Exploration Geophysics - Volume 32, Issue 3-4, 2001

Forward modelling of potential field data, combined with new geological mapping and deep seismic reflection transects acquired by the Australian Geodynamics Cooperative Research Centre (AGCRC) and New South Wales Department of Mineral Resources, has led to iterative testing of models of crustal architecture of the eastern Lachlan Orogen in New South Wales. This integrated analysis has led to new conclusions about the subsurface that are unlikely to be deduced solely from any of the individual data sets used.

Conclusions supported by the consideration of these data include:

• Presence of lower crust in the eastern Lachlan Orogen, characterised by higher than average crustal density, high P- wave velocities, and repeated, stacked bands of strong reflectivity. This crust is interpreted to be a stacked pile of metaturbidites and modified oceanic crust (greenstones).
• Presence of large volumes of Ordovician volcanic rocks underlying many areas of Silurian-Devonian basin rocks.
• Evidence for extensive, deep-cutting blind thrust faults and detachments throughout the crustal section. Major movements on these faults during the early Silurian appear to have significantly thickened the whole crust.
• Evidence for many high-level upper crustal slivers, mostly formed during the Carboniferous.
• Differences between the western Ordovician Junee-Narromine Volcanic Belt and the eastern Ordovician Molong Volcanic Belt. The former is quite dense, and is inferred to have a large volume of lavas and intrusive rocks. Its structural style is predominantly that of an imbricate stack around a deeper-rooted core. The latter has lower bulk density, and a higher volume of volcanicla tic material. It is now entirely composed of thin, imbricate slices. These differences suggest that the eastern belt may be the rifted off forearc or apron of the western belt which may be the original magmatic centre.
• Evidence for different styles of granite intrusion, reflected in different intrusive geometry of Silurian, Devonian and Carboniferous granites.

A model-based approach for quantifying post-stack seismic amplitudes was conducted for the Alpha and Omega prospects within the Sampang PSC, East Java, Indonesia.

Deterministic tuning curves were generated by analysing the effects of porosity variations, fluid content and reservoir thickness on the seismic wavelet for various siliciclastic and limestone units for the prospects. The tuning curves generated are able to assist with prediction of reservoir parameters at undrilled prospects.

The results of this modelling study suggest the high amplitude seismic anomalies observed over the Alpha and Omega prospects are bright in response to the presence of gas.

An AVO modelling scheme is proposed in which we create a 3D volume of modelled CDP gathers by varying two physical parameters, one in the in-line direction, and one in the cross-line direction. This 3D volume is then processed using conventional AVO analysis techniques, and the results are interpreted using time or structure slices. Three illustrative examples are presented, the first involving P-wave velocity change versus S-wave velocity change, the second involving P-wave velocity change versus thickness change, and the third involving porosity versus water saturation change.

The Elang formation produces a weak event which is virtually impossible to pick on conventionally processed seismic sections. Its top is difficult to define on wireline logs. A rigorous approachto modelling its reflectivity is as a superposition of reflections from a sequence of layers at ≤ 2 m thickness. This formation response verifies that Amplitude Variation with Offset (AVO) can be derived for interfaces near the top of the Elang if Vp, Vs and ρ are averaged over at least 10m above and 10 m below this interface. Using a depth of 3264 m for Top Elang this method predicts a polarity reversal at ray parameter, p = 0.105 s/km (about mid-streamer).

However, at near-to-mid streamer offsets (p < 0.105 s/km) the reflectivity is too sensitive to depth to derive an AVO representative of the Elang. At Laminaria East, the full depth model predicts a peak/trough wavelet at the TWT near top Elang (3215 ms) which is grossly influenced by the sidelobes of boundaries beneath the reservoir.

At far offsets (p > 0.105 s/km) a consistent trough develops in the τ-p formation response which extends out to an equivalent 8 km source/receiver offset. This trough is clear on the full model seismogram when multiples and guided waves are suppressed. It may provide the basis for defining a more interpretable signature for the Elang by weighting far offsets traces and perhaps acquiring data out to very long offsets.

In this study we investigate variations of P-wave reflection coefficients with offset (AVO) and with azimuth (AVA) for a transversely isotropic medium with a horizontal axis of symmetry (HTI) in contact with water, and also with another HTI medium. Aligned, equidistant, vertical fractures, inbedded in an isotropic background, were used to represent an HTI medium. To grasp the effect of fractures their contribution was computed separately and then added to the isotropic background. The total response of a medium was then obtained by adding the two.

