Abstract:
Modern petroleum industry operates huge volumes of data and requires modeling techniques for exploration,
production and refining of hydrocarbons for the purpose of business process optimization.
This presentation considers the problem of modeling physical properties of a hydrocarbon reservoir by means of stochastic simulations.
This type of modeling enables optimizing drilling and development processes as well as deriving the probability distribution of recoverable
resources and as consequence evaluating the associated with the project economic risks.
From the point of view of the probability theory, stochastic reservoir modeling is a kind of conditional simulation of a spatial random process.
Wide sense stationary random processes are classically used for this purpose due to the lack of data for estimating higher than the second order moments.
A generalization of conventional probabilistic model for the new generation of computer reservoir models is presented.
|
