Diagenetic and sedimentary controls on porosity in Lower ...

Marine and Petroleum Geology 27 (2010) 1395e1410

Contents lists available at ScienceDirect

Marine and Petroleum Geology journal homepage: www.elsevier.com/locate/marpetgeo

Diagenetic and sedimentary controls on porosity in Lower Carboniferous fine-grained lithologies, Krechba field, Algeria: A petrological study of a caprock to a carbon capture site P.J. Armitage a, *, R.H. Worden a, D.R. Faulkner a, A.C. Aplin b, A.R. Butcher c, J. Iliffe d a

Department of Earth and Ocean Sciences, University of Liverpool, Liverpool L69 3GP, UK School of Civil Engineering and Geoscience, Drummond Building, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK Intellection Pty Ltd, FEI, PO Box 2269, Milton BC, Queensland 4064, Australia d BP Exploration, Chertsey Road, Sunbury upon Thames, Middlesex, UK b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 November 2009 Received in revised form 9 March 2010 Accepted 30 March 2010 Available online 14 April 2010

Fine-grained siliciclastic lithologies commonly act as sealing caprocks to both petroleum fields and host reservoirs for carbon capture (CO2 sequestration) projects. Fine-grained lithologies are thus of great importance in controlling fluid flow and storage in the subsurface. However, fine-grained rocks are rarely characterised in terms of primary sedimentary characteristics, diagenesis and how these relate to their flow properties (i.e. sealing or caprock quality). Seventeen samples from Lower Carboniferous estuarine caprock to a gas field (also to be used as a carbon capture site), have been analysed using a range of petrological and petrophysical techniques. The rock unit that represents the caprock to this gas field was found to be predominantly silt grade with porosity values as low as 1.8%. In these rocks, caprock quality (porosity) is controlled by intrinsic and extrinsic factors linked to primary mineralogy and diagenetic processes. Depositional mineralogy was dominated by quartz, detrital mica, detrital clay (likely Fe-rich 7Å clay and illiteesmectite) with minor feldspar and oxide phases. Diagenetic processes included compaction, minor feldspar dissolution and kaolinite growth and the more important processes of chlorite, siderite and quartz cementation, as well as the likely transformation of smectite into illite. Caprock quality is controlled by the primary quantity of illite-muscovite in the sediment and also by the localised extent of chlorite and quartz cementation. Deposition in an estuarine environment led to highly heterogeneous distribution of primary and diagenetic minerals, and thus caprock quality, between and within the samples. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Caprock CO2 Carbon capture CO2 sequestration Porosity Fine-grained siliciclastic lithologies Diagenesis Characterisation Tight gas Caprock quality Seal

1. Introduction The concept of petroleum entrapment and how caprocks act as seals is well established. Capillary sealing mechanisms have been presented and reviewed by Berg (1975) and Schowalter (1979). Schlomer and Krooss (1997) stated that, except for seismicallyinduced seal rupture or tectonic deformations, the retention of petroleum by overlying seals is controlled by capillary entry pressure, permeability, relative permeability and diffusive losses through the fluid saturated pore space. Capillary forces resist entrance of the buoyant petroleum in the underlying reservoir into the caprock. The capillary entry pressure is the threshold pressure at which this happens. The capillary entry pressure is dependent upon three factors; the radius of the largest * Corresponding author. Tel.: þ44 1517945149; fax: þ44 1517945196. E-mail address: [email protected] (P.J. Armitage). 0264-8172/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpetgeo.2010.03.018

connected pore throats, wettability and the petroleum e water interfacial tension (Downey, 1984). Once the capillary entry pressure has been exceeded, permeability and relative permeability will control transport through the caprock (Downey, 1984). Diffusion is considered to be marginal for oil migration, but may be significant for natural gas (Downey, 1984). Permeability data from tight lithologies are relatively rare due to the difficulties of direct measurement (Yang and Aplin, 2007). As a result, permeability is often assumed to be a function of other, more easily-measured physical properties, such as porosity. To substantiate this assumption, Dewhurst et al. (1999) used the few data available, previously collated by Neuzil (1994), with additional data from Dewhurst et al. (1995, 1998), Katsube and Williamson (1994) and Schlomer and Krooss (1997) to identify a logelinear relationship between porosity and permeabilties in finer grained clastic lithologies. Thus we use porosity in this paper as a proxy for permeability and caprock quality.

