Case Study: shear sonic, density properties help define rock matrix
Detecting the water cut
RMS amplitude map over the study area. Well symbols are positioned at the take points for the reservoir.
The abundance of 3D data has made the identification of potential reservoirs a more robust process. This has made the exploration phase of geoscience much more manageable, but what about the second phase, the exploitation of the reservoir?
The acoustic properties of the unconsolidated sediments of the offshore areas change dramatically when hydrocarbons fill the pore spaces. These large changes cause a strong amplitude, or bright spot, at the interface of the reservoir with the sealing rock.
This amplitude can be four or five times larger than amplitude associated with a change in lithology only. With the fluid type remaining constant in the reservoir, we equate the variations in the magnitude of amplitude to changes in the quality or quantity of the reservoir rock. This assumption should make the exploitation phase very simple: drill the strongest amplitude, find the best reservoir.
Vp (red) and density (blue) curves for the B1 well. Note the blocky nature of the Vp log compared to the gradational density log. Here Vp change indicates massive sand, while density shows a fining upward sequence.
A study of amplitude values over an existing field in the Mississippi Canyon area, Gulf of Mexico, highlights this all-too-common problem: amplitude values do not directly correlate to reservoir properties. Four companies drilled a total of 16 wells into the amplitude anomaly, of which only 10 were producers. The 16 wells showed a strong variation in net thickness from 160-0 ft that is divided between two potential reservoir sands. With the exception of two of the locations, all the wells appear to drill into the high amplitude anomaly. The A5 well encountered the reservoir but was water wet, while the B1 well had only a small amount of pay on water. These two take points indicate a water level significantly updip from the level indicated on the amplitude map.
Prior to the acquisition of the current 3D seismic dataset, the A1 well, watered out sooner than any reservoir model predicted. The well is not significantly lower on structure than the other wells that continue to produce, and could possibly represent a stronger stratigraphic control on the reservoir than was anticipated.
Density contrast map. Larger numbers represent greater density contrast between the reservoir and the caprock. Values below 2 are due to lithology variations only.
Seismic data is the measure of three main acoustic properties: Vp (compressional sonic), Vs (shear sonic), and density. It is the variations in these rock properties that control the amplitude of the seismic trace at a given horizon; small changes yield reflectors, and large variations yield bright spots. Of the three rock properties Vp exhibits the largest overall change associated with the presence of hydrocarbons.
The large variations in Vp are often found to be nonlinear with many of the reservoir parameters that we are trying to understand. Hydrocarbons in minor saturation can produce the same Vp as higher saturation. Once a reservoir is charged, lower porosity rocks (such as silts) can produce velocities very similar to higher porosity sands.
In contrast, density will show a linear decrease proportional to the amount of fluid in place. The porosity or cleanliness of the reservoir interval will therefore be the primary influence on density variations. Vs will vary inversely proportional to the density, and with any strong changes in the rock matrix. This strong dependence on the rock matrix makes shear wave information very valuable as a reservoir tool, and has received a great amount of attention due to the increase in shear wave acquisition in the Gulf of Mexico.
The assumption that stronger amplitude will correlate to better reservoir often fails because of the magnitude of the Vp change associated with any amount of hydrocarbons, and its insensitivity to rock matrix parameters.
This leaves stacked amplitudes as a useful tool for the identification of potential reserves, but not very robust in the exploitation phase. Impedance inversion suffers from the same fate because of its strong dependence on Vp as well.
Predicted net reservoir thickness derived from the seismic data only.
Amplitude versus offset (AVO) solved one of the early problems with bright spots: distinguishing between amplitudes caused by strong lithology variations from those caused by the presence of hydrocarbons. The driving force behind AVO analysis is identifying strong impedance contrasts that also have strong Vp/Vs ratio changes across the acoustic boundary. In its most simplistic form, AVO does this very well, but it too suffers from the domination of Vp in reservoir delineation. AVO will also be anomalous in zones of low hydrocarbon saturation and in poor quality reservoirs.
Recent advances in many aspects of geophysics have allowed us to push AVO further. Density and Vs are much more stable reservoir parameters than Vp. Of the two, density is more often measured directly. Unlike stacked data, which is the measure of one variable (amplitude), AVO measures many variables. Proper interpretation and manipulation of the data can yield maps that are dominated by one or more of the rock property changes.
Maps that are processed to highlight areas of strong density contrast look very different from standard amplitude maps. The density contrast map over the study area looks very different from the amplitude map. The A5 and B1 wells that appeared to be updip of the water contact are on the outer edge of the density anomaly. The A1 well that watered out also sits in a zone of poor density contrast, also visible is a strong N-S trend in the anomaly indicating strong stratigraphic variations along this trend. The wells with the cleanest sands sit along the trends with the strongest density contrast.
A time-isopach map created from the top and base of the reservoir can show how thick (in time) the reservoir may be. To get the information into depth, an interval velocity must be incorporated. The outcome is a good estimate of the gross reservoir thickness. However, the economics of a field are based on the net ft of pay in place, and the difficulty of producing the hydrocarbon from the rock. The density contrast map helps with some of the questions about ease of production and hints at net thickness variations.
Gross thickness also fills in some of the gaps, but variations in net/gross ratio commonly range from 80% to 20%, the later being unproductive. The amplitude and temporal thickness of a seismic response are closely linked. Amplitudes are strongest when the top and bottom of a bed are resolved. By performing the two analyses together and incorporating AVO information where available, an estimate of net thickness variation is extracted from the data.
The net thickness map is an uncalibrated product generated from the seismic data only. Although the absolute numbers do not tie perfectly, the strong variation in net thickness found in the field is predicted from the seismic data. The thickest and thinnest net areas are properly identified on the map, and an N-S trend similar to the one found on the density map, is also visible.
Given the tools that are available today to the explorationist, reservoir identification and exploitation is possible with a higher degree of accuracy than ever before. Using only amplitude to choose the best five drilling locations, a total of 260 ft of net pay would have been found in four producing wells. By incorporating the AVO and net thickness information, the first five wells would have found 502 ft of net pay in five producing wells, an increase of almost 2:1 over just amplitude analysis.
Tougher times require that we use all the tools at our disposal when evaluating acreage for production. Amplitude analysis alone does not have the ability to discriminate the complexities of a reservoir. No one tool really does. Amplitude analysis integrated in with other geophysical tools, like AVO and thickness work, can provide us with a much clearer picture of the suburface geology.