Exploration Geophysics - Volume 20, Issue 1-2, 1989

We have computed EM coupling responses for the dipole-dipole IP array over realistic 3-D bodies using a volume integral equation approach. The objectives of these simulations are to gain insight regarding negative EM coupling over good conductors, to analyze the phase extrapolation method of removing EM coupling, and to determine whether interpretation of the EM coupling can help resolve the resistivity structure. Our results show that a good conductor, especially if it is shallow, can produce a complex pattern of negative and positive coupling in a pseudosection. Extrapolating three-frequency phase data to zero frequency eliminates EM coupling in most cases; however, residual EM effects may be present in extremely low-resistivity environments. We are studying ways of utilizing EM coupling to provide information about 3-D bodies beyond that produced by DC resistivity data.

Most of the gold deposits in the Barberton Greenstone belt of South Africa are relatively small and in structurally complex geological areas.

The mise-a-la-masse electrical technique, where a current electrode is earthed in a mineralised zone, was used on one of our exploration projects consisting of a sulphides/gold-bearing carbonaceous banded iron formation within a succession of mafic, ultramafic and sedimentary rocks. The technique was successful in delineating individual mineralised units within a broad lithological sequence. During the survey, electrical potential measurements were recorded on surface, in underground drives and in twenty five boreholes. Measurements were also repeated by earthing the mineralised zone in a number of boreholes. Major discontinuities were recognised within the ore zones and used to interpret geological structures. These were then used to define specific units for ore reserve calculations and the application of selected mining techniques.

Geoelectrical resistivity soundings conducted along the north-south expressway in the Tangkak-Pagoh area during its construction enabled the mapping of a marine clay which was present in alluvium. The apparent resistivity curves of the marine clay area, were characterised by low apparent resistivity values falling to almost zero, and a characteristic shape which was different from areas of weathered metasediments and weathered granite adjacent to this area. Apparent resistivities for both metasediments and granite were intermediate to high, with a distinct upward convex curve for metasediments and a flat upward convex curve for granite. Quantitative interpretation was able to distinguish between an upper and a lower clay horizon, which have different consistency and fine sand content. The thickness of the upper clay horizon, which varies between 8.0 m to 9.0 m according to borehole data showed good agreement with that obtained by resistivity soundings. The base of the lower clay horizon remained undetected by the resistivity soundings because of its very low specific resistivity. Using the thickness of weathered layers as observed along the road cuttings as a guide, good agreement between field data and calculated values was obtained in determining the specific resistivities and individual layer thicknesses for the granite and metasediment weathering profiles.

Two CSAMT case history studies are presented in which modelling has been applied to data in order to produce geo-electric depth sections. These sections then assist in developing an idealized geologic model used as part of a geological exploration programme. At the geological planning level, one can define certain key geologic features which need to be identified as part of an exploration programme. Modelling may give some indication as to the success of a particular planned geophysical programme.

At the interpretation level, modelling can be used to confirm or disprove geologic ideas previously developed, and aid in quantifying features that survive the test. In addition, modelling may give new insights on how to get more information from the geophysical data and give rise to new ideas.

The case history information comes from the Golden Cross mine development programme near Waihi, New Zealand, and from the Togi exploration programme conducted near Kanazawa on the island of Honshu, Japan. While the distance separating the two survey sites is great, they are both epithermal gold prospects. CSAMT was used to obtain high resolution structural mapping both for near-surface geology and geology at depth, which are two important tasks.

Modelling of the CSAMT data obtained at Golden Cross and Togi is provided by three different computer programmes: CSINV is a one-dimensional general EM field approach using a finite number of layers where the survey geometry is specified; a variation of CSINV uses an infinite number of one-dimensional layers; and EM2D is a two-dimensional approach using a finite number of bodies/layers based on plane-wave EM theory. All three modelling approaches provide resistivity values varying as a function of depth, however useful results with EM2D are limited to modelling frequencies where data satisfies plane-wave criteria.

