b'AEGC 2023Short abstractsSaint Arnaud Group facilitated fault development. FaultEnergy is focused on sustainable development of its ~3 600 reactivations in the Late Ordovician focused mineralising fluidkm2 Valhalla Project area. Few wells have been drilled on flow along these interfaces. the Lennard Shelf and basin margins bounding the deep Fitzroy Trough depocenter, but a number have encountered Stawell Gold Mine operates on the Magdala Dome. A suitesignificant gas shows and have tested conventional and of structurally related domes have been identified north- unconventional resources. The BME tight gas acreage has northwest along the corridor, and several have been confirmed1.5 trillion cubic feet (TCF) of contingent gas resources and to bear economic-grade gold ores, such as the Wildwood and11.8 TCF of prospective gas resources. These resources are Lubeck domes. However, due to the increasing thickness ofcontained within the prograding Laurel marine siliciclastics Murray Basin sedimentary cover northwards, these have notand carbonates that were deposited in a transitional ramp previously been explored in detail. setting. Multiple tectonic phases have altered this basin, This project contributes to understanding gold-bearingcreating an extensive set of both transpressional and structures in the Stawell Corridor by modelling newly acquiredextensional fault systems.ground gravity data across the Magdala, Wildwood andAhead of the planned 2024 appraisal programme, the 132 line Lubeck domes. We resolve their extents and geometries.km Odin 2D seismic programme, currently under acquisition, Additionally, a machine learning approach is used to train aseeks to delineate these deeper gas resources and provide visual recognition algorithm to recognise these structuresfurther structural definition to the southwest portion of the at depth from a regional gridded dataset, so that additionalpermit. This presentation and paper highlights some of the key domes might be identified. aspects of the Basin Centered Gas Play as defined by the 2022 2D seismic programme. We investigate the interpretation of Deep learning and natural language processing tokey structural elements as relates to prospectivity for future analyse legacy text reports and enhance petroleum,unconventional wells.minerals and clean energy exploration workflows. We are excited to bring our expertise to Australia, emerging as an industry leader in responsibly developed and Russell Menezes environmentally conscious natural gas supplies. We believe 80% of data in the resources industry is unstructured, that is,Project Valhalla will create energy security, socio-economic they exist in the form of PDFs, images and text files (includinguplift to the project areas Traditional Owners, local employment scanned and handwritten). Extracting information from this isopportunities, connections to southeast Asian LNG markets, and time consuming and expensive. Hence, under 2% of this dataa leading cost advantage for WAs manufacturing industry as we is ever used by the industry. This data is, however, extremelybuild Australias next great energy companyvaluable, as it contains information collected over the last 200 years and contains the expertise of thousands of subject matterLeveraging the value of historical AEM datasets with experts who have contributed to these reports.quantitative inversion.Going through these reports and being able to analyse and extract information efficiently and accurately forms the basisShane Mul 1 and Ross Brodie2of exploration in the petroleum, minerals and clean energy1 CSIRO industries. Can we leverage the power of AI in the resources2 Geoscience Australiaindustry to enhance our exploration workflows by automating some of the laborious tasks? Large holdings of airborne electromagnetic (AEM) data exist in the public archives of government agencies. The majority In this presentation, we will explore state-of-the-art deepof these are historical high resolution surveys, however due learning and natural language processing techniques (NLP) andto the challenges involved in working with these data they see how they can improve workflows in the entire oil and gasare seldom used and integrated into modern projects. Many value chainfrom data management to seismic interpretation,have only been interpreted as profile data to identify discrete petrophysics, sedimentology, and reservoir engineering. Forconductor anomalies, and some may have been subjected the mining industry, we will explore case studies where we canto parametric plate-modelling. Most datasets from prior to leverage NLP techniques to find unrecognised mineral deposits.the turn of the century will not have been interpreted with Finally, we will look at, how using these same techniques andfull non-linear 1D inversion methods that are now used as a datasets, we can explore for alternative energy sources likeroutine practice.hydrogen and geothermal.Since historic datasets were not routinely quantitatively Unlocking the Canning Basin: Gas resources forinverted at the time, there was minimal priority on the capturing of accurate system specifications and calibration. As Western Australia. a result, many datasets contain unknown, uncertain, or poorly Jarvis Moore 1, Perry Richmond1, Sean Taylor1, Samanthameasured system parameters and undocumented or poorly Richardson2 and Robert Hull1 documented metadata. Additionally, most of these surveys were collected with now obsolete systems or system variants. 1 Black Mountain Oil and GasDespite these impediments, we have been motivated to 2 Black Mountain Energy ascertain if further value can be extracted from historic datasets through quantitative inversion techniques.The Canning Basin, located on the northern coast of Western Australia, is one of Australias largest basins at over 400 000We have investigated various techniques in an attempt to km2. Within the northern portion of the Basin, Black Mountainovercome these impediments. This includes using inversion 121 PREVIEW FEBRUARY 2023'