partners for the allowing us to publish the

well data. Comments by the AAPG reviewers S. Bloch, R. E. Larese, and K. O. Stanley are sincerely appreciated. We also thank Olav Walderhaug andN. R. Rothwell for constructive comments during the work. Statoil a.s isthanked for encouraging us to publish the results.

The Effect of Grain-Coating Microquartz on Preservationof Reservoir Porosity1

Nils Einar Aase, Per Arne Bjørkum, and Paul H. Nadeau2

importance of microquartz coatings for porositypreservation during burial. We also present athermodynamic explanation for why coatings ofmicroquartz crystals on the quartz grains tend toinhibit development of macroquartz cementsourced from stylolites.

Observations of microquartz have been pub-lished in numerous studies (see McBride, 1989;Ramm and Forsberg, 1991; Vagle et al., 1994; andreferences therein). The fact that microcrystallinequartz coats quartz grains has been known forsome years, but its source, distribution, and effecton high porosities in deeply buried sandstones hasnot been fully evaluated or understood. This pro-cess can be critically important for hydrocarbonexploration because it provides good reservoirquality at depths far below the economic basementdefined on the basis of typical sandstones that lackmicroquartz coatings. This paper presents resultsof a study from the Ula and Gyda fields on theNorwegian continental shelf of sandstones thatexhibit anomalously high porosities.

GEOLOGICAL SETTINGS

The Gyda and Ula fields are located in blocks 2/1and 7/12, respectively, along the Ula trend in theNorwegian part of the southern North Sea CentralGraben (Figure 1). The reservoirs of both fields areUpper Jurassic shallow-marine sandstones, with theGyda Sandstone in the Gyda field and the UlaSandstone in the Ula field (Spencer et al., 1986;Home, 1987). The Gyda Sandstone reservoir is rela-tively deep (3650–4165 m), hot (155°C at 4155 m),and overpressured (605 bar) (Rothwell et al.,1993). The Ula Sandstone is located at a depth of3400–3800 m, with a reservoir temperature of143°C and a pressure of 483 bars at 3450 m(Spencer et al., 1986; Home, 1987). The UpperJurassic sandstones are predominantly fine andmedium grained, moderately well sorted, relativelyhomogeneous, and variably argillaceous, and weredeposited in offshore marine-shelf settings (Brownet al., 1992; Rothwell et al., 1993). The thickness ofthe Upper Jurassic section varies considerably as a

Aase et al. 1655

Figure 1—Map of the Central Graben area located in the southern part of the Norwegian shelf with main structuralelements and major discoveries. Study wells, marked with letters, are located in blocks 1/3, 2/1, 7/11, and 7/12(modified after Brown et al., 1992).

1656 Effect of Microquartz on Porosity

Table 1. List of Inspected Samples from 15 Wells and Estimated Values Based on SEM Investigations

Well Depth Porosity Permeability Grain SizeWell Symbol (m RKB*) (%) (md) (mm) Microquartz** Macroquartz** Clay** Carb.**

2/1-9 A 4067 16.7 0.7 0.1 2 2 4 12/1-9 A 4068.68 2.2 0.01 0.1 0 0 1 52/1-9 A 4075 13.3 2.4 0.12 0 3 3 32/19 A 4106 13.2 0.9 0.12 0 5 4 12/1-9 A 4120 11.3 1.25 0.12 0 5 3 12/1-9 A 4121.67 10.8 0.33 0.12 0 5 4 32/1-9 A 4122 4.5 0.02 0.12 0 5 3 42/1-9 A 4123 23.7 468 0.12 3 2 2 02/1-9 A 4126 23.5 322 0.12 3 2 2 02/1-9 A 4130.33 23.3 199 0.12 4 2 3 02/1-9 A 4133.67 23 191 0.12 4 2 4 02/1-9 A 4136.67 21.4 40 0.12 4 2 4 02/1-9 A 4142.33 25.4 532 0.12 4 2 2 02/1-9 A 4148.33 24.3 157 0.12 4 2 3 02/1-9A B 4147.7 14.6 40.3 0.1 1 4 2 12/1-9A B 4152.7 17.9 87.7 0.1 1 4 2 12/1-9A B 4153.35 16 33.4 0.1 1 4 2 12/1-9A B 4169.35 19.5 138 0.1 3 3 3 12/1-9A B 4184.7 23.2 549 0.1 3 2 2 12/1-9A B 4199.35 26 808 0.1 3 1 1 12/1-9A B 4223.7 3.2 0.02 0.1 0 0 0 52/1-9A B 4233.7 23.2 34 0.1 5 2 2 12/1-9A B 4255.35 18.3 5.4 0.1 2 3 4 12/1-9A B 4273.7 23.4 41.3 0.1 5 2 2 12/1-9A B 4287.35 22.2 13.2 0.1 5 1 4 12/1-9A B 4299.7 12.6 0.11 0.1 5 1 5 12/1-9A B 4307.35 25.7 1530 0.1 4 2 2 12/1-9A B 4315.8 8.2 0.25 0.1 2 5 2 12/1-6 C 4217 24 204 0.1 3 2 2 12/1-6 C 4225.35 23.7 164 0.1 4 2 2 22/1-6 C 4230.35 22.3 184 0.1 4 1 2 22/1-6 C 4251 9.7 1.5 0.1 0 4 2 12/1-6 C 4308 9.2 0.63 0.1 0 4 2 12/1-6 C 4328.65 21.4 44 0.1 4 2 3 2– D 4183.4 13.5 0.17 0.07 5 2 4 3– D 4192.7 12.1 0.17 0.08 5 2 4 3– D 4200.55 15.8 1.3 0.08 5 2 4 3– D 4208.65 20.1 9.3 0.1 4 2 3 2– D 4213.6 23.5 139.5 0.1 3 2 2 1– D 4215.4 26.3 310 0.1 3 1 2 1– D 4224.7 21.5 77 0.1 3 3 3 1– D 4246.5 26.8 1019 0.15 3 2 1 02/1-2 E 3332 14.8 0.26 0.08 4 2 4 22/1-2 E 3334.5 15.7 0.21 0.08 4 2 4 22/1-2 E 3336.15 16.8 0.52 0.08 4 2 3 22/1-3 F 3828.87 13.8 11 0.1 1 3 1 12/1-3 F 3841.12 23.7 253 0.1 3 2 1 12/1-3 F 3843.62 25.6 418 0.1 4 2 1 12/1-3 F 3851.75 20.4 11 0.1 2 2 2 12/1-3 F 3855.5 18.9 0.48 0.1 1 3 3 12/1-3 F 3860.87 20.6 4.3 0.1 1 3 3 12/1-4 G 4080.35 13.2 3.2 0.1 0 4 2 12/1-4 G 4093.7 15.2 2.8 0.1 0 4 2 12/1-4 G 4097.35 20.5 23 0.1 3 2 4 22/1-4 G 4102.7 21.7 32 0.1 3 2 4 22/1-4 G 4105.7 22.8 21 0.1 3 2 4 22/1-4 G 4117.35 15.1 0.32 0.08 5 1 4 22/1-8 H 3907.23 17.9 1.8 0.1 3 1 2 22/1-8 H 3911.97 25.5 356 0.1 4 1 2 2

result of erosion during the middle–late Volgian,and also as a result of differential subsidence.

SAMPLES AND METHODS

Our results are based on scanning electronmicroscopy (SEM), x-ray diffraction (XRD), andthin-section analyses. The data set includes SEMobservations of 100 selected core material samplesfrom 15 wells in Ula and Gyda fields and the sur-rounding area. The wells are from blocks 1/3, 2/1,

7/11, and 7/12. Identification for three of the wellscannot be provided (Table 1). All samples wereselected from cores for which routine core analyses(porosity, permeability, and grain density) wereperformed. To avoid the effects of primary sandquality variation on the reservoir quality, weexcluded low-porosity zones resulting from highclay contents or carbonate cements from the dataset. The data set was also selected such thatchanges in grain sizes were minimal.