Modelling at Golden Cross provides cross-sectional geo-electric information which extends known drill-hole geology. This provides broader detail to known information, and suggests additional potentially interesting features. Modelling at Togi was directed at providing plan view coverage of certain geologic structures. Here one survey objective of the CSAMT survey was to provide structurally related control to be used with interpretation of geochemical data.

Even though targets are geologically similar, planning and interpretational goals differed. CSAMT modelling optimized the interpretational use of CSAMT data in each case history.

The magnetometric resistivity (MMR) method was used to locate palaeodrainage channels in the Eastern Goldfields of Western Australia. Drilling results showed the method does not distinguish between the palaeochannel and porous weathered basement containing saline water. This distinction however was not essential to the success of the program as the weathered basement was generally symmetrical around the channel.

This paper outlines the development of a cross hole seismic tomography package by The Broken Hill Proprietary Co. Ltd. (BHP), as a tool for mineral exploration and mine planning. The methodology of cross hole seismic tomography, field procedures, instrumentation, processing software, and field trials are described.

Explosives are principally used as the source of seismic energy. A repetitive source, based on rapid hydrogen-oxygen combustion, has also been developed. Signals are detected by geophone-based detector strings, and recorded by a data acquisition system developed by BHP. Tomographic imaging is conducted by the Algebraic Reconstruction, Back Projection and Simultaneous Iterative Reconstruction techniques.

Surveys have been conducted in a number of different geological environments, and include: lead-zinc, iron ore and manganese exploration leases and mines to locate mineralisation and overburden interfaces; underground coal mines to locate regions of mining induced stress; and open cut and underground coal mines to locate coal and overburden contacts. The results of these surveys are discussed.

Geophysical techniques have been applied to petroleum exploration since early in the 20th Century. More recently geophysical methods have been applied in detail to mineral and coal exploration. As a generalisation, geophysical techniques have not been applied in the areas of mine planning, development and production.

A variety of geophysical methods have been improved or adapted within BHP to provide accurate, cost effective services to the mine manager on time scales that are realistic for day to day planning and production. Considerable success has been achieved with in-seam seismic, cross-hole seismic and surface seismic techniques. Electrical and magnetic methods have also been beneficial for specific applications.

he identification and evaluation of mineral deposits increasingly uses a range of advanced geophysical techniques. Geophysical techniques are now also emerging as key factors in mine planning and production. The purpose of this paper is to show how BHP is developing a variety of geophysical techniques to improve the efficiency of exploration, mine planning and production both for minerals and coal. Emphasis is placed on the benefits of these advanced geophysical techniques on day-to-day mine operations. This, of course is only one company's perspective viewpoint, but since BHP has such a wide diversity of operations, this viewpoint may have general applicability.

BHP has had a long history of using geo-expertise in a wide range of operations over the past 40 years. This expertise developed in the minerals and coal industries but has subsequently developed into the petroleum industry. In regard to the coal industry alone, several notable geophysics firsts can be attributed to the coal geology groups within BHP. These firsts include:

1. The application of surface seismics to coal exploration;
2. Geophysical logging – BHP were instrumental in bringing BPB Instruments Ltd to Australia;
3. Radar – early experiments were undertaken at Cook Colliery;
4. Development and application of high resolution surface seismics in Queensland and New South Wales;
5. Development and routine application of in-seam seismics;
6. Cross-hole seismic/in-seam seismic tomography – application of a production oriented package to coal and metalliferous mines.

In the development of these techniques for the mining industry, a number of common factors are present which have resulted in them being commercially successful. BHP's background as a large resources company has obviously provided the initial impetus to develop smarter geophysical techniques, but this is only one factor which has made them successful. The old adage of a new product or technique being 1% inspiration and 99% perspiration also applies to the development of these techniques.

Probably the most important single factor to consider for the successful development of innovative geophysical techniques is that they require a multi-stage team effort over at least two years, (typically 4-5 years for the more complex developments) and that failures can be expected throughout this period. Also the expectations of production personnel are often too great during this developmental stage, which leads to a perception that the technique in question is not useful even after all the 'bugs' in the system have been removed. The onus is on researchers to clearly outline both the potential benefits and possible failures of a new technique during its developmental stage, so that it will subsequently be more readily accepted in the mining production environment.

