Figure 1. Contour map (a) of a horizon interpreted from 3-D data in Colorado showing large-scale trends in the time surface. The figure on the right is the same surface, with contour lines overlaying the actual time surface. Subtle trends are observed within contour intervals. (Figures courtesy of SMT)

Advances in Windows-based hardware and software have produced affordable computer resources that allow everyday interpreters the same access to exploration tools previously held only by large organizations. This “leveling of the field” has created a diverse dynamic industry of innovative science that exploits resources in ways unimaginable to previous generations.

The objective of visualization is to graphically display large volumes of data in a meaningful way to aid in the exploration and development of hydrocarbon resources.

Visualization techniques are used in every facet of petroleum E&P, including seismic acquisition, seismic processing, geological well logging, interpretation, reservoir engineering, management, and financial review/planning. This article reviews some general visualization methods common to geoscience interpretation and the use of these techniques to determine regions of possibly overlooked pay. The following examples and analysis were conducted using SMT’s KINGDOM software on a Microsoft Windows platform.

Need for visualization

Figure 2. The left image (a) shows a relief shading of the depth surface. Note the subtleties in the hues using only gradational shading between white, purple, and black. It is easy to detect faults, channels, and surface relief in fault blocks. Figure b is a shaded relief of the depth surface. Sun angle 30? above the horizon to the southeast. Note the striking appearance of the northeast-trending faults. If the sun angle is moved to different locations, other features will brighten.



Geological and geophysical acquisition generates large volumes of information to be processed and interpreted for single projects. Data volumes today can exceed 4,500 sq miles (11,650 sq km) aerially and consist of tens of thousands of wells. An integrated visualization workstation is critical to effectively interpret such volumes. Such a system:
• Has the ability to quickly scan data;
• Conveys visual information in a form meaningful to the interpreter;
• Simultaneously displays different categories of data in the same environment;
• Contains tools to manipulate the visual data; and
• Provides a foundation for communication among various disciplines and management/investors.

The evolution of affordable Windows-based workstations has allowed interpreters access to large datasets previously in the domain of large companies with far-reaching resources. Of particular interest to the explorationist is the interface of well log information (high vertical resolution, low lateral control) and seismic reflection data (good lateral control, poor vertical resolution).

The actual selection and manipulation of visual displays are dependent on several factors. The most important decisions include the objectives of the project, the available data, and the target audience to which the results are communicated. After these criteria are established, the interpreter pursues a workflow that dynamically responds as the project evolves.

Visualization issues peculiar to geoscience interpretation Color.

Figure 3. Relationship of well log information to the seismic dataset. The color traces are synthetic traces derived from well logs at this location. The interpreted horizon (yellow) is tied to a formation top via a time-depth chart taken in this well.



The ability to extract and view meaningful data relies nearly as much on the careful choice of viewing parameters as to the attributes being displayed. For example, Figure 1a shows a contour map of a horizon of interest in Colorado. The actual presentation of contours depends on many parameters, including choice of contour interval, range, and fill color.

The colors in the map change stepwise across contour lines to show the gross changes in the surface reflection time. This produces an overall understanding of major trends within the horizon of interest. An understanding of subtle secondary trends within the horizon may also require a gradational change of colors within the larger structure. Figure 1b displays the same surface, this time overlaying the contour lines from Figure 1a on the actual time surface. Different color scales may be tested to determine optimal interpretive parameters.

Relief. Sometimes a reliance on color alone may not yield sufficient information about the interpreted horizon of interest. Figure 2a is a grid representation of the interpreted horizon converted to depth using a relevant time-depth chart. The surface has been highlighted in relief shading and the colors held to white, purple, and black. Additional perspective may come from moving the light source in both direction and angle above the horizon. In Figure 2b, the angle of the sun is from the southeast and approximately 30? above the horizon. Such views are effective in displaying micro-trends within fault blocks and the location of possible seals and migration pathways as well as in identifying individual fault blocks for reservoir interests. Relief shading is also valuable as a quality-control tool to review data processing.

The horizon displayed in Figure 1 was extracted via a 2-D hunt algorithm on a reflector of interest. The surface was chosen by simultaneously examining a formation top of interest chosen from well logs. Because the well logs are recorded in depth and the horizon in time, a time-depth conversion must be applied based on velocity information taken from a well in the area of interest.

