Knowledge management for images in petroleum exploration: a case using petrography ontology

21th PNEC International Conference on Petroleum Data Integration, Information and Data Management - 2017


Petroleum exploration professionals are taking advantage of the wide range of image equipment available to document their reservoir characterization process. As a result, a large set of microscopy, core, and SEM pictures must be selected, indexed and published for consultation by interpreters. The challenge in developing a tool for image management comes from the fact that the high-value of the documented image is not explicit in the picture itself, like in a document or database record, but it requires interpretation and documentation done by the geologist.

We describe here the workflow for systematic capturing, cataloging, auditing, sharing and consulting rock image documentation in a petroleum company and the tools that supports the whole process. The workflow gets advantage of a domain ontology of petrography that supports the documentation of relevant rock features for reservoir characterization.

Petrologists describe thin sections or well cores capturing pictures during the process. The terminology of petrography ontology is applied both, for producing a standard rock description and to label the most important features in the rock image with the support of an ontology-based knowledge system. The petrologist stores the rock documentation (formatted description and images) in the central repository and selects the images that are more representative of the features that affect the reservoir quality. These images are then submitted to the shared image review system. A senior petrologist certifies the picture both in quality and in content and authorizes its distribution. Another application provides the search over the certified image repository using the petrography terminology, such as, type of rock, mineralogical composition, sedimentary structures and presence of a particular cement or pore type. The consultation retrieves all pictures in the repository that contain the requested terminology. Images that historically were stored in personal computers inside the company and isolated from any automated search mechanism are now accessible to the whole company. The knowledge management workflow transformed images into knowledge assets.

This workflow is being in use for seven years and petrographers of geographically distributed laboratories access the image repository to understand or explain reservoir characteristics. Images generated by other processes or partners can also be imported to the central repository. In order to help geologists in the understanding of important petrographic features, a set of illustrative pictures of the ontology terminology is available on the free open-access portal at

Authors: Alexandre Lorenzatti(1), Carlos Eduardo Santin(1), Vânia Campinho(2), Mara Abel(3) & Luiz Fernando De Ros(4)

(1) ENDEEPER – Brazil
(2) Petrobras – Brazil
(3) Institute of Informatics – Universidade Federal do Rio Grande do Sul (UFRGS) - Brazil
(4) Geosciences Institute – Universidade Federal do Rio Grande do Sul (UFRGS) - Brazil