Summary This case history describes a complete methodology from reservoir characterization to reservoir management for a producing field. A two-stage geostatistical approach was applied to distribute the reservoir heterogeneity coherently over the field. Ten 3D facies distributions were simulated. In a second step, 10 porosity and permeability images were geostatistically simulated on the basis of facies distributions. After suitable upscaling, all 10 reservoir images were initialized in the dynamic simulation model. The production history match was the discriminating criterion for the selection of the reservoir image to be used in the forecast phase. Introduction The study was aimed at piloting the drilling of the future 30 wells of the second and third development platforms in the northern area of an offshore field. The first platform wells had been producing since 1988, while six wells from the second and third platforms had a production history of 8 months. To obtain a reliable tool for reservoir management, the study was developed with the following goals:to preserve the effect of the petrophysical reservoir heterogeneity in the fluid-flow simulation model andto minimize the "adjustments" to the static model necessary to match the production history during the dynamic simulation phase. It is widely accepted that the quality of a flow simulation model (and the resulting forecasts) depends strongly on the accuracy of the static model on which it is based. When a dynamic simulation model is built, the reservoir characterization must focus on describing and quantifying the reservoir heterogeneities that control the fluid flow to reproduce the reservoir production behavior as accurately as possible. Unfortunately, to achieve the history match, the static model can sometimes be modified so severely that, after the final match, its results can be radically changed. Taking these considerations into account, the multidisciplinary team in charge of this study chose an integrated approach based on a combination of geostatistical and dynamic simulations. A geostatistical approach is particularly suitable to reproduce reservoir heterogeneities. It allows the use of all available geological knowledge, hard and soft, and provides n reservoir images, all respecting the input data and all equally probable from a static point of view. Consequently, only the past production behavior of the field can provide a guide for the selection of the reservoir image that is closer to reality. The n equiprobable images must therefore be initialized in the dynamic model, and the production history must be run for all images before a selection can be made. The reservoir image that best reproduces the production history should need only slight "adjustments" to obtain the final match and can therefore be considered a reliable basis for forecasts and decisions. Geological Setting and Reservoir Characteristics. The field structure is an anticline with a major axis that is northwest/southeast-trending. No remarkable tectonic phenomena have occurred, and no relevant faults have been recognized in the reservoir. The reservoir formation is made up of a complex lithology defined as "gray silty sandy dolomites passing to medium fine-grained quartz sands with dolomitic cement, with intercalations of tight dolomites and dolomitic limestones."

The paper presents the experience of a multilateral well campaign for a viscous oil field and its benefits in terms of production. The field is located in the North Slope of Alaska. The production start up was in January 2011. The main hydrocarbon reservoir is a Cretaceous sedimentary series developed in a sand lobe environment within the Schrader Bluff formation. The structure is a monocline gently dipping to NE with a gross thickness of about 30–35 ft. The oil presents different characteristics due to the instauration of a biodegradation process. Oil densities vary between 16°-19° API with associated viscosities in the range of 50–200 cP at reservoir conditions. The crude quality and drill pad constraints (one offshore, one onshore) drove the development concept which consisted of a waterflood linedrive with horizontal producers and water injector wells located side by side. The lateral sections are 6,000′–10,000′ long through the reservoir with undulating counter-phase trajectories across the two main sand bodies. A challenging horizontal reach/depth step-out ratio around 6.5 has been achieved. The completion string was equipped to provide the capability of integrating a complex reservoir monitoring plan including distribute temperature surveys, inflow control devices and polymer tracers. The choice came out from the need of minimizing the cost and mitigating risk with the highest respect for the environment and the local communities. With the goal of increasing reservoir contact, the multilateral concept consisting of undulating counter-phase dual-laterals was tested in August 2013. This initial test was followed by a limited campaign. As a result, well production doubled with respect to the average before the intervention. Similar productivity improvements were also observed in the new producers. At the time production was sustained by 20 producers and 15 injectors. Since then, eight additional laterals were added to existing producers and three new development wells were drilled and completed. The fieldwide step-change was evident with an overall production enhancement around plus 82% in May 2014. After the first encouraging results, the multilateral campaign has been extended to all the producers and is now standard practice for the remaining development wells in this field.

No abstract

The proposed methodology utilizes the concept of facies as the basis for reservoir characterization. Facies represent the elementary units of the reservoir. They are objectively determined through a log-based cluster analysis, and are fully characterized with petrophysical properties from cores. A stochastic model is used to distribute facies and related petrophysical parameters (porosity, permeability and irreducible water saturation) within the reservoir volume. Several equally probable reservoir realizations are generated from the same model and tested in a flow simulator. The realization providing the best history match is selected for production forecasts.