1396

P.J. Armitage et al. / Marine and Petroleum Geology 27 (2010) 1395e1410

The evolution of porosity in clastic units, whether they are sandstone, siltstone or mudstone, results from the interplay of depositional and diagenetic processes. At the time of deposition, mud has up to 70% porosity while sand has up to about 45% porosity (e.g. Worden and Burley, 2003). Over time this depositional porosity is lost due to the combined diagenetic processes of compaction and cementation. Porosity is an essential attribute of petroleum reservoirs, consequently a great deal of effort has been spent in documenting, trying to understand and then to predict porosity in deeply buried reservoir rocks. Significantly less effort has been expended on studies of caprock porosity even though the physical properties of these typically fine-grained rocks play a major role in controlling petroleum column height and ultimately whether a trap structure is effectively sealed or whether it leaks. If present, faults and fractures in caprocks could act either as pathways (e.g. Screaton et al., 1990; Brown et al., 1994; Clennell et al., 1998) or as barriers to fluid flow (e.g. Smith, 1980; Morrow et al., 1981, 1984; Dewhurst et al., 1996a,b) and therefore might have a significant effect on the sealing properties of a caprock. However, the whole core samples examined in this study were devoid of fractures (hand-specimen scale). This observation has been substantiated at the thin section scale since no fractures have been observed in thin section. Core plugs taken for analysis in this study therefore only detail matrix porosity. There are new motives for developing a better understanding of porosity in caprocks and other low permeability sedimentary rocks. Tight gas sands are a new exploration target (e.g. Olson et al., 2009). Although these rely on fracture porosity to flow to well-bore, the (relatively diminutive) matrix porosity remains important for overall reserves. Similarly shale gas is increasingly important as a resources (e.g. Ross and Bustin, 2008) again combining fracture-related flow to well-bore with ultimate reserves largely sitting within microporosity in the shale matrix. Caprock properties have been studied for their ability to retain petroleum fluids but there is a new imperative for the study of caprocks; to assess a caprock’s ability to retain CO2 in the various carbon capture (CO2 sequestration) projects that are being developed (e.g. Baines and Worden, 2004). Carbon capture (also known as geological sequestration of CO2) in depleted petroleum reservoirs and aquifers is currently considered to be a viable option for reduction of CO2 emissions to the atmosphere (Holloway, 1997), and is under investigation in field trials such as Weyburn in Canada (Riding et al., 2003), and Sleipner in the North Sea (Torp and Gale, 2003). The effects of CO2 storage on the host storage system (analogous to the petroleum reservoir) have been the subject of some recent research. However most research has been focussed at the host reservoir, with significantly less work reported on caprocks. CO2 addition to saline formation waters may lead to elevated bicarbonate concentrations and reduced pH. CO2 injection may induce watererock (geochemical) reactions in the caprock (Gaus et al., 2005). The lower interfacial tension of CO2-brine system than the original pore fluids will result in a lower capillary sealing pressure (Li et al., 2006). Storage of CO2, with its potential to increase fluid pressure, reduce effective stress and thus weaken the rock, and its potential to dissolve in formation water and change watererock equilibrium conditions, may thus significantly alter the geomechanical and geochemical properties of caprocks which could plausibly facilitate the release of CO2. In 2004, a CO2 capture and storage project was initiated at the Krechba gas field, in Algeria (Fig. 1; Riddiford et al., 2003). Krechba is a joint venture between Sonatrach, the Algeria National Energy Company, BP and Statoil and produces 9 billion cubic metres of natural gas per year. Approximately 10% of the gas in the reservoir is CO2. Rather than venting the CO2 to atmosphere, it is being compressed and injected into Lower Carboniferous sandstones into

