| An Integrated Image-Analysis Framework for Automated Extraction of Displacement Parameters from Micromodel Videos |
| Paper ID : 1041-UGESR |
| Authors |
|
Mozhdeh Sajjadi *1, Farzam Nazeripour2, Amirhossein Gholamzadeh2, Faezeh Allahdadi2 1Institute of Petroleum Engineering, University of Tehran, Technical Faculty Campus, Kargar Shomali St., North of Jalal Al Ahmad Intersection, Tehran, Iran 2Institute of Petroleum Engineering, School of Chemical Engineering, College of Engineering, University of Tehran |
| Abstract |
| Visualization of pore scale phenomena using micromodels is a common way to study the subsurface displacement processes. The analysis of the results is usually conducted based on the video recording or consecutive snapshots of the tests. Although in these experiments the porous medium’s hydrodynamic properties, the displacement profile, and pore scale interfacial phenomena are interrelated, yet in most of the studies the analyses are focused on one of the named aspects. This paper introduces an integrated approach for automatic detection of the micromodel’s static properties and the displacement’s dynamic features. The framework consists of a marker based watershed segmentation scheme that captures the porosity, pore size distribution and throat size distribution of the porous medium integrated with a trend-based detector that locates sweeping milestones and provides temporal variation of saturation profiles. The algorithms are validated against a numerical simulation using a commercial software with known properties and extracted saturation profiles. The porosity measurement is exact for the synthetic binary profile and the saturation profile remains within a 0.04 margin from the reference profile. Despite the algorithm’s validated performance on clean pictures of micromodels, the challenges regarding processing of non-ideal pictures and how the results may be affected are discussed at the end of the paper. |
| Keywords |
| Immiscible displacement; Micromodel; Pore-scale analysis; Sweep efficiency; Automated image analysis; |
| Status: Accepted (Oral Presentation) |