ASEG Extended Abstracts - ASEG2013 - 23rd Geophysical Conference, 2013

Scintrex GRAVILOG borehole gravity meters are now successfully deployed by Abitibi Geophysics crews in boreholes for mining exploration in Canada and USA, and by Scintrex and Micro-g LaCoste crews in wells used for CO2 injection, sequestration and leaching in Canada, USA and Europe. In 2013, GRAVILOG systems and crews will be available in Europe, Russia, Brazil and Australia. Scintrex has developed and will be testing a dual sensor GRAVILOG probe in early 2013.

A brief review of the GRAVILOG development and specifications is presented, followed by several recent case histories. The result from Donner Metals / Xstrata Zinc’s Bracemac KT Zone in the Matagami region of Quebec is the first documented use in mining exploration of borehole gravity to measure excess mass coincident with a borehole EM conductor. Forward modeling and inversion of GRAVILOG data from multiple holes has proven effective in outlining the mineralization and estimating the tonnage of the Virginia Mines’ Lens 44 on the Coulon property in Quebec. Apparent bulk density measurements from multiple holes at the Labrador Iron Mines’ James South Extension iron ore deposit near Schefferville, Quebec reduce drilling costs and time needed to obtain this information.

A major source of error in bulk density measurements of thin beds is minimized by the dual sensor GRAVILOG system. In addition to eliminating the error in the depth interval between the sensors, common mode noise rejection improves the gravity difference data.

Assuming without evidence that magnetic sources are magnetised parallel to the geomagnetic field can seriously mislead interpretation and can result in drill holes missing their targets. I present two new methods for providing information about magnetisation of anomaly sources, independent of the geometry of the causative bodies. The first method is based on analysis of magnetic gradient tensor data. Integral moments of tensor invariants locate the horizontal and vertical centres of magnetisation and estimate the magnetisation direction. The depth estimate allows correction of the integral moments for the finite range of integration, which can accordingly be restricted to the main part of the anomaly. This reduces interference from neighbouring sources. This method provides information on location, total magnetic moment (magnetisation × volume), and magnetisation direction of a compact source, without making any assumptions about its shape.

The second method employs a single combined gradiometer/magnetometer, operating in base station mode within a magnetic anomaly of interest. The response to geomagnetic time variations allows the contributions of induced magnetisation and remanence to the anomaly to be separated. This method allows remote estimation, prior to drilling of (i) the total magnetisation direction of the source, which is a key to accurate modelling (ii) the remanence direction, which can provide geological information such as age of intrusion or alteration, (iii) the Koenigsberger ratio Q, which is indicative of the magnetic mineralogy of the source. If the source is compact, the method also provides a direct indication of the direction to its centroid.

Recent noise reductions in airborne electromagnetic (AEM) systems have allowed detection of conductors at great depths, but systems now have also become sensitive to superparamagnetic (SPM) effects. We distinguish SPM effects in airborne electromagnetic survey data from the response of good conductors. In electromagnetic data processing, off-time data can be accurately represented as amplitudes of a set of basis functions that are comprised of decays that decrease exponentially as a function of time. The SPM impulse response can be approximated by a decay that is proportional to time to the inverse power, a time dependence associated with magnetic viscosity. We identify the presence of SPM effects, as distinct from the decay of good conductors, by using inverse power- law decays as additional basis functions in constrained least-squares fitting. Application of the method to airborne time-domain electromagnetic (TEM) surveys shows that the method allows correction of SPM and hence aids significantly in conductive target identification.

The 11GA-YO1 deep seismic reflection survey reveals information on the crust down to ~66 km depth, imaging the crust-mantle boundary and the upper mantle. The seismic survey traverses the Yilgarn Craton, Officer Basin and Musgrave Province yielding information on their key structures and boundaries.

Interpretation of the seismic reflection data was complemented with forward modelling and 3D inversion of gravity and magnetic data. This allowed the geological structures interpreted in the seismic data to be investigated and extended into 3D space.

The 3D gravity and magnetic inversions reveal information on the geology and structure of the crust in the Yilgarn-Officer-Musgrave (YOM) region. In particular, information on the nature of dipping bodies beneath the Officer Basin and the boundary between the Yilgarn Craton and Musgrave Province.

One of the first usages of spectral methods in EM was by Peter Annan in development of the algorithm for EM modelling program “PLATE”, popular in the days of mainframe VAX computers. Using recent developments in spectral methodology and parallel computing on GPU boards, it is predicted to be possible on a desktop to run geologically plausible AEM forward models in microseconds, and inversions in milliseconds. In practice, the limitations of noise, imperfectly known waveforms, coupled with the effects of dielectric permittivity, viscous magnetism (superparamagnetism) and induced polarization conspire to limit the accuracy and speed of the process.

