Monte Carlo Simulation Operation
Almost any parameter and date in the resources and economic model can be defined as a distribution.
Monte Carlo simulations can be run in three situations:
- Monte Carlo simulation can either be performed on the volumetric calculations in the Reservoir Resources. In this case, the simulation is initiated on the Reservoir - Reserves Tab by selecting the
- Run Reservoir MC link.
- The other option is to conduct Monte Carlo simulation on the entire project model including the economic model. In this case. The full-cycle Monte Carlo Simulation is started by pressing the button in top button row in Plan or Econ. Invoking the full-cycle simulation will cause the entire model including the Timeline and economic model to be updated in each iteration cycle.
- In PetroVR Portfolio, a Monte Carlo simulation can be run to calculate probabilistic results. This option is only available if one of the models included in the portfolio has full-cycle Monte Carlo results available and furthermore, at least one Monte Carlo result parameter must be routed to the portfolio. The Monte Carlo Simulation is started by pressing the button in top button row.
- The Monte Carlo on Scenarios and Monte Carlo on Contracts tools also provide options to run specific MC simulations.
The screen layout is similar for all three simulation options.
However, before running any Monte Carlo simulation, uncertainty must be defined on at least one input parameter. This is done by selecting the distribution button next to an input variable to open the Probability Distribution Selection dialog.
Distributions do not affect cloned wells and facilities differently, i.e., all replications receive the same values within each Monte Carlo iteration. This is so because the concept of cloning is used to model the behavior of a set of identical wells in a given scenario, and not variation within this set - for which the Duplicate command is best used.
The simulation time of Monte Carlo with complex projects or with a large number of iterations can be significantly improved by means of Multi-Core Processing.