Sedimentary Basin Models
Geodynamic Models
Geodynamic models focus on basins produced from large-scale thermal and mechanical deformation of the crust and lithosphere. The models are either kinematic or dynamic, and we use both forward and inverse models. We use mechanics to simulate deformation of the crust and lithosphere in response to a given stress and temperature field. We either assume the properties and distribution of materials in the crust and lithosphere or derive them from forward model solutions. We commonly apply mathematical/mechanical analogs (e.g., we model the lithosphere as an elastic beam or mantle deformation as fluid flow within a pipe). Most geodynamic models are concerned with time scales of millions of years, and with spatial scales on the order of tens to hundreds of kilometers.
Geodynamic models — especially when we include thermal and maturation models — are among the most extensively developed models in the petroleum industry. They are used so routinely that often their results are presumed accurate and precise. Yet, like all models, their accuracy, predictability and resolution are limited.
Burial history, geohistory and backstripping plots are common types of inverse geodynamic models. In these models, we start with an observed stratigraphic column or cross-section, and, by progressively removing each layer, reconstruct the geometry of the basin through time. Our goal is to estimate the influence on basin formation of each of the major contributors to deformation — tectonic, thermal, sediment and water loads. We must assume several variables when generating these reconstructions: the strength and mechanics of deformation of the crust and lithosphere; the mass of the remaining sediment pile; the thickness and mass of sediment removed and redistributed during the formation of unconformities; water depth (if marine) or topographic elevation (if non-marine) of the sediment surface at each time interval of the calculation; porosity variations with depth (equivalent to the compaction history of the sediments); absolute sea-level elevations through time; and the age of the strata at each time interval.
In the case of a rift or passive margin, we calibrate that portion of subsidence induced by thermal decay to the general thermal subsidence curve of oceanic lithosphere. In an ideal case using oceanic lithosphere, the observed departure from the ideal, calculated thermal decay curve is D300 m (Figure 1, Subsidence of the north Atlantic.
The contours enclose two-thirds of the depths around the mode. The dashed and thick solid lines show subsidence curves for the cooling half-space and the cooling-plate model, respectively.) and (Figure 2, North pacific oceans).
Compared with oceanic crust and lithosphere, continental crust and lithosphere are more heterogeneous, lack a direct relationship between age and thickness, and vary more in rheology. Therefore, in the absence of indirect measurements of crustal and lithospheric properties, our estimates of deformation and subsidence in sedimentary basins formed on continents will have errors greater than those for subsidence calculations on oceanic crust. The error bars on a subsidence curve for most continental sedimentary basins are at least D300 m.
Both the mass of the sediment pile and the decompaction calculations are dependent on assumed porosity/depth/time relationships. Conventionally, we use empirical relationships between porosity and depth derived from contemporary basins using the decompaction function. However, these estimates do not account for changes in diagenetic history (early cementation will reduce compaction), or the occurrence of overpressured zones within the stratigraphic column (overpressured zones reduce or delay compaction). In addition, empirical curves show considerable scatter in porosity values from shallow depths, where much of the compaction occurs. Assumed or observed values of topographic elevation, paleobathymetry and absolute sea level can easily have error bars of at least D300 m. Our knowledge of the ages of strata varies considerably, but often the resolution is no better than 1 My. Finally, most commonly used models do not consider likely modifications to basin geodynamics that other variables (e.g., topography, dynamic fluid flow, differential sediment supply) undoubtedly cause.
Thermal and maturation models constitute another type of common geodynamic model. Maturation models have been applied for at least two decades, but have been in common use in the industry only since the 1980s. Two different classes of forward maturation models are used today — one is based on Lopatin’s (1971) method and the other, termed kinematic models, is based on Tissot’s (1969) work (see Wood, 1988, and Waples, 1989).
In Lopatin’s method, we construct a burial depth-versus-age plot, and then add maturation levels, assuming a knowledge of the temperature history of that basin. Until recently, we used existing stratal thicknesses in these plots; that is, the sediments were not decompacted. To estimate the temperature history in these forward models, we must assume a thermal gradient and conductivity of the sediments. Conventionally, we have assumed constant thermal gradients equal to contemporary thermal gradients (which are measured from wells and have their own inaccuracies). Even if the basement and surface temperatures never changed through time, uplift and erosion of previously deposited strata at unconformities may produce major changes in the maturation history; we usually ignore these problems. We usually ignore variations in thermal conductivity of the sediments as well. Lopatin models become increasingly less accurate as maturation time and temperature increase. Despite these limitations, Lopatin models have an important advantage in that we can compare and calibrate their predictions using observed maturation indicators, and we can readily adjust the models by trial-and-error until we achieve a best fit between predicted and recorded observations.
Kinematic models predict maturation using organic material reactions. In these models, we must specify the types, mass and proportions of original organic matter in a sediment. Then, we simulate the maturation of these hydrocarbon precursors at specified reaction rates as a function of time and temperature. The models predict the types and volumes of hydrocarbons generated, usually expressed as a fraction of the kerogens’ original source potential. To run the model, we must specify a time/temperature burial history of the sediments. One important limitation of kinematic models is that they do not provide a method for calibrating calculated and measured maturities. Kinematic models predict the progress of hydrocarbon generation, rather than indirect measures of maturation level (e.g., vitrinite reflectance). There are no accepted standards for measuring hydrocarbon generation directly, so we can not use kinematic models to check the accuracy of our assumed geologic and geothermal histories.
We perform inverse maturity modeling (Lerche, 1989) by minimizing the differences in information provided by multiple data types, each of which gives some estimate of the maturation history of sediments at one or more sample localities in a basin. We can use numerous types of data to infer either a time/temperature function (specific times and temperatures are inferred) or a time/temperature integral (the cumulative effect of time and temperature are inferred). Common data types are vitrinite reflectance, fission tracks, isotope ratios, clay mineral transformations, and fluid inclusions. Each data type requires different assumptions, has a given level of accuracy and confidence and leads to different limitations in the resulting maturation history. By inverting multiple observations from multiple data types, and by minimizing differences among inferred time/temperature/maturity histories through comparison with a forward model, we can produce a more accurate representation of the true basin maturation history.