A swept-frequency borehole seismic source has been constructed and tested that consists of a portion of the borehole isolated from the remainder and driven to build up pressure oscillations by resonance.

The length of the isolated portion is changed to vary the resonant frequency. The source radiates an approximately isotropic P-wave whose total energy is comparable in magnitude to that created by a surface vibrator truck. Strong shear waves are also generated.

The source has been tested over a frequency range of 30 to 120 Hz, but the design can be operated from 15 to 500 Hz. Because the driven section of the borehole is isolated, strong tube waves are not generated. No damage to the casing-cement bond has been observed after prolonged operation of the source at a fixed depth.

This source has the strength and uniform radiation pattern to suit it for both inverse VSP and cross-borehole surveying.

Two dimensional seismic surveying is commonly used in the coal mining industry to assist the mining and development of coal deposits by seismically imaging coal seams. A specialised three dimensional seismic surveying technique has recently been performed over coal mining leases in South Australia and New South Wales, to trial its applicability to coal mine planning and extraction operations.

The first two case histories of its trial in Australia are presented, and the conclusion drawn that the specialised three dimensional technique developed to date offers the ability to image coal seams in three dimensions and thereby improve mine planning in regions of complex faulting.

In the past, seismic reflection surveys on land in Japan have usually been carried out along straight lines irrespective of surface terrain. However in Japan, where most of the areas are covered by mountainous terrain, it may require considerable time and cost to carry out field operations by the straight line method.

From August to October in 1988, a reflection seismic survey was carried out at the Sarufutsu area in Hokkaido, Japan. The area is in a mountainous region where the straight line method was very difficult to conduct. A trial of the crooked line method was introduced to the area by laying out seismic lines along roads and valleys.

Four test borings with VSP and extensive field geological surveys were carried out in this area by NEDO (New Energy and Industrial Technology Development Organization) prior to 1987, and the results show that there is a north-south trending syncline in

The data acquisition system adopted was a 48-trace IFP digital recording system with one-millisecond sample interval.

The result of the survey gives a good correlation with existing geological information.

It has been concluded that seismic reflection surveys by the crooked line method are applicable for coal exploration in such a mountainous terrain.

A new short-pulse radar system (RAMAC) developed by ABEM AB has now been in operation for three years during which more than 100 km of borehole logging has been performed. The bulk of the surveys have been in granites and gneisses.

The RAMAC system operates at centre frequencies in the interval 20 to 60 MHz. At those frequencies single-hole reflection ranges of 50 to 150 m are normally obtained in gneissic and granitic rock. Cross-hole ranges have in some cases exceeded 300 m. The large probing range in combination with resolution of the order of a few metres makes borehole radar a unique technique for investigation of fracture zones in crystalline rock.

Case histories illustrate application of the RAMAC system in three different configurations (single-hole reflection, cross-hole reflection, and cross-hole tomography) and demonstrate how combination of these three can yield consistent 3D models of fracture zones and other structures.

Fireholes at the top of the thick Latrobe Valley brown coal seams pose a geotechnical hazard to overburden dredges and reduce coal reserves. Overburden thickness (typically 10 to 15 m) is up to 50 m in the fireholes, which are from 20 to hundreds of metres in diameter and are infilled with baked clays, soft lacustrine clays and alluvial deposits.

Given the complexity of firehole geometry and overburden geology, firehole definition prior to overburden stripping, by drilling alone, is expensive and is not definitive. To improve firehole exploration, geophysical methods were tried in a test area with good borehole control (115 holes), near Morwell open cut.

Grid geophysics (20m.x20m., 2805 grid stations) using gravity, EM34 20 m loop conductivity and high resolution magnetics gave very good results. Shallow seismic reflection methods were not successful.

Residual gravity defined overburden thickness variations best with gravity highs of up to 6.5 micrometres/sec² over the fireholes. EM conductivity showed reasonable correlation with overburden thickness, with EM conductivity highs over fireholes infilled with lower resistivity lacustrine clays and silts. High resolution magnetics using a TM-3 caesium vapour magnetometer, despite high cultural interference, showed broad, low amplitude highs over fireholes where higher susceptibility baked clays are thickest.