In electromagnetic exploration using broadband sensors and systems, joint recording of B and dB/dt data derived from the same physical sensor allows for a greater number of “signal” bits that lie above sensor noise. Optimum choice of a crossover frequency where the gains of B and dB/dt data match may double the number of useable bits. Field data confirms general expectations that B mode of operation is good at low frequencies, dB/dt operation good for high frequencies, but that both can be collected from the same ARMIT sensor.

The critically important steps to get best value from your gravity gradiometry data, assuming your contractor has done his job well in designing and acquiring the data, is the preparation of the representation of the potential field gradients. The ~200 m resolving power of existing gradiometer systems approaches what is necessary for minerals applications. In particular, beyond the aircraft, the topographic surface represents the largest and most proximal density contrast encountered in an airborne survey. Hence terrain effects can have significant impact on AGG data. The critical steps are:

• Terrain correction and determining ‘best’ terrain density
• Gridding, using all the measured gradients to constrain the interpolation
• Smoothing/de-noising by using the 3 rd order tensor constraints
• Anti-alias filtering of the gradient signals so that wave lengths are properly represented in all directions
• Transformation of the gradients by integration to estimate the gravity or magnetic field

Terrain corrections are a necessary step in the processing of observed AGG data in rugged terrain, in order to highlight subsurface density variations with a minimal overprint from the terrain. We propose a simple and rapid AGG tensor-based method to estimate an optimum bulk terrain density for subsequent terrain-correction. Each of the currently deployed systems for acquiring gradiometry is evolving driven by competition and the users’ needs. Mining applications of the technology to directly detect ore-bodies that show up as anomalies can now be successful provided the dimensions are of the order of 200 m or more. High resolution 3D geology models of operating mines can be used to calibrate gradiometry surveys

A synthetic rock analogue with simple microstructure was used to advance our understanding of the influence of cracks and pore fluids on seismic properties. The glass beads with ~ 300 μm diameter were sintered near the glass transition with average 1~2% porosity and subsequently quenched from high temperature into water at room temperature to introduce cracks with uniformly low aspect ratio α ~ 0.0007. Jackson-Paterson attenuation apparatus, with independently controlled confining and pore-fluid pressure systems, was used for both torsional and flexural mode forced oscillations at seismic frequencies to extract shear and Young’s modulus respectively, with or without the presence of pore fluids (e.g. argon, water) of varying viscosities. By perturbing the pore pressure at either end of the cracked specimen by ~5 MPa, permeability was obtained by analysing the evolution curve for pore fluid re- equilibration. Shear modulus is found lower with longer oscillation periods for the cracked and argon pore fluid saturated material possibly indicating the transition from the saturated isolated to saturated isobaric regime, with minimal strain-energy dissipation 1/Q < 0.003. The averaged elastic moduli for different oscillation periods and permeability are discovered to be extremely sensitive to variation of effective pressure. The crack closure effects can be observed easily at the effective pressure level at ~ Eα, consistent with the theoretical prediction.

Australia wide maps have recently been generated and released by CSIRO and Geoscience Australia using the 14 band satellite-borne ASTER sensors. Seventeen map products related to surface composition have been developed, based on spectral absorption features representing either abundance of mineral groups, specific minerals and their chemistry, vegetation cover or regolith related characteristics. This study aims to test the geoscience mapping capabilities of these products, individually, and integrated with airborne geophysics and DEMs over the semi-arid Mt Fitton area, South Australia, and within the agricultural Wagga Wagga region, New South Wales. The robustness of these techniques is evaluated by comparing the results from these two areas with different geological exposure and cultivation.

Various image processing, statistical and principal component analytical (PCA) techniques were utilised. ASTER map products including Ferric Oxide Content, Ferrous Iron Index, AlOH Group Content, MgOH Group Content, MgOH Group Composition, Ferrous Iron Content in MgOH, and Silica Index products proved useful discriminating previously mapped host or altered units of Mt Fitton. PCA correlation statistics proved helpful in devising potentially useful RGB composite imagery incorporating ASTER and geophysics. 2 D scatterplots between different products also proved useful for classifying surface composition. However ASTER map products within the Wagga area proved strongly affected by agricultural and floodplain cover. Slope per cent products generated from Wagga DEMs were useful in focussing ASTER products in areas of hilly and piedmont outcrop or eroded exposures.

