Michael C. Kelly

AVO inversion, Part 2: Isolating rock property contrasts

Part 1 of this article (TLE, March 2001) described a method for extracting rock property contrasts from seismic data. This has opened many doors previously thought to have been closed to the explorationist. This paper will address issues such as determining hydrocarbon saturation, production effects, nonbright pay, and S-wave seismic from P-wave acquisition and processing. These are pertinent problems in todays industry, and many require great expense to solve.

The effect of reflector dip on AVO analysis

AVO (amplitude variation with offset) and seismic attribute analyses often assume a flat model for the calculation of reflection angles from the offset distance. If the dip is not considered, the calculated AVA (amplitude variation with angle) will be incorrect in steeply dipping areas. Raytrace modeling and AVO analysis in this study indicates substantial muting of AVO attribute values for steeply dipping beds compared to more flat lying beds with similar rock properties. If dip above about 20° is present on a prospect, modeling should be done to understand its influence on AVO and other seismic attributes and the AVO analysis should include the structure. With increased dip, certain rock property contrasts produce reduced HCI and AVO signatures that could be mis-interpreted because they do not appear anomalous.

Non-linear AVO equations and their use in 3-parameter inversions

Summary The exact Zoeppritz equations have been simplified to involve non-linear contributions from the elastic rock property contrasts. This approximation is more accurate than the linearized equation widely used by the industry. The non-linear equations are a result of expanding the exact equations to second order in the rock property contrasts. Consistent with the original linearizations provided by Bortfeld and others, no explicit small angle assumptions have been made using the ray parameter p or other small angle related parameters. The resulting equations are more accurate than the linearized equations, especially for large rock property contrasts and large angles of incidence. Inversion, based upon these non-linear equations, provides more accurate and stable results and allows amplitudes corresponding to very large angles found in modern long cable acquisition to be used in inversion. A number of examples illustrating the improved accuracy of these equations compared with the traditional linear equations will be provided. A number of examples of inversions, using the non-linear equations applied to both synthetic and real data, will also be provided.

Quantitative AVO analysis

Performing AVO analysis with the objective of producing quantitative attributes or rock property contrasts, which can be directly compared with well data is possible. The seismic attributes and rock property contrasts derived from seismic data that has been properly calibrated can be directly compared with well data using either cross-plots or a log curve format. Amplitude calibration resulting from applying amplitude corrections consistent with average amplitude background behavior commonly use rock property relationships that describe background event behavior. For the calibration process to be successful it is necessary that processing steps do not modify the dynamic range of the data. The calibration process uses information readily available within basins around the world. In this paper we will describe the calibration process and provide examples derived from actual data.

Subsalt AVO effect of shooting direction in areas of dipping salt: Violation of amplitude reciprocity

Summary The well known reciprocity principle in seismic acquisition states that the raypath of seismic energy is unaffected by the switching of source and receiver. Less well known are the amplitude effects of the switching of source and receiver. In this study, raytrace modeling results of subsalt reflectors below dipping salt bodies show a marked AVO difference between a line shot in one direction and the same line shot in the opposite direction. Calculations of Zoeppritz equations confirm the difference in amplitudes caused by partitioning of seismic energy at the salt/sediment interfaces. The effect should be considered when processing and interpreting subsalt reflections and when modeling subsalt areas.

AVO Rollover: A saturation sensitive direct hydrocarbon indicator

Introduction of small or large quantities of hydrocarbons into a porous low impedance reservoir has been known for some time to produce an amplitude increase with moderate offset. This direct hydrocarbon indicator signature has been successfully exploited for some time. The drawback of this indicator is its inability to clearly distinguish between fully and partially saturated reservoirs. What will be highlighted in this paper is a robust direct hydrocarbon indicator, the “AVO rollover,” which is sensitive to hydrocarbon saturation. Unlike the traditional amplitude increase with offset, the rollover shows itself at large rather than intermediate offset. A number of Gulf of Mexico examples will be provided to illustrate the effect. Statistics indicating the widespread nature of this effect will also be provided. Further, the physical basis for this effect and why it is sensitive to saturation will be shown using both interface and 1-d modeling. Non-low impedance reservoirs also exhibit a specific AVO rollover signature, but the focus of this paper will be on the low impedance case.

