JOURNAL OF SEDIMENTARY RESEARCH,VOL. 73, NO. 2, MARCH, 2003, P. 146–156 Copyright  2003, SEPM (Society for Sedimentary Geology) 1527-1404/03/073-146/$03.00 THE EFFECT OF STYLOLITE SPACING ON QUARTZ CEMENTATION IN THE LOWER JURASSIC STØ FORMATION, SOUTHERN BARENTS SEA OLAV WALDERHAUG AND PER ARNE BJØRKUM Statoil ASA N-4035 Stavanger, Norway e-mail: OWALD@statoil.com ABSTRACT: The shallow marine Lower Jurassic quartz arenites of the Stø Formation in the southern Barents Sea comprise (1) intervals where dispersed detrital clay is absent, and where the spacing between clay-rich laminae that evolved into stylolites upon burial is exception- ally large, up to several meters, and (2) intervals where minor detrital clay matrix occurs, clay laminae are very common, and stylolite spac- ing is typically less than a centimeter. Point counting of thin sections and cathodoluminescence micrographs shows that quartz cement con- tents are far lower in the intervals where stylolite spacing is excep- tionally large, 4–11%, versus 10–20% outside these intervals. There is also a correlation between distance to nearest stylolite and volume of quartz cement. Samples located a centimeter or less from a stylolite contain 10–20% quartz overgrowths, for distances of 3–20 cm quartz cement content is 4–10%, and only 3–8% when the closest stylolite is more than 20 cm distant. Modeling of quartz cementation with the Exemplar diagenetic modeling program indicates that the observed trend of decreasing quartz cement abundance outwards from stylolites is not caused by variations in grain size, degree of grain coating, or content of quartz grains, i.e., the trend is not due to more quartz sur- face area being available for overgrowth formation close to stylolites. On the contrary, the modeling suggests that the samples situated more than 20 cm from stylolites contain 5–8% less quartz cement than what would have been the case given a more normal stylolite abundance. This study indicates that sandstones with exceptionally few clay-rich or micaceous laminae and without clay or mica at individual grain contacts will be significantly less quartz cemented and more porous than other sandstones with similar temperature histories. However, such sandstones seem to be highly unusual on the Norwegian conti- nental shelf. This suggests that exceptionally low abundance of stylolite precursors may be of only local importance for preserving reservoir quality at elevated temperatures, and that it is normally not necessary to include stylolite spacing and distance to the nearest stylolite as var- iables in quantitative models of quartz cementation. INTRODUCTION Quartz cement is the dominant diagenetic mineral and the main control on reservoir quality in deeply buried quartz-rich sandstones in many basins (Blatt 1979; Land and Fisher 1987; McBride 1989; Bloch et al. 1990; Ehrenberg 1990), but there is still not a general consensus regarding sources of quartz cement and mechanisms of quartz cementation (McBride 1989; Worden and Morad 2000). Some workers favor transport of dissolved silica into sandstones from external sources (Riches et al. 1986; Burley et al. 1989; Land and Milliken 2000), whereas others regard the extent of fluid flow required for transporting significant amounts of dissolved silica as prohibitive, and favor internal sources of quartz cement (Bjørlykke 1980; Walderhaug 1994; Bjørkum et al. 1998). Among possible internal sources, the most commonly cited are dissolution of quartz grains at stylolites (Heald 1955; Sibley and Blatt 1976; Olaussen et al. 1984; Bjørlykke et al. 1986) or individual grain contacts (Waldschmidt 1941; Lowry 1956; Sibley and Blatt 1976; Houseknecht 1988). Whether dissolution at stylolites and grain contacts is a pressure-solution process where dissolution rate is a function of the stress on the grain contacts (Weyl 1959; Dewers and Or- toleva 1990; Mullis 1991) or rather a pressure-insensitive process where quartz dissolution is a result of the catalytic effect of the clay or mica at the contacts (Bjørkum 1996; Oelkers et al. 1996; Walderhaug 1996) is also a matter of continuing debate. Several recent quantitative models of closed-system quartz cementation consider the total quartz cementation process to comprise three steps: (1) dissolution of quartz at stylolites or individual grain contacts containing clay or mica, (2) short-range diffusion (millimeters–centimeters) of silica, and (3) precipitation of dissolved silica as quartz overgrowths on quartz grain surfaces. The models either model all three steps of the process (Oelk- ers et al. 1996; Bjørkum et al. 1998) or assume that the precipitation step is rate-limiting and therefore model only the precipitation step (Walderhaug 1996; Lander and Walderhaug 1999). One of the simplifications resulting from assuming precipitation rate control is that the spacing between sty- lolites and distance to nearest stylolite are not necessary input parameters when calculating quartz cementation in a sample. In other words, the de- crease of silica supersaturation and reduction of quartz precipitation rate per unit surface area outwards from stylolites are assumed to be so small that they can be ignored, and silica supersaturation is assumed not to reach significantly higher values between closely spaced stylolites compared to between stylolites with greater separation. These assumptions have been shown to be valid for typical sandstones from the Norwegian shelf (Wald- erhaug 2000; Walderhaug et al. 2000) and from the Gulf of Mexico and the Baltic (Lander and Walderhaug 1999). However, as stylolite spacing increases and reaches unusually large values, the simplified closed-system quartz cementation models become less accurate and overpredict quartz cementation, especially far from stylolites. This study attempts to determine the effect of stylolite spacing and distance to nearest stylolite on quartz cementation by measuring volumes of quartz cement as a function of these two parameters, and also compares the measurements with quartz cement volumes calculated on the basis of assumed precipitation rate control. The Stø Formation cores from well 7120/6-1 in the Norwegian sector of the Barents Sea are exceptionally well suited for this study because they con- tain intervals of unusually clean sandstone where stylolite spacings reach several meters. GEOLOGICAL SETTING Well 7120/6-1 is located in the Hammerfest Basin in the southern part of the Barents Sea approximately 150 km northwest of Hammerfest in northern Norway (Fig. 1). The Lower Jurassic Stø Formation is 84 m thick in well 7120/6-1 and comprises very fine-, fine-, and medium-grained sand- stones plus a few thin intervals of shale, and conglomeratic sandstone and phosphate nodule lag deposits (Fig. 2). The sandstones are shallow marine and include strongly bioturbated intervals deposited in a lower-shoreface setting, cross-bedded upper-shoreface sandstones, and a thick interval of extremely clean and largely structureless sandstones which has been con- sidered to be a possible tidal bar complex (Bjørgen 1985), or possibly a delta mouth bar. The top of the Stø Formation is located at 2386 mRKB (meters below the rotary table) corresponding to 2049 m below the sea floor. However, as is typically the case in the Hammerfest Basin, Paleocene sediments are overlain by Pleistocene deposits in well 7120/6-1, probably reflecting large- scale Tertiary erosion and uplift (Berglund et al. 1986; Nyland et al. 1992). The incomplete Upper Cretaceous stratigraphy suggests that erosional events may also have occurred during this period. The Stø Formation is