In order to evaluate the use of shallow seismic technique to map the bedrock up to the depth of 20-25 meters, three high resolution seismic reflection profiles were carried out. The data were acquired using a Strata Visor with 48- channel, 40 Hz geophones and a weight drop system as seismic source. Seismic reflection data were recorded using a CMP (common mid-point) acquisition method. The results show that the bedrock lies at about 18-25 meters depth. The bedrock related horizon observed here is of low frequency, its depth is almost similar in all three seismic lines and thus giving us the enough confidence in results and also following the subsurface structure. Reflection line 3 is been crossed by reflection line 1 and reflection line 2. To confirm the structure and same statics, I did tie these lines to confirm and the reflectors are exactly matching, hence no need to give any shifting. There is high frequency loss due to high attenuation in near surface. A good structural image of subsurface is visible from the seismic sections and for interpreter it’s easy to mark structure and integrate it with other methods. With the proper equipment, field parameters and particularly great care in data collection and processing, we can image reflections from layers as shallow as 25 meters.

Eureka: (def.) “An interjection used to celebrate discovery.” Literally: “I have found it!” Reaching a ‘eureka moment’ is the result of much thought, effort, success, failure, perseverance, patience, and fortune. It usually occurs as the result of a team effort - although one person may lead the way. With regard to gravity gradiometry - and its ‘coming of age’ - the “eureka moment” comes when an explorationist realizes that something that couldn’t be done before can now be accomplished.

Gravity gradiometry surveys have been commercially available since 1999. Over the past 14 years, the capability has grown to a point of what could be called “adolescence.” Adolescence (from Latin: adolescere meaning “to grow up”) is a transitional stage of physical and capability development occurring during the period from youth to adulthood The period of adolescence is most closely associated with the teenage years. While gravity gradiometry doesn’t retain human qualities and characteristics, the analogy is used here to review and discuss the advances and maturity of the capability. Improvements and growth in system performance, operational readiness, survey volume, and value of information will be addressed in this review.

Airborne time-domain electromagnetic (EM) surveys are effective tools for mineral exploration, geologic mapping and environmental applications. 3D inversion of airborne electromagnetic data is a challenging computational problem. The size of the surveys and the spatial resolution required to adequately discretize the transmitters and receivers results in very large meshes. Solving the forward problem repeatedly on such a mesh can quickly become impractical. Fortunately, using a single mesh for both the forward and inverse problem for all of the transmitters is not necessary. The forward problem for a single source or a small group of sources can be solved on different meshes, each of which need only be locally refined with fine cells close to the selected transmitters and receivers. Away from the selected transmitters and receivers, the mesh can be coarsened. The forward problem can then be broken into a number of highly parallel problems. Each forward modelling mesh is optimized specific to the selected transmitters and receivers and has far fewer cells than the fine inversion mesh. In this abstract, we present an implementation of this idea using a finite volume discretization on OcTree meshes. We demonstrate our approach on a VTEM data set over a porphyry deposit in Canada.

Reservoir simulation is a key component for reservoir monitoring and flow prediction. Unfortunately such simulations and further flow predictions can fail due to the lack of knowledge of reservoir parameters such as porosity and hydraulic conductivity. To alleviate this difficulty we develop a new technique that combines reservoir simulation and electromagnetic imaging. We show how to use flow simulation in order to better invert borehole electromagnetic data and how in turn, the inversion results can improve flow simulation and prediction.

The equivalent source is usually calculated by an iterative method, designed to converge on the observed potential field, magnetic or gravitational. Each iteration involves two Fourier transforms, one forward and one inverse. As the number of grid point increases the time penalty may become prohibitive. However, the number of transforms can be dramatically reduced by the a priori transformation of the observed potential field to the frequency domain. We describe this method which reduces the number of Fourier transforms from n to two, and demonstrate its use on a simple magnetic anomaly. An extension of this equivalent source technique is also given for the case of a draped magnetic survey that requires leveling correction. The observed data is separated into layers according to the height changes, and the equivalent source is iteratively calculated via comparisons of the upward continued data. Whilst this process is computationally intensive, it is designed to scale with the complexity, and hence the discretisation, of the topographical changes. This gives the method a speed advantage compared with similar Taylor series approaches. A synthetic example containing multiple dipole sources is used to test the method, and to illustrate the advantages and differences of draped surveys and the need to reduce the data to a common datum.

Easily recognized bright events on seismic data can infer a host of phenomena, ranging from lithology interplay to indicating lucrative prospects in relatively greenfield settings. The availability of seismic angle stacks integrated to a geological model afford the petroleum explorationist more options to reduce obviously incorrect conclusions and help to identify the more likely cause(s) of these bright events within the seismic data.