The SEM examinations were made using a JeolJSM 840 on freshly fractured, stub-mounted samples

Aase et al. 1657

2/1-8 H 3912.72 25.5 755 0.1 3 1 3 22/1-8 H 3914.85 22.1 30.4 0.1 3 1 3 22/1-8 H 3916.53 19.7 9.7 0.1 2 1 2 22/1-8 H 3921.25 19.4 5.7 0.1 4 1 3 27/12-4 I 3465.35 21.9 1648 0.17 0 3 1 07/12-4 I 3479.3 20.8 85 0.1 2 3 2 07/12-4 I 3488.8 22 328 0.1 3 3 1 07/12-4 I 3499.1 22.6 309 0.1 2 2 1 07/12-5 J 3850.7 12.5 0.44 0.1 0 3 4 17/12-5 J 3860.7 12.8 7.5 0.1 0 4 4 17/12-5 J 3880.31 11.1 0.59 0.1 0 4 4 17/12-5 J 3889.98 11.9 1.4 0.1 0 4 4 17/12-5 J 3902.01 12.8 0.48 0.1 0 4 4 1– K 4172.77 24.6 50 0.1 5 1 3 1– K 4181.3 23.8 32 0.1 4 2 3 1– K 4183.9 27.4 320 0.1 4 0 2 1– K 4195.29 15.4 1013 0.4 4 3 1 1– K 4202.73 10.8 0.14 0.1 2 5 3 1– K 4210.65 12.9 0.25 0.1 2 5 3 1– L 4114.65 24.8 171 0.08 5 3 1 1– L 4123.25 20.5 30 0.08 5 3 2 2– L 4128.61 21.8 536 0.12 3 3 2 1– L 4129.25 20.8 1353 0.2 3 2 2 2– L 4148 14.1 0.83 0.08 0 2 5 17/12-6 M 3429.03 20.5 2144 0.25 0 1 1 17/12-6 M 3442.71 24.4 2620 0.25 0 1 1 17/12-6 M 3473.2 21.9 2478 0.2 0 1 1 17/12-7 N 3811.35 15.8 6 0.1 0 4 2 27/12-7 N 3817.65 22.2 49.4 0.1 2 3 2 17/12-7 N 3827.65 16.5 2.3 0.1 0 4 2 27/12-7 N 3831 17.8 195 0.2 1 3 2 37/12-7 N 3831.65 24 1720 0.2 3 1 2 17/12-7 N 3836.35 14.8 2.5 0.1 0 4 2 27/12-2 O 3401.78 19.6 10 0.1 4 2 3 27/12-2 O 3410.05 19.5 127 0.1 3 2 3 17/12-2 O 3418.34 21.7 2236 0.2 0 1 1 17/12-2 O 3437.3 24.3 1461 0.15 0 1 1 17/12-2 O 3444.22 20 917 0.15 0 1 1 17/12-2 O 3456.95 27.9 933 0.1 3 1 2 17/12-2 O 3474.15 20 140 0.1 1 1 2 1

*RKB = referenced to kelly bushing.**Classification amounts: 0 = not observed, 1 = very minor, 2 = minor, 3 = medium, 4 = much, 5 = very much.

Table 1. Continued.

Well Depth Porosity Permeability Grain SizeWell Symbol (m RKB*) (%) (md) (mm) Microquartz** Macroquartz** Clay** Carb.**

Figure 2—SEM and thin-section photomicrographs of sandstones with different types and amounts of quartz cement. (A)No microquartz cement is present, but large amounts of euhedral macroquartz cement are present. Porosity is typicallyaround 10%. (B) Minor amounts of microquartz are present, but macroquartz cement content is substantial. The micro-quartz does not coat grains completely and macroquartz cement may develop on available surfaces. Porosity is typically10–15%. (C and E) Optimal amounts of microquartz cement are present, coating all quartz grains with a thin layer. Macro-quartz cement is essentially absent and porosity is usually above 20%. Note some larger crystals on photomicrograph Cand the fibrous diagenetic illite on photomicrograph E. (D and F) Close-up of microquartz cement coatings. Typical crys-tal size is 0.2–0.8 µm, with larger crystals ranging in size from 5 to 20 µm commonly observed. (G) Optimal to excessivemicroquartz cementation. The microquartz crystals are coating all surfaces, but also precipitate as aggregates in openpores. Aggregates in the center of the photomicrograph may be the remains of dissolved sponge spicules. Porosity ishigh, but permeability may be reduced if aggregates are blocking pore throats. (H) Excessive microquartz cementation.Silica precipitates both as cryptocrystalline quartz and as aggregates of microcrystalline quartz. Both porosity and perme-ability are usually low. (I) Thin-section photomicrograph of an optimal amount of microquartz cement (arrows) coatingall quartz grains with a thin layer. (J) Excessive microquartz cementation. Silica precipitates both as cryptocrystallinequartz and as aggregates of microcrystalline quartz. Note the cryptocrystalline layer thickness of approximately 10 µm.

coated with gold. Based on SEM observations, wemade a preliminary classification of the sand-stones related to the type and amount of quartzcement. We also visually estimated grain size, claycontent, and carbonate cement content. Thesedata are presented in Table 1. A set of photomicro-graphs of the sandstones shows the differentamounts of microquartz observed (Figure 2).

In wells A and B, XRD analyses were done togive better control on the primary sand composi-tion, particularly regarding clay and carbonatecontents. The XRD samples were run on a PhilipsPW 1710 using Cu Kα radiation at a scan rate of 1°2θ/min in the 2–50° 2θ range. SEM and XRD workwas carried out by Statoil’s geochemical laborato-ry. XRD analyses data are presented in Table 2.One microphotograph was selected for point-counting analyses (500 counts, two times). Thepurpose of these analyses was solely to estimatethe optimal amount of grain-coating microquartzobserved in a sample with regard to porositypreservation.

SEM OBSERVATIONS

The amounts of microquartz and macroquartzcement in all of the samples studied by SEM havebeen given a relative value from 0 to 5, where 0 =not observed, 1 = very minor, 2 = minor, 3 = mod-erate, 4 = high, and 5 = very high. In this classifica-tion, a moderate amount of microquartz was deter-mined to be optimal with regard to reservoirquality. We designated sandstone samples as havinga moderate amount of microquartz when there wasa thin and continuous rim of microquartz aroundall the quartz grains and no or little pore-filling“aggregates” (see the following section). Theamounts of clay (including diagenetic illite) andcarbonate have been estimated in the same way.Grain size was also estimated by SEM (Table 1). Theoccurrence of quartz cement within the sandstonesand corresponding reservoir quality can be dividedinto four groups, as shown in Table 3. The optimalamount of microquartz cement was estimated tobe 2–3% from point counting of a selected thin

Aase et al. 1659

Figure 2—Continued.


Tab

le 2

. X

RD

Dat

a fr

om

Wel

ls A

an

d B

Dep

thD

epth

Qu

artz

K-F

eld

.P

lag.

Cal

cite

Do

l.P

yrit

eH

alit

eK

aol.

Mic

a-Il

lite

Ch

l.-B

erth

.T

ota

lP

or.

Per

m.

(RK

B)*

(TV

D)*

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(%)

(md

)

Wel

l A40

67.0

040

54.0

074

.49.

07.

51.

51.

50.

60.

00.

02.

23.

310

0.0

16.7

0.70

4068

.68

4055

.67

46.3

8.9

6.5

35.6

0.7

0.6

0.0

0.0

1.3

0.0

100.

02.

20.

0140

75.0

040

62.0

072

.012

.87.

53.

11.

10.

40.

00.

01.

81.

210

0.0

13.3

2.37

4106

.00

4093

.00

70.7

13.3

10.8

0.5

1.9

0.3

0.3

0.0

1.3

1.2

100.

