Datapages, Inc.Print this page

Quantitative Assessment of Hydrocarbon Charge Risk in New Venture Areas - Are We Fooling Ourselves?

Noelle B. Schoellkopf
Chevron Energy Technology Company, San Ramon, CA

As greater computing power enables basin modelers to run ever larger 3D models at their desks, and modeling software allows for increasingly quantitative simulation of natural processes, there is an increased expectation that modelers will provide predictive hydrocarbon charge volumes, estimates of API gravity, and GOR values, for prospects ahead of drilling. New ventures business decisions, whether new country entry, exploration block acquisition, or partnerships, rest on the expectation that sufficient resource volumes of the right oil or gas composition and quality are present in the area being considered. These resource volumes, expressed as statistical distributions, and compositional predictions, also have an influence on field development scenarios and economic assessments. The question is, are we favoring quantifiable statistical distributions at the expense of more qualitative, but perhaps more critical, parameters affecting hydrocarbon charge?

In the typical risk assessment process, risk categories considered are:
• hydrocarbon charge,
• reservoir,
• structural integrity,
• and seal.

Some companies have fewer categories, some more, but the process overall is fairly similar for most.

The hydrocarbon charge component of risk examines such topics as source presence, distribution, richness, extent and timing of generation, migration pathways and effectiveness/likelihood of charge, and expected oil or gas composition. Our predictive capabilities have been greatly enhanced by past and ongoing research and case histories, especially in geochemical topics such as kerogen kinetics, kerogen transformation processes and oil stability. Models now provide tools that allow us to consider fluid phase, and apply PVT concepts to consider phase changes between subsurface and surface conditions. We can choose to do Monte Carlo or similar simulations, within which we may statistically vary parameters such as fetch area, depth, source thickness, TOC, HI, temperature, gradient or heat flow, source kinetics, migration losses and timing, and even in some cases lithologic or sealing parameters. Although these tools require some expertise to use, they can result in an increased understanding of which parameters most strongly affect the model.

These statistical tools are designed to function for a given geologic model, best for a simple single source-rock and single reservoir situation. However, they begin to become unmanageable in multiple source-rocks, multiple reservoirs cases. They require geologic and geochemical judgment of how much each parameter can vary and what statistical function best represents the distribution of that parameter. They especially do not address the core question of whether the underlying geologic assumptions are valid.

These problems are compounded in new venture areas. Whereas in producing areas the models can be carefully calibrated to wells, using geochemical parameters and fluid compositions, in new venture areas seismic and well control are typically sparse and models remain poorly calibrated. The new ventures basin model is much more conceptually-driven, whether for predicting basin stratigraphy, thermal history or source rock and reservoir distribution. Thus the level of uncertainty is considerably greater.

While statistical distributions of possible resources provide a means to compare one prospect against another, or to evaluate whether sufficient charge is even possible, it is important to remember that they represent only one possible geological scenario. The real test in new ventures risk assessment is in determining whether the guiding geological concept is the right one, and whether alternate geological scenarios exist. Even if various scenarios are considered, the eventual assigning of risk should rely strongly on geological judgment, even when the model parameters are constrained by a reasonable degree of certainty.

It is the basin modeler’s role to provide some clarity to management of the impact of parameter uncertainty on the final resource volumes. While thorough statistical analysis is often beyond the scope of many projects, and sometimes not even desirable, some assessment of the big-impact parameters is essential. For instance, the basin modeler can assess if there is a risk of under-maturity due to uncertainty in the heat flow model, or the impact of kerogen type on product character, or if the likely carrier bed lithologies and fault behavior can enhance or inhibit migration.

However, equally important, and sometimes difficult for the basin modeler to assess, is whether the geologic model as provided is accurate and rigorous, and what the impact on the model outcome would be if it were different. Do alternative models exist? For instance, what if the section interpreted as Oligocene-age source is really Eocene? Or Miocene? What is the impact? What if the predicted marine source rock facies turn out to be proximal instead of distal? How confident are we in the overall geologic model?

So, in conclusion, while modern basin models provide a valuable means to quantitatively assess a given geologic scenario, the geologist’s knowledge of regional basin history and tectonics, stratigraphy and source-rock depositional models is equally as important in weighing alternative scenarios and assigning hydrocarbon charge risk.


AAPG Search and Discovery Article #90091©2009 AAPG Hedberg Research Conference, May 3-7, 2009 - Napa, California, U.S.A.