Numerical modelling of abnormal fluid pressures in the Navarin Basin, Bering Sea

Abstract

The Navarin Basin, located in the Bering Sea, Alaska, is composed of three major units: (1) a thick Late Eocene and Oligocene overpressured shaly section; (2) a Miocene sandy section; and (3) a Late Miocene to Pliocene section, characterized by high porosity and a possible zone of abnormal pressure due to the presence of diatomaceous shales. The basin was affected by folding during the early stages of basin filling, whereas strike-slip faulting occurred until mainly Miocene time. To model the fluid flow and compactional history of the basin, a cross-section was used, controlled by four wells for which porosity and pressure data were available. A first approximation used a one-dimensional model to bracket the range of parameter values which govern the fluid flow/compaction equations, using data from the Arco COST No. 1 well. A sensitivity analysis then allowed: (1) determination of the best set of parameters of the model; (2) the behaviour of each parameter and its relative importance to be assessed; and (3) the role of an erosion event in influencing the evolutional history of the basin to be tested. With a slight readjustment of the parameters to allow for lateral fluid flow, a two-dimensional model was then constructed. The model reproduced correctly the variation of porosity and pressure data with depth for the four wells. The main overpressure in the deeper part of the basin arose during the earliest stages of basin filling by shale because of strong undercompaction. In the shallower part of the basin, high porosity values were obtained in agreement with the data, whereas some overpressure was developed in association with the undercompaction and low permeability of the diatomaceous shales. Different assumptions have been investigated, such as the permeable or impermeable nature of faults, the role of a rock fracturing coefficient and the anisotropy of permeability but, because of the location of the wells on structural highs, no major variations were observed relative to the well data, so that the impact of such different assumptions on the total dynamic history of the basin cannot be determined from presently available data.