ASEG Extended Abstracts - 2nd Australasian Exploration Geoscience Conference: Data to Discovery, 2019

The study of incoming seismic waves from teleseismic earthquakes can be used to investigate the deep structure of the Earth’s crust and upper mantle. Resolving the deep crustal structure and locating potential mantle-tapping structures is critical for understanding the metallogeny of mineral deposits at the surface. From 2014 to 2018, a network of 83 seismic monitoring stations was moved across the Capricorn Orogen in Western Australia. This survey was intended to complement the previous deep crustal seismic reflection lines, and although the resolution was lower, the survey covered the majority of the orogen in 3D and provided different geophysical information from the reflection surveys.

Results from studies of receiver functions provide the depth to Moho and average composition of the crust for the orogen. The distribution of more felsic crust and a deeper Moho outline the extent of the Archean Glenburgh Terrane in the central part of the orogen. Common conversion point (CCP) studies provide a view of compositional layering in the crust, and have led to a revised interpretation of the 10GA-CP2 seismic reflection line. Intriguing features within the upper mantle obtained by bodywave tomography have yet to be interpreted within the context of the tectonic evolution of the orogen.

The project geology has been compiled reviewed and a new, modern, resource database has been developed for the Barrytown New Zealand, Ilmenite Garnet Gold Zircon project.

The work required existing databases to be updated to include all historic mineral sands and gold data, and consolidated and checked versus original records.

The NZ Petroleum and Minerals (NZPAM) Mineral Report (MR) documents for the Barrytown project were checked and the data was compared to that contained in the project databases compiled prior to this recent 2018 work. Additional information was extracted and added into the databases including omitted holes, collars, size fractions, magnetic fractions, lithology, and especially gold assays. Within the project budget, as much additional data as could possibly be extracted was found by being persistent and systematically working through the contained information. The NZPAM system, although not ideal, does contain a wealth of information.

Loop is a new open source 3D geological and geophysical modelling platform in full development.

The new platform consists of 4 main work packages:

1. Knowledge Management: use of AI techniques for knowledge extraction from literature, maps and reports using geological ontology. Geological rules will be encoded to ensure proper knowledge extraction.
2. Geological Event Management: Loop is a time-aware geological modelling platform and the event manager is capturing topological and time relationship between geological objects and structural events
3. Forward and inverse structural modelling: we will encode structural geological rules in a time-aware context to account for folds (including overprinting), faults, shear zones, unconformities and intrusions. The modelling is based on probabilistic modelling and allows for the definition of an objective function for geology and quantification of uncertainty via posterior probabilities.
4. Uncertainty characterisation and modelling: using stochastic simulations or the result of Bayesian modelling, Loop allows for characterisation and quantification of 3D uncertainty.

In order to improve the South West Hub Carbon Capture and Storage (SW Hub) site characterisation and reduce uncertainties around CO2 capacity, injectivity and containment, conceptual fault hydrodynamic models are defined and numerical simulations are carried out to investigate the impact of faults on CO2 plume distribution. These simulations model a worst case scenario with a plume reaching the interface between the top of the proposed injection unit and a secondary containment unit. The conceptual fault hydrodynamic models are defined to incorporate host rock and fault properties accounting for fault zone lithology, cementation, and cataclastic processes.

Numerical flow simulations of CO2 injection are performed to quantify cross- and up-fault migration of CO2 in case of plume-faults interaction. Here CO2 is intentionally injected close to a fault and >1,000 m shallower than the proposed large-scale injection depth, to force the supercritical CO2 to reach the discontinuity and the interface between injection and a secondary containment unit.

Results show that if a CO2 plume reaches the top of the injection unit none of the compartment bounding faults in the SW Hub would critically control CO2 flow nor would they act as primary leakage pathways. CO2 flow at the top of a faulted injection unit is primarily controlled by the sedimentological morphology of the secondary containment unit.

