A New Approach towards Precise Planar Feature Characterization Using Image Analysis of FMI Image: Case Study of Gachsaran Oil Field Well No. 245, South West of Iran

• Main Authors: mosayeb shafiezadeh, mansour ziaee, behzad tokhmechi
• Format: Article
• Language: English
• Published: Reaserch Institute of Petroleum Industry 2015-07-01
• Series: Journal of Petroleum Science and Technology
• Subjects: fmi; stratums; carbonate reservoirs; dip; azimuth
• Online Access: https://jpst.ripi.ir/article_506_157a7ae607069aaaf6a078f5c49edc53.pdf
• ISSN: 2251-659X; 2645-3312

<span style="font-size: medium; font-family: Calibri;">Formation micro imager (FMI) can directly reflect changes of wall stratums and rock structures. Conventionally, FMI images mainly are analyzed with manual processing, which is extremely inefficient and incurs a heavy workload for experts. Iranian reservoirs are mainly carbonate reservoirs, in which the fractures have an important effect on permeability and petroleum production. In this paper, an automatic planar feature recognition system using image processing was proposed. The dip and azimuth of these features are detected using this algorithm to identify more precise permeability and the career of fluid in reservoirs. The proposed algorithm includes three main steps; first, pixels representing fractures are extracted from projected FMI image into location matrices <em>x</em> and <em>y</em> and the corresponding value matrix <em>f</em>(<em>x</em>, <em>y</em>). Then, two vectors X and Y as the inputs of CFTOOL of MATLAB are produced by the combination of these three matrices. Finally, the optimum combination of sine function is fitted to the sine shape of pattern to identify the dip and azimuth of the planar feature. The system was tested with real interpretation FMI rock images. In the experiments, the average recognition error of the proposed system is about 0.9% for the azimuth detection and less than 3.5% for the dip detection and the correlations between the actual dip and azimuth with the determined cases are more than 90% and 97% respectively. Moreover, this automatic system can significantly reduce the complexity and difficulty in the planar feature detection analysis task for the oil and gas exploration.</span>