A new tool for exploration is a data volume of rock properties in a form compatible with seismic data, but derived from well information. Such volumes can be prepared for fundamental rock properties such as shale P-wave velocity, or for derived properties such as gas sand shear wave velocity, normal incidence reflectivity at a shale-gas sand interface, or Poisson reflectivity of an oil sand.
One of the perennial problems in exploration is putting well information into a form where it can be easily related to seismic data. This paper describes a technique for generating data volumes of rock properties in SEG Y format from well logs. These volumes are compatible with all standard seismic interpretation systems, and can be used by the interpreter to constrain interpretation of seismic data by giving the probable properties of rocks in an undrilled prospect. How the volumes are generated The starting point for the rock property volumes is the standard suite of well logs recorded in most wells. The rock properties of greatest interest in a dominantly clastic sequence are supplied by, or can be computed or interpreted from, velocity, density and resistivity logs. Properties computed using fluid replacement require measurements or assumptions about properties of the fluids, especially density of the oil and gas, and salinity of the water. Temperatures can be based on the temperature measurements made while logging, and formation pressures estimated from drilling mud weights: both temperature and pressure are strongly correlated with depth, and do not have discontinuities with depth. A petrophysicist classifies the rocks penetrated by each well, separating intervals into water-filled sand, shale, and all other lithologies (salt, coal, limestone, hydrocarbon-filled sand, etc.). These last intervals are excluded from the analysis. Wells are then divided into uniform depth intervals, using an interval large enough to contain significant quantities of both shale and water-filled sand, but small enough to adequately describe systematic variations. If the chosen interval is too small, many of the intervals will contain only sand, or only shale. If the interval is too large, there may be significant differences in rock properties from top to bottom, due to the difference in compaction, and more depth samples will include rocks with widely varying depositional environments. For the Gulf of Mexico, the interval chosen is 200 ft (61 m). For each interval, the averages of the fundamental properties of sand and shale are computed, along with the amounts of sand and shale within the interval and the variation of each property within the interval (recorded as standard deviation). Additional rock properties can be computed from the fundamental properties using standard procedures such as the Greenberg-Castagna technique[Greenberg and Castagna (1992)] for computing shear wave velocities, inverse Gassmann’s equation [Gassmann (1951)] for computing dry rock properties, and Gassmann’s equation along with the dry rock properties to compute the properties of hydrocarbon-filled sands [Hilterman et al.(1999), Hilterman (1990), Hilterman et al. (1998)]. Once the well database is constructed, the SEG Y data volumes can be generated.