Method for generated log parameter volumes from seismic

Wire line logs have been used by geologists, geophysicists and reservoir engineers to identify parameters such as lithology type, fluid type, degree of invasion, degree of hydrocarbon saturation, and porosity of subsurface formations. The reliability and accuracy of well logs are such that they are considered ground truth. The limitation of logs as a tool for understanding the subsurface is, of course, that a hole must have been drilled in which the logging tool can be run, limiting its use to a post drilling evaluation tool rather than allowing it to be used as a predrill exploration tool such as seismic. As an evaluation tool, the sparse surface sampling resulting from the measurements restricted to a well bore, although offering excellent vertical sampling, is limited compared with seismic which offers dense surface sampling but has, compared with logs, limited vertical sampling. The disadvantages of seismic are that amplitudes and attributes are generally not considered as directly correlated with the important geologic and geophysical parameters routinely derived from logs and they are not as easily and accurately interpreted.

Quality controlling AVO inversion products: Examples of amplitude decomposition into rock property contrasts

Rock property contrast cubes derived from AVO inversion have brought the industry closer to extracting the physical properties of the subsurface from seismic data. VanKoughnet et al. (2002) showed the value of this type of application to complex reservoir problems. This study takes on the challenge of demonstrating the potential accuracy and existing pitfalls of this technique through a process termed amplitude decomposition. Examples derived from both wireline and real seismic data will be used to demonstrate the technique and support the conclusions.

P-S AVO Attributes And Cross-Plotting

Summary P-P AVO has proven to be one of the most effective geophysical risk reduction methods available. With the widespread availability of multi-component data and improved multi-data type true amp processing methods it becomes possible to accurately calculate PS AVO attributes. These attributes, used alone or jointly with P-P attributes, provide a simpler, more direct connection to the underlying rock property contrasts resulting in quantities which are easier to interpret and more sensitive to fluid type, saturation and reservoir quality variations. Cross-plotting multiple attributes provides even more discrimination leverage and sensitivity than traditional plotting methods. It will first be shown that Equation 1 provides a linearized form of the exact equation. This form is accurate for most exploration cases. The D0 vs. D1 cross-plot displacement signatures associated with each rock property contrast will be shown in Figure 1. In addition, it will be shown that the crossplot background trend (BT), associated with reflections from layers following local average rock property trends, can be calculated from simple relationships. The cross-plot displacement from the BT, associated with hydrocarbon substitution, can then be easily predicted. It will also be shown that P-S attribute cross-plots can provide information about the variation of rock properties associated with two spatial locations or for a single location but at two times (T1 and T2) if time lapse data is used.

P-P and P-S Angle Stack Inversion

Summary Angle stack amplitudes have proven to be one of the simplest and most effective tools for determining qualitative AVO signatures. The AVO response of an event, as revealed through angle stacks, has proven to be a good discriminator and successful at risk reduction. In spite of these positives, those familiar with the use of amplitudes from angle stacks are acutely aware that a single signature, such as an increase in amplitude with angle of incidence for the P-P case, does not uniquely identify zones of fully saturated pay with good reservoir quality. To go beyond the qualitative information provided by angle stacks it is necessary to establish the relationships between angle stacks and the underlying rock property contrasts. We will show how these relationships can be calculated for both PP and P-S data. Examples of these relationships for specific angle ranges will be shown. We will also invert these relationships providing expressions that give the rock property contrasts in terms of a collection of angle stack amplitudes. Real data examples showing predicted rock contrast maps will be shown and compared with well data. Effects of anisotropy will be considered.