The three geophysical data sets and overburden data were gridded (5m.x5m.) and the grids dumped to a MicroBrian image processing system. Conventional image processing analysis was carried out to compare, enhance, filter, display and classify the complementary data sets. A classification scheme for overburden type based on geophysical responses plus a routine firehole exploration methodology using residual gravity, EM, magnetics, progressive drilling data and the image processor was devised to reduce drilling costs and increase exploration confidence. The case history presents the results of the grid geophysics and image processing approach.

Two separate seismic reflection surveys have been conducted over Aberfoyle Resources Hellyer orebody and the enclosing Cambrian Que-Hellyer volcanics. The initial survey, conducted as a test case to determine if seismic reflection techniques could detect massive sulphide orebodies at depth, successfully identified the Hellyer orebody, which was found to have a characteristic seismic expression. Identification and mapping of stratigraphic units within the Que-Hellyer volcanics was also possible due to the good geological control available.

The second survey did not successfully identify the Hellyer orebody, although some information relating to the structure of the enclosing basin could be inferred from the seismic data. Due to data degradation conventional seismic interpretation techniques could not be applied to the second survey.

High resolution seismic surveys can detect massive sulphide orebodies if low velocity overburden problems are eliminated and high frequency data is collected. High resolution surveys can also provide structural and stratigraphic information between widely spaced drill holes in volcanic terrains.

The cost of shot hole drilling is often more expensive than using vibratory energy sources in high resolution seismic surveying. However, such costs are often accepted since conventional vibrators cannot always provide the extreme imaging capacity required in high resolution work. Conventional seismic vibrators sweep in a range from 5 Hz to 250 Hz – the range of which is limited by the vibrator. The impulse train of the high resolution wacker used by the MiniSOSIE¹ system is also band limited, causing a reduction in imaging resolution. The ideal solution is to sweep a broad range of frequencies from the lower seismic range to as high as 500 Hz. This could offer a cost effective solution to the acquisition of broad band high resolution data.

In high resolution seismic profiling, explosives are commonly used as the source. Small charges below the weathered layer produce the highest frequency content (Ziolkowski and Lerwill, 1979). Unfortunately, the cost of drilling shot holes is a major component of the survey costs. For oil exploration in Australia/New Zealand, dynamite surveys average 43% more than Vibroseis² surveys in dollars per kilometre (Montgomery, 1987), despite more hardware being required for Vibroseis recording.

The MiniSOSIE system is also used for some high resolution surveys, because it is relatively cheap. However, this does not achieve equivalent results to small explosives. It will give worse results as the soil becomes softer, as the rebound from an impact takes longer, and hence the wavelet is broader.

An alternative to these sources is the hydraulic powered vibrator, which has sometimes been used for high resolution coal work in Europe. With vibrators, the spectrum is controllable within certain limits. The Vibroseis system can also produce zero phase wavelets, if used properly with its controllable frequency wave-train sweep; and with repeatable multiple sweeps this results in enhancement of signal to noise ratio together with the promise of the highest frequency returns. Zero phase wavelets have slightly better resolution than the same bandwidth minimum phase wavelets as produced by impulsive sources. The breadth (t) of a zero phase Klauder wavelet with a boxcar spectrum can be predicted from the sweep start (fS) and end (fE) frequencies by the approximation:

τ =1/(fS + fE)

By sweeping 50 to 500 Hz, a wavelet 1.8 ms wide should result, which is the resolution required to locate faults with a throw of less than two metres. In practice, a wider wavelet may be obtained, due to absorption of the high frequency energy.

The Vibrator Seismic Source (VSS) is presented here in its first application of this new hydraulic powered vibratory source, which operates under different mechanical and electronic control than used heretofore by conventional vibratory sources. The VSS has been developed continuously since 1980 when an initial grant was received from NERDDC. During the intervening years till 1987, two more grants were received (Stewart, 1988).

