AN INTEGRATED VELOCITY MODELING WORKFLOW TO PREDICT RELIABLE DEPTHS IN TRAT FIELD, GULF OF THAILAND

Abstract

Trat field is located within Pattani Basin of the Gulf of Thailand. The reservoirs in this field are predominantly Lower Middle Miocene fluvial sands. Traditionally, nearby check shot data have been used for depth prediction of proposed wells and it is common to observe significant difference between predicted and actual depths of key markers. In this study, different velocity sources have been used to build accurate velocity model for depth prediction. Accuracy of these models was tested by eliminating velocity information of some wells, through blind test analysis for both single wells and for whole platforms. Predicted depths from this blind test analysis were compared with their actual depths. This study shows that the most accurate depth prediction is obtained by integrating multiple velocity sources such as check shot data, T-D tables of synthetic seismograms, stacking velocities and pseudo velocities calculated from depths of picks and time of respective horizons. Moreover, these velocities were interpolated along the interpreted horizons to incorporate the structure constraints in the model. This new proposed workflow of velocity modeling significantly reduces the error in depth prediction.