Stochastic Reservoir Characterisation Integrating Seismic and Petrophysical Data in the Agbada Formation

This study develops an original stochastic reservoir characterisation workflow integrating 3D seismic data and petrophysical well measurements for the Agbada Formation, Niger Delta, with explicit uncertainty quantification relevant to production forecasting. Reservoir characterisation uncertainty significantly affects subsurface confidence for drilling decisions and field development planning, yet most practical reservoir characterisation workflows in the Niger Delta are deterministic, obscuring the range of geologically plausible models consistent with available data. A stochastic approach that generates multiple equiprobable realisations of reservoir properties respecting both seismic and petrophysical constraints provides the uncertainty model needed for robust decision analysis. This study develops a workflow combining seismic inversion (deterministic and probabilistic), model-based co-simulation of porosity and saturation using iterative joint simulation, and geostatistical integration of well-derived variograms with seismic-derived trend information. The workflow is implemented in SGeMS, PETREL, and Python, using 3D seismic and well data from a producing field in the central Niger Delta. Ensemble Kalman filter updating integrates dynamic production data to reduce stochastic model uncertainty. Findings reveal that stochastic models generate a portfolio of realisations encompassing a 35 percent range in connected hydrocarbon volume, compared to the single deterministic estimate. Production forecast uncertainty from dynamic simulation of 50 stochastic realisations spans 28 percent of the expected 20-year cumulative production. Key uncertainty drivers are channel connectivity and shale barrier lateral extent, which seismic data alone cannot resolve below the tuning thickness. The study contributes an original integrated stochastic workflow applicable to Agbada Formation field development characterisation.

Keywords: stochastic reservoir characterisation, seismic inversion, Agbada Formation, uncertainty quantification, geostatistics.