The Cooper and Eromanga Basins of South Australia and Queensland are the largest onshore hydrocarbon producing region in Australia. Igneous rocks have been documented infrequently within end of well reports over the past 34 years, with a late Triassic to Jurassic age determined rom well data. However, the areal extent and nature of these basaltic rocks were largely unclear. Here, we integrate a variety of subsurface datasets to clarify the nature and origin of igneous rocks preserved within Eromanga Basin sequences overlying the Nappamerri Trough of the Cooper Basin. We recognise mafic monogenetic volcanoes that extend into tabular basalt lava flows, igneous intrusions and, more locally, hydrothermally altered compound lava flows. The volcanic province covers ~7500 km² and is proposed to have been active between ~180-160 Ma. We name this Jurassic volcanic province the Warnie Volcanic Province after the Warnie East 1 exploration well, drilled in 1985. The distribution of extrusive and intrusive igneous rocks is primarily controlled by basement structure, with extrusive and intrusive igneous rocks elongate in a NW-SE direction. The magmatism is interpreted as a product of extension and intraplate convective upwelling above the subducting Pacific Slab.

Rift-related magmatism resulting in widespread igneous intrusions has been documented in various basins, including the Faroe Shetland Basin (UK), Voring and More Basins (Norway) and the Browse and Carnarvon basins of the NW Shelf of Australia. Seismic mapping, combined with field work, has resulted in greater understanding of subsurface intrusive plumbing systems, but knowledge of emplacement style and the mechanisms by which they propagate is limited. The interpretation of a 3D seismic dataset from the Exmouth sub-basin, NW Shelf Australia, has identified numerous igneous intrusions where a close relationship between intrusions and normal faults is observed. These faults influence intrusion morphology but also form pathways by which intrusions have propagated up through the basin stratigraphy. The steep nature of the faults has resulted in the intrusions exploiting them and thus manifesting as fault-concordant, inclined dykes, whereas in the deeper parts of the basin, intrusions that have not propagated up faults typically have saucer shaped sill morphologies. This transition in the morphology of intrusions related to fault interaction also highlights how dykes observed in outcrop may link with sills in the subsurface. Our interpretation of the seismic data also reveals subsurface examples of bifurcating intrusions with numerous splays, which have previously only been studied in outcrop.

To evaluate the extent of a low salinity aquifer within the continental shelf offshore New Jersey, USA, we adapt a Bayesian trans-dimensional Markov chain Monte Carlo (McMC) algorithm for jointly inverting surface towed controlled source electromagnetic (CSEM) and seafloor magnetotelluric (MT) data. Owing to the relatively flat seafloor topography and layered underlying geology, our inversion is parametrized to produce an ensemble of 1D models at each site along the tow line. The results identify a high probability region of high resistivity of varying depth and thickness in the upper 500 m of the continental shelf, corroborating results from a previous study that used a regularized, gradient-based 2D inversion method. Using a combination of synthetic studies and real data we also evaluate the joint model parameter uncertainty in comparison to the uncertainty obtained from the individual data sets and demonstrate quantitatively that joint inversion offers reduced uncertainty. In addition, we show with a Monte Carlo scheme, how the Bayesian model ensemble can subsequently be used to derive quantitative uncertainty estimates of pore water salinity within the low salinity aquifer. The ability to produce meaningful non-linear uncertainty estimates on both conductivity as well as related parameters is a strong motivator for the use of Bayesian inversion methods when computationally feasible.

The Carlow Castle deposit is located 40 kilometres east of Karratha, Western Australia. The deposit is shear hosted in the Ruth Well Mafic Formation, part of the Roebourne Group. It is interspersed with metagabbro intrusions and brecciated chert cataclastites. Carlow Castle contains Au, Cu and Co mineralisation. Data from a Sub-Audio Magnetic (SAM) survey was acquired in both galvanic and inductive configurations to map structural responses over known mineralisation. On-time Magnetometric Conductivity (MMC), Total Magnetic Intensity (TMI) and Off-time Galvanic Source Electromagnetic (GSEM) data were extracted from the galvanic dipole configuration. Off-time Inductive Source Electromagnetic (ISEM) data were extracted from the loop source configuration. MMC data were combined, modelled and interpreted to rank structural targets around existing Au and Cu mineralisation. This was used in conjunction with Co soil geochemistry as a vectoring tool for further cobalt mineralisation. The MMC data was successful in increasing level of detail over the project. ISEM and GSEM were also compared with previously highlighted VTEM anomalism. Five high and seven medium priority targets were identified for follow up with further drilling.

