Graduate Seminar - The Value of Information

Friday, October 12, 2012 from 03:30 PM to 04:30 PM

Please join us at 3:30 p.m. Friday, October 12, 2012 in Keplinger Hall, room #M2, for the McDougall School of Petroleum Engineering's Graduate Seminar featuring Pierre Delfiner, PetroDecisions.


The acquisition of additional data for a development project, be it a well test, a 4D seismic for field monitoring, or a multi-vessel survey such as a wide azimuth for subsalt imaging, is a difficult decision to make. While the immediate cost is known the expected benefits are uncertain at the time of decision. Will the data be of good quality? Will the information be relevant? Will the interpretation be reliable? In other words, is it really worth it?

The Value of Information (VoI) is a concept from Decision Analysis that supports this decision making process. Its principle is simple: compare the expected value of the project with and without the new information. If the information is useful it will lead to better development decisions and the value of the project with the information will be higher than without it. The difference represents the maximum price one should be willing to pay to acquire the information. If acquisition costs are included in the economic model of the field (so as to account for specific fiscal and contractual terms) the difference is the added value of the information, which may be negative.

A VoI study is a cross-disciplinary exercise involving all stakeholders in the project. The first and most important step is problem framing, i.e., define the issues, the key drivers, the alternative solutions, identify the key risks and uncertainties, and accept to work within these boundaries. Beyond economic results, the main benefits of a VoI analysis reside in the process itself, which makes the decision more rational and systematic. Case studies will illustrate the VoI approach.

On the technical side, a VoI analysis involves two quite different types of probabilities: priors and likelihood functions. The prior captures the chance of success based on all available information except the proposed acquisition, while likelihood functions capture the physics of the problem. An example will be presented in which the likelihood functions can be derived from a geophysical model combined with a classification scheme.

Rami Younis
(918) 631-2426