Porosity prediction from seismic attributes of the Ordovician Trenton-Black River groups, Rochester field, southern Ontario O. C. Ogiesoba ABSTRACT This article integrates three-dimensional (3-D) seismic attri- butes and log data to determine porosity distribution of the Ordovician Trenton-Black River groups within the Rochester field, southern Ontario. The rocks are composed of tight lime- stone, parts of which were dolomitized to form porous reser- voir rock. Previous studies of the Trenton-Black River dolomite reservoirs have indicated a close relationship between faulting and reservoir development, but few published studies have attempted to examine these relationships using 3-D seismic data. This study explores the stratigraphy and structure of the Rochester fault-related dolomite reservoir using 3-D seismic data and neural networks to predict porosity. By predicting porosity using seismic attributes, vertical and lateral distribu- tions of porosity that can be used to guide development and exploration drilling for optimal hydrocarbon recovery were obtained. The sites of highest porosity were found to be along and within the fault zones. Faults extending from the basement into overlying Paleozoic rocks are composed of several short- plane, vertical, and subvertical fault segments. However, some of these faults appear to have originated and died within the Paleozoic rocks; they cannot be traced to the basement because of little or no offset where they penetrate the basement. Al- though the five identified attributes are considered important in exploration for fault-related dolomite reservoirs, the single most important attribute to employ is the amplitude envelope because the other attributes are mathematically related to it. Furthermore, the sags that are seen in the Rochester field are AUTHOR O. C. Ogiesoba  Earth and Planetary Sci- ences Department, McGill University, Montreal, Quebec, Canada; present address: Bureau of Economic Geology, University of Texas at Aus- tin; osareni.ogiesoba@beg.utexas.edu Osareni (Chris) Ogiesoba received a B.Sc. de- gree in geology from the University of Benin, Nigeria, and an M.Sc. degree in geophysics from Imperial College of Science and Technology, London. After working as a prospect generator with Mobil Oil for 20 years, he received an M.Sc. degree in applied seismology from the University of Calgary, and a Ph.D. in earth and planetary sciences from McGill University, at which point he joined the Bureau of Economic Geology, Uni- versity of Texas at Austin (2007), as a research associate. While at Mobil, he was instrumental in the discovery of more than 1 billion bbl of oil and more than 500 bcf of gas, and he proposed the first horizontal well to be drilled by Mobil Producing Nigeria. At McGill, he worked on seis- mic exploration methods for exploring hydro- thermal dolomite reservoirs within the Trenton- Black River groups, and in Calgary, he developed an algorithm for extracting velocity ratios from converted-wave data. Since joining the Bureau, he has poststack processed data sets from the Campos Basin in Brazil and the Gulf Coast to eliminate both high-angled coherent and random noise. He has also interpreted data sets from several Gulf Coast subbasins, in which he has identified numerous prospects. He is currently working on a gas-shale project in south Texas. ACKNOWLEDGEMENTS The project was jointly funded by Talisman En- ergy Inc. and Natural Sciences and Engineering Research Council of Canada. I thank Talisman Energy for supplying the data and allowing us to publish the results. Support of this research by Landmark Graphics Corporation via the Land- mark University Grant Program is gratefully ac- knowledged. I also thank Hampson and Russell for the use of their software. Publication was authorized by the Director, Bureau of Economic Geology. The AAPG Editor thanks the following reviewers for their work on this paper: David E. Eby, Jeff Lonnee, and Langhorne B. “Taury” Smith. Copyright ©2010. The American Association of Petroleum Geologists. All rights reserved. Manuscript received February 12, 2009; provisional acceptance April 3, 2009; revised manuscript received December 26, 2009; 2nd revised manuscript received July 15, 2009; final acceptance April 6, 2010. DOI:10.1306/04061009020 AAPG Bulletin, v. 94, no. 11 (November 2010), pp. 1673 – 1693 1673