Summary
This paper discusses the integration of dynamic reservoir data at the flow-unit scale into the reservoir management and reservoir simulation efforts of the Takula field. The Takula field is currently the most prolific oil field in the Republic of Angola.
Introduction
The Takula field is the largest producing oil field in the Republic of Angola in terms of cumulative oil production. It is situated in the Block 0 Concession of the Angolan province of Cabinda. It is located approximately 25 miles offshore in water depths ranging from 170 to 215 ft.
The field consists of seven stacked, Cretaceous reservoirs. The principal oil-bearing horizon is the Upper Vermelha reservoir. This paper discusses the data acquisition and integration for this reservoir only.
The reservoir was discovered in January 1980 with Well 57- 02X. Primary production from the reservoir began in December 1982. The reservoir was placed on a peripheral waterflood in December 1990. Currently, the Upper Vermelha reservoir accounts for approximately 75% of the production from the field.
Sound management of mature waterfloods has been identified as a key to maximizing the ultimate recovery and delivering the highest value from the Block 0 Asset.1 Therefore, the objective of the simulation effort was to develop a tool for strategic and dayto- day reservoir management with the intent of managing and optimizing production on a flow-unit basis. Typical day-to-day management activities include designing workovers, identifying new well locations, optimizing injection well profiles, and optimizing sweep efficiencies. To perform these activities, decisions must be made at the scale of the individual flow units.
In general, fine-grid geostatistical models are developed from static data, such as openhole log data and core data. Recent developments in reservoir characterization have allowed for the incorporation of some dynamic data, such as pressure-transient data and 4D seismic data, into the geostatistical models. Unfortunately, pressure-transient data are acquired at a test-interval scale (there are typically 3 to 4 test intervals per well, depending on the ability to isolate different zones mechanically in the wellbore), while seismic data are acquired at the reservoir scale.
The reservoir surveillance program in the Takula field routinely acquires data at the flow-unit scale. These data include openhole log and wireline pressure data from newly drilled wells and casedhole log and production log (PLT) data from producing/injecting wells. Because of the time-lapse nature of cased-hole log and PLT data, they represent dynamic reservoir data at the flow-unit scale. To achieve the objectives of the modeling effort and optimize production on a flow-unit basis, these dynamic data must be incorporated into the simulation model at the appropriate scale. When these data are incorporated into a simulation model, it is typically done during the history match. There are, however, instances when these data are incorporated during other phases of the study. The objective of this paper, therefore, is to discuss the methods used to integrate the dynamic reservoir data acquired at the flow-unit scale into the Upper Vermelha reservoir simulation model.
Reservoir Geology
The geology of the Takula field is described in detail in Ref. 2. The aspects of the reservoir geology that are pertinent to this paper are elaborated in this section.
Reservoir Stratigraphy.
The Takula field consists of seven stacked reservoirs. The principal oil-bearing horizon is the Upper Vermelha reservoir. This reservoir contains an undersaturated, 33°API crude oil.
For reservoir management purposes, 36 marker surfaces have been identified in the reservoir. Flow units were then identified as reservoir units separated by areally pervasive vertical flow barriers (nonreservoir rock). This resulted in the identification of 20 flow units. The thickness of these flow units ranges from 5 to 15 ft.
Reservoir Structure.
The reservoir structure is a faulted anticline that is interpreted to be the result of regional salt tectonics. Closure to the reservoir is provided by faults on the southwestern and northern flanks of the structure and by an oil/water contact (OWC) on the eastern, western, and southern flanks of the structure. A structure map of the reservoir is presented in Fig. 1.
Data Acquisition in the Takula Field
Openhole Log Program.
Most original development wells were logged with a basic log suite of resistivity/gamma ray and density/ neutron logs. In addition, the vertical wells drilled from each well jacket were logged with a sonic log and, occasionally, velocity surveys.
All wells drilled after 1993 were logged with long spacing sonic and spectral gamma ray logs. In many wells drilled after December 1997, carbon/oxygen (C/O) logs have been run in open hole to distinguish between formation and injected water.3
A few recent wells have been logged with nuclear magnetic resonance (NMR) logs. The NMR log data, when integrated with data from other logs, have been of value in distinguishing free water from bound water, formation water from injection water, and reservoir rock from nonreservoir rock.
Implementation of artificial intelligence and support vector machine learning to estimate the drilling fluid density in high-pressure high-temperature wells
