Maptek said it was commissioned last year by Aurora Gold to carry out 3D modelling of the Hidden Valley deposit in the Morobe Province, 250km north of Port Moresby. Aurora is the project manager and a joint venture partner in the Morobe project. Maptek worked closely with the site geologists to produce stratigraphic, structural and mineralised envelope models of the deposit to be used for block modelling by Aurora in Perth, Western Australia.

There was no Vulcan user at the pre﷓feasibility project and deadlines left no time to train existing staff. Using a Maptek consultant was therefore the ideal solution.

At Hidden Valley, the Kaindi Metamorphics flanks the Morobe Granodiorite; porphyry dykes intrude both granodiorite and metasediments. Gold mineralisation is controlled by the moderately dipping, sub﷓parallel Hidden Valley and Lost/No Save fault zones, as well as the overlying Kaindi metasediments contact. Most of the gold and silver mineralisation occurs within the granodiorite between the two fault zones.

Vulcan’s power was evident in the efficient manipulation of data. A substantial re﷓logging of all available core and an aggressive infill drilling program resulted in a very detailed dataset that was consistently logged — an enviable situation for a 3D modeller.

All logging codes were reviewed with site geologists to ensure compatibility before importing geological data into Vulcan. The ability to separately identify fault footwall and hanging wall structures led to greater flexibility in 3D modelling downstream. Automatic checks in Vulcan during initial database importing identified minor inconsistencies requiring data modification. Ongoing update and integration of further data was accomplished easily in Vulcan.

Site geologists undertook initial fault interpretation on 2D sections spaced between 12.5m and 50m apart. The footwall and hanging wall were imported into Vulcan and the unique down﷓hole position was tagged with an appropriate structural explanation (such as cleavage or brecciation). The superior visualisation features of Vulcan proved useful for accurate structural modelling.

Each fault was displayed in 3D, and if adjustments were indicated to either the footwall or hanging wall the database was easily altered. Vulcan’s 3D environment allowed several faults or sections of faults to be merged to produce more continuous and therefore more accurate structures.

Timing had previously been identified as an important control on gold and silver mineralisation. Working with the site geologists helped clarify the relative timing of structures, in particular which faults terminated against others. In total, 16 faults were modelled and a further seven smaller structures were identified but not included in the final model.

Mineralised envelopes are cut by both barren and mineralised faults. Rather than create many individual envelopes, it is preferable to model each envelope as a continuous structure, then use priorities later in block modelling to ensure cross﷓cutting relationships are logical.

Once the faults were interpreted, 24 mineralised envelope polygons (of more than 1 gram per tonne gold) were created on 2D sections, digitised by snapping to the drillholes and then modelled. Polygon end points were amended in Vulcan to hold them in the intended plane when converting from 2D to 3D, where calculation of accurate volumes was ensured. A simple unconstrained block model was created before modelling the mineralised envelopes. Grade shells helped identify possible grade domains. Using unconstrained block models, the grade shells may reflect the drillhole traces, and are therefore suitable for first﷓order interpretation only. The distribution of grades indicated that some faults may be domains constraining mineralisation.

The result of the pre﷓feasibility study increased the grade at Hidden Valley by 25% (with a reduction in tonnes of 25%). The Morobe total insitu mineral resource at the end of September, 1999, above 1pt gold, was 60.5 million tonnes at 2.3gpt gold and 28gpt silver.