An approximate equation is used to compute the partitioning of seismic energy at layer interfaces. The solution compares well with the results obtained in physical modelling experiments. To analyse AVO and AVA effects we used 3-D multi-azimuth transmission and reflection surveys. The results show that AVA effects are significant even at moderate angles whilst they are dominant at large incident angles. Routine 2-D AVO analysis, as commonly performed, is clearly unsuitable for HTI media and should be preceded by AVA analysis.

The Buffalo Oil Field is located in Production Licences WA-19-L and WA-21-L in the Northern Bonaparte Basin, approximately 300 km off the coast of northernWestern Australia. The discovery well, Buffalo-1, was drilled in September 1996 and encountered a 45 metre oil column within sandstone reservoirs of the Callovian Elang Formation.

The Top Elang boundary is typically very difficult to interpret on seismic data. Modelling has shown this boundary to have a Class IIAVO response, i.e. where a polarity reversal of the seismic event occurs across the CMP gather, resulting in attenuation of the event during stacking.

In addition to this fundamental geophysical problem, the Buffalo field underlies a seafloor carbonate bank (‘Big Bank’), which rises from the seafloor, 300 m below sea level, up to a depth of 27 m below sea level. Interpretation of the underlying geology on conventionally processed seismic data is made very difficult by severe ray-path distortion from the steep sides of this bank and the large velocity contrast with the surrounding water. At top reservoir level the very poor seismic data quality is a consequence of poor signal penetration, poor reflectivity, faulting, multiples and severe imaging problems, which also give the reservoir section the appearance of being highly faulted.

A 3D seismic survey was acquired in 1996 over the field, and a vast improvement in data quality and reliability was obtained through advanced processing techniques incorporating Wave Equation Datumming (WED). The remaining depth-structure uncertainty was quantified by a geological-model-based interpretation and mapping process to provide a range of realistic possible structural outcomes. A key assumption - that velocity changes across the field would be a relatively minor contributor to the depth uncertainty - was confirmed by development drilling results, which demonstrated the success of the WED method in removing most of the undesirable effects of the carbonate bank.

This paper focuses on the WED processing and interpretation process which reduced the uncertainty of oil-in-place calculations for Buffalo to an acceptable level and allowed development of the field to proceed without the need for further appraisal drilling.

The 3D data set over the Elang Field, Northern Bonaparte Basin, was reprocessed to provide greater confidence in mapping the structural configuration of the field and to investigate infield development opportunities. The original (1994) data set, which was acquired and processed in the inline (strike) direction, suffered from poor resolution and reflection coherency and multiple contamination at the reservoir level.

The reprocessing strategy included rebinning and horizon velocity analysis (HVA) in the dip direction followed by radon demultiple applied close to the HVA velocities. The result was a significant improvement in resolution of reservoir horizons and faults. Interpretation of the reprocessed data yielded a simpler structural picture of the Elang Field with the crestal area interpreted as a simple rollover with little or no erosion of the reservoir section. The fault interpretation, aided by edge detection and coherency cube results, differs from the original 3D interpretation and is now characterized by a simple east-west fault pattern at the reservoir level.

Reservoir mapping confirmed significant attic oil potential in the vicinity of Elang-1. This potential was tested in January 2000 by Elang-1/ST1 which intersected the reservoir 28 m up dip of Elang-1 and close to the prognosed depth. Production from the well has exceeded pre-drill predictions and added substantial incremental reserves to the field. This has increased value to the project and extended the life of the Elang production facility.

A three dimensional (3D) seismic refraction survey carried out across a shear zone shows that there is an increase in the depth of weathering and a decrease in seismic velocity in the sub-weathering associated with the shear zone. Although the shear zone is generally considered to be a two dimensional (2D) feature, the significant lateral variations in both depths to, and seismic velocities within the refractor in the cross-line direction, indicate that it is best treated as a 3D target. These variations are not predictable on the basis of a 2D profile recorded earlier.