013

.20.

9041

20.0

041

07.0

076

.710

.28.

60.

02.

80.

40.

00.

01.

40.

010

0.0

11.3

1.25

4121

.67

4108

.67

72.2

10.9

9.6

0.5

4.9

0.3

0.0

0.0

1.7

0.0

100.

010

.80.

3341

22.0

041

09.0

059

.613

.211

.40.

712

.40.

90.

00.

01.

90.

010

0.0

4.5

0.02

4123

.00

4110

.00

73.8

11.6

9.6

0.0

2.7

0.5

0.0

0.0

1.7

0.0

100.

023

.746

8.00

4126

.00

4113

.00

75.1

11.5

8.3

0.0

1.4

0.5

1.8

0.0

1.4

0.0

100.

023

.532

2.00

4130

.33

4117

.30

73.2

12.0

10.5

0.0

2.3

0.5

0.0

0.0

1.5

0.0

100.

023

.319

9.0

4133

.67

4120

.67

71.6

10.4

10.7

0.0

3.1

0.5

0.5

0.0

2.6

0.6

100.

023

.019

1.00

4136

.67

4123

.67

68.9

11.0

11.8

0.0

3.9

0.6

0.0

0.0

3.0

0.8

100.

021

.440

.00

4142

.33

4129

.33

68.9

15.2

9.2

0.0

3.3

0.6

0.7

0.0

2.1

0.0

100.

025

.453

2.00

4148

.33

4135

.33

73.7

11.6

8.9

0.7

2.8

0.6

0.0

0.0

1.6

0.0

100.

024

.315

7.00

Wel

l B41

47.7

039

73.1

873

.011

.89.

61.

41.

30.

30.

00.

71.

80.

010

0.0

14.6

40.3

041

52.7

039

76.1

575

.011

.28.

91.

31.

60.

20.

00.

01.

10.

610

0.0

7.9

87.7

041

53.3

539

76.5

473

.112

.58.

91.

11.

40.

40.

30.

01.

30.

910

0.0

16.0

33.4

041

69.3

539

86.0

068

.314

.510

.31.

41.

80.

40.

30.

62.

40.

010

0.0

19.5

138.

0041

84.7

039

95.0

074

.013

.38.

70.

81.

60.

50.

00.

01.

10.

010

0.0

23.2

549.

0041

99.3

540

03.4

869

.414

.410

.50.

71.

90.

20.

00.

01.

71.

210

0.0

26.0

808.

0042

23.7

040

17.4

235

.98.

16.

645

.71.

00.

20.

00.

02.

50.

010

0.0

3.2

–42

33.7

040

23.0

868

.19.

310

.22.

22.

30.

70.

50.

04.

32.

410

0.0

23.2

34.0

042

55.3

540

35.2

165

.28.

912

.11.

42.

20.

80.

90.

06.

22.

410

0.0

18.3

5.40

4273

.70

4045

.38

67.2

9.5

11.8

0.9

2.0

1.1

0.5

0.0

5.6

1.5

100.

023

.441

.30

4287

.35

4052

.87

64.9

8.6

11.2

0.9

2.0

1.2

0.7

0.0

9.5

1.1

100.

022

.213

.20

4299

.70

4059

.60

62.3

6.8

11.2

0.8

2.2

1.8

0.5

0.0

9.6

4.7

100.

012

.60.

1143

07.3

540

63.7

572

.612

.08.

02.

01.

70.

30.

40.

02.

40.

710

0.0

25.7

1530

.00

4315

.80

4068

.32

79.5

8.1

6.4

1.9

1.5

0.6

0.2

0.0

1.5

0.3

100.

08.

20.

25

*RK

B =

ref

eren

ced

to k

elly

bus

hing

, TV

D =

tota

l ver

tical

dep

th.

section from a high-porosity sample (Figure 2I).This percentage is probably a little low due to thegeneral underestimation of quartz cements whenpoint counting thin sections, and 3–5 vol. % micro-quartz is estimated to be optimal with respect topreserving reservoir quality. A rim of microquartzwith a thickness of 1–2 µm (Figure 2F), coatinggrains having an average size of 0.1 mm (Table 1),corresponds to 2–5 vol. % microquartz. Theamount of microquartz or cryptocrystalline quartzprecipitated will be approximately equal to theamount of sponge spicules dissolved.

No microquartz or minor amounts, optimalamounts, and excessive amounts of microquartzwere commonly observed (Table 3), but intermedi-ate stages also exist among these four classes, withstages between class 3–4 and class 5 common.Typically in a situation with high levels of micro-quartz cementation, both grain-coating micro-quartz and aggregates or clusters of microquartzare present. These aggregates tend to precipitate inopen pores and may represent remains of dissolvedsponge spicules. In this case, porosity is still highwhile permeability will be lower. The permeabilityof deeply buried sandstones is usually more affect-ed by the presence of illite than by quartz cement;however, these Upper Jurassic sandstone samplescontain small to moderate amounts of illite, andmeasured permeabilities are basically controlled bythe amount of macroporosity. These class designa-tions, therefore, are generally valid for measuredpermeability values.

Of the wells examined, three (wells A, B, and C,with depths ranging from 3970 to 4320 m) fromthe Gyda field area are presented here as examples.We selected these wells because differences in

reservoir quality caused by inhibited macroquartzgrowth are most evident at these depths.Examination of the gamma-ray log curves fromthese wells also indicates relatively homogeneoussandstone compositions. Well A is particularlyimportant because it contains an oil/water contact,which also allows evaluation of the effect of hydro-carbons on the reservoir quality. Well B is fullyhydrocarbon bearing, whereas well C is fromwater-bearing sands. In Figure 3, the porosities ofthe three selected wells have been plotted on atrue vertical depth (TVD) scale, along with thegamma-ray (GR) response and the sonic log (DT).

Well A

The cored interval ranges from approximately4030 to 4140 m TVD. The oil/water contact is at4115.0 m. In this well, two intervals exhibit signifi-cantly different reservoir quality (Figure 3). The topinterval (4030.0–4110.0 m) consists of sandstoneswith porosities of approximately 12% and perme-abilities of about 1 md. The lower interval (4110.0–4140.0 m) consists of sandstones with porositiesabove 20% and permeabilities above 100 md. SEMobservations indicate that the top interval iscemented with macroquartz cement (Figure 4A, B),whereas the lower interval (marked by verticalarrows in Figure 3) lacks extensive macroquartzcementation. Further detailed SEM work on 14samples from both intervals (see Table 1 for identi-fication and distribution of samples) shows thatgrain-coating microquartz is absent in the upperpart of the interval, but optimally present (moder-ate amounts) in the lower part of the interval

Aase et al. 1661

Table 3. Occurrence of Quartz Cement and Corresponding Reservoir Quality

Cement Reservoir Porosity FigureClass Present Quality at Depth Permeability Reference

0 No microquartz cement is present and Poor 8–16% at Low 2Amacroquartz cement contents are usually high >4 km depth

1–2 Minor amounts of microquartz present Poor 10–20% at(<1 vol. %), but macroquartz cement contents >4 km depth Varies 2Bare substantial

3–4 Optimal amount of microquartz cement Good 20–27% at High 2C–F, I(2–5 vol. %). All grains are coated with a thin >4 km depthlayer of microquartz cement. Macroquartzcementation is essentially absent

5 Excessive microquartz cementation 12–25% Varies and is 2G, H, J(>10 vol. %). Silica precipitates both as unusually low atcryptocrystalline quartz (chalcedony) and as >2 km depthaggregates of microcrystalline quartz

(Figure 4C, D). The presence or absence of grain-coating microquartz correlates exactly with thehigh- or low-quality reservoir zones at depth 4110.0m TVD (4123.0 m RKB or referenced to kelly bush-ing). Note also the relatively uniform and low claycontent as reflected in the low gamma-ray activityand XRD data (Table 3). Grain size and carbonatecontent are constant throughout the cored section,and as such do not correlate with variations inreservoir quality. The amount of illite is also con-stant through the cored interval (Table 2). SEMobservations indicate moderate amounts of diage-netic illite (Figure 4). The sonic log reflects thechange in porosity and is also useful as an indicatorof carbonate cementation.

