NSG 2024 5th Conference on Geophysics for Mineral Exploration and Mining

The completion of the GREENPEG project marks the culmination of a four-year effort to advance pegmatite exploration through ground geophysics. The Geological Survey of Norway (NGU) focused on evaluating Electrical Resistivity Tomography (ERT) and Ground Penetrating Radar (GPR) methods at the Tysfjord test site. Initial tests, supported by borehole networks and surface outcrop mapping, showed promising results in distinguishing pegmatite from host rock, aiding exploration crucial for reducing EU’s dependency on critical raw material imports. Preliminary findings suggested pegmatite might be associated with extremely high resistivity clusters and minimal reflectivity. However, further surveys at two additional sites with less ground truth information emphasized the complexity of interpreting geophysical data in pegmatite-rich regions without support from additional information. Testing reaffirmed that high resistivity alone is inconclusive, and low GPR reflectivity doesn’t confirm pegmatite presence. While ERT can delineate pegmatite shape near the surface and GPR can outline boundaries in depth, interpretation challenges arise without any ground truth data. Therefore, collaborative efforts between ground geophysics and ground truth data collection are imperative for reliable interpretations in pegmatite exploration. Having ensured that, ground geophysics can offer valuable insights into characterizing both buried and outcropping pegmatite.

Shear zones, characterized by intense crustal deformation, serve as vital conduits for mineral-rich fluids, playing a pivotal role in mineral exploration endeavors. This study introduces a comprehensive predictive model integrating remote sensing, geophysical techniques, and field-based observations to delineate hydrothermal alterations associated with gold mineralization along the Fatira shear zone within Egypt’s Eastern Desert.

Leveraging ASTER satellite data, the study delineates hydrothermal alteration zones, highlighting distinctive spectral signatures indicative of various alteration types, including argillic, propylitic, phyllic, and iron-bearing alterations. Aeromagnetic data, processed through the CET porphyry method, identifies circular anomalies signifying hydrothermal alteration features.

Findings underscore the extensive distribution of hydrothermal alterations along the Fatira shear zone, particularly in proximity to granitic formations. These alterations, characterized by changes in mineral composition, texture, and color, provide valuable insights into mineralization processes. The study underscores the significance of shear zones as focal points for mineral deposition and underscores the efficacy of remote sensing techniques in delineating alteration patterns, thereby advancing mineral exploration methodologies.

P&S data acquisition within unconsolidated sections of an open borehole are often limited due to the unstable ground conditions. In such cases, P&S logging manufacturers recommend the logging tool to be used within grouted PVC cased boreholes for soil and soft rock sediments. The process of grouting the annulus between the borehole wall and the PVC casing increases the stiffness of the borehole wall, thereby altering the physical conditions of the borehole. Additionally, grouting the boreholes in the nearshore/offshore environment are often challenging, thereby affecting the project cost and schedule.

It is therefore important to investigate the correlation in properties obtained from the in-situ measurements within the open hole and the ungrouted PVC cased condition. This would determine the success of performing P&S logging measurements within ungrouted PVC cased boreholes in soft soils/unconsolidated sediments. The P&S logging tools manufactured by Geovista and Robertson Geologging are best suited to acquire data within such borehole conditions.

This paper presents a comparative study of the P&S logging data collected within the open section condition and an ungrouted PVC cased condition of the same borehole, performed at two completely different geological strata.

Geophysical well logging is a useful method for estimating subsurface characteristics and is used in a variety of fields. Among the geophysical logging methods, gamma-gamma density logging provides in-situ formation bulk density. Measured count values from a detector of density tool are converted to the bulk density through several calibration processes. In these processes, the ratio of the atomic number (Z) and atomic weight (A) is required for accurate conversion, but it is difficult to obtain the ratios for all depths. Therefore, a limestone-based conversion equation is typically used. It can be cause of error. In this study, we derived an equation that is suitable for a test borehole to obtain an accurate density log. The matrix density was estimated by using a pycnometer and the Z/A ratio was calculated from X-ray fluorescence analysis results. Then, the electron density was calculated by these values and used for derivation of the equation. Compared to density laboratory test results for drilling core, the accuracy of the density log using a derived equation has been improved.

Anomaly response curves are widely used for interpretation and modeling of borehole electromagnetic (BHEM) data in mineral exploration. Plate modeling in a vacuum setting is one common application. Obtaining proper vacuum anomaly residuals from measured data is key for these types of applications. Failing to handle the effects of the surrounding host on the electromagnetic signal may lead to wrong conclusions about conductor locations and orientations. An electromagnetic wave traveling through a conductive medium will experience phase lag and amplitude attenuation. For a BHEM survey with the transmitting source at surface, these effects grow stronger at depth. By operating in the frequency domain, we can both identify and manage these effects. At depth, typical vacuum response signatures from simple conductor shapes like plate models may no longer hold. Here, we explore a polarization ellipse-driven Model Normalization approach for obtaining geometrically sound anomaly residuals, where predicted attenuation and phase lag due to a conductive background are accounted for.

