Model‐uncertainty quantification in seismic tomography: method and applications. (19th August 2013)

• Record Type: Journal Article
• Authors: Osypov, Konstantin; Yang, Yi; Fournier, Aimé; Ivanova, Natalia; Bachrach, Ran; Yarman, Can Evren; You, Yu; Nichols, Dave; Woodward, Marta
• Is Part Of: Geophysical prospecting. Volume 61:Number 6(2013:Nov.)
• Page Start: 1114
• Page End: 1134
• Publication Date: 2013-08-19
• Subjects: Anisotropy -- Pore‐pressure prediction -- Rock physics -- Uncertainty -- Velocity model building; Prospecting -- Geophysical methods -- Periodicals; 622.15
• DOI: 10.1111/1365-2478.12058
• Languages: English
• ISSNs: 0016-8025

Abstract

ABSTRACT: Uncertainty is inherent in every stage of the oil and gas exploration and production (E&P) business and understanding uncertainty enables mitigation of E&P risks. Therefore, quantification of uncertainty is beneficial for decision making and uncertainty should be managed along with other aspects of business. For example, decisions on well positioning should take into account the structural uncertainty related to the non‐uniqueness of a velocity model used to create a seismic depth image. Moreover, recent advances in seismic acquisition technology, such as full‐azimuth, long‐offset techniques, combined with high‐accuracy migration algorithms such as reverse‐time migration, can greatly enhance images even in highly complex structural settings, provided that an Earth velocity model with sufficient resolution is available. Modern practices often use non‐seismic observation to better constrain velocity model building. However, even with additional information, there is still ambiguity in our velocity models caused by the inherent non‐uniqueness of the seismic experiment. Many different Earth velocity models exist that match the observed seismic (and well) data and this ambiguity grows rapidly away from well controls. The result is uncertainty in the seismic velocity model and the true positions of events in our images. Tracking these uncertainties can lead to significant improvement in the quantification of exploration risk (e.g., trap failure when well‐logging data are not representative), drilling risk (e.g., dry wells and abnormal pore pressure) and volumetric uncertainties. Whilst the underlying ambiguity can never be fully eradicated, a quantified measure of these uncertainties provides a valuable tool for understanding and evaluating the risks and for development of better risk‐mitigation plans and decision‐making strategies.