Multimineral diagenetic forward modeling for reservoir quality prediction in complex siliciclastic reservoirs

Abstract

Reservoir quality (RQ) prediction models for sandstones in use by the oil and gas industry primarily emulate mechanical compaction, quartz cementation, and cementation by a few select aluminosilicate minerals during burial. The modeled cements are treated on a kinetic basis using independent Arrhenius-style rate equations, and nonmodeled cements are constrained by empirical observations and data from analog rocks. The nonmodeled cements pose a significant modeling challenge in complex lithic and arkosic sandstones that contain carbonate, clay, or zeolite cements when analog data are not available for constraint. Twenty-nine samples covering a broad range of sandstone compositions were selected from four fields in four different sedimentary basins to investigate the potential of using modern reactive transport models (RTM) to simulate a more comprehensive suite of minerals involved in sandstone diagenesis. A commercially available RTM, GWB® X1t, was configured to model chemical diagenesis in a time–temperature burial framework conceptually similar to that employed in standard industry RQ forward models. Strategies were developed to consistently constrain kinetic parameters, fluid compositions, and fluid fluxes with the goal of standardizing these parameters to reduce configuration and calibration time. Results show that RTM using such standardized parameters can reproduce relative timing and volume changes caused by mineral dissolution and precipitation that are accurate within the uncertainty of the petrographic data constraint. The results suggest that incorporation of RTM into current industry models could provide a valuable improvement to siliciclastic RQ prediction in frontier plays with complex mineralogies wherever calibration data are sparse or unavailable.