Abstract
Geothermal reservoirs hosted in volcanic rocks, like the Rotokawa Geothermal Field in the Taupo Volcanic Zone (TVZ), New Zealand, typically contain fracture networks that control fluid flow. Realistic discrete fracture network (DFN) models have the potential to improve geothermal resource management. However, the spatial distribution and geometries of fracture networks are often poorly understood due to limited data and complex deformation histories including lava emplacement, subsequent burial and faulting.
To understand better the distribution of fractures formed during lava emplacement, we study andesitic flow exposures from Mt Ruapehu, at the southern end of the TVZ. Terrestrial Laser Scanner (TLS) acquisition on three 50-200 m2 outcrops provided large 3D point clouds of the shape of the outcrop. Delineation of thousands of individual fractures has been semi-automated using local geometrical constraints and a shape detection algorithm detecting planar and curved surfaces. Fracture orientation, length, area, linear (P10) and areal (P20) densities from the TLS data provide input parameters for the DFN models. Fracture detection is validated using high-resolution panoramic photographs (GigaPan) and manual scanline measurements. Cooling joints are highly connected via sub-horizontal joints that are aligned with vesicular layers. UCS tests show a mechanical anisotropy between vertical and horizontal samples. Most of the cooling joints terminate within or at the brecciated margins of individual flows which contrast mechanically with the massive flow interior. Thus, highly connected and curved fractures are mostly confined to lava flows.
This study provides a framework for developing DFNs for geothermal reservoirs hosted in andesitic flows based on empirical observations of intrinsic fracturing and mechanical anisotropies of the host lithology. Fractures in individual lava flows may be interconnected in the reservoir by a combination of cooling joints, subsequent tectonic fault zones and hydrothermal fractures. The combination of TLS, Gigapan and manual scanlines is widely applicable to constraining DFNs other geological settings.
To understand better the distribution of fractures formed during lava emplacement, we study andesitic flow exposures from Mt Ruapehu, at the southern end of the TVZ. Terrestrial Laser Scanner (TLS) acquisition on three 50-200 m2 outcrops provided large 3D point clouds of the shape of the outcrop. Delineation of thousands of individual fractures has been semi-automated using local geometrical constraints and a shape detection algorithm detecting planar and curved surfaces. Fracture orientation, length, area, linear (P10) and areal (P20) densities from the TLS data provide input parameters for the DFN models. Fracture detection is validated using high-resolution panoramic photographs (GigaPan) and manual scanline measurements. Cooling joints are highly connected via sub-horizontal joints that are aligned with vesicular layers. UCS tests show a mechanical anisotropy between vertical and horizontal samples. Most of the cooling joints terminate within or at the brecciated margins of individual flows which contrast mechanically with the massive flow interior. Thus, highly connected and curved fractures are mostly confined to lava flows.
This study provides a framework for developing DFNs for geothermal reservoirs hosted in andesitic flows based on empirical observations of intrinsic fracturing and mechanical anisotropies of the host lithology. Fractures in individual lava flows may be interconnected in the reservoir by a combination of cooling joints, subsequent tectonic fault zones and hydrothermal fractures. The combination of TLS, Gigapan and manual scanlines is widely applicable to constraining DFNs other geological settings.
Original language | English |
---|---|
Publication status | Published - 2015 |
Event | AGU Fall Meeting 2015 - Moscone Center, San Francisco, United States Duration: 13 Dec 2015 → 15 Jan 2016 |
Conference
Conference | AGU Fall Meeting 2015 |
---|---|
Country/Territory | United States |
City | San Francisco |
Period | 13/12/15 → 15/01/16 |