Qualitative measures of azimuthal anisotropy are obtained from the seismic velocities and the time-depths computed from the traveltime data with the generalised reciprocal method (GRM) algorithms and from the head wave amplitudes. These three methods give similar, consistent results, with the direction of the greater seismic velocity being approximately parallel to the direction of the dominant geological strike over most of the survey area. Conversely, all three methods show that the direction of the greater seismic velocity is approximately orthogonal to the direction of the dominant geological strike in a small region adjacent to the shear zone.

The amplitudes of the refracted signals are approximately proportional to the ratio of the specific acoustic impedances between the upper layer and the refractor, and they provide a convenient and detailed qualitative measure of azimuthal anisotropy or rock fabric. The amplitudes also contain additional useful geological information, although some of the cross-line amplitudes could not be completely explained.

The in-line results show that both accurate refractor depths and seismic velocities can be computed with moderate cross-line offsets, say less than 20 m, of shot points. These results demonstrate that swath shooting with a number of parallel recording lines would be adequate for 3D surveys over targets such as highways, damsites and pipelines. Only a modest increase in the number of shot points over the requirements for the normal 2D program would be required in the cross-line direction to measure azimuthal anisotropy and rock fabric with amplitudes.

The results of this study demonstrate that simple and efficient 3D refraction methods using a GRM approach can provide more useful geological interpretations than would be the case with detailed 2D approaches.

Multi-component OBC data relies on mode conversion from p- wave to s-wave in the subsurface to obtain information about the elastic properties of the earth. Since the energy converted to s- wave is now missing from the p-wave, an alternative to recording OBC multi-component data is to examine p-wave data for the missing energy. The conversion is dependent upon the incident angle and occurs most at wider angles of incidence; thus, estimation of shear wave properties can be achieved by wide angle or long offset AVO analysis. Note that since s-wave velocities are typically less than p-wave velocities, then for a given conversion point, the p-wave energy reaches the surface at a longer offset than the corresponding s-wave. Consequently, if we are to obtain the same information from wide angle AVO as from multi-component recording, we need longer offsets for pure p-wave than for multi- component recording.

A non-linear, wide-angle (including post critical) AVO inversion has been developed that allows elastic properties to be extracted from long offset p-wave data. In order to extract amplitudes at long offsets for this inversion it is necessary to image the data correctly, including correcting for higher order moveout.

In addition, the higher order moveout may itself be inverted to yield additional information about the anisotropy of the sub-surface.

Tomostatics have been recently applied to two types of data sets. The first is from the Middle East region where the near surface layer is exhibiting high-velocity basalt outcrops and the second is a multi-component (4C) OBC data set containing lowvelocity gas clouds. Near-surface velocities estimated from turning-ray tomography provide useful information for geologic interpretation and structural imaging. Tomostatics have shown advantages over traditional refraction statics in regions where (1) no refractors can be easily identified, and (2) high velocity materials (e.g. basalt) are overlaid on top of the low velocity sediments immediately below the topography.

P-wave velocities estimated from turning-ray tomography can be used to calculate the traveltime contours. Correlating P-wave and converted-wave traveltimes would allow us to estimate Vp/Vs ratios, leading to a practical approach for converted-wave longwavelength statics solution.

We have used numerical modelling to improve understanding of seismic reflection in basalt covered regions. Our models are based on hydrocarbon prospects in the Denison Trough (Queensland, Australia).

Reflectivity modelling has been used to assess the influence on reflection events, of a range of model and source parameters. Models that include a single near-surface basalt layer generally result in relatively noise-free reflection signals, provided the basalt is reasonably attenuative. Reflection quality is poorest for models with buried high-velocity basalts, or for multi-layered basalts interspersed with lower-velocity material. Such models result in strong reverberatory noise, apparently propagating between the surface and basalt, or within the multi-layered basalts. In these situations, reflection strength is significantly improved if the source can be positioned below the basalt.

Finite-difference modelling permits analysis of models incorporating lateral variations in basalt geometry. Shot records generated with this approach exhibit basalt-related features seen routinely in real field data. Simple stacking of the finite-difference records indicates that reasonable sections can be obtained in areas of near-surface, or thin, basalts. Poorer stack quality is associated with thicker, buried basalts although deeper reflectors may be imaged by undershooting the basalt

Conventionally, geotechnical information comes from cored drill holes. However, coring is expensive and most boreholes are drilled without, or with very limited coring. Alternative ways of obtaining geotechnical information need to be found. One approach is by geophysical logging, which is carried out routinely at coal mines. It measures various in-situ petrophysical parameters which are usually correlated with rock types and can be used for rock mass characterisation, litho-stratigraphic interpretation, orebody delineation and grade estimation.

