GRIF Petro
Improving the modeling of Oil Production Systems

A project presented by Stéphane Collas & Nicolas Clavé

stephane_collas_exploration_production_total
Stéphane Collas

Best Innovators

With economic and environmental constraints growing stronger every day, engineering high-performance system architectures at an acceptable cost is of critical importance. The ability to quickly model the various alternatives under consideration in order to calculate their production availability plays a key role here. This means having the necessary resources on hand to shorten time-to-decision as much as possible once a project team has come up with a new architecture option. With this in mind, we have developed a unique method for modeling complex and dynamic multi-flow systems called GRIF Petro.

Our systems modeling expertise

For nearly 40 years, we have been developing modeling software to determine the reliability, availability and/or production availability of safety and production systems. This expertise was the starting point for our proprietary GRaphical Interface for reliability Forecasting (GRIF) software platform.

GRIF comprises a number of advanced modules, including the Petri module for building Petri nets. Because production availability calculations have to take planned and unplanned events into account, as well as related logistical support, the Petri module incorporates all types of complications likely to arise in an analysis, including all:

  • Unplanned events that can impact production, such as equipment failures and the related maintenance downtime;
  • Scheduled operations, such as five-year turnarounds, safety tests and preventive maintenance of rotating machinery;
  • Operational constraints/data, such as maintenance downtime, logistical timeframes, weather conditions, flaring constraints and spare parts management.

Quick, User-Friendly Decision Support

However, the Petri module’s high level of abstraction has been a major drawback, as the models it generates can be excessively complex to handle and check. What’s more, it can take several days or weeks to model and analyze the results for each architecture under consideration.

This prompted us to innovate with the GRIF Petro module. Using the same Petri net technique, the Petro module offers a more intuitive and user-friendly approach to modeling production systems. Users can create inter-connectable block diagrams that are easy to identify thanks to a specific métier-type graphic treatment. Once the model has been finalized and all the characteristics have been entered, the new interface can automatically generate Petri nets and the different flows within the system. The proprietary MOCA-RP (Monte Carlo Petri Nets) engine developed by Total then produces a large number of automatically post-treated results, including:

  • Production availability and temporal, mean or periodic production for each flow and each of the system outlets
  • The contribution of each system, subsystem and piece of equipment to system production and/or injection losses
  • Resource use/consumption
  • Flared gas estimates
  • Shutdown causes and duration

GRIF Petro is the only software module of its kind in our industry. It has significantly modified the working methods of our reliability engineers and field development engineers, providing effective decision support. GRIF Petro has enabled us to test and validate proposals that have reduced capex by more than $300 million over the past two years. New uses for these models are likely to emerge as modeling times are drastically reduced and the solution is connected to our Pangea supercomputer.

Best Innovators

The 10 Innovations Rewarded in 2018