Bulletin of the Geological Society of Malaysia, Volume 68, December 2019, pp. 91 – 97
Ismailalwali A. M. Babikir*, Ahmed M. A. Salim, Deva P. Ghosh
Centre of Excellence in Subsurface Seismic Imaging & Hydrocarbon Prediction (CSI), Department of Geosciences,
Institute of Hydrocarbon Recovery, Universiti Teknologi PETRONAS,32610 Seri Iskandar, Perak, Malaysia
* Corresponding author email address: email@example.com
Abstract:The waveform classification is a machine learning method for pattern recognition, aims to classify areas of comparable waveforms, along a seismic horizon. It excels in mapping the subtle changes in seismic response and identifies facies and reservoir properties in greater detail compared to other seismic attributes. The waveform classification was applied to identify the stratigraphic architecture and the depositional elements of the coal-bearing Group E in the Northern Malay Basin. The studied interval is characterized by thin sand reservoirs, shale, and significant occurrence of coal beds. Although coal is a major source rock in the Northern Malay Basin and offers good marker horizons for structural seismic interpretation, it introduces uncertainty in seismic attributes analysis due to its masking effect on seismic data.
The generated waveform classification maps revealed that the interval is deposited in a channel-dominated deltaic setting. Depositional elements such as distributary channels, distributary mouth bars, and subaqueous levees were identified on the maps. Well calibration indicated that the distributary channels and the distributary mouth bars are good sand reservoirs.
Keywords: Seismic attributes analysis, seismic geomorphology, waveform classification