Well B

In well B, we studied 14 samples by SEM andXRD from depths of approximately 3970.0 to4070.0 m TVD. The gamma-ray log and XRD data

indicate uniform clay content (Figure 3). The topinterval (3960.0–3980.0 m) contains large amountsof macroquartz cement (Figure 5A), but also tracesof microquartz cement. The microquartz cementhas a limited or discontinuous distribution, and(normal) macroquartz cementation is extensivelydeveloped. Observations indicate that quartz grainsurfaces were readily available for nucleation andgrowth of normal quartz cement.

The underlying interval, from 3980.0 to 4025.0 m,represents the best reservoir zone in this well andhas porosity and permeability values on the order of24% and 500 md, respectively (Figure 3). In thiszone we observed an optimal amount of grain-coatingmicroquartz cement and very small amounts of nor-mal macroquartz cement (Figure 5B).

The interval from 4025.0 to 4070.0 m varies inthe amount and distr ibution of microquartzcement. Intervals with optimal amounts of grain-coating microquartz cement (Figure 5D) andintervals containing very high levels of micro-quartz cement (Figure 5C) were documented.


Figure 3—Gamma-ray log (GR),sonic log (DT), and helium porosity variation through coredintervals in the three selectedwells (A, B, and C) (see Figure 1)from the Gyda field. The 20%porosity line is marked on eachwell. Arrows on the side of eachwell mark intervals where anoptimal amount of microquartzcement has been observed.Depths are true vertical depthsand thus do not correlate withthe sample core depths listed inTable 1. Note that the OWC(oil/water contact) in well A doesnot correlate with the shift inreservoir quality.

Thin carbonate-cemented zones are also present.Considerable pore space in the very high micro-quartz cemented zones is actually secondary poros-ity after diagenetic dissolution of sponge spicules(Figure 5C). Vertical arrows in Figure 3 outlineintervals with optimal amounts of grain-coatingmicroquartz.

Well C

Six samples in this well have been examined bySEM. The cored interval down to approximately4175.0 m and below 4320.0 m TVD has not beeninvestigated due to high clay contents (Figure 3). Inthe remaining intervals, both the gamma-ray log

and the grain size are relatively uniform. In fact, thegamma-ray data are slightly lower and grain size is alittle coarser in the middle interval (see the follow-ing section). The porosity and permeability values,however, vary significantly. The well can be dividedinto three intervals with respect to varying degreesof reservoir quality. The uppermost interval,4175.0–4220.0 m, has about 20% porosity and10–100 md of permeability. The middle intervals,from 4220.0 to 4240.0 m and from 4275.0 to 4295.0 m,have average porosities (below 10%) and permeabili-ties (below 1 md). The lowermost interval, 4295.0–4320.0 m, contains porosity ranging from 12 to 20%and a permeability of less than 50 md.

In the uppermost interval, optimal to highamounts of grain-coating microquartz cement and

Aase et al. 1663

Figure 4—SEM photomicrographs of sample from well A. (A) Sample depth is 4075.00 m RKB (referenced to kellybushing), 4062.0 m TVD (true vertical depth). Porosity is 13.3%. Large amounts of euhedral macroquartz cement arepresent. (B) Sample depth is 4120.00 m RKB, 4107.0 m TVD. Porosity is 11.3%. Large amounts of euhedral macro-quartz cement are present. (C) Sample depth is 4126.00 m RKB, 4113.0 m TVD. Porosity is 23.5%. Optimal micro-quartz cement coating. No macroquartz cement is present. (D) Sample depth is 4130.33 m RKB, 4117.3 m TVD.Porosity is 23.0%. Optimal microquartz cement coating. No macroquartz cement is present. Minor amounts of col-lapsed illite are also observed.

negligible macroquartz cement were observed(Figure 6A, B). The middle intervals contained notraces of microquartz cement, but did possess alarge amount of macroquartz cement (Figure 6C,D). In the lower interval, high grain-coating micro-quartz contents were observed with little macro-quartz cement.

Crystal Size Distribution of Microquartz

Three randomly chosen areas from selected sam-ples containing well-developed grain-coatingmicroquartz cement were selected for detailedmeasurements of the crystal size distribution.

Image analysis pictures were taken of the areas atmagnification 7000–9000× and printed out on full-scale paper. The diameter of every observed crystalwas measured and visually estimated. Figure 7shows the frequency distribution (diameter inmicrons) of the microquartz crystals. Most of thecrystals have a radius of 0.1–0.4 µm (Figure 2F),and the distribution approximates a Poisson fre-quency distribution (Davis, 1973). Larger quartzcrystals of 5–20 µm in diameter are commonlyobserved (Figure 2C, D). Normal macroquartz over-growths are commonly 100 µm in diameter andlarger. Vagle et al. (1994) reported chalcedonicquartz crystals with average diameter of 0.1 µm,and microquartz crystals in the range of 5 to 10 µm.


Figure 5—SEM photomicrographs of well B. (A) Sample depth is 4147.70 m RKB, 3973.0 m TVD. Porosity is14.6%. Large amounts of euhedral macroquartz cement are present. (B) Sample depth is 4184.70 m RKB, 3995.0m TVD. Porosity is 23.2%. Optimal microquartz cement coating. Minor macroquartz cement is present. (C) Sam-ple depth is 4287.35 m RKB, 4053.0 m TVD. Porosity is 22.2%. Excessive microquartz cementation. Largest poresare secondary pores after diagenetic dissolution of sponge spicules. Macroquartz cement is not present (D) Sam-ple depth is 4307.35 m RKB, 4064.0 m TVD. Porosity is 25.7%. Optimal to high amounts of microquartz cementare present.

DISCUSSION

Porosity and Permeability

The distribution in porosity of these reservoirscannot be explained solely by variations in primarysand composition. Both clay and carbonate con-tents, grain size, and sorting are relatively uniformthroughout the intervals of interest, and are notcorrelated with the main variations in porosity. Themajor factor related to variations in porosity is thedifference in the amount of quartz cement. Quartzcementation in the lower interval of well A mayhave been stopped by the hydrocarbon emplace-ment, and hence the interval had good reservoir

Aase et al. 1665

Figure 6—SEM photomicrographs of well C. (A) Sample depth is 4217.00 m RKB, 4191.0 m TVD. Porosity is 24.0%.Optimal microquartz cement coating. Minor amounts of macroquartz cement and illite are present. (B) Sampledepth is 4225.35 m RKB, 4200.0 m TVD. Porosity is 23.7%. Optimal microquartz cement coating. Minor macroquartzcement is present. (C) Sample depth is 4251.00 m RKB, 4225.5 m TVD. Porosity is 9.7%. No microquartz cement, butsubstantial amounts of macroquartz cement are observed. (D) Sample depth is 4308.00 m RKB, 4283.0 m TVD.Porosity is 9.2%. No microquartz cement, but substantial amounts of macroquartz cement are observed.

Figure 7—Frequency distribution of microquartz crystalsizes. Note the Poisson distribution.

quality. The oil/water contact, however, does notcorrelate with the boundary between the low-quartz-cemented and high-quartz-cemented inter-vals. The oil/water contact is located 5 m belowthe top of the good reservoir zone at 4115.0 mTVD, and there are no changes in reservoir qualityabove or below this contact. Furthermore, thewater-bearing sand in well C contains a 75-m-thicklow-porosity sandstone interval in the middle oftwo high-porosity sandstones. The preservation ofporosity in these separate intervals probably couldnot be explained by the presence of hydrocarbons.

