Deep-water reservoirs consisting of turbiditic sandstones moderately to heavily reworked by bottom currents are common in canyon-and trough-filling deep-water (bathyal) Palaeocene-Eocene sequences of the Campos Basin, offshore southeastern Brazil. A number of wells with conventional logs, together with cores, provided the database for the study. Seismic data provide additional support, but low resolution and noise hamper detailed analysis. The sandstones presenting better reservoir quality in these sequences are interpreted as being deposited by turbidity currents, as suggested by the dominance of unstratified normally graded sandstones, with grain sizes ranging from fine to coarse sand, and low clay-matrix content. Sandstones interpreted as bottom-current deposits (mid-water contourites) form poor-quality reservoirs, or baffles and barriers. These rocks are commonly moderately to heavily bioturbated, with variable, frequently high, clay-matrix content. Common trace fossils include Planolites, Palaeophycus and Zoophycos.Locally, these sandstones show faint horizontal stratification and planar cross-stratification. Contourites with thickness ranging from a few decimetres to several metres occur intercalated with turbiditic sandstones. Because they present distinct reservoir qualities, the mapping of the limits between turbidites and contourites is critical for adequate reservoir characterization. Most of this mapping has been performed using well information, constrained by outcrop analogues. The currents responsible for reworking turbiditic sands are interpreted to be deviated geostrophic currents, with velocity enhanced in narrow canyons and troughs.

Horizontal well geosteering involves careful monitoring of the well during the drilling phase. To this end, relevant geophysical and geological tools are used to build a consistent geological model. Real-time information enables constant updating of this model providing the necessary information for geosteering decisions. These geosteering procedures allow for close control of the well trajectory in the target layer, reducing the risks of redrilling. Introduction One of the main factors that have made the development of several deep-water Campos basin fields feasible was the advance in the horizontal well technology. Horizontal wells potentially improve the productivity / injectivity index and also optimize the sweep efficiency for a given drainage plan. The challenges faced to drill these types of wells in deep- waters, sometimes within thin layers, have pushed the implementation of solutions forward, reducing the risks and costs of these well projects. The horizontal or high angle drilling can be divided into three main phases: planning, geosteering and completing. At the planning phase, the role of the well within the reservoir drainage plan must be understood by the multidisciplinary team involved in the enterprise, in order to optimize the decision-making process during the drilling of the well. To perform detailed well planning, all geological, geophysical and reservoir engineering data available are taken into account, and, the depth converted seismic volumes, well sections, correlation logs and 3D geological model play the most important role. Horizontal well geosteering involves careful monitoring of the well during the drilling phase. To this end, relevant seismic and geological tools are used to build a consistent geological model, and geosteering software uses the comparision of simulated and acquired logs to confirm or adjust the constructed model. This paper presents some of these geosteering tools used during the development phase in a Campos Basin deep-water field. Horizontal Well Planning and G&G Interpretation Horizontal or high angle well planning requires a detailed knowledge of the structural and stratigraphic framework in the target area. The framework is defined from the geological and geophysical interpretation of the available data. Among the relevant data are seismic sections and maps of different seismic attributes, like amplitude, acoustic and elastic impedance, and coherency. Moreover, the geological input for consistent well planning includes electric and radioactive logs, lithological descriptions as well as reservoir pressure data. With this information available, a detailed definition of the reservoir model in the area of the well is obtained, including the external geometry and the reservoir architecture, in order to properly define the main well targets, see Figures 1 and 3. Prior to the horizontal well, a pilot well is drilled to support the horizontal well planning and monitoring. This well is quite often a deviated well, with penetration angles no greater than 45 degrees, aiming at getting relevant information to help with the horizontal well project design and reservoir management activities. This pilot well provides information like the actual reservoir top and base depths, oil-water contact, sand-shale distribution throughout the reservoir and pressure data, revealing the vertical and horizontal connectevity.

fax 01-972-952-9435. AbstractCaratinga giant oil field is located in the central part of Campos Basin, Southeast of Brazil, in water depths around 1,000 meters. The total reserve is around 290 MM boe in these turbidite sandstones, 78% of which are within the lower Oligocene Reservoir (CRT100) and 22% are within Oligocene/Miocene and Eocene/Paleocene reservoirs. The CRT100 is a turbidite submarine fan of maximum thickness of 40 m, that was further cut by a NW-SE, Lower Oligocene submarine canyon, that have segmented the reservoir in two blocks: North Block and South Block. This canyon was further filled by the Oligocene MRL600/700 and the Oligocene/Miocene MRL330 sandstones. These canyon-filling sandstones constitute the Central Block. The development strategy for this field included a Pilot Phase in which three producer wells, one in each block, produced to a FPSO. The history match of this production data did not consider connectivity between the North and South Blocks of the CRT100 reservoir; although, seismic data have suggested possible reservoirs connection through the Central Block. Despite these features were represented in the 3D reservoir modeling, the transmissibility multipliers used kept these connections closed. The definitive production system, with 12 producers and 8 injection wells, started in February of 2005. The extensive use of pressure down hole gauges and a dedicated reservoir team allowed the observation of very important new information about the reservoir hydraulic behavior. Among several other issues that have arisen, a good communication across the Canyon was confirmed, connecting the North and South Blocks. This effective hydraulic communication observed can be explained by the sand to sand juxtapositions between off-canyon and canyon-filling sandstones in some regions, whereas the low thickness reservoirs, associated to geological faults, can justify the fluid flow behavior in other regions. These issues have been considered in the flow simulation model. The results are showing an increment on the total recoverable oil, since some portions of the canyon sandstones that used to have a low recovery volume, are now being swept by the water flooding.