Fig. 1. Locality map of the Krechba field within the Timimoun Basin in Algeria, North Africa.

the water leg; the deeper parts of the structure. Around one million tonnes of CO2 are being injected into the reservoir every year (Armitage, 2008). The caprock to the Lower Carboniferous sandstone reservoir must be sealing since there is petroleum gas in the structure but the caprock lithology has been reported to be relatively coarse thin-bedded siltstone and highly heterogeneous (Armitage, 2008). There is a renewed need to characterise caprocks when CO2 is being injected as part of a carbon capture project. The main generic questions being addressed by this research are: (1) What controls the quality (porosity) of caprock to this carbon capture site? (2) How variable is the caprock quality between (and possibly within) samples? (3) What is the mineralogy of the carbon capture site caprock? The local (specific) issues to be addressed for the Krechba caprock study include detailing the depositional environment of the Lower Carboniferous rocks at Krechba, their grain size and depositional mineralogy, the early and burial diagenetic cementation and porosity-reduction processes, and defining the importance of compaction processes for reduction of porosity. The overall aim is to define the main control on caprock quality in these Lower Carboniferous clastic rocks. 2. Geological background The main Krechba field comprises a NWeSE orientated, fourway dip closure anticline. The samples for this study come from the Lower Carboniferous (Tournasian) C10.2 and C10.3 (Fig. 2) units of the Krechba field, taken from two wells, KB6 and KB501. The main reservoir target is the lower part of the C10.2 unit capped immediately above by the upper C10.2 and C10.3 units, with non-


1397

bedsets of fine to very fine-grained, trough cross bedded to massive sandstones, with rare planar lamination. Plant fragments are abundant, large (up to 5 cm in length) and tend to define stratification. In the massive sandstones, these fragments are randomly arranged. These are interpreted to be the result of deposition from semi confined (channelized) flows close to a vegetation source, possibly a salt marsh and are the target rock units for CO2 sequestration. The tidally influenced muds comprise metre scale units that coarsen upward from heterolithics into dm thick very fine-grained sandstones containing mud-draped ripples, climbing ripples and extensive bioturbation. These deposits may have formed as sand prone shoals deposited from flood or ebb tides. The mud sheets are laterally-extensive and will act as caprocks (or at least barriers to flow) to the incised channel fill sands. The source rock for the Tournasian and Devonian reservoirs in Krechba was probably the regional Silurian hot shale although there are Devonian (Givetian) intervals with source rock potential (Hirst et al., 2001). The Lower Carboniferous (Tournasian) C10.2 and C10.3 units are currently at approximately 1800 m depth (Fig. 3). They reached a maximum burial of approximately 3300 m soon after burial, by about the Westphalian (Pennsylvanian). After this they were partially inverted during Hercynian orogenic events, followed by renewed burial to about 2200 m throughout the Jurassic with a final uplift happening during the lower part of the Tertiary (Palaeocene or Eocene). The heat flow history is not well known although the intracratonic setting will have led to muted fluctuations in thermal gradient. A maximum temperature of about 100 to 110 C can be estimated for the Lower Carboniferous rocks in Krechba, using an assumed geothermal gradients of approximately 30 C typical of marginal intra-cratonic basins (Allen and Allen, 1990). 3. Materials and methods Several wells have been drilled in the area, two of which have provided sample material: KB501 and KB6. Well KB6 is an older well that was drilled towards the centre of the main field. Well KB501 was drilled approximately 2 km eastenortheast of KB501 as an inclined pilot for a horizontal CO2 injector well located outside

Fig. 2. Stratigraphy of the reservoir and caprock of the Krechba field. The interval of interest is the Lower Carboniferous Tournasian units (C10.2), after Hirst et al. (2001).

