
Full text loading...
Construction of accurate velocity models for depth conversion is always a critical task in hydrocarbon exploration and production. In areas of extremely complex geology and poor well control, this activity becomes challenging as small variations in the interpreted picks and layer velocities can result in huge variations of the predicted depths, away from the control points. The present paper describes the use of 3D seismic data to help constrain a well derived velocity model, in region of the Colombian Foothills with very scarce well control and an incredibly complex geology. The methodology uses PSTM velocities to derive average velocities, which are in turn scaled to match the well velocities at the well locations and finally produce a stable velocity model for the whole area that has a lower uncertainty. This velocity model was built in 2008 and supported the development plan of the fields. The average velocities obtained from the seismic data showed tendencies consistent with the well data, with deviations less than 10%, thus supporting a valid medium for extrapolating velocities in the zones without control. The model was tested with the drilling of two new wells, away from the control points, resulting in a mismatch from the model of 10% or less in the predicted the top of the structures with 10% error in depth. This outcome can be explained by the anisotropy behavior of rock velocities due to abrupt changes in structural attitude of the principal structures which affects the real velocity field and can be sub estimated on the modeling.