A three-dimensional approach to migration modelling through faulted structures is described that allows the effects of stratigraphical uncertainty on potential hydrocarbon accumulations to be assessed. Deterministic, three-dimensional fault seal analysis typically produces results that are extremely sensitive to input parameters such as structural architecture and stratigraphical variation. These parameters can be amongst the most poorly defined of modelling inputs because subsurface structural detail and stratigraphical variation are often below the limit of seismic resolution and are not sampled by well data.
The new technique is fully integrated with three-dimensional fault seal analysis and hydrocarbon flow pathway modelling to give a workflow that predicts the likelihood of fault-controlled hydrocarbon accumulations, given these uncertainties. Multiple deterministic realizations of the model are used to highlight specific uncertainties in stratigraphy to which predicted hydrocarbon accumulations are sensitive. The results of these realizations are incorporated into, and used to condition, a stochastic model to risk predicted accumulations based on their likelihood of occurrence and volume.
This technique has many advantages over either purely deterministic or purely stochastic approaches. A single deterministic realization places over-optimistic faith in the accuracy of the geological model because of the high sensitivity of fault seal analysis to input parameters. Multiple realizations allow specific input parameter uncertainties to be investigated, and the resulting common traps can be considered low risk, but accumulations exclusive to individual realizations cannot be risked. Moreover, a purely stochastic simulation based on all uncertainties will, at best, reduce efficiency by modelling uncertainties to which the result is insensitive or, at worst, may bias results with geologically implausible, stochastically-generated trials.