A recent further NERDDC grant was received in 1988 jointly by ACIRL, Curtin University and University College ADFA for areal coal seam mapping by three-dimensional seismic reflection surveying, with an emphasis on high resolution imaging of faults.

The novelty of the VSS lies in the use of a single flow path for hydraulic oil through the flow stage of a Servo Popper Valve (SPV) (Stewart, 1986). This powers the vibrator by application of the oil to only one side of a piston in the linear actuator which produces the forced output of the vibrator on the surface of the earth. Conventional vibrators use a spool valve to alternately reverse the flow of oil into opposite chambers of a double acting cylinder. Hence the VSS has a fluid power advantage over conventional vibrators and this is evident by better performance at the higher frequencies. The VSS can sweep typically from 50 Hz to 500 Hz, and was initially field tested as a high resolution energy source. Innovations in both mechanical and electronic control systems are presented and results of the initial field trials of the VSS are compared to explosive seismic source results.

¹Trade Mark of CGG ²Mark of Conoco

Coal due to its low conductivity and high electromagnetic contrast with surrounding rocks is an attractive medium for study by ground radar. Results of trials in Australian coal mines show that ground radar can be a useful tool for horizon control, locating old underground workings and mapping geological structure both from the surface and within mine roadways.

Diffraction tomography is an approach to seismic inversion which is analogous to f-k migration. It differs from f-k migration in that it attempts to obtain a more quantitative rather than qualitative image of the Earth's subsurface. Diffraction tomography is based on the generalized projection-slice theorem which relates the scattered wave field to the Fourier spectrum of the scatterer. Factors such as the survey geometry and the source bandwidth determine the data coverage in the spatial Fourier domain which in turn determines the image resolution. Limited view-angles result in regions of the spatial Fourier domain with no data coverage, causing the solution to the tomographic reconstruction problem to be nonunique. The simplistic approach is to assume the missing samples are zero and perform a standard reconstruction but this can result in images with severe artefacts. Additional a priori information can be introduced to the problem in order to reduce the nonuniqueness and increase the stability of the reconstruction. This is the standard approach used in ray tomography but it is not commonly used in diffraction tomography applied to seismic data.

This paper shows the application of diffraction tomography to crosshole and VSP seismic data. Using synthetic data, the effects on image resolution of the survey geometry and the finite source bandwidth are examined and techniques for improving image quality are discussed.

The Tasman Project of Seafloor Magnetotelluric Exploration (TPSME) took place between December 1983 and April 1984 (Filloux era/., 1985; Ferguson era/., 1985; Lilley etal., 1989). Seven magnetotelluric and two (additional) magnetometer sites spanned a range of tectonic features across the Tasman Sea. Initial analysis by Ferguson (1988) indicated large-scale three-dimensional induction effects to be present in the data. It was concluded that the most probable causes were the continental margin effect and changes in bathymetry.

In the present paper, a method is presented of modelling the salt water of the Tasman Sea and adjoining oceans as a thin sheet of variable lateral conductance, which overlies a series of uniform layers representing the solid Earth. The theory and a suitable computer algorithm were developed in a group led by J. T. Weaver at the University of Victoria, B.C., Canada. Many of the features present in the TPSME data are reproduced by this method, and with a greater understanding of induction processes in the ocean which is thus obtained, it is possible to remove three-dimensional effects from observed data. The TPSME data are then solely a measure of the response of the Earth directly beneath the observing sites, and one-dimensional modelling techniques may be used to determine the conductivity structures.

The substantial gravity data base in Tasmania has been used to formulate a regional crustal model. This was derived by array modelling techniques for geological sources of crustal scale. A simultaneous solution for mantle, basement and granite forms was created by this means within a framework of realistic and internally consistent assumptions. The regional field derived from this geological model (including the ocean basins) is not dependent on any filtering or smoothing procedure and thus the magnitude and sign of any residuals is absolute. The residual map was produced by removing the effect of the crustal model at individual data points. The resultant map enables detailed and reliable modelling of upper crustal features as well as revealing crustal character hitherto concealed beneath post Carboniferous cover. An important example of the value of the residual separation is shown by the structural relationships exposed in NE Tasmania which involve gold mineralisation.