A key aspect of geophysical inversion is the ability to model the earth with a low dimensional representation. There exist various approaches to solve the inverse problem. However, most methods do not automatically adapt inverse model complexity or the number of active model parameters as dictated by data noise and sparse receiver coverage, do not quantify inverse model uncertainty or do not work equally well for 1D, 2D or 3D earth models. Low frequency electromagnetic (EM) inversion for example, can require for 3D problems upward of 10⁶ cells to forward model. Only a small fraction of these cells is effectively resolvable and there are significant trade-offs between them. To address these limitations, we present a novel approach to earth model parametrization by using a Gaussian Processes (GP) machine learning (ML) technique, coupled with a parsimonious Bayesian trans-dimensional (trans-D) Markov chain Monte Carlo (McMC) sampling scheme. One aspect that sets our approach apart from recent spatial dimension agnostic algorithms in the trans-D or ML literature is the ability to specify inversion property priors directly, as opposed to doing so in a transform domain of the property. Finally, we note that our method falls in the category of ML approaches that do not attempt to learn the physics of the process, but instead learn the representation of parameter values through a misfit function. We apply the trans-D-GP method to a 1D controlled source electromagnetic (CSEM) and 2D non-linear regression problem, using actual field data from the Northwest Australian Shelf for the former. The key advantages in using our method are the simplicity of prior specification, parsimonious low dimensional representations, and ease of representing large-scale models in 1D, 2D or even 3D with the same parametrization and computer code.

The Queensland Mining Journal represents a wealth of information relating to mining activities in the state, from 1800 to the present day. This material has been scanned and made available in the public domain as high-quality image files. We have applied Google Vision optical character recognition, parsed the output JSON files through a domain-specific filter and indexed the content for presentation via an analytics dashboard based on mineralogy and which can be filtered by commodity age and location. The dashboard further incorporates mine site information from Queensland’s Digital Exploration (QDEX) Data System allowing prospective miners to drill down into QDEX content based on mineral occurrence, mine status and deposit size. We plan to build on this activity using Elastic Search to improve our association of content from articles to spatial location.

This activity supports individuals and mining companies with an interest in Queensland to rapidly identify locations of interest. It offers a pragmatic approach using freely available information to support the Queensland authorities in attracting investment to their region through lowering the barrier in terms of effort level for companies looking to explore in the state. This comes at a time of heightened competition for investment and could ultimately lead to increased exploration activity and success, resulting in brownfield development of historic prospects and mine sites, minimising environmental impact of new exploration and mining yet generating income for the state through local activity and tax revenue on any mineral extraction.

The Exmouth Sub-basin forms part of the intra-cratonic rift system of the Northern Carnarvon Basin, Australia. It has undergone a complex tectonic history with multiple phases of uplift, erosion, inversion, and regional tilting. Hydrocarbon exploration has resulted in discovering a variety of oil and gas accumulations; however, the distribution of hydrocarbons and charge from different petroleum systems is still poorly understood. The new basin-wide, long-offset, broadband Exmouth 3D Multiclient seismic data (MC3D) allowed for an updated understanding of the structural and depositional evolution.

We built a regional 3D basin and petroleum system model that integrates this updated structural and stratigraphic framework and results from potential field modelling to a) improve our understanding of the thermal history and petroleum charge, b) assess critical parameters and processes impacting the petroleum systems, and c) evaluate the associated exploration risks such as biodegradation. The thermal model is based on a) a short period of extension in Latest Triassic (Rhaetian) as observed in the deeper image of the recent MC3D seismic survey, and b) a Latest Jurassic to Earliest Cretaceous magmatic event documented by potential field modelling and interpreted igneous intrusions on seismic data. The 3D basin model was calibrated to data from more than 30 wells in the sub-basin. Model results predict that burial of the Early Cretaceous Barrow Group controlled hydrocarbon generation and expulsion over large areas of the sub-basin; towards the south, however, results suggest that hydrocarbon expulsion from shallower Jurassic source rocks continued into Late Cretaceous time, a period when the regional Early Cretaceous Muderong Formation is predicted to be an efficient seal rock. This implies that vertical, short-distance migration may contribute significant petroleum charge to the accumulations in the southern part of the Exmouth Sub-basin.

