Simply defined, seismic interpretation is the science (and art) of inferring the geology at some depth from the processed seismic record. While modern multichannel data have increased the quantity and quality of interpretable data, proper interpretation still requires that the interpreter draw upon his or her geological understanding to pick the most likely interpretation from the many “valid” interpretations that the data allow.
The seismic record contains two basic elements for the interpreter to study. The first is the time of arrival of any reflection (or refraction) from a geological surface. The actual depth to this surface is a function of the thickness and velocity of overlying rock layers. The second is the shape of the reflection, which includes how strong the signal is, what frequencies it contains, and how the frequencies are distributed over the pulse. This information can often be used to support conclusions about the lithology and fluid content of the seismic reflector being evaluated.
The interpretation process can be subdivided into three interrelated categories: structural, stratigraphic, and lithologic. Structural seismic interpretation is directed toward the creation of structural maps of the subsurface from the observed three-dimensional configuration of arrival times. Seismic sequence stratigraphic interpretation relates the pattern of reflections observed to a model of cyclic episodes of deposition. The aim is to develop a chronostratigraphic framework of cyclic, genetically related strata. Lithologie interpretation is aimed at determining changes in pore fluid, porosity, fracture intensity, lithology, and so on from seismic data. Direct hydrocarbon indicators (DHI, HCIs, bright spots, or dim-outs) are elements employed in this lithologic interpretation process.
This article discusses a basic three-step methodology, which, when followed, provides for a more complete and accurate geological interpretation from seismic data.
Step one: Interpretation plan
Accomplishing a successful interpretation requires that the interpreter first carefully consider the following questions.
What are my objectives?
An interpreter should clearly understand what conclusions are required from the data. Because so much information is available on the seismic, it is important to focus maximum attention on extracting the data pertinent to completing the objective task. Does the objective require evaluating the entire dataset from first sample to last, one stratigraphic sequence, or just one specific amplitude anomaly? This dictates what combination of the three basic interpretation types should be used, when the interpretation should be completed, and what supporting databases are required.
What are the regional tectonic, structural, and depositional trends?
It is important for the interpreter to have a basic understanding of what tectonic influences and depositional systems occur within the area of the seismic survey to be investigated. Although this preconceived earth model may be vague and incomplete, particularly in frontier basins, it provides interpreters with insight and constraints as to what types of structures, faulting, and stratigraphic geometries may exist. The interpretation of fault styles, structural geometries, and facies patterns must be consistent with regional tectonic forces and basin infilling.
What seismic patterns should I be looking for?
Perhaps the most common interpretational pitfall, and certainly one of the most dangerous, is the mapping of events, amplitude, or AVO changes without qualification as to what geological analog they represent. To prevent this mistake, it is critical that all types of available geological data be gathered and merged with the seismic data. Key to this merging are well-constructed synthetic seismograms, vertical seismic profiling (VSP) data, and/or seismic models (see Synthetic seismograms, Checkshots and vertical seismic profiles, and Forward modeling of seismic data). This verifies the seismic signature of the target, the location of the mapping horizon, and the adequacy of the time-depth functions. Varying the synthetic seismogram or model parameters allows for the prediction of seismic responses for various lithologics and fluid types.
Step two: building and merging datasets
After developing an interpretation plan, the next step is to begin assembling the complete dataset. An inventory of available seismic data of all vintages is made, including p wave seismic, shear wave seismic, well data, velocity surveys, and VSP. The data are scanned for quality and suitability. At this point, a determination can be made whether the available data can reasonably support the goals of the project.
Available well, core, test, paleontology, and outcrop data are gathered and organized for integration with the seismic data. Where available, gravity and magnetics data should be tied to the seismic data to identify the location of basement, salt bodies, igneous intrusives, and shale masses. Another type of data that sheds light on the geological conditions of a specific reservoir is pressure and production history data. These data can provide information on the presence and proximity of faulting and the size of fault blocks.
If digital data are available, a decision must be made whether to use a workstation or proceed with a “paper interpretation.” Generally, a workstation offers the interpreter a valuable edge achieving a “correct” interpretation where detail is important (see Two-dimensional geophysical workstation interpretation: generic problems and solutions). For much regional work, paper is often still the most used medium due to the display limitations of the workstation screen. Seismic interpretations on paper, however, can always be digitized later for computer mapping or incorporation into a workstation project.
Step three: interpretation
The process of interpreting seismic data eventually comes down to putting pencil to paper or cursor to screen. After building an exploration analog by integrating the available geological data, it is advisable to scan the dataset to observe the basin setting, major structural components, and major stratigraphic components, such as reefs, shelf breaks, and major sequence boundaries. While scanning, major faults can be picked as a guide to establishing the dominant structural style.
After scanning, detailed mapping begins by working outward from a point where geological information exists, preferably a well location with a synthetic seismogram. The horizons selected for mapping and observed fault cuts are correlated from the well to the seismic. The interpreter then begins to pick these same events away from the well on the seismic, being careful to tie at all other well locations.
Critical to the interpretation process is comparing how horizons and faults tie at line intersections. Significant effort is expended correcting misties of faults, horizons, and sequence boundaries at every line intersection. In this regard, closing the interpretation in loops around the seismic grid is a particularly effective technique. On a workstation, a quick way to check for misties is a contourmap. Misties will be evident by groups of unreasonable contours. In addition, workstations can be very helpful for working out the misties among varying vintages of two-dimensional data by applying time and phase shifts automatically.
Tying all lines in both 2-D and 3-D data sets is the only way to reliably construct a three-dimensional model of the subsurface using two-dimensional images. Tying around data loops is also the best way to correlate from fault block to fault block. Otherwise, faults must be jumped using reflection character, sequence analysis, or additional well control.
After all lines are picked and tied, the results of the interpretation are then summarized and presented as maps. Basically, any observation that can be made using seismic data can be posted on a base map and mapped. Maps that are routinely made include
- Time structure maps with faults
- Depth structure maps
- Seismic facies maps for reservoir, source, or seal analysis
- Seismic amplitude maps for DHI analysis
- Thickness maps inferred from seismic tuning analysis
- Fault plane maps
- Fault plane maps with cross-fault sand juxtaposition for seal analysis
- Isochron or isopach maps showing growth or thinning in a stratigraphic interval
- Seismic velocity maps for lithology determination or depth conversion
In addition, many combinations of these maps can be made, such as seismic amplitude plotted on top of structure. The only limitations in constructing these maps are the imagination and skill of the interpreter.
The overall aim of seismic interpretation is to aid in constructing the most accurate earth model or reservoir description possible. This can best be accomplished when the seismic data are merged with petrophysical, geological, and engineering databases. While the process of interpreting seismic data is basically the same on paper or in a workstation environment, the workstation offers advantages in data management, manipulation, and display and it allows for a more convenient integration of other data types.