A combination of simultaneous seismic inversion, and far and near angle stack comparisons gave more confident deductions on certain recognized seismic features as well as identifying potential interesting prospects to pursue and study in the context of petroleum exploration in offshore Myanmar. In the Bay of Bengal Rakhine Outer Fold Belt, exploration 3D seismic acquisition was carried out in 2010 and a significant gas discovery made in 2012. We present some examples of features that were identified on seismic data and how the range of geological hypotheses could be narrowed by seismic inversion.

Aeromagnetic data over eastern Australia reveal a pattern of domains defined by systematic regional highs and lows, emphasised by low-pass filtering, over which are superimposed shorter (<20 km) wavelength anomalies related to mappable geology and its inferred subsurface continuation. Long-baseline levelling by Geoscience Australia has clarified the definition of these magnetic domains, and confirmed that they are not an artefact of grid merging. Geothermal and teleseismic data indicate that neither variation in Curie depth nor upper mantle magnetisation can produce the long-wavelength pattern. Hence, domain-wide variations in magnetisation at the middle to lower crustal level are presumably the cause of these long-wavelength features. Although reversed polarity remanence could contribute to deeply sourced negative magnetic anomalies, the correspondence of magnetic low domains with the Proterozoic Curnamona Craton and the Ordovician Macquarie Arc, and of a high domain with the western Lachlan Orogen floored by Cambrian ocean crust, suggests that the control may be simply stark contrasts in lower to middle crustal susceptibility. Moho thickness determined by the AusMoho model mimics the pattern of susceptibility domains, suggesting a relationship between tectonic history and mid to lower crustal composition. Implicit in this analysis is a division of the domains into continental and oceanic basement, with implications for the tectonic evolution of the Tasmanides and the distribution of mineral systems in eastern Australia. The mid to lower crust below the Macquarie Arc appears to be continental, and the Thomson Orogen is a compound feature, comprising both attenuated continental/arc crust and accreted oceanic crust.

We present a Gauss-Newton-based 3D inversion method for airborne ZTEM (Z-axis Tipper Electromagnetic) data to define resistivity structure relating to uranium deposits in the Athabasca Basin of Saskatchewan, Canada. The geophysical targets in this region can be represented by conductive plunging dykes in a resistive basement beneath a thick, more resistive overburden. We demonstrate using synthetic examples the effectiveness of the inversion method for detecting and delineating the target dykes and discuss how the inversion results are affected by various factors. It is shown that the dykes can be well imaged to depths more than 2 km even for the data from 200-m receiver height, provided the flight line is oriented perpendicular to the strike, and that the inversion results are relatively robust to the choice of the starting model. It is also shown that topographic effects are not serious for detecting the dykes at depth, because topographic effects are more significant at higher frequencies, while the sensitivity to the dykes increases with decreasing frequencies. One important finding is that if the flight line is oblique to the strike, the dependence of the starting model increases and the overall resolution decreases, compared to the 2D case, due to 3D effects.

Vibrators are the preferred sources for onshore vertical seismic profile (VSP) surveys. Truck-mounted units are especially useful as they are more road mobile than buggy-based units. In particular, they can be driven directly to the wellsite, resulting in improved response times, simplified logistics with fewer vehicles and personnel at the wellsite and improved transport safety.

Acquiring broad-bandwidth, high-quality VSP data, using state-of-the art vibrator equipment, however, has generally required the use of the latest buggy-mounted vibrators, compromising well-site logistics and transportation safety for VSP operations. To deliver optimum performance, two new truck vibrator models specifically designed to take advantage of all of the most recent developments in vibrator technology while retaining the logistical advantages of truck-mounted units have been developed.

These new truck-mounted vibrators offer greatly improved data quality with tests showing that they can transmit a signal with a frequency bandwidth of 6.6 octaves compared to previous models which can only transmit 2.5 octaves. Being mounted on modern truck chassis these vehicles retain all the mobility and HSE advantages of truck vibrators.

The vertical resolution of conventional seismic data limits the illumination and definition of thin sand reservoirs offshore North West Australia. With the availability of more wells within a field it is justified to use a well-centric stochastic seismic inversion approach to characterize the reservoirs. The stochastic AVO seismic inversion yields multiple triplet realizations of acoustic impedance, Poisson’s ratio and density at the fine enough vertical sampling for a more relevant reservoir modelling procedure. The paper will discuss the stochastic seismic inversion approach, followed by a structured analysis workflow which condenses the generated multiple realizations to more interpretable data volumes.