The hardness of rocks in the subsurface varies with factors such as depth of burial, compaction, porosity, lithological composition and the presence of hydrocarbons. In a seismic survey, the reflection strength (seismic amplitude) is a measure of the contrast in properties between two subsurface layers. Using seismic inversion, the measured reflection strength can be converted to relative impedance, indicating the relative change of the rock property impedance, or hardness, as the interpreter moves from one layer to the next.
A map of amplitudes associated with an interpreted seismic pick might be interpreted in terms of lateral variations of the geology. In order of magnitude from strongest to weakest, rock property changes that might result in amplitude variations that might be seen and interpreted on seismic are typically lithology, porosity, presence of gas and presence of oil.
The amplitude variations most difficult to detect are those caused by the presence of direct hydrocarbon indicators (DHIs), oil in particular being a very subtle effect. In a modern seismic inversion workflow, the inversion is typically performed prestack, which allows the impedance to be represented as amplitude vs. offset (AVO) anomalies.
A very significant risk factor with amplitude analysis (and AVO interpretation) is that thickness-related tuning effects severely distort amplitudes, complicating analysis in general and introducing issues of reliability and uncertainty in the identification of genuine amplitude anomalies. Tuning is defined as constructive or destructive interference resulting from two or more layers around a quarter of the dominant wavelength, and it particularly affects the detection of AVO anomalies, including DHIs. It is very common for amplitude maps to be analyzed and interpreted while completely ignoring tuning effects. Thickness-related tuning effects are the single largest factor affecting amplitudes in seismic data, larger than any other effects due to geological changes, including lithology or hydrocarbon fluid effects.
The significance of the effect of tuning can be demonstrated using a simple constant property “wedge” model. An example for a soft sand reservoir in the Kadanwari Field, Pakistan, is shown in Figure 1. Initial rock physics modeling indicated that gas-charged reservoir sand would be expected to be about 25% brighter than a brine sand, so an initial geological assumption was that amplitude maps at reservoir level would indicate the presence and distribution of gascharged sands.
The model in the top panel of Figure 1 is of a wedge of brine-saturated sand. All of the layers have constant rock properties, and only the thickness is changing, thinning from left to right. The center panel shows the modeled seismic relative impedance response. The reservoir interval appears as the yellow- and green-colored (low-impedance) layers, but note that the amplitude varies from yellow (left), through green (bright) in the middle and then decays again to the right. There are no property changes in the model; the amplitude change is only caused by the thickness variation.
The lower panel plots the picked amplitude for the yellow-green sand event as a function of the time thickness of the wedge. Note how the amplitude reaches a maximum at about 13 ms and then falls off to about half the peak amplitude at about 35 ms. The amplitude change as a function of thickness (the tuning effect) is a variation of +/- 100% and is much larger than the expected amplitude change of about 25% due to the presence of gas.
Until recently, the only way to detune seismic amplitudes was to interpret a seismic event, extract the thickness and amplitudes in the form of maps and then detune the mapped response. A map-based approach is slow, potentially inconsistent and may be subject to interpreter bias.
Direct detuning of seismic traces
In response to this, Earthworks Reservoir has researched and developed a detuning methodology called DT-AMP that is capable of directly detuning seismic traces in 2-D or 3-D surveys. By detuning the entire impedance seismic section output, false amplitude indicators are suppressed and genuine anomalies highlighted. The method has been successfully trialed with clients over the last two years.
The DT-AMP algorithm can use reflection seismic or relative or deterministic inversion results as input (Figure 2). The outputs are detuned impedance datasets. The method works on both prestack and post-stack seismic inversion products. Through the application of this technique:
• Tuning curves can be analyzed without any seismic picking, thus accelerating the analysis and detuning of the seismic traces;
• Because the seismic trace data themselves and not maps are detuned, comparison of pre- and post-detuning responses significantly accelerates interpretation and mapping of amplitude anomalies by focusing interpretation effort on genuine amplitude and AVO anomalies, significantly decreasing subsequent interpretation effort and, of course, risk;
• Seismic amplitudes can be de-risked, and the input data to subsequent interpretation or seismic attribute work are not contaminated by tuning effects;
• Robust identification of potential DHIs from seismic data by removing potentially false DHI signatures caused by thickness related tuning can be enabled; and
• More consistent amplitude and thickness interpretation can be enabled, allowing picking and mapping using ancillary attribute products from DT-AMP.
Example applications for these include generation of regional play fairway maps and rapid identification of amplitude anomalies.
FIGURE 2. In this figure the top panel shows the original seismic relative impedance. Downdip Well 4 has higher average amplitude in reservoir (blue/purple) than Well 5 updip even though Well 5 has a much higher net:gross ratio. The lower panel shows DT-AMP-processed detuned relative impedance: Average amplitudes now correctly indicate the change in net:gross at the wells. The inset graphic shows the amplitudes relative to the 100% sand line before and after detuning. (Source: Earthworks)
The DT-AMP detuning algorithm has wide application and has been used successfully in frontier basin exploration offshore southern Africa; the U.K. Continental Shelf southern gas basin for both well risk reduction and acreage relinquishment decisions; identifying seismic net pay directly on 3-D seismic and subsequently proving up the prediction with drilling in North Africa; gas reserve identification, delineation and quantification over multiple reservoirs offshore the Nile Delta; and for shallow gas hazard amplitude de-risking.
In summary, seismic trace detuning is a significant development in seismic amplitude analysis (Figure 3). It accelerates seismic amplitude analysis, generates corrected amplitudes for interpretation and de-risks amplitude anomaly analysis and direct hydrocarbon detection from seismic data.