This paper presents examples of automated geotechnical characterisation from geophysical logs to identify the key strata responsible for caving behaviours during longwall coal mining at Southern Colliery in Central Queensland. The method is based on computer program ‘LogTrans’. Conventional logs such as density, natural gamma and sonic derived UCS (Uniaxial Compressive Strength) logs and full waveform sonic data are examined. The geotechnical units predicted from the geophysical logs are well matched with the original strata classification and core photographs. The results could assist site geologists, planning and production engineers predict and manage mining conditions on an ongoing basis.

We present results obtained from a microseismic monitoring at Pit 20DU at Moura Mine in central Queensland. The objective of this study was to investigate the feasibility of using the microseismic method to map roof fracturing associated with highwall mining for highwall mining stability assessment. 14 triaxial geophones were installed in 7 boreholes across the highwall bench, covering an area of 300 m by 400 m. The experiment was earned out from Jun 12 to Jul 25, 1999. More than 7,000 events were recorded.

Generally, the recorded events were weak in seismic energy. Many of the events only triggered the nearest one or two geophones. The events were classified into three types on the basis of then’ frequency content and seismic duration. Each type appears to be associated with different fracturing mechanisms.

The first type consisted of two groupings of events. One was located in the main sandstone roof and constrained by existing faults. They may be caused by the release of a localised stress concentration. The second group was dominated by events located in the floor of the DU Seam near the highwall bench. They appeal to be controlled by the fault and floor stress concentration near the highwall face.

The second type of events occurred across the mining area and followed the general sequence of the mining. These are inferred to represent a general ground response to the mining. Events of the third group were located mainly in the immediate roof of the DU Seam. They were found to be associated with minor fracturing near the mine entries.

During the monitoring period, about 30 entries were mined and no significant entry or pit stability problems were encountered. Therefore, the fracturing giving rise to the microseismic activity observed at 20DU did not have an adverse impact on the highwall mining stability.

Borehole to borehole electromagnetic tomography, also known as the Radio Imaging Method (RIM2), has been used at Callide Coal Mine since 1996 for mapping fault structures behind high walls. The RIM2 system has been used at frequencies from 12.5 kHz to 302.5 kHz at borehole separations of 50 to 120 m. A high level of correlation with the logged geology is obtained by using close spaced transmitter and receiver spacings and by the choice of the highest frequency at which signal penetration still produces a good dynamic range. The use of RIM2 was instrumental in improving the geologic model of the structures behind the highwalls at Trap Gully. The RIM2 system has imaged three main structural styles including reverse faulting and monoclinal folding, and horst block faulting with seam offsets of less than 5 m to greater than 20 m. The tomographic images produced have since been confirmed by the use of other geophysical methods and geotechnical drilling.

The resource industry seeks to forecast accurately the likely outcome of an exploration program - in particular minimum and maximum and average size of a success; also the chance of achieving a success.

Commonly such predictions are employed on a project-byproject basis. When a group of projects are combined in a portfolio, a higher level of evaluation is possible by quantifying the correlation coefficients among the projects. The outcome is that the portfolio total is not necessarily the sum of the individual parts. Combining the best individual projects usually does not produce the most efficient (lowest uncertainty) portfolio. Including some ‘risky’ projects (larger uncertainty) can lower the overall uncertainty of the portfolio.

Ranking projects by common measures such as Expected Monetary Value or Expected NPV are not the best approach to building a portfolio.

The approach in this paper draws on financial portfolio theory. This is routinely employed in the financial industry, which faces a similar challenge of forecasting outcomes (from investments). It is suggested that combining the better elements of resource and financial evaluations produce a more definitive prediction of the outcome of an exploration program.

Analysing projects within a portfolio structure reduces the uncertainty in the range of deposit success-case sizes that may be encountered. It allows selection of the most efficient group of projects. This maximises the expected NPV value for the total portfolio, while at the same time minimising the uncertainty in the range of NPVs that could occur.

Furthermore, if projects are selected with low correlation coefficients with each other (‘diversification’), then the chance of obtaining one success is increased, compared to projects that are more positively correlated.