--> Abstract: Identifying and Quantifying Uncertainty in Basin Modeling, by P. J. Hicks, Jr., J. D. Shosa, C. M. Fraticelli, M. J. Hardy, and M. B. Townsley; #90091 (2009)

Datapages, Inc.Print this page

Identifying and Quantifying Uncertainty in Basin Modeling

Paul J. Hicks, Jr.1, Jennifer D. Shosa2, Carmen M. Fraticelli1, Martine J. Hardy3, and Michael B. Townsley4
1ExxonMobil Exploration Co., Houston, TX
2ExxonMobil Upstream Research Co., Houston, TX
3ExxonMobil International Limited, Surrey, UK
4ExxonMobil Technical Computing Co., Houston, TX

The number of input parameters required for a basin model is large. Most, if not all, of these parameters have associated uncertainties. It is difficult and time consuming to accurately quantify and propagate these uncertainties through a model in order to assign risks, probabilities and / or error bars to the output parameters of interest. We discuss a workflow to high grade the key uncertainties in the input parameters, so the uncertainties can be better quantified if feasible, and then propagate these uncertainties through the model to adequately evaluate the pertinent results.

Basin models typically require a wide range of inputs and many of these inputs can be functions of time and/or space. Examples include isopach thickness, ages, amount of eroded or missing section, post-depositional salt movement, rock properties (porosity, thermal conductivity, density, …), seal capacity and boundary conditions such as surface temperature and the bottom boundary condition of heat flow or temperature. These models are used to address a number of business questions concerning oil & gas generation, pressure, oil quality and reservoir quality. In many cases the identification of reasonable outcomes, permissible given the data constraints, is as important as predicting the most likely result.

This work flow involves the following steps:

1. Develop a preliminary most-likely or base-case scenario
• The objective of this step is to build a notional base case model that will be used in the subsequent sensitivity analysis.

2. Estimate the uncertainty in selected input parameters and perform screening simulations to identify which parameters need to be worked harder
• In the screening step, the parameters whose uncertainty has a significant effect on the result of interest are identified. The “result of interest” should be selected with care and in light of the overall objective of the modeling exercise. For example, given the same model, the key uncertainties are likely to be significantly different depending on whether prediction of overpressure, hydrocarbon charge, or oil quality has motivated the construction of the model. The method involves conducting a set of deterministic simulations wherein each parameter is independently varied from the minimum to the maximum value. The results are plotted using a tornado or similar plot to aid in identification of parameters whose uncertainty has the greatest effect on the desired result. In this evaluation it is important to consider potential nonlinearities in the parameter behavior and to re-consider the notional base case model when the evaluation is complete.

3. Evaluate the uncertainty in the key parameters identified in the previous step
• Once the key parameters have been identified, the uncertainties in these parameters are quantified. Quantification typically involves selecting and populating a distribution (uniform, triangular, normal, …) for each input of the key input parameters identified in step 2.

4. Propagate the quantified uncertainty through to the key output parameter(s)
• The final step is running a Monte Carlo simulation wherein values for the input parameters identified in step 3 are randomly selected from the distributions assigned in step 4. The key results from each realization are saved for subsequent evaluation.

We will demonstrate the manner in which this workflow may be used to identify the key uncertainties on generated hydrocarbon volume and generation timing. The input parameters studied include:
• Rock parameters (compaction, etc.)
• Time to depth conversion
• Source generation kinetics
• Surface temperature and
• Basal heat flow.

Limitations of the workflow are discussed including how to be cognizant of nonlinearities in the system and how to avoid falling into the “garbage in – garbage out” trap.

 

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