Secondary migration of petroleum as a self-adjusting colloidal flow: II. Numerical tests and implications

The hypothetical model for colloidal secondary migration, presented in part I, is tested here with numerical models to examine its viability and to determine the conditions under which it becomes ineffective. The main assumptions are that petroleum migrates as a Pickering emulsion of individual nanodroplets (a few tens of nanometres in size) and groups or ‘flocs’ of nanodroplets. These are nanodroplets are protected from coalescence by coatings of silica, asphaltenes and clay fines. Migration is achieved by diffusion (Brownian motion) of the nanodroplets, and advection of the flocs, working together cooperatively. The cases tested here with numerical models are: (1) Oil migration into an anticlinal structure (e.g. Ghawar Anticline, Saudi Arabia); (2) Gas migration into an anticlinal structure (e.g. Ghasha Anticline, United Arab Emirates); (3) Migration within a tight gas sandstone in a foreland basin (e.g. Niobrara gas field, Rock Island gas field, USA); (4) Migration within a tight oil sandstone in a foreland basin and its effects on a tight (shale) gas reservoir (e.g. Powder River Basin, USA); (5) Migration of heavy oil in a foreland basin (e.g. Western Canada Sedimentary Basin); and (6) The role of colloidal migration in reservoir diagenesis. The main implications of the model in these situations are: (1) and (2) Colloidal migration is highly efficient in the conventional oil and gas windows and is generally orders of magnitude faster than Darcy migration. (3) The mechanism breaks down rather abruptly in good carrier beds in the gas window, typically at a pore-throat size of c . 1 µm. It provides a satisfactory explanation for the filling of unconventional tight gas sandstones and their low water saturations. (4) With lower-quality carrier beds, the mechanism breaks down in the late oil window, leading to tight oil carrier-bed plays. (5) The colloidal mechanism can migrate heavy oils relatively fast and easily, compared with Darcy flow, because the main resistance is the viscosity of the porewater rather than that of the petroleum. (6) Migrating Pickering emulsions provide an effective means of transporting inorganic matter long distances into traps. This has strong implications for reservoir diagenesis. For example, the mechanism can account for the observed trends of quartz cementation in petroleum traps and the timing of petroleum fluid inclusions in quartz overgrowth cements. If this hypothesis is substantiated by direct observation of the proposed petroleum nanodroplets, many traditional concepts of petroleum systems will have to be revised.