Effectively, anomaly residuals can be observed and modelled as vacuum responses under suitable conditions.

In the past, the use of 3D reflection seismic acquisition for mapping ore bodies in mining exploration has been limited. This can be attributed to concerns about its applicability in geologically complex, hardrock environments and the high cost of conventional 3D acquisition systems. However, recent years have seen a shift, with the introduction of smaller, more cost-effective nodal systems, prompting the mining industry to re-evaluate its potential. There is an increasing interest in utilizing 3D seismic from the initial exploration phase through to the advanced extraction planning on mature sites. We show how velocity models derived from first-break refraction tomography for one particular high-density survey assist in the 3D mapping of the overburden/bedrock contact of a magmatic intrusion and provide extra detail to the extent of the low velocity ore bodies beneath. The potential for cross-hole tomography surveys to be complimented by first arrival data from a surface deployed 3D grid of continuously recording sensors is also examined.

We have collected high resolution P-wave and S-wave tomograms at the Veracio drilling test site at Brukunga, South Australia. Grouted geophone sensor strings in nearby drillholes demonstrate the feasibility of opportunistically using the diamond drilling process as a source to gather high resolution 2D seismic velocity data in a hard rock environment. These geophone sensor strings and the seismic while drilling methodology envisage “life of mine” sensing and geological/rock-mass characterisation. Various modifications of the seismic string installation with respect to 1-C and 3-C sensor pods and grouting method were trialled.

Both P- and S-wave modes with energy from 10 Hz to over 500 Hz were used in tomographic reconstruction at distances of up to 200m from the diamond drilling. A configuration of two boreholes with permanent sensors and the drilling of another hole in line with the first two produced good travel-time data. Tomograms were produced via a first-arrival time, cross-well tomography algorithm using either drill-cutting or drill string hammer as a seismic source. Tomography in a hard rock environment, similar to many base-metal and gold mineral deposits, proved feasible able to resolve formations as thin as 5m over 100m scale lengths.

The Geological Survey of Sweden has developed two data acquisition platforms for total magnetic field and electromagnetic fields measurements onboard an electric UAV. We present the result of two pilot studies in Sweden to show the efficiency of the systems and accuracy of the acquired data. In area one we compare the apparent resistivity maps from the UAV-EM data with the data collected in a routine fixed wing airborne survey. Comparison between resistivity models from 2D inversion of ground very low frequency data and the UAV-EM data in the same frequency range show very similar resistivity models in resolving the presence of mineralization at the abandoned Enåsen Au-Cu deposit. The second example is from the Blötberget Fe-oxide deposit where the highly detailed resistivity/susceptibility models from 2D/3D inversion combined with 3D geological modelling reveal a NE-continuation of the main ore body which is segmented by a cross-cutting NW-striking fault system.

Geological structure is distinctive and boundaries between different units are often sharp. Geophysical models are often smooth and the geological meaning is interpreted. These interpretations are often subjective and the model can be unconstrained and not always influenced by prior information. Presenting these results to a non-geophysicist can be difficult. Typically, boundaries are sketched or overlain on the geophysical model to communicate the results. Information on structure orientation is often ambiguous. With prior information, we can select more geologically relevant models using regularization. However, prior information is not commonly available. We propose an automated approach to infer structure in geophysical models by applying image segmentation methods borrowed from computer vision. A transformer network is used to perform the segmentation on a physical property model obtained through inversion. We can infer the orientation of units through principal component analysis of the segmentation. The orientation is then incorporated into the model norms smoothness term. This provides a more clearly defined targeting system for structural orientations of intrusive units or interfaces like unconformities which are often associated with mineralization zones. The results provide an outlook on using machine learning and foundational models for structural interpretation of geophysical models.

It is nowadays accepted that the airborne electromagnetic (AEM) measurements are sensitive to Induced Polarization (IP) effects when collected over polarizable halfspaces. It has also been shown how using a frequency-dependent resistivity for the AEM modelling allows to properly retrieve the halfspace resistivity model as well as to map the airborne chargeability. With this work we propose a novel approach that points to actively use the airborne chargeability in the interpretation workflow and to the design a ground DCIP follow-up survey based on the airborne models. This approach aims to use the airborne chargeability as vector for the exploration to reduce costs and risks and to increase the targeting. To do this, we first define a novel modelling approach to extract the IP from AEM data reducing the equivalencies in the “IP-expanded” model space. Then, we integrate the obtained airborne resistivity and chargeability models with the ancillary information to design the ground DCIP follow-up. The ground IP modelling has been done using the same approach used for the AIP and consistent results has been obtained.