Évaluation de la qualité et du potentiel d'utilisation des données géospatiales acquises par des systèmes lidar mobiles dans une mine souterraine en production

Authors: Drolet, Michel
Advisor: Grenon, MartinMostafavi, Mir AbolfazlLarouche, Christian
Abstract: Open-pit and underground mines are dynamic production environments. The geomechanical properties of the rock mass control the behavior of excavations. Excavation planning is an important phase of mine operation in order to work safely and profitably. An adequate 3D modeling of the mine site geometry allows an estimation of the excavated material and a mapping of the structural regime of the rock walls. Typically, volume calculations are performed using GNSS receivers (in case of open-pit) and lidar systems and geological structures are characterized manually using a compass. Although so far successful, using a manual method does not always allow the characterization of the entire rock face and it also requires a lot of time on site. However, the development of lidar technologies is constantly growing and there are several instruments available on the market offering fast and remote measurements. On the other hand, there is little information that evaluates the quality of the geospatial data and validates the potential applications of these lidar technologies in mining environments. As outlined by Devillers (2004), users of geospatial data need to be aware of the quality of the data they are handling in order to reduce the risk of misuse. Hudson and Harrison (1997) mention that discontinuities can play a critical role in the stability of underground structures. Misuse of the point cloud data to interpret the behavior of a rock mass could have important consequences. The main objective of this master's thesis is to evaluate the quality of geospatial data acquires with mobile lidar systems (MLS) in an underground mine. The data acquisition was done at the Eldorado underground mine in Val d'Or. The reference instrument used is the Faro Focus S70 static scanner and the mobile lidar evaluated are the Zeb-Revo and the uGPS Rapid Mapper. The secondary objective is to evaluate the potential use of these mobile scanners for volume calculation and structural regime mapping of both open-pit and underground rock faces. Data acquisition was also done on an open-pit setting. The same instruments were used with the exception that the uGPS Rapid Mapper, unavailable at the time of the surveys, was replaced by an iPad Pro 12. This research project evaluated the quality of SLM geospatial data in the underground environment. The estimated error of the Zeb-Revo is ± 0,03 m and ± 0,15 m for the uGPS. The results show that the Zeb-Revocan be used to calculate the volumes of an underground excavation. The volume difference between the Zeb-Revo and the reference value produced by the Faro Focus S70 scanner is at most 2% and the volume difference between the uGPS and the Faro Focus S70 scanner is more than 40%. However, special attention should be given to the drift of the navigation system of the mobile scanners. These analyses were not done with the iPad in the underground mine, but the error estimated from the open-pit surveys is in the order of 0,04 m. The research also demonstrated the possibility of measuring the orientations of geological structures in a rock mass using mobile lidar. In the open-pit environment, measuring the orientation of geological structures based on an SLM survey has satisfactorily quantified the orientation of geological structures. Depending on the method and the lidar scanner, angular deviations of set orientations ranged from 5° to 27° for Discontinuity Set Extractor(DSE) (Abellán, 2018), 5° to 12° for PointStudio (Maptek, 2021), and 10° to 30° for the kd-tree plugin (Dewez et al., 2016) from the reference compass survey. In the underground environment, automatic methods clearly do not distinguish between a discontinuity and the orientation of the rock face excavation. The automatic DSE method gives similar results for the three lidar systems and are partially different from the reference stereonet. The angular deviations of the orientations of the sets range from 8° to 73°. In addition, two sets seem to correspond to the orientation of the drift excavation rather than the natural geologic structure. A manual method can visually identify the discontinuities one by one, provided the quality of the lidar data is adequate. The manual PointStudio method combined with the Faro lidar were able to identify most of the sets with an angular deviation between 10° and 30° from the reference survey. The quality of the mobile lidar tested in the underground environment is not sufficient for manual extraction of discontinuities. In summary, DSE and PointStudio software do well what they were designed for, i.e. extract planes from a point cloud. However, users must be careful with the data they are processing and must have a sufficient knowledge of the site.
Document Type: Mémoire de maîtrise
Issue Date: 2021
Open Access Date: 20 December 2021
Permalink: http://hdl.handle.net/20.500.11794/71246
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
37542.pdf13.74 MBAdobe PDFThumbnail
All documents in CorpusUL are protected by Copyright Act of Canada.