9:00 AM - 9:15 AM
[STT34-01] Extraction of geological structure from a density model generated by 3D inversion analysis of airborne gravity gradiometry data
Keywords:airborne gravity gradiometry, three-dimensional analysis, density model, geothermal resource exploration, High Horizontal Derivative Line
Japan Organization for Metals and Energy Security (JOGMEC), an independent administrative agency, introduced Airborne Gravity Gradiometry (AGG) for the first time in Japan in 2012 for the purpose of the geothermal resource exploration. Since then, the AGG survey has been carried out in 19 areas nationwide until FY2022. In addition to features of airborne geophysical surveys (dense survey stations and homogeneous data acquisition regardless of such topography as ocean and land areas), AGG has advantages of gravity gradient survey methods (emphasis on the density structures in shallower depth ranges from 1 to 2 km than conventional ground gravity surveys, and acquisition of the six gravity gradient tensor components allowing more accurate analyses). We will report a new method that has enabled us to obtain more accurate geothermal structural models than the previous ones. The contents of this report have been published as JOGMEC (2024).
A geothermal structural model is used as a guideline for geothermal resource exploration and is considered to show the properties, scale, and distribution of three elements: reservoir, heat source, and fluid (e.g. Noda, 1997; Ogawa, 1987). Regarding the reservoir structure, through such filter analyses of the Bouguer anomaly as vertical and horizontal derivatives, parts with abrupt changes in the gravity anomaly are conventionally interpreted as faults and fracture zones, which are considered to be drilling targets. However, this approach does not represent the dip direction and angle or the depth of faults and fracture zones.
On the other hand, by using a density model obtained by three-dimensional analysis of the gravity gradiometry tensor, it is possible to interpret the dip direction and angle of faults and fracture zones, as well as their continuity in the depth direction, in 3D. However, since models are generated to avoid abrupt changes in the density values between the adjacent cells, the width of the "abruptly changing parts" is approximately 1 km. Generally, the width of a fault fracture zone is thought to be several hundred meters or less at most. The one-kilometer-wide "rapidly changing part" is too large to be considered as a drilling target.
To solve this problem, we have introduced the idea of high horizontal derivative lines; the horizontal derivative of the density distribution in the model is computed and the ridges of the high horizontal derivative zones are traced. The width of the trace is approximately 100m. By interpreting the traces as faults and fracture zones, we are able to locate the drilling targets within a narrow range. The trace is called a High Horizontal Derivative Line (HHDL).
The Musadake region of Hokkaido will be introduced as an example. The HHDLs extracted in the Musadake region showed good agreement with the known geological structures observed on the surface, such as faults running in the NS to NE-SW direction and faults in the NW-SE direction, and they can be well explained using a regional stress field. Furthermore, the HHDLs clarified the direction and dip angles and the continuity of those structures along the depth. Those structures were interpreted as the reservoir for the geothermal fluid. A high-temperature rock body or magma deep below Musadake volcano was interpreted as the heat source. For the fluid, the low resistivity zone identified by time-domain airborne electromagnetic survey (JOGMEC, 2017) was interpreted as the cap lock. These three elements were combined to construct a geothermal structure model.
The use of the first horizontal derivative is a standard and relatively simple analysis method in gravity surveys. Once a three-dimensional density model is obtained by 3D analysis of AGG data, it is possible to understand the 3D reservoir structure from the HHDL. The proposed approach is expected to be very effective in locating drilling targets.
A geothermal structural model is used as a guideline for geothermal resource exploration and is considered to show the properties, scale, and distribution of three elements: reservoir, heat source, and fluid (e.g. Noda, 1997; Ogawa, 1987). Regarding the reservoir structure, through such filter analyses of the Bouguer anomaly as vertical and horizontal derivatives, parts with abrupt changes in the gravity anomaly are conventionally interpreted as faults and fracture zones, which are considered to be drilling targets. However, this approach does not represent the dip direction and angle or the depth of faults and fracture zones.
On the other hand, by using a density model obtained by three-dimensional analysis of the gravity gradiometry tensor, it is possible to interpret the dip direction and angle of faults and fracture zones, as well as their continuity in the depth direction, in 3D. However, since models are generated to avoid abrupt changes in the density values between the adjacent cells, the width of the "abruptly changing parts" is approximately 1 km. Generally, the width of a fault fracture zone is thought to be several hundred meters or less at most. The one-kilometer-wide "rapidly changing part" is too large to be considered as a drilling target.
To solve this problem, we have introduced the idea of high horizontal derivative lines; the horizontal derivative of the density distribution in the model is computed and the ridges of the high horizontal derivative zones are traced. The width of the trace is approximately 100m. By interpreting the traces as faults and fracture zones, we are able to locate the drilling targets within a narrow range. The trace is called a High Horizontal Derivative Line (HHDL).
The Musadake region of Hokkaido will be introduced as an example. The HHDLs extracted in the Musadake region showed good agreement with the known geological structures observed on the surface, such as faults running in the NS to NE-SW direction and faults in the NW-SE direction, and they can be well explained using a regional stress field. Furthermore, the HHDLs clarified the direction and dip angles and the continuity of those structures along the depth. Those structures were interpreted as the reservoir for the geothermal fluid. A high-temperature rock body or magma deep below Musadake volcano was interpreted as the heat source. For the fluid, the low resistivity zone identified by time-domain airborne electromagnetic survey (JOGMEC, 2017) was interpreted as the cap lock. These three elements were combined to construct a geothermal structure model.
The use of the first horizontal derivative is a standard and relatively simple analysis method in gravity surveys. Once a three-dimensional density model is obtained by 3D analysis of AGG data, it is possible to understand the 3D reservoir structure from the HHDL. The proposed approach is expected to be very effective in locating drilling targets.