A basin model describes the thermal evolution of a basin to estimate when its source rocks expel hydrocarbons and of what type. Basin models do so by evaluating a number of relationships, such as heat flow from beneath the basin, heat generation, heat transfer through the sediments and the loss of heat at surface.
These relationships need to be anchored to provide somewhat accurate results for the unknown past. One obvious anchor is the present-day temperature measured at the surface and in wells. This fairly reliably anchors one end of the modelled history but leaves large uncertainty for the entire modelled history. We need other indirect thermal indicators informing us about the thermal conditions in the past. The most widely used one is the reflectance of light off the surface of woody particles, called vitrinite.
Vitrinite is the principal component of coal but is also present in trace amounts in other hydrocarbon source rocks. As vitrinite thermally matures, its internal structure rearranges, leading to an increase in its ability to reflect light. Pioneered by the coal industry, vitrinite reflectance (VRo) was eventually tied to time and temperature through lab heating experiments and field investigations. Initial VRo models relied on statistical correlations between VR and temperature-time exposure. By the 1980s and 1990s, models such as Sweeney & Burnham’s VITRIMAT introduced kinetic reaction principles. These models treated vitrinite transformation as a series of temperature-dependent reactions, significantly improving predictions over simple empirical relationships. EASY%Ro, a simplified version of VITRIMAT, in particular, became widely adopted due to its ease of implementation and broad applicability.
The 2000s saw further refinements with models incorporating variable activation energies and detailed kinetic reaction schemes. These refinements were driven by growing computational speed, better calibration sample sets and the need to explain the mismatch of VRo models with some VRo data. More extensive calibration data proved an older idea that VRo evolution during burial is affected by the presence of liptinite organic matter. Vitrinite from woody coals increases its reflectance gradually with increasing temperature and time. Vitrinite from liptinite-rich samples (i.e. with high hydrogen index) display delayed reflectance at lower temperatures and faster reflectance increase at higher temperatures. This caused two phenomena observed in common VRo data: 1) the “dog leg” kink in the VRo trend at depth and 2) apparent “VRo suppression” in high-quality source rock samples (with high HI) relative to EASY%Ro model. This also means that basin models calibrated to the EASY%Ro model tend to overestimate the thermal history of a basin in order to fit the “suppressed” VRo calibration data.
An updated model, EASY%Ro-DL, solved this problem by calibrating its parameters to higher-HI samples and by adjusting the frequency factor of the reaction equations to yield more consistent results over a range of heating rates. Various uncertainties remain in VRo application for basin modelling, but I think that, at the very least, we now have a sufficiently accurate VRo model to interpret available VRo data.
