WP 3

History matching inversion by coupled flow and geomechanics

WP 3 Lead: NGI

Joonsang Park

The objective of this task is to explore multiple strategies for using forward coupling flow-geomechanical models to rigorously invert for pressure-saturation plume evolution and other physical-mechanical properties. Previous studies using surface deformation alone, in the absence of other data, can lead to a poorly-constrained inversion process.

In this task, we will expand the current deformation inversion techniques to include additional constraints based on gauge pressure measurements, injection rates, and other available monitoring data. In addition, we apply a type of reduction of dimension method (i.e. Poroelastic Linear Vertical Deflection) with combining the conceptual/theoretical SMART models (developed in WP2) for realistic-scale inversion application of pressure and surface heave history-matching in near real time.

We expect to provide both a fast converging gradient method (local optimizer) and a thorough Bayesian (or ML) approach (global optimizer).

Example: SPE10-based benchmark. Sketch of the simulated domain and horizontal permeability field (a). Snapshots of the simulation results at t = 98 days (b, c, d). Note that in (c) injection and production wells are indicated in blue and red, respectively.

Click on the image to view a larger version.

Click on the image to view a larger version

Participating persons:
Tore Ingvald Bjørnarå: Tore.Ingvald.Bjornara@ngi.no
Antoine Soulat: antoine.soulat@ifpen.fr
Audrey Estublier: audrey.estublier@ifpen.fr
Joshua A. White: white230@llnl.gov