In general, no clear relationships between hydrocar-bon fill and sandstone reservoir quality are observed in the Norwegian continental shelf and North Sea.Ramm and Bjørlykke (1994) concluded that porosity is not substantially higher in hydrocarbon-saturated than in water-saturated reservoir sand-stones. This view is also supported by the study ofGiles et al. (1992) in which the analyses of aregional data set from the United Kingdom sectorshowed no differences in porosity distributionbetween oil- and water-bearing Brent sandstones.Walderhaug (1990, 1994a) observed numeroushydrocarbon-bearing f luid inclusions withinquartz overgrowths, indicating that quartz cemen-tation processes are operative in the presence ofhydrocarbons.

Factors other than hydrocarbon emplacementapparently control the porosity distribution in theUpper Jurassic sandstones. The fact that high-porosity zones correlate exactly with the presenceof grain-coating microquartz is far more significant.Based on these observations, we conclude that thevariations in reservoir quality are related to the dis-tribution of microcrystalline quartz. We furtherpropose that its presence inhibits development ofnormal macrocrystalline quartz cement, which ismainly sourced internally from stylolite dissolutionsurfaces (e.g., Bjørlykke et al., 1992). Ramm and

Forsberg (1991) presented XRD and thin-sectiondata from well C. Those data also indicated that themain difference between good-quality sandstonesand poor-quality sandstones is the amount ofquartz cement. Ramm and Forsberg (1991) pointcounted 8–18% (average 13%) quartz cement in thepoor-quality sandstones and only 0–2% quartzcement in the good-quality sandstone. Ramm andForsberg (1991) also observed that all grains in thehigh-porosity sandstones were coated with micro-quartz crystals, and that this inhibited the growthof macroquartz cements.

The correlation can be observed on Figure 3,wells A and C, where the intervals are dominatedby either an optimal amount of grain-coating micro-quartz cement, resulting in high-porosity intervals(>20%), or are extensively cemented by normalmacroquartz, resulting in poor-porosity zones. WellB contains more than these two end members ofquartz cement types. Both very minor and veryhigh microquartz contents give rise to better than“normal” porosities, but not as high as if theamounts were optimal. This subtle relationship isnot so readily detected.

The frequency distribution plots of porosity(Figure 8) from the wells also show a clear bimodaldistribution in wells A and C, with a distinct distri-bution for well B caused by the presence of micro-quartz. All clean sandstone samples have been plot-ted and placed into two groups based on thepresence or absence of microquartz. Shaded sam-ples, those without microquartz, basically plot withporosities in the range of 5 to 15%, whereas blacksamples, those with microquartz, plot with porosi-ties in the range of 15 to 26%. Well B, which con-tains a wide range of microquartz contents (oftenvery high), shows a corresponding variation inporosity with more overlap between the groupsthan is observed in wells A and C, and with a differ-ent type of frequency distribution.


Figure 8—Frequency distribution of all porosityvalues from wells A, B, and C. Shaded bars indicatesamples without microquartz and solid barsindicate samples withmicroquartz. Note thebimodal distribution inwells A and C caused bybasically two distinct classes of quartzdistribution.

In Figure 9 the frequency distribution of porosi-ty and permeability within different microquartzclasses is more distinct. The relationship betweenan optimal amount of grain-coating microquartz inclass 3–4 and high reservoir quality is readilyapparent. The improved reservoir quality due tothe microquartz, in contrast to the normal cement-ed samples (class 0), is higher on average by 10porosity units and a factor of 100 in millidarcy per-meability units. Classes 1 and 2, representing veryminor and minor amounts of microquartz, haveintermediate values. Class 5, with very high con-tents of microquartz, shows comparably highporosity values for a given permeability (see thefollowing discussion).

Commonly in reservoir quality studies, a bimodalporosity distribution indicates a distinct change infacies. This change can be reflected in grain size orsorting changes, such as from sandy to silty litholo-gies, or major changes in clay or carbonate con-tents. This data set, however, was selected suchthat changes in grain sizes were minimal, and mostof the clay- or carbonate-r ich samples wereremoved. The bimodal porosity distributionsobserved in these wells are related to the complexdistribution of quartz cements. The distribution ofthese cement types (grain-coating microquartz,chalcedony, pore-filling aggregates of microquartz,and normal macroquartz) and their interrelation-ships control reservoir quality in these sandstones.

Aase et al. 1667

Figure 9—Frequency distributionof porosity and permeability of theinvestigated samples grouped intodifferent classes based on theobserved amounts of microquartz.Class 0 = not observed, class 1 =very minor, class 2 = minor, class 3 = moderate, class 4 = high, class 5 = very high. Notehighest porosity and permeabilityin class 3–4 and lower values(especially permeability) in class 5.

SEM observations indicate the presence of mod-erate amounts of delicate diagenetic illite in thesewells (e.g., Figure 2E). Furthermore, these observa-tions of air-dried samples indicate that illite wouldreduce the routine core analysis permeability mea-surements to some extent, and, more importantly,that the effects of illite on subsurface permeabili-ties need to be considered (cf. Pallatt et al., 1984;de Wall, 1989; Ehrenberg and Nadeau, 1989).Routine permeability measurements still would bemainly controlled by the amount of intergranularmacroporosity. Therefore, the data would be influ-enced by the presence of microquartz coatings andtheir general inverse relationship with macroquartzcement. Furthermore, permeability is more affect-ed than porosity when the amount of microquartzis excessive. This difference may be explained bythe formation and distribution of pore-filling aggre-gates of microquartz within the intergranularregion, and the creation of ineffective secondaryporosity by the partial dissolution of biosiliceousgrains. Commonly, the microquartz aggregates areintergrown with some diagenetic illite. These typesof samples contain complex clusters with micropo-rosity (cf. Hurst and Nadeau, 1995) and secondaryporosity from dissolution of sponge spicules. Totalporosity may still be high, but permeability is clear-ly reduced due to the plugging of pore throats bymicroquartz aggregates. This type of porosity–permeability relationship is shown in Figure 10.The symbol size in each plot indicates the amountof microquartz and macroquartz cement, respec-tively. A clear relationship exists between highreservoir quality and the optimal amount of micro-quartz; however, no clear relationship existsbetween poor reservoir quality and the amount of

macroquartz cement. Note that very high contentsof microquartz cement (largest symbols on Figure10A) result in high porosity, but lower permeabilityvalues compared with those for an optimal amountof microquartz (Figure 10B).

Dissolution of even higher concentrations ofsponge spicules will result in the formation of largeamounts of cryptocrystalline or microcrystallinequartz with resultant poor reservoir quality (Figure2J). In these types of rocks, porosity may be com-pletely inverted; primary pores are filled withcement, and sponge spicules dissolve and createsecondary, but basically ineffective, porosity.

Figure 11 shows the normalized frequency distri-bution of all porosity and permeability values fromthe different wells. The data set excludes clay- andcarbonate-rich samples and depths shallower than3600 m RKB. The grouping of samples into micro-quartz coated or not coated samples is based onthe study of the selected samples and the assump-tion that reservoir quality of the samples in closeproximity is related to the same processes. Theaverage porosity of samples containing grain-coating microquartz cement is 20.3%, and the averagepermeability is 161 md. The corresponding porosityand permeability of the samples without microquartzis 11% and 6.4 md, respectively. The average depth ofthe microquartz-coated samples is 4186 m, 130 mdeeper than the samples without grain-coating micro-quartz. When normalized to the same depth, the dif-ference in reservoir quality (porosity) between thetwo groups is approximately 10%.

