Drilling and sampling: learning the orebody’s language

Explorers and miners have an array of different drilling techniques at their disposal. There are differences, however, in the information provided.

How the information provided can be used throughout the course of an exploration program, in the delineation of resources or production grade control, can have a marked impact on project timing and the need for additional drilling as a project progresses.

Exploration and resource evaluation are, first and foremost, the process of collecting pertinent, geological observations that may be used to determine how a deposit was formed, the character of mineralisation and where extensions or repetitions of identified resources may exist. Observations should be unbiased and unconstrained by preconceptions of how a discovery may be ultimately mined. Understanding the controls on mineralisation development and distribution is critical.

I like to think of mineralisation being able to be classified, from a primary, structural viewpoint, as falling into two discrete groups:

1. Structured mineralisation, where the controls on mineralisation distribution may be observed or reliably interpreted from mesoscale (drill core, hand specimen or outcrop scale) observations; and,
2. Unstructured mineralisation, where mesoscale observations do not provide a clear basis for understanding mineralisation distribution.

The simplest example of structured mineralisation is, arguably, coal, probably followed by evaporites that occur along palaeosurfaces. The material of economic interest occurs within a stratigraphic sequence where other components have little or no economic value. Stratiform basee metal mineralisation is another good example where bedding has a profound impact on mineralisation distribution. Not all structured mineralisation is controlled by bedding: far from it. Secondary structures, including cleavage and faults, can impose tight controls on mineralisation distribution, as can alteration which may be both a product of the mineralising process itself or responsible for creating distinct geochemical and physical conditions under which mineralisation develops.

Unstructured mineralisation may include some styles of gold mineralisation, but that may be more a function of gold not being visible, but other associated minerals define different styles of mineralisation. Porphyry Cu-Au mineralisation is probably also an example of unstructured mineralisation due to the boundaries between mineralogically distinct parts of the mineralised system being diffuse and gradational rather than sharp and distinct.

Diamonds are an interesting case where the problem of sampling sparsely distributed minerals becomes an overriding concern. In my view, more diamondiferous kimberlites are examples of structured, rather than unstructured mineralisation due to significant grade variation occurring between mineralogically and texturally distinct rock types in the volcanic system in which they occur, which has profound implications for collecting representative samples.

Ask the question “can I put my finger on a point at which the lithology, mineralogy or fabric of the rock and the mineralisation it contains changes”? If the answer is yes, the mineralisation you’re dealing with should be considered to be structured.

What does this mean for drilling and sampling?

Structured mineralisation needs to be sampled to the visible domain (lithological, mineralogical, structural) domain boundaries. Failure to honour these boundaries results in materials with potentially different characteristics being mixed. This could adversely affect the determination of their actual grade, mineralogy, metallurgical and even rock mass (geotechnical and hydrogeological) characteristics. The simplest impact is that samples may incorporate mineralisation and dilution.

Is this something I should be concerned about?

Yes. Dilution and ore loss form part of the suite of modifying parameters applied in converting Mineral Resources to Ore Reserves. Samples collected on a regular interval basis, for example, from RC or other forms of non-cored drilling, could contain significant proportions of unmineralised material. Defining boundaries for geological modelling become based on analytical data (grades or hyperspectral mineralogy) which will almost invariably contain diluting material which will reduce the grade obtained from samples or vary the concentrations of deleterious elements or minerals evident in analytical data.

These issues will be further compounded when dilution and ore loss factors are applied during even conceptual mining studies if the project gets that far with a pessimistic suite of grades. Potential for selective mining may go unrecognised, again with potentially significant implications for a project. Waste rocks adjacent to mineralisation, similarly, may be incorrectly classified as requiring more onerous environmental management and remediation than is actually required.

These impacts are less evident when dealing with unstructured mineralisation which needs to be interpreted solely using analytical data, and may ultimately be mined using similar data. Mining of structured mineralisation may be controlled visually, improving mining efficiency, reducing dilution and ore loss and minimising costs.

I prefer to use RC