Extending the limits of technology to explore below the DHI floor; successful application of spectral decomposition to delineate DHI's previously unseen on seismic data

Abstract

The field of interest is one of ExxonMobil's development assets in offshore West Africa discovered in the late 90's. A main development well was drilled subsequently in an attempt to confirm additional reserves in an adjacent fault block. The fault block was located above what was thought of as a field‐wide oil‐water contact, but the well surprisingly encountered wet sands. Application of Spectral Decomposition helped us explain this dry well as well as delineate a new group of anomalies never seen before on the seismic data. These anomalies were deeper than any penetrated interval in the field and below what was thought of as the DHI floor or the limit of our DHI detection. The anomalies were on the opposite side of the fault from the dry well but on the same side of the fault as the original discovered reservoirs, thus could be reached easily from the platform. The anomalies showed class 2 DHI responses at low frequency which was consistent with the structure. A new development well was finally drilled two years ago to test this interval and discovered 160 meters of net oil sands thus adding significant amount of reserves to the field and confirming our prediction. The field is comprised of deepwater channel complexes draped over a complexly faulted high‐side closure.. The discovery well was drilled in late 90's by the “Well A” exploration well and penetrated three oil bearing channel sand complexes comprising of Class III to a Class II type DHI responses, from shallowest to deepest, respectively. Near‐field development drilling concentrated on adjacent fault blocks and similar depth anomalies due to earlier drilling results and previous rock property predictions for the area limiting the DHI play, i.e. the “DHI Floor”, to depths around those of first exploration wells. However, in our attempt to explain the results of the dry well and look for more potential in the area, one of the technologies we experimented with was Spectral Decomposition. The application of Spectral Decomposition in this area was partially responsible for understanding the difference between the wet from hydrocarbon response on the seismic data. It also allowed us to extend the concept of “DHI floor” to a new limit as new and additional anomalies not previously imaged on the seismic data were delineated. Spectrally Decomposed Far‐Offsets and Spectrally constrained AVO datasets revealed the expected response for Class II type direct hydro‐carbon indicators, showed clear reservoir amplitude conformance to structure, improved the stratigraphic details for higher confidence interpretation and ultimately supported the presence of thick high‐quality reservoir section beneath the existing discovered reservoirs help potentially helping to extend the life of the field. Based on these results a new development well was designed and drilled two years ago to a target picked on the basis of the spectral decomposition analysis. The well penetrated over 160 meters TVD depth of net oil sands, confirming the presence of thick, high‐quality oil sands deeper than any encountered oil pays in the field. In this paper we will summarize our pre‐drill analysis and our post‐drill results.