Typically, average Brent Sandstone porosity is13% at 4200 m depth (Bjørlykke et al., 1989, 1992;Giles et al., 1992), whereas average Haltenbankenporosity data at 4.2 km are closer to 10% (Bjørlykke


Figure 10—Porosity vs.permeability plots, withsymbol size from smallestto largest representingincreasing amounts of (A) microquartz and (B) macroquartz cementfor wells A, B, and C. Carbonate- cemented samples and clay samplesare excluded.

et al., 1986; Ehrenberg, 1990). The average porosity –depth trend (continuous line based on black cir-cles, Figure 12A) from the Upper Jurassic sand-stones of the “normal” samples without grain-coat-ing microquartz is parallel to and slightly lowerthan the Brent Sandstone trend (dashed line) andsimilar to the Haltenbanken trend. The sampleswith grain-coating microquartz (Figure 12B) showan almost constant porosity vs. depth trend atabout 20%, which is 9% higher than the “normal”Upper Jurassic sandstone porosity at 4200 m and7% higher than the Brent Sandstone porosity at thisdepth. This difference will probably increase withfurther burial of the sandstones.

The permeability (Figures 9, 11) of these microquartz-coated Upper Jurassic sandstones(depth >4000 m) is in the range of 10 to 1000 md,which is approximately two orders of magnitudehigher than the normally cemented Upper Jurassicsandstones. These values are also much higher thanpermeabilities usually observed in Brent orHaltenbanken sandstones at these depths and tem-peratures. Ramm and Forsberg (1991) reported in astudy of Upper Jurassic sandstones of the Gyda areathat sandstones containing microcrystalline quartzovergrowths had higher porosities (around 20–22%at 4.1–4.2 km) than sandstones that containedmacroquartz overgrowth. In addition, Spark andTrewin (1986) reported porosity preservation in theTen Foot Sandstone of the Claymore field, North

Sea, which they attributed to early formation of dis-crete quartz crystals. Spark and Trewin (1986) alsosuggested that the sandstones were supersaturatedwith respect to silica because discrete crystalsrather than syntaxial quartz overgrowths devel-oped. They considered the source of silica to be aresult of dewatering of adjacent shales. Recently,Vagle et al. (1994) discussed the formation ofmicrocrystalline quartz cement in the UpperJurassic Brora Formation in relation to the occur-rence and abundance of Rhaxella spongespicules.

Preservation of anomalous high porosity due toformation of other grain-coating minerals, namelychlorites, is commonly observed around the world,as well as on the Norwegian shelf in some Halten-banken and North Sea sandstones (Ehrenberg,1993). Average porosity in these chlorite-coatedintervals is also 10–15% higher than would be pre-dicted from regional trends of mean porosity vs.depth. Both these grain-coating processes preserveporosity by inhibiting the growth of macroquartzcement.

Preserving High Porosities

Amorphous silica sponge spicules are regardedas being the source of the cryptocrystalline andmicrocrystalline quartz. Such quartz morphologies

Aase et al. 1669

Figure 11—Normalized frequency distribution ofporosity and permeabilityfrom all wells. Samplesshallower than 3600 m,carbonate-cemented samples, clay-rich samples,and coarse-grained samples have been excluded from the data set.A and B are samples withmicroquartz present, and C and D are samples without microquartz.

suggest rapid crystallization from locally saturatedsilica solutions (Williams et al., 1985). Spongespicules composed of hydrated silicon dioxidehave a solubility of 120–140 ppm, which is 15–20times higher than the solubility of quartz (6–10ppm) at 25°C. In the temperature range of 0 to100°C, chalcedony is approximately twice as solu-ble as quartz (Gislasson et al., 1993). Rhaxellaspicules themselves are often recrystallized tomicroquartz and moldic pores lined by fine quartzcrystallites. This process can proceed with little tono net reduction in total porosity.

One mechanism that could explain the inhibi-tion of normal quartz cementation is that the equi-librium concentration (i.e., solubility) of a givenmineral is a function of its crystal size (Enüsten andTurkevich, 1960). The degree of supersaturationrequired for small crystals to grow can be calculat-ed by applying the Ostwald-Freunlich equation,which states that

(1)

where Kr = solubility for a crystal with radius r; K∞= solubility of an infinitely large crystal; v = molevolume of the mineral (for quartz = 22.688cm3/mol); σ = interfacial free energy for quartz inwater (3.6 × 10–5 J/cm2); R = 8.3143 J/K mol; T =temperature in kelvins (373K or 100°C).

For a crystal radius of 0.5 µm, solving equation 1gives 1.011. For a crystal radius of 0.25 µm, solving

equation 1 gives 1.022. The solubility of quartz(K∞) at 100°C is 60 ppm. Hence, the solubility ofthe small crystals with r = 0.5 and r = 0.25 µm is60.6 and 61.2 ppm, respectively. This value is likelyto be higher than the amount of dissolved silicasourced from stylolite dissolution surfaces, whichare very sensitive to silica saturation (Oelkers et al.,1996). In this case, the normal supply of silica fromstylolites is inhibited, which does result in quartz-cementation porosity loss, implying that, in gener-al, microquartz crystals will stop growing whenthe amorphous silica source is exhausted, and any redistribution of silica from the smaller to thelarger crystals within the microquartz populationwill not result in a net reduction in total porosity.Further mechanical compactional porosity loss ofthe partly cemented sand is likely to be minor.Finally, any late-diagenetic silica from feldspar dis-solution associated with illitization also will notresult in a net reduction in total porosity (Bjørkumet al., 1993). It follows, then, that the porositywould be relatively insensitive to depth after thispoint in the burial history. If all the quartz surfaceswere not coated with microcrystalline quartz andsilica supersaturation was insufficient to inhibitquartz dissolution at stylolites, normal macro-quartz cementation would proceed. The rate ofthe overall porosity reduction would then be afunction of the temperature (Oelkers et al., 1992,1996; Walderhaug, 1994b) and of the availability ofclean quartz surfaces for the growth of macro-quartz cement.

The coating of quartz grains by microquartzcement will be reasonably efficient with initial

log.

K

K

v

r RT

r

∞

= σ1 152


Figure 12—Plot of porosityvs. depth for the 15 investigated wells in theCentral Graben area. × =points are individual values, • = well averages.(A) Samples without microquartz, (B) sampleswith microquartz. Blacksolid line on each plot represents a best-fit trendbased on the plotted data;dashed lines represent the Brent porosity vs.depth trend (Bjørlykke et al., 1989).

sponge spicule contents of 1–5%. If the contentof sponge spicules exceeds 5%, the probabilitythat all quartz grains will be completely coatedwith microcrystalline quartz is close to unity.Figure 13 shows the calculated solubility ofquartz for any given crystal radius. If microcrys-talline quartz crystals become much greater than2–3 µm in radius, the effect of size is likely to beminor and the risk of further quartz cementationincreases.

PALEOENVIRONMENTAL CONTROLS

Rhaxella spicules can be found within a largerange of different rock types. In Brora, Scotland,Rhaxella spicules are found in Jurassic sedimentsfrom Sinemurian to Kimmeridgian in age and rocktypes that include shales, siltstones, fine-grainedsandstones, and carbonates (Vagle et al., 1994).Offshore, sponge spicules are found in the MorayFirth in silty sands and fine-grained sandstones ofmiddle Oxfordian age. Spicules also are common inthe Alness Spiculite Formation and in the UpperJurassic shallow-marine Fulmar Sandstones in theClyde field.

The distribution of sponge spicule abundanceplotted against depth is related to primary deposi-tional and geological controls. A comparison ofthese distributions in established reservoir zoneswould show a close correlation, but would notexplain their overall distributional patterns.

To develop a predictive model for spongespicule distribution, a correlation of these distribu-tional patterns would have to be made to a genericdepositional model; that is, the distribution ofsponge spicules would be correlated to sandstonehorizons viewed in a time-stratigraphic sense.

