Carbonates make up about half of the sedimentary rock in the Western Canada Sedimentary Basin (WCSB) and range from Cambrian to early Jurassic in geologic age. Large oil and gas reserves exist in Devonian and Early Carboniferous car- bonate formations. Middle and Upper Devonian carbonate rocks alone have known reserves of about 15 billion barrels of oil and 35 trillion cubic feet of natural gas. A variety of car- bonate reservoirs exist in the WCSB, mostly reef and platform types. These reservoirs are formed by porous limestones or platforms that were dolomitized prior to basin-wide hydro- carbon maturation and migration. Fractured reservoirs were formed by natural fractures due to regional or local stress regimes or by local overpressure. Until recently, seismic analysis of data from carbonate reservoirs relied mainly on interpreting zero-offset (stacked) volumes. Common knowledge within the world of AVO sug- gests that zero-offset information is often insufficient to dif- ferentiate shale from carbonate porosity, or to discriminate gas-saturated from brine-saturated reservoirs. However, in the last a few years, great efforts have been made to apply AVO analysis to carbonate reservoir characterization but several issues must be addressed in investigating the feasibility, poten- tial, and sensitivity of the response of carbonate rock proper- ties to porosity and fluid. First, a lack of carbonate rock property information is con- sidered an obstacle in applying AVO to carbonate reservoir characterization. Second, the differences between clastic AVO and carbonate AVO need to be clarified. Third, procedures and calibration in seismic data processing and interpretation need to be developed. The situation has been greatly improved due to recent significant acquisition of dipole sonic logs that sam- ple a variety of reservoirs and formations directly associated with carbonates. This newly acquired information provides in-situ reservoir and nonreservoir measurements that are beyond laboratory and theoretical rock property predictions. On the seismic side, experience in both AVO processing and interpretation, including both amplitude analysis and inverted rock property analysis, has also been growing. In this article, some issues facing the application of AVO in carbonate reser- voirs (such as physical relationships between rock properties, fluid sensitivity of the carbonate rock property, data process- ing, and calibration and interpretation) will be reviewed and discussed. Carbonate rock properties. Figure 1 shows a set of dipole well logs from the Foothills of the WCSB in which the mudrock line for clastics is V S = 0.862V P - 1172.4. A line with the rela- tionship of V S = 0.4878V P + 230.0 is fitted to the carbonate lithol- ogy cluster. Similar to Castagna’s definition, we call this linear relationship a carbonate line. It can be seen that the carbon- ate line deviates from the clastic mudrock line with a slope significantly less than that of clastic rocks. Notice that, oppo- site to clastic rocks, the V P /V S ratio increases with increasing velocity or decreasing porosity. For this typical log data set, the carbonate rocks consist mainly of water-saturated lime- stone. In Figure 1, as is always observed, the data points of the gas sand in these two wells shift away from the clastic rock cluster and have a low V P and a low V P /V S ratio in compari- son with water-saturated sandstone. Fluid effects in carbonates, especially gas effects, are con- tentious but of great interest. The common wisdom is that flu- ids have little or no effect on carbonate rock properties because carbonate rocks have very high moduli. In other words, the high velocity of the carbonate rock matrix causes seismic waves to travel primarily through the matrix where they are little influenced by pore fluids. However, an analysis of the dolomite data from the Williston Basin by Rafavich (1984) indi- cates that gas does influence carbonate rock properties and its effect is significant (Figure 2). Further evidence of this can be seen in an analysis of a large data set of lab measurements on carbonate rocks from the WCSB. This data set includes lime- stones and dolomites. It represents a wide range of carbonate reservoirs and nonreservoirs. An analysis of this data set indi- cates that the result is consistent with the data set of the Williston Basin (Figure 3). Notice that the behavior of dolomite rocks due to gas saturation is similar to that of sandstones. Namely, P-wave velocity and V P /V S ratio decrease, and S- wave velocity increases slightly due to decreasing density. In addition, the rocks are more sensitive to fluid with increasing 670 THE LEADING EDGE JULY 2003 Recent applications of AVO to carbonate reservoirs in the Western Canadian Sedimentary Basin YONGYI LI and JONATHAN DOWNTON, Core Laboratories Reservoir Technologies Division, Calgary, Alberta, Canada BILL GOODWAY, EnCana Corporation, Calgary, Alberta, Canada Figure 1. Velocities and V P /V S ratio of dipole well logs from Foothills, the WCSB. Figure 2. Gas effect of dolomite rock properties for the data set from Williston Basin. Figure 3. Gas effect of dolomite rock properties for a data set from the WCSB. Downloaded 01/21/15 to 5.32.45.6. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/