Figure 4. Structural cross section of the Sooner field (a). The D Sand is the yellow stippled body in the middle of the stratigraphic column. Stratigraphic cross section by flattening of the top of the D Sand (b).

An overlay of the log data on the trace display is the key to understanding this relationship to the trace volume. Figure 3 shows the relationship that was used to determine the hunt parameters for the seismic horizon. In this display, a synthetic trace has also been generated that calculated a reflectivity trace based on velocity and density logs, convolved the result with a wavelet chosen by the interpreter, and is tied to the seismic dataset in time and phase. The resulting knowledge not only ties the well data to the seismic but also indicates whether the reflectors are in the correct phase to accomplish later attribute calculation and interpretation.

Geologic interpretation. Visualization is also effective as an interpretation tool via creative use of cross sections. The ability to rapidly change parameters and view the results has advanced this field well beyond the paper-based displays of previous generations. A good visual display will incorporate information from location, logs, formation tops, and a priori geological knowledge to make a display that communicates the objective to the target audience.

Figure 4a shows a structure cross section through the D Sand, Sooner field, Colorado. This sand has been a source of significant oil and gas production for many years.

Figure 4b shows a simple parameter change in which the section is flattened on the top of the D Sand. This cross section yields valuable information regarding the gross thickness of the sand for reservoir management purposes. Rapid parameter changes such as this have markedly reduced interpretation time and shortened exploration to production cycles, subsequently leading to improved economic return on investment.

Use of seismic attributes. After seismic horizons have been tied to well data, it becomes important to visualize as much information as possible from attributes taken from seismic trace data. Attributes can be classified according to the classes found in Table 1.

Figure 5. Time slice extraction of a volume of spectral decomposition followed by dip of maximum similarity. The spectral decomposition bandwidth is centered at 40 Hz. Here one can easily see compartments of like seismic stratigraphy within discrete areas across the sand surface. Blue colors indicate low rate of change of dip, while yellow indicates a rapidly changing dip along this time slice.

Each seismic trace attribute from Table 1 is useful under specific circumstances, although the relevant attributes for a particular survey may not be apparent until after interpretation commences. Attributes can be used as valuable indicators of geologic and structural changes in a seismic volume; the combinations of attributes may provide more information regarding specific exploration and production strategies.

An example of the combination of spectral decomposition followed by dip of maximum similarity shows the structure taken along a time slice at the D Sand (Figure 5). The spectral decomposition uses a bandwidth of a convolved wavelet to the trace data on a sample-by-sample basis within a moving window. The dip of maximum similarity calculates the rate of change of the dip. Taken together, one can interpret channel boundaries to guide well-path planning.

Using attribute analysis to detect possibly overlooked pay. An alternate view of the spectral decomposition/dip of maximum similarity combination is to render the data in a 3-D view. Figure 6 shows the data from the same volume used for the slice display in Figure 5, but now the attribute is displayed in volume mode. Regions of relatively little change in dip have been made transparent to interpret the 3-D structure of channel boundaries displayed in yellow. Because the interiors of the channels are now transparent, it is easy to observe locations where previously drilled wells penetrate them. Shaded lathe renditions of gamma ray logs can be easily followed into the channels and used to calibrate rendering parameters, sand quality, and sand type within individual sand units. Noteworthy, however, are several regions that appear to be physically separate from production blocks that remain untapped by wells. In this case, a reservoir management/drilling program can be derived from this valuable information, and the field can be economically exploited.

Figure 6. Volume rendering of the same dataset as in Figure 5. Gamma ray lathe plots clearly show that production has come from specific sand units of relatively flat dip (transparent) bounded by channel edges (yellow rendering). It is noteworthy that several channel reservoirs can be interpreted from this view that have not yet been drilled.

Conclusions

Visualization is a maturing technology that continues to make rapid advances based on ongoing improvements in visual display and the speed of data processing. Within petroleum exploration, an effective integrated interpretation software system based on a reliable hardware platform allows the interpreter few limits to the imagination. Benefits include improvement in exploration targets, reservoir management, risk analysis, and economic forecasting.