CONCLUSIONS

During burial and with increasing temperature,sponge spicules dissolve to precipitate as micro-quartz overgrowth on quartz grains. The presenceof this grain-coating microquartz cement appearsto inhibit the development of normal macroquartzcement, resulting in anomalously high porosity incertain deeply buried sandstones. The fact thatmicrocrystalline quartz cement can coat quartzgrains has been known for some years, but the dis-tribution of microcrystalline quartz cement inUpper Jurassic sandstones in the North Sea CentralGraben area and the role of this cement in preserv-ing high porosities in deeply buried sandstoneshave not been fully appreciated.

The principal working hypothesis, consistentwith the observations, is that pore-water silica con-centrations in equilibrium with the microcrys-talline quartz are slightly higher (1–2 ppm) than inequilibrium with quartz. These concentrations aresufficient to inhibit the normal supply of silica fromquartz dissolution along stylolites, which, in theabsence of microquartz, would ordinarily assist inmacroquartz cementation. This argument is similarto that proposed by Bjørkum et al. (1993) for theeffect of illitization on porosity loss by quartzcementation. The silica in equilibrium with quartzdissolution along stylolites is likely to be less thanthat required for the small (1 µm) microquartz crys-tals to grow. Hence, when the amorphous silicasources have been exhausted, normal quartzcementation processes are halted in a metastablestate. If some quartz grain surfaces are available orthe microquartz crystals have grown to be largerthan a critical size, macroquartz cementationsourced from stylolites may develop, but the poros-ity loss with time would be less than normal incases where these surfaces are small compared tothe total surface area of the sandstone.

Our data set indicates that samples with grain-coating microquartz preserve porosity of 20% to atleast burial, which is approximately 9% higher thanthe “normal” Upper Jurassic sandstone porosityfound at 4200 m, and about 7% higher than theaverage Brent Sandstone porosity at this depth andtemperature. Average permeabilities of these sam-ples are higher than normal because the permeabil-ity is mostly controlled by the amount of intergran-ular macroporosity. Routine core permeabilitymeasurements, however, may be somewhat higherthan subsurface permeabilities due to the presenceof small to moderate amounts of delicate diageneticillite in most of the wells.

Optimal coatings of microquartz are thought tobe about 2–5 vol. %. This amount approximatelycorrelates with the amount of sponge spicules dis-solved. Higher concentrations of sponge spicules

Aase et al. 1671

Figure 13—Plot of crystal size and solubility of quartz.

yield excessive amounts of microquartz (or cryp-tocrystalline quartz) cement, which reduces reser-voir quality. Mapping sponge spicule-prone sedi-mentary facies and their likely reworking pathsinto sand-rich depositional systems is importantbecause the content and distribution of spongespicules are controlled by sedimentology. Such sys-tems could then be incorporated into sequence-stratigraphic models of sea level fluctuation, andresult in a better understanding of possible loca-tions of optimal reservoir-quality sandstones in theCentral Graben area of the North Sea.

REFERENCES CITEDBjørkum, P. A., 1984, Et studium av de mellom-jurassiske sand-

steinene i kjerne 30/6-7, Oseberg, med særlig vekt på diagenese:Master’s thesis, University of Bergen, Bergen, Norway, 131 p.

Bjørkum, P. A., 1994, How important is pressure in causing disso-lution of quartz in sandstones? (abs.): AAPG Annual MeetingAbstracts, p. 105.

Bjørkum, P. A., 1996, How important is pressure in causing disso-lution of quartz in sandstones?: Journal of SedimentaryResearch, v. A66, p. 147–154.

Bjørkum, P. A., O. Walderhaug, and N. E. Aase, 1993, A model forthe effect of illitization on porosity and quartz cementation ofsandstones: Journal of Sedimentary Petrology, v. 63, p.1089–1091.

Bjørlykke, K., P. Aagaard, H. Dypvik, D. S. Hastings, and A. S.Harper, 1986, Diagenesis and reservoir properties of Jurassicsandstones from the Haltenbanken area, offshore mid-Norway,in A. M. Spencer, E. Holter, C. J. Campell, S. H. Hanslien, P. H. H.Nelson, E. Nysæther, and E. G. Ormaasen, eds., Habitat ofhydrocarbons on the Norwegian continental shelf: London,Graham and Trotman, p. 275–286.

Bjørlykke, K., M. Ramm, and G. C. Saigal, 1989, Sandstone diagen-esis and porosity modification during basin evolution:Geologische Rundschau, v. 78, p. 243–268.

Bjørlykke, K., T. Nedkvitne, M. Ramm, and G. C. Saigal, 1992,Diagenetic processes in the Brent Group (Middle Jurassic) reser-voirs of the North Sea: an overview, in A. C. Morton, R. S.Haszeldine, M. R. Giles, and S. Brown, eds., Geology of the BrentGroup: Geological Society Special Publication 61, p. 263–287.

Bjørnseth, H. M., and J. G. Gluyas, 1991, Petroleum exploration inthe Ula trend, in H. Lein, ed., Celebration of 25 years explo-ration in Norway (abs.): Norwegian Petroleum Society (NPF),Proceedings Abstracts, p. 23.

Brown, A., A. W. Mitchell, I. R. Nilssen, I. J. Stewart, and P. T.Svela, 1992, Ula field: relationship between structure andhydrocarbon distribution?, in R. M. Larsen, H. Brekke, B. T.Larsen, and E. Talleraas, eds., Structural and tectonic modellingand its application to petroleum geology: NorwegianPetroleum Society (NPF) Special Publication 1, Elsevier,Amsterdam, p. 409–420.

Burley, S. D., 1993, Models of burial diagenesis for deep explo-ration plays in Jurassic fault traps of the Central and NorthernNorth Sea, in J. R. Parker, ed., Petroleum geology of northwestEurope: Proceedings of the 4th Conference, London, TheGeological Society, p. 1353–1375.

Cazier, E. C., A. B. Hayward, G. Espinoza, J. Velandia, J.-F. Mugniot,and W. G. Leel, Jr., 1995, Petroleum geology of Cusiana field,Llanos basin foothills, Colombia: AAPG Bulletin, v. 79, p. 1444–1463.

Davis, J. C., 1973, Statistics and data analyses in geology: NewYork, John Wiley and Sons, 646 p.

de Wall, J. A., 1989, Influence of clay distribution on shaly sandconductivity: SPE Formation Evaluation, v. 4, p. 377–383.

Ehrenberg, S. N., 1990, Relationship between diagenesis and reser-voir quality in sandstones of the Garn Formation, Haltenbankenarea, mid-Norwegian continental shelf: AAPG Bulletin, v. 74, p. 1538–1558.

Ehrenberg, S. N., 1993, Preservation of anomalously high porosityin deeply buried sandstones by grain-coating chlorite: exam-ples from the Norwegian continental shelf: AAPG Bulletin, v. 77, p. 1260–1286.

Ehrenberg, S. N., and P. H. Nadeau, 1989, Formation of diageneticillite in sandstones of the Garn Formation, Haltenbanken area,mid-Norwegian continental shelf: Clay Minerals, v. 24, p. 233–253.

Enüsten, B. V., and J. Turkevich, 1960, Solubility of fine particlesof strontium sulphate: Journal of the American ChemicalSociety, v. 82, p. 4502–4509.

Giles, M. R., S. V. Stevenson, S. J. Martin, S. J. C. Cannon, P. J.Hamilton, J. D. Marshall, and G. M. Samways, 1992, The reser-voir properties and diagenesis of the Brent Group, a regionalperspective, in A. C. Morton, R. S. Haszeldine, M. R. Giles, andS. Brown, eds., Geology of the Brent Group: Geological SocietySpecial Publication 61, p. 289–327.

Gislasson, S. R., P. J. Heaney, D. R. Veblen, and K. J. T. Levi, 1993,The difference between the solubility of quartz and chal-cedony: the cause?: Chemical Geology, v. 107, p. 363–366.

Gluyas, J. G., 1985, Reduction and prediction of sandstone reser-voir potential, Jurassic, North Sea: Philosophical Transactionsof the Royal Society of London, Series A, v. 315, p. 187–202.