Onshore exploration technology continues to evolve with the arrival of new airborne instrumentation systems. Central to this has been the need to also evolve quality control processes that ensure useable signal is being captured during the surveying process, even though the true value add occurs at a later time. Gravity gradiometry is now well established, and able to provide independent mapping detail to wavelengths of less than 400 m. Airborne electromagnetic data is also starting to provide cross-sections that are reflecting actual geology bodies in terms of dips and thicknesses.

The quality control (QC) technology applied across the industry is not uniform, and sometimes inappropriate for new datatypes being acquired. Government contract specifications can help. Also improved software tools being generally available and having trained operators, is an emerging requirement. This critical aspect includes fit for purpose geophysical gridding.

This research presents a new method for assessing magmatic fertility using pXRF in altered terranes using Zr and Y ratios. A study of global batholith related mineralised porphyries, highlights a low Zr fractionation trend. The depletion in Zr is associated with early crystallisation of titanite. Yttrium depletion is related to early crystallisation of both titanite and amphibole, an indication of hydrous melts. Previously differentiation of hydrous, potentially ore-forming (fertile) porphyries relied on Sr/Y from whole rock assay of least altered rocks. Finding samples in porphyry terranes where Sr can be demonstrated to be immobile is difficult given the ubiquity of hydrothermal alteration associated with porphyry emplacement. The requirement for unaltered samples is based on the mobility of Sr which precludes the widespread use of pXRF as quantitative assessment of LOI% (as a proxy for alteration) is beyond the current capabilities of the technology. Using comparatively immobile Zr and Y overcomes issues with alteration and provides a more reliable new indicator of magmatic fertility. In this case study, using a systematic workflow pXRF was successful in identifying the mineralising intrusions in the Northparkes intrusive complex using a Zr/Y vs. Y fertility indicator.

The North West Shelf of Australia is a volcanic rifted margin. Jurassic-Cretaceous breakup of Gondwana resulted in the intrusion of significant volumes of igneous material into sediments of the Northern Carnarvon Basin. The last major review of the regional distribution of this intrusive material was published over 20 years ago. Since, there have been major advances in the understanding of intrusions in sedimentary basins, and step changes in the quality of 3D and reprocessed 2D seismic data available in Northern Carnarvon basin.

We present preliminary findings of a new study building on these advances that investigates the regional spatial and temporal distribution of intrusions in the basin and the impact of this magmatism on petroleum systems. In this paper, we have mapped plumbing systems of intruded igneous complexes, interpreted on 3D surveys within the Exmouth Sub-basin, and on the Exmouth Plateau.

Within the study area we see (1) typical ‘saucer shaped’, and ‘stepped’ intrusion geometries; (2) intrusions exploiting faults and cross cutting stratigraphy as they rise to the surface; (3) interconnected intrusions, forming continuous magmatic plumbing systems >50km in length; (4) intrusions mainly present within in the Triassic Mungaroo Formation, and (5) intrusions possibly sourced from a hotspot that was located beneath the current Cape Range Fracture Zone during breakup.

Australia is in the top six global producers for a wide range of mineral commodities of abundant metals (Fe, Mn, Al, Ti), base metals (Ni, Cu, Zn, Pb), precious metals (Au, Ag), scarce metals (REE, Li, Sb), energy metals (U, Th) and industrial/ precious minerals (salt, diamonds, zircon). Australia also has a wide range of uranium deposit types but their future is controlled by government policy. There are very few mineral deposit types that are not present or minor in the Australian continent. These include PGE deposits in large layered intrusions and Witwatersrand-type gold deposits, Carlin-type gold deposits, low-sulfidation-type epithermal Au-Ag deposits, Zambian-type copper deposits and MVT deposits. For most of these, appropriate depositional environments are extremely rare or absent. The future of the Australian mineral industry will continue to depend on a combination of brownfield exploration in known mineral districts and greenfields exploration, increasingly under deeper cover, for deposit classes within these mineral districts. Craton margins that are intruded by basic and/or felsic magmas are particularly attractive tectonic targets. They will continue to be fertile exploration environments for IOCG deposits and intrusion-hosted Ni-Cu deposits, and submarine gold-rich porphyry-to-VMS systems may have been neglected as exploration targets in the past. The future potential for pegmatite-hosted lithium deposits is high.

