PARAMETERS CONTROLLING SONIC VELOCITIES IN A MIXED CARBONATE-SILICICLASTICS PERMIAN SHELF-MARGIN (UPPER SAN-ANDRES FORMATION, LAST-CHANCE CANYON, NEW-MEXICO)

We have measured the acoustic properties and mineralogic composition o f 48 rock specimens from mixed carbonate-siliciclastic outcrops of the Permian upper San Andres formation in Last Chance Canyon, New Mexico. The goals were: (1) identify and model the parameters controlling the sonic velocities; (2) assess the influence of postburial diagenesis o n the acoustic velocities. The variation in sonic velocity in the 0 to 25% porosity range is primarily controlled by porosity, and secondly by the ratio of carbonate-siliciclastic material. Linear multivariate fitting resulted in a velocity-porosity-carbonate content transform th at accurately predicts sonic velocity at different effective stresses. The slope of the velocity-porosity transform steepens with increasing carbonate content, which may be explained by the higher velocity of c arbonate minerals. Another reason may be the property of carbonate min erals to form more perfect intercrystalline boundaries that improve th e transmission properties of acoustic waves and are less sensitive to changes in effective stress. The velocity ratio V-p/V-s is an excellen t tool to discriminate between predominantly calcitic lithologies (rat io between 1.8 and 1.95) and predominantly dolomitic and quartz-rich l ithologies (ratio between 1.65 and 1.8). Gardner's experimental curve overestimates, and the velocity-porosity transforms by Wyllie and Raym er underestimate, the observed sonic velocities, probably because they do not account for variations in texture, carbonate mineralogy, and p ore geometry. Petrographic observations show that postburial diagenesi s is minor and does not seem to significantly affect porosity. Therefo re, the outcrop data set can be regarded as a proxy for the subsurface analog. These findings underline the significantly more complex acous tic behavior in mixed carbonate-siliciclastic sedimentary rocks than i n pure siliciclastics where mineralogic composition explains most of t he observed relationships between porosity and sonic velocity.