Capturing Reservoir Behavior by Simulating Vertical Fracture and Super-K Zones in the Ghawar Field
- Robert E. Phelps (Saudi Aramco) | Jonathan P. Strauss (Baker Atlas)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- August 2002
- Document Type
- Journal Paper
- 333 - 340
- 2002. Society of Petroleum Engineers
- 5.1 Reservoir Characterisation, 5.8.6 Naturally Fractured Reservoir, 4.1.5 Processing Equipment, 5.5 Reservoir Simulation, 4.3.4 Scale, 5.5.3 Scaling Methods, 3 Production and Well Operations, 1.6.9 Coring, Fishing, 5.6.4 Drillstem/Well Testing, 3.3.2 Borehole Imaging and Wellbore Seismic, 5.4.1 Waterflooding, 5.5.8 History Matching, 5.1.5 Geologic Modeling, 3.3.1 Production Logging, 5.1.2 Faults and Fracture Characterisation
- 2 in the last 30 days
- 445 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
A dynamic simulation model containing several million cells shows geological details used to simulate the vertical fractures and high stratiform permeability of the Ghawar field. Observed performance data have shown that well intersections in the Ghawar field exhibit thin intervals of extremely high productivity. These can be related to high-permeability layers (stratiform features) or to subvertical faults and fractures. To capture observed field behavior in a dynamic simulation model, these two features must be incorporated in the simulation model. This paper discusses how vertical fractures and high stratiform permeability were simulated in the world's largest oil reservoir.
It has long been known that well intersections in the Ghawar field show thin intervals of extremely high productivity. These can be related to high-permeability layers (stratiform features) or to subvertical faults and fractures. This paper discusses the characterization of these features and their inclusion in a field-scale simulation model. All high-permeability features (both matrix- and fracturerelated) are incorporated into the "fracture" component of a dualporosity, dual-permeability (DPDP) simulation. Fine-grid modeling is used to validate the approach and to gain an understanding of the physical processes taking place within the reservoir. Practical considerations and difficulties encountered in setting up a large-scale (16×17 km) sector model using the DPDP properties are discussed. Results for several runs that include 40 years of production history show that the methodology has the potential to improve the simulation modeling of this field.
The Ghawar field is the largest oil field in the world, stretching some 230 km in length and 25 km in width. The reservoir is located in the typically highly porous and permeable carbonates of the Arab-D formation. The reservoir has been under waterflood for almost 40 years. Attempts to history match the field with singleporosity models have had difficulty in matching the speed of water breakthrough and reproducing the irregular nature of the flood front. With the addition of horizontal wells and the use of borehole image logs, the identification and measurement of fractures in wellbores has added to the understanding of the Ghawar field. Individual fracture traces were picked from the image data and classified according to appearance (orientation, relative size, degree of mineralization, etc.).
Production logs frequently show thin zones that provide a large percentage of the total influx (Super-K zones). Examples of both stratiform (high-permeability layer) and fracture-/fault-related Super- K have been encountered. It is clear that both types of Super-K need to be incorporated into reservoir simulations to realistically model fluid flow in this field.
The characterization of various geological features and the assessment of their contribution to flow represents a major study1,2 that will be summarized here only briefly. The study relies on the analysis of data from an area of 23×27 km covering the northern part of the Uthmaniyah reservoir, which is part of the Ghawar field (Fig. 1). These data include image logs (17 wells), production logs (137 wells), conventional openhole logs (253 wells), core data (46 wells), and seismic fault and lineament maps.
In addition to statistical analysis of the data, various small-scale finite-element and finite-difference simulations were conducted to understand the role and interaction of the various features. The finite-element simulations were used to determine single-phase composite permeabilities for networks of discrete features (stochastic fractures and permeable matrix layers). The finitedifference reservoir simulations were used to investigate multiphase effects and to calibrate the permeability of individual and composite features within a permeable matrix against observations.
Geological Flow Components.
Based on observations and modeling, the reservoir can be broken into the following geological components (Fig. 2):
Faults and fracture clusters.
Highly permeable matrix layers (k-spikes).
Note that k-spikes include Super-K as a subset but also include more common features with lower permeability.
Investigation of the role of these components shows that substantial fracture contribution to permeability is confined to faults and fracture clusters. Within fault zones, the transmissivity can reach magnitudes similar to the most permeable matrix layers. In one example, a single fault zone provides 15,000 STB/D of flow from a 3-ft well intersection, approximately 75%of the total well rate. The calibrated transmissivity for this zone is 30 D·ft.
Typically, background fractures do not greatly enhance horizontal permeability, but they can, in conjunction with permeable matrix layers, significantly increase vertical permeability. Fluid transfer to k-spikes can be shown to depend both on the permeability of the surrounding matrix and on the density of background fractures.
Both k-spikes and faults/fracture clusters represent highpermeability features that hold a relatively small fraction of the total pore volume and, as such, are allocated to the fracture component of the DPDP description. In addition, the fact that individual fracture and matrix features reach similar magnitudes of permeability suggests that they should be grouped together. The "matrix" component is made up of the remaining, or background, matrix. The background fractures do not directly form part of either component; instead, their effect is included as a modifier to the transfer parameters.
The implication of the fact that only the larger-scale features such as faults and fracture clusters appreciably influence permeability is that these are the features that need to be characterized. Unfortunately, these features are rarely intersected by vertical wells, thus greatly reducing the statistical base from which to work. On a more positive note, these features are more likely to be visible on seismic, thus allowing the variation in their abundance and orientation to be mapped out.
|File Size||3 MB||Number of Pages||8|