reservoir layers acting as barriers to flow in the C10.3 unit. Some rock samples that are effectively reservoir-like (i.e. considerable visible porosity in thin section) are included in this study to elucidate continuity of processes which control the caprock petrology across the storage domain (Armitage, 2008). Armitage (2008) and Sutcliffe and Sabaou (1993) have characterised the depositional environments of the Krechba field, the summary that follows is largely derived from their analyses. The C10.2 unit has different sediment architecture in KB6 and KB501, reflecting a change in depositional setting. The C10.2 unit in the main Krechba field (e.g. KB6) is composed of well-connected, laterally-extensive, fining-up sand bodies with mudclast conglomerate lags at the base. These possibly denote tidal sandsheets, tidal channels and other shallow marine deposits that may be transgressive events with transition towards more distal settings. The C10.2 unit in the off-structure KB501 is composed of two distinct types of depositional element, incised channel fills and tidally influenced muds. The incised channel fills are sharp based and comprise stacked and amalgamated fining-upward (dm to m scale)

Fig. 3. Burial history of the Krechba reservoir and caprock interval.

1398


the main Krechba field. Fifteen samples from KB501 and two samples from KB6 were analysed in this study. The main reservoir target, the heterolithic C10.2, is capped by C10.3. The C10.2 unit contributes porosity to the gas reservoir in the crestal part of the Krechba field (at well KB6). The C10.2 in KB501 is down dip and below the gasewater contact and forms host reservoir units interlayered with laterally continuous caprocks or flow barriers to CO2 injection. C10.2 is capped by more homogeneous, shaley, C10.3. A total of 17 samples were taken from cores through the heterolithic reservoir and caprock to the Krechba accumulation (Fig. 4). Samples for petrological analysis were oriented perpendicular to bedding in order to reveal depositional and diagenetic features. A variety of lithologies were collected to represent the variable nature of the sedimentary rocks including silty units with subtle bedding, finely interbedded silt and mudstones, bioturbated intervals as well as (on the thin section scale) unbedded silty-sandstones. Whilst most of the Krechba samples are relatively fine-grained and have low porosity some of the samples are siltstone verging on moderately-porous, very fine-grained sandstone. There are essentially no variations in mineralogy with grain size for the Krechba samples with the implication that the occurrence and order of diagenetic processes have been similar irrespective of grain size. Uniquely defining some diagenetic phases is more straightforward in siltstones (and certainly in very fine-grained sandstones) than low

porosity mudstones; for example identifying quartz cement and grain-coating minerals. We have thus included the siltstone and fine-grained sandstone samples to help confirm the occurrence and order of diagenetic processes in the finest grained, lowest porosity samples. Samples 1 and 2 were taken from the C10.3 unit in KB6 (Figs. 2 to 4). Samples 3 to 9 were taken from the C10.3 unit in KB501. Samples 10 to 17 were taken from the C10.2 unit in KB501. Light optical examination was made in transmitted plane and polarized light and reflected light with a Meiji 9000 microscope fitted with an Infinity 1.5 camera. Images were stored and processed using Infinity Analyser software. Back-scattered electron (BSE), SEM-CL and EDAX analyses were carried out on a Philips XL30 SEM fitted with a K.E. Developments Ltd cathodoluminescence detector (D308122) and an Oxford Instruments Secondary X-ray detector. BSE images were collected at 20 kV and spot size 5. SEM-CL images were collected at 10 kV and spot size 7. They were collected by integrating the signal of 16 frames using a slow scanning raster. SEM-CL images took up to eight minutes to collect. Mineral analysis was performed using X-ray diffraction (XRD) and Fourier Transform Infra Red spectroscopy (FTIR). Samples were prepared from drill core, with at least 5 mm from the outer edge of the core removed to minimise contamination from drilling fluids. Samples were crushed and sieved to