Most of the wells drilled in deep-water Gulf of Mexico only have basic logs such as gamma ray, neutron, density, and resistivity. Lithology interpretation using these logs is often simplified with binary models???for example, sandstone versus shale or net versus nonnet. Log-based rock typing in complex gravity flow sediments can be influenced by interpreters??? bias as a result of the simplifications, approximations, and confusions between geological and petrophysical terminologies. To minimize subjectivity, a log-facies classification scheme is established in a cored well in the lower Miocene reservoirs in the Vito field based on the calibration of log analysis, core description, and borehole image interpretation. A similar data set in the Wilcox Formation in the Great White oil field is used to illustrate the rock facies and log response variations across a wider region of the Gulf of Mexico. This calibration not only makes visual comparisons between the cores, borehole images, and log responses of the penetrated rocks but it is also a factual-based multidisciplinary evaluation of the facies classification and associated uncertainties. The classification scheme is then applied in two appraisal sidetrack wells with different available log suites but without cores in the Vito field as a mini-case study. Practical methods are demonstrated to facilitate identification and evaluation of the key facies such as thinly bedded turbidites and transitional debrite-turbidites using the conventional logs. Porosity and net-to-gross variation in the thinly bedded
