The sealing potential of volcanic rocks in hydrocarbon systems: A case study from the Rosebank Field

Abstract

Hydrocarbon exploration in frontier regions has resulted in a series of discoveries within intra- and sub-volcanic basins along the NE Atlantic margin. The majority of these basins are blanketed by varying thicknesses of subaerial lavas and volcaniclastic sequences that form the Paleogene-aged North Atlantic Igneous Province. During Exploratory drilling in 2004, the 213/27-1Z well encountered two oil and gas accumulation with a total pay thickness of 52 metres, within the Late Palaeocene-Early Eocene aged Colsay Member sandstones within the Rosebank Field. These hydrocarbon-bearing reservoir intervals comprise numerous fluvial clastic sequences that are inter-layered between sub-aerially erupted basaltic lavas and volcaniclastic sediments. This discovery gave rise to a new hydrocarbon play concept. In this study, the sealing potential of the volcanic and volcaniclastic sequences is investigated using multi-scale subsurface datasets (with a particular emphasis on the value of high resolution borehole image logs; FMI) and field analogues from the Faroe Islands.
Through the combined use of wireline logs (Bivariate cross-plots) and high resolution FMI images, the volcanic and volcaniclastic rocks in the Rosebank Volcanic Formations (RVFs) were characterised, and a detailed volcanic stratigraphic section of the RVFs was constructed, to understand how these lithologies vary vertically and laterally between wells. Bivariate cross-plots (photoelectric factor versus bulk density) indicates that there are two distinct compositions of lava flows: (a) Lava Type I is characteristic of the Rosebank Lower Volcanics, and has a higher Iron/Magnesium variety, suggesting that it was emplaced during the initial phase of volcanism; and, (b) Lava Type II is characteristic in the Upper and Lower Rosebank Middle Volcanics (RMV_U and RMV_L) and the Rosebank Upper Volcanics, and is depleted in Iron and Magnesium. The FMI images revealed that the RVFs comprise a diverse and complex suite of volcanic and volcaniclastic rocks (e.g. pillow lavas, hyaloclastite, sub-aerial lavas and inter-basaltic rocks), some of which exhibit complex diagenetic overprints produced during post-volcanic process. Detailed interpretation of FMI images and integration with wireline logs showed that the ratio of volcanic, volcaniclastic and siliciclastic rocks varied between studied wells, for example the southernmost wells in Rosebank Main (205/1-1, 213/26-1, 213/26-1z and 213/27-1Z) comprise 57-62% sub-aerially erupted pahoehoe lava, ~25% volcaniclastic rocks, and ~15% siliciclastic rocks, while the wells in northern Rosebank Main (213/27-4 and 213/27-2) comprise ~35% sub-aerially erupted lava, ~50% volcaniclastic rocks, and ~20% siliciclastic rocks. These ratios indicate that the sub-aerially erupted lavas are thinning to the North of the field.
Fractures interpreted in the volcanic rocks in the RVFs mainly occur in the more brittle units, for example, in the massive cores of sub-aerially erupted lava flows. Fractures distribution in the RVFs is strongly controlled by the type of lithology (e.g. simple lava flow, pillow lava, hyaloclastite etc.), their thickness and the degree of secondary degradation and alteration they have undergone. In comparison to the Colsay reservoir intervals, the RVFs are more brittle and highly fractured and thus much more prone to fracturing during drilling. Their sealing potential decreases from southern Rosebank Main to northern Rosebank Main, predominantly because the RVFs are much thinner to the North, and contain large, connected fractures. The inter-layering between volcanic and volcaniclastic rocks in southern Rosebank Main acts as an efficient vertical seal because the volcaniclastics tend to be clay-rich and do not fracture as easily. In terms of sealing potential, all the RVFs except RMV_U are low risk seals, suggesting that in some parts of the field, particularly in the northern most part of Rosebank Main, the Colsay 1 and 2 reservoir intervals may be connected. Evidence of faulting in the RVFs is infrequent in the FMI images, although where inferred, they tend to occur at contacts between lava and volcaniclastic rocks, usually at the transitional interface between the Colsay reservoir intervals and the RVFs. Similarly, faults were interpreted in the Ocean Bottom Node (OBN) seismic data, although they are generally small and discontinuous. It is therefore predicted that if faults are present in Rosebank, they are likely to be low-strain with minimal displacement.
Integration of field analogues from the Faroe Islands with the subsurface datasets (FMI and recovered core) suggests that the majority of the fractures interpreted from the FMI images, and any possible faults within the RVFs, are likely to be sealing. Vein infill at the studied outcrops indicate that cementation rates were low, and therefore many fractures were still potential conduits. In Rosebank, however, due to high cementation rates, most of the fractures are completely healed. This also provides a significant clue on hydrocarbon migration within the Colsay reservoirs. It is evident that none of the fractures or the volcanic rocks recovered either through full core recovery, or side-wall cores, have any hydrocarbon shows, and therefore, the fractures present in the RVFs are likely to have been completely filled, during hydrocarbon migration into the Colsay reservoirs. It is therefore suggested that, hydrocarbons laterally migrated into the Colsay sands from the sands in the Flett Basin, and because the fractures in the RVFs did not act as conduits for fluid flow, the hydrocarbons were successfully trapped within their respective reservoirs.