Skip to content
Inicio > Simulating Hydrocarbon Migration Using Agent-Based Modeling (ABM)

Simulating Hydrocarbon Migration Using Agent-Based Modeling (ABM)

Instructors: Dr. Hernán Merlino (FIUBA), Dr. Silvia Barredo (FIUBA)

In petroleum systems, understanding how hydrocarbons migrate through heterogeneous porous media remains one of the fundamental challenges in petroleum geology. Traditionally, these processes are modeled using differential equations that describe system behavior at a continuous scale. However, many key phenomena—such as percolation, pore connectivity, and the formation of accumulations—emerge from local interactions that are difficult to capture using classical approaches.

Agent-Based Modeling (ABM) offers an innovative perspective to address this problem. Instead of representing the system as a continuous field, ABM allows for the simulation of thousands of discrete entities—in this case, droplets or packets of hydrocarbons—that move through porosity and permeability networks following simple physical rules such as buoyancy, capillary pressure, and flow resistance.

From these local interactions, the model reproduces, in an emergent way, key processes such as preferential migration pathways, percolation, trapping, and hydrocarbon accumulation.

This course will introduce the conceptual foundations of Agent-Based Modeling and its application to geological problems, developing a conceptual simulator in Python using the Mesa library and three-dimensional visualization tools such as PyVista. Participants will be able to modify parameters of the porous medium—connectivity, heterogeneity, or petrophysical properties—and observe in real time how these changes affect migration dynamics and trap formation.

The course combines concepts from petroleum geology, computational modeling, and complex systems simulation, providing participants with an innovative tool to explore subsurface fluid migration processes.

By the end of the course, attendees will have gained a clear understanding of the potential of Agent-Based Modeling as a complementary tool to classical petroleum system modeling approaches, as well as the foundational knowledge needed to develop their own conceptual models applied to geological problems.

  • Format: In-person (with the possibility of virtual participation)
  • Minimum Enrollment: 10 participants
  • maximum Enrollment: 40 participants
  • Total Duration: 8 hours (1 day)