The Pillara Zn-Pb deposit is located on the northeast margin of the Canning Basin, Western Australia where it is hosted by a series north-northeast-trending, en echelon, conjugate normal faults that bound a central graben plunging gently to the northeast. Kinematic indicators on all fault surfaces indicate dominantly dip-slip movement. Extension increases from approximately 15% in the south to as much as 50% in the north of the graben structure. The complex array of faults is consistent with development in a northwest-southeast directed extensional stress field with a vertical σ1 prior to minor tilting of the sequence to the northeast. Northwest-southeast extension was localised within a northeast-trending transfer structure.

Local controls on the development of mineralisation plunging shallowly to the northeast include dilation where faults steepen or overlap and where antithetic faults intersect the main bounding Western and Eastern Faults. Areas of greatest dilation are largely controlled by the orientation of individual fault segments relative to the local stress field and the intersection of the faults with a competent fenestral limestone unit (Unit 5) and lower platform and bank facies (Unit 1). Both units display low uniaxial compressive strengths compared to other limestone samples within the sequence. The intersection of north and northeast trending fault segments defines a steep, northerly plunge that appears to control high grade orebodies in the F10 system (A Lode and B Lode).

Mineralisation is dominantly contained within tectonic breccia zones, open-space fill along faults, and extensive hanging-wall hydraulic breccia zones localised by fault relays and intersections with antithetic faults. The latter grade outwards from rubble-breccia through mosaic breccia to crackle breccia and stockwork vein zones. Veins and breccias may show a history from early marine calcite cements through to pre-ore marcasite and calcite-cemented crackle breccia (B0), to ore-stage breccias characterised by sulphide-cemented rubble breccias (OB1) at times incorporating sulphide breccia clasts (OB2), to post-ore breccias cemented by sparry calcite cement (OB3). The ore textures and paragenetic sequence are consistent with introduction of the ore fluids during active fault movement, and rapid precipitation into open cavities created during faulting.

Downhole electromagnetic (DHEM) logging was undertaken within four holes drilled at the Las Cruces VMS body in Spain. The geology of the deposit is geophysically complex inasmuch as the background and overlying geology is conductive. A number of loop positions were pre-modelled in order to optimise coupling with the sulphides. The objective in all cases was to locate additional off-hole massive sulphides, belonging to both the main sulphide body and a smaller satellite body. The DHEM was undertaken with a variety of loop positions and at a number of base frequencies. Modelling was undertaken using Maxwell EM modelling software. In three of the holes, by using a number of thick plates and various conductive layers and half-spaces, the DHEM results could be explained by the known sulphides. In the case of a fourth hole, off-hole anomalies were located at depth. A new borehole was recommended based on these data. This borehole was later drilled and an extension to the satellite body intersected.

Subtle, and not so subtle, features in our increasingly large and complex datasets are being missed because of our reliance on a single number, typically chi-squared based, to describe the distance between a plausible model and the data. Additional measures need to be used. Here I propose that the autocorrelations of the residuals between the data and the model responses be minimized as part of the objective function during inversion. I demonstrate the use on a toy example and three 1D magnetotelluric examples, one a synthetic “impossible” problem and two real ones.

The Sinclair Caesium Deposit, discovered in 2016 by Pioneer Resources Limited, is Australia’s first mining operation to commercially extract the caesium-rich mineral pollucite. Economic caesium deposits are extremely rare, with only three mining operations having produced commercial quantities of pollucite: the Bernic Lake Mine (or Tanco Mine, Manitoba, Canada), the Bikita Mine (Zimbabwe) and now the Sinclair Mine (Western Australia).

The formation of pollucite only occurs in extremely differentiated lithium-caesium-tantalum (LCT) pegmatites and given their size and rarity, it can be assumed that caesium-rich deposits globally are either challenging to explore for or form surpassingly rarely during the emplacement of LCT pegmatites.

Common with the Bernic Lake and Bikita Deposits, the Sinclair Deposit’s host LCT pegmatite consists of an outer pegmatite wall zone that is coarse grained, and dominated by plagioclase feldspar, muscovite and quartz with other accessory minerals; and an inner core zone composed of quartz, albite (cleavelandite), lepidolite, pollucite, petalite, zinnwaldite, eucryptite, beryl and amblygonite. The core zone is ‘capped’ by a thick (~35-40 m) monomineralic potassium feldspar zone.

This paper will present details on the discovery and geology of the Sinclair Caesium Deposit.