The sensitivity of exploration decisions to variations in several input parameters for maturity modeling was examined for the MITI Rumoi well, Hokkaido, Japan. Decisions were almost completely insensitive to uncertainties about formation age and erosional removal across some unconformities, but were more sensitive to changes in removal during unconformities which occurred near maximum paleotemperatures.

Exploration decisions were not very sensitive to the choice of a particular kinetic model for hydrocarbon generation. Uncertainties in kerogen type and the kinetics of different kerogen types are more serious than differences among the various kinetic models.

Results of modeling using the TTI method were unsatisfactory. Thermal history and timing and amount of hydrocarbon generation estimated or calculated using the TTI method were greatly different from those obtained using a purely kinetic model. We strongly recommend use of the kinetic R[o] method instead of the TTI method.

If we had lacked measured R[o] data, subsurface temperature data, or both, our confidence in the modeling results would have been sharply reduced. Conceptual models for predicting heat flow and thermal conductivity are simply too weak at present to allow us to carry out highly meaningful modeling unless our input is constrained by measured data. Maturity modeling therefore requires the use of more, not fewer, measured temperature and maturity data.

The use of sensitivity analysis in maturity modeling is very important for understanding the geologic system, for knowing what level of confidence to place on the results, and for determining what new types of data would be most necessary to improve our confidence. Sensitivity analysis can be carried out easily using a rapid, interactive maturity-modeling program.