Gluyas, J. G., A. J. Leonard, and N. H. Oxtoby, 1990, Diagenesisand oil emplacement: the race for space—Ula trend, North Sea:13th International Sedimentological Congress 193,International Association of Sedimentologists, p. 193.

Gluyas, J. G., A. G. Robinson, D. Emery, S. M. Grant, and N. H.Oxtoby, 1993, The link between petroleum emplacement andsandstone cementation: is it causal or casual?, in J. R. Parker,ed., Petroleum geology of northwest Europe: Proceedings ofthe 4th Conference, London, The Geological Society, p. 1395–1402.

Heald, M. T., and R. E. Larese, 1974, Influence of coatings onquartz cementation: Journal of Sedimentary Petrology, v. 44, p. 1269–1274.

Home, P. C., 1987, Ula, in A. M. Spencer, E. Holter, C. J. Campbell,S. H. Hanslien, P. H. H. Nelson, E. Nystæher, and E. G.Ormaasen, eds., Geology of the Norwegian oil and gas fields:London, Graham and Trotman, p. 143–151.

Hurst, A., and P. H. Nadeau, 1995, Clay microporosity in reservoirsandstones: an application of quantitative electron microscopyin petrophysical evaluation: AAPG Bulletin, v. 79, p. 563–573.

McBride, E. F., 1989, Quartz cement in sandstones: a review:Earth-Science Reviews, v. 26, p. 69–112.

Oelkers, E. H., P. A. Bjørkum, and W. M. Murphy, 1992, The mech-anism of porosity reduction, stylolite development and quartzcementation in North Sea sandstones, in Y. K. Kharaka and A. S. Maest, eds., Water–rock interactions: Rotterdam, Balkema,p. 1183–1186.

Oelkers, E. H., P. A. Bjørkum, and W. M. Murphy, 1996, A petro-graphic and computational investigation of quartz cementationand porosity reduction in North Sea sandstones: AmericanJournal of Science, v. 296, p. 420–452.

Pallatt, N., M. J. Wilson, and W. J. McHardy, 1984, The relationshipbetween permeability and the morphology of diagenetic illitein reservoir rocks: Journal of Petroleum Technology, v. 36, p. 2225–2227.

Ramm, M., 1991, Porosity depth trends in reservoir sandstones. Aquantitative study on effects of variations in pore pressure tem-perature and sandstone petrography on reservoir quality: Ph.D.thesis, University of Oslo, Oslo, Norway, 308 p.

Ramm, M., 1992, Porosity depth trends in reservoir sandstones:theoretical models related to Jurassic sandstones offshoreNorway: Marine and Petroleum Geology, v. 9, p. 553–567.

Ramm, M., 1994, Porosity depth trends in Upper Jurassic reser-voirs, Norwegian Central Graben: an example of porosity


preservation at deep burial by grain-coating micro-quartz(abs.): AAPG Annual Meeting Abstracts, p. 241.

Ramm, M., and K. Bjørlykke, 1994, Porosity/depth trends in reser-voir sandstones: assessing the quantitative effects of varyingpore-pressure, temperature history and mineralogy: Norwegianshelf data: Clay Minerals, v. 29, p. 475–490.

Ramm, M., and A. W. Forsberg, 1991, Porosity vs. depth trends inUpper Jurassic sandstones from the Cod Terrace area, centralNorth Sea, in M. Ramm, ed., Porosity depth trends in reser-voir sandstones: Ph.D. thesis, University of Oslo, Oslo,Norway, 308 p.

Robinson, A., and J. Gluyas, 1992a, Model calculations of loss ofporosity in sandstones as a result of compaction and quartzcementation: Marine and Petroleum Geology, v. 9, p. 319–323.

Robinson, A., and J. Gluyas, 1992b, Duration of quartz cementa-tion in sandstones, North Sea and Haltenbanken basins: Marineand Petroleum Geology, v. 9, p. 324–327.

Rothwell, N. R., A. Sørensen, J. L. Peak, K. Byskov, and T. A. M.McKean, 1993, Gyda: recovery of difficult reserves by flexibledevelopment and conventional reservoir management: SPE26778, Proceedings of the 1993 Offshore Europe Conference,p. 271–280.

Selley, R. C., 1978, Porosity gradients in North Sea oil-bearingsandstones: Journal of the Geological Society of London, v. 135, p. 119–132.

Sommer, F., 1978, Diagenesis of Jurassic sandstones in the VikingGraben: Journal of the Geological Society of London, v. 135,

p. 63–67.Spark, I. S. C., and N. H. Trewin, 1986, Facies-related diagenesis in

the main Claymore oilfield sandstones: Clay Minerals, v. 21, p.479–496.

Spencer, A. M., P. C. Home, and V. Wiik, 1986, Habitat of hydro-carbons in the Jurassic Ula trend, Central Graben, Norway, inA. M. Spencer, E. Holter, C. J. Campbell, S. H. Hanslien, P. H.Nelson, E. Nysæther, and E. G. Ormaasen, eds., The habitat ofhydrocarbons on the Norwegian continental shelf: London,Graham and Trotman, p. 111–127.

Vagle, G. B., A. Hurst, and H. Dypvik, 1994, Origin of quartzcements in some sandstones from the Jurassic of the InnerMoray Firth (UK): Sedimentology, v. 41, p. 363–377.

Walderhaug, O., 1990, A fluid inclusion study of quartz-cementedsandstones from offshore mid-Norway—possible evidence forcontinued quartz cementation during oil emplacement: Journalof Sedimentary Petrology, v. 60, p. 203–210.

Walderhaug, O., 1994a, Temperatures of quartz cementation inJurassic sandstones from the Norwegian continental shelf—evidence from fluid inclusions: Journal of SedimentaryResearch, v. A64, p. 311–323.

Walderhaug, O., 1994b, Precipitation rates for quartz cement insandstones determined by fluid-inclusion microthermometryand temperature-history modelling: Journal of SedimentaryResearch, v. A64, p. 324–333.

Williams, L. A., G. A. Parks, and D. A. Crerar, 1985, Silica diagene-sis, I. Solubility controls: Journal of Sedimentary Petrology, v. 55, p. 301–311.

Aase et al. 1673

Nils Einar Aase

Nils E. Aase received his M.S.degree in geology from the Uni-versity of Oslo in 1990. He thenjoined Statoil in Stavanger, where heworked at the geological laboratoryfor four years, examining diageneticprocesses affecting reservoir qualityand their impact on exploration andproduction operations. He is nowworking as a staff geologist in thepetroleum technology group, on thedevelopment of the Sleipner Vest field, Norwegian conti-nental shelf.

Per Arne Bjørkum

Per Arne Bjørkum serves as atechnical advisor in petrology with-in Statoil’s Exploration TechnologyDivision. He is also a professor ofreservoir geology at Rogaland Uni-versity Center. His research interestsinclude quartz, clay minerals, andcarbonate cementation in sand-stones. Understanding and quantify-ing these diagenetic processes byapplying chemical kinetics and thermodynamics have been the main emphasis of thiswork. More recent collaborative activities concern basin-scale model development for fluid pressure, migration,and hydrocarbon entrapment.

Paul H. Nadeau

Paul Nadeau serves as staff geol-ogist in Statoil’s Exploration Tech-nology Division. Paul’s main areasof responsibility include geologicmodels for reservoir quality, riskevaluation, and hydrocarbon pre-diction, both for Norwegian andinternational operations. Morerecent collaborative activities con-cern basin-scale model develop-ment for fluid pressure, migration,and hydrocarbon entrapment by applying his back-ground in clay mineralogy, silicate diagenesis, rockproperties, and exploration concepts. He resides withhis wife, Marion, and their son and daughter in Forus,Norway.

ABOUT THE AUTHORS

The Effect of Grain-Coating Microquartz on Preservation of ...

Documents