The use of offset-dependent time-shifts to characterize dynamic overburden effects in 4D seismic data

Abstract

Time-lapse seismic surveying is used for the monitoring and management of hydrocarbon fields in order to evaluate production-related subsurface changes that occur in the reservoir system. The time-lapse seismic method conventionally analyses 4D attributes of the reservoir which are generated from stacked versions of the base and monitor datasets which, in turn, are imaged using a common velocity model. A drawback of this approach is that changes that are observed in the time-lapse data that occur in the overburden above the reservoir unit may not be corrected for in the seismic workflow. This lack of update degrades the quality and reliability of the resultant reservoir time-lapse analysis. This research project investigates the effects of dynamic changes in the overburden, i.e., variations in conditions in the medium above the producing reservoir, which occur between the acquisition of the baseline and monitoring survey datasets. The objective of this research is to evaluate the dynamic overburden effects on the monitoring programme of the target time-lapse reservoir, to interpret dynamic overburden effects through the use of 4D time-shift attributes and design methodologies to compensate for dynamic overburden variations. The focus of this research is the Shearwater field, a high-pressure, high-temperature central North Sea field, which exhibits a common dynamic overburden system of extensional stress-arching as a reaction to compaction in the Jurassic reservoir unit. A synthetic modelling study of a variety of dynamic overburden features shows variability in the magnitude of time-shift responses as a function of the source-receiver offset at common midpoint locations beneath the overburden anomalies. These pre-stack, 4D, time-shift variations are found to be sensitive to the geometry and distribution of the 4D overburden anomalies, according to the relative exposure of the seismic ray-paths that transect the 4D effect. Dynamic overburden effects in the Shearwater field are interpreted via the derivation of pre-stack time attributes of time-shift intercept and time-shift gradient, which are generated via least-squares fitting of time shifts as a function of offset derived from the base and monitor datasets. There is agreement between the pre-stack time-shift attributes and the established overburden extension system. A weak negative gradient of time-shift is noted at the Top Fulmar reservoir. These attributes are also found to agree with those from an analogue at the South Arne field, in which a decrease in time-shift is reported from near to far offsets. The interpretation of the pre-stack, time-shift attributes for the Shearwater field indicates the value that can be achieved through analysis of pre-stack 4D data, as its use can enable the characterisation of 4D anisotropy velocity effects and the derivation of geomechanical attributes such as the stress path parameter. Two techniques are developed to derive the perturbation velocity from the pre-stack time-shift. The perturbation velocity is defined as the change in seismic velocity between the base and monitor surveys. Derivation of the perturbation velocity offers the opportunity to compensate for dynamic overburden effects that are traditionally ignored in the seismic workflow, via monitor survey imaging. The first method utilises bi-linear stacking in the offset domain and relocation of the 4D effect to its implied subsurface location, based on a geometrical relationship. The application of this method to the Shearwater dataset enables the derivation of a model that shows alignment to the overburden extensional slow-down and local variations that coincide with fracture closures in the Hod formation. The second method involves linear least-squares tomography of pre-stack time-shifts. Application of this technique to Shearwater leads to the derivation of a model that is aligned with a vertical strain field generated from a Geertsma model produced from post-stack time-shift data. This project demonstrates the value of pre-stack inversion in 4D seismic methods and its potential to improve accuracy in 4D analysis and to deliver information from post-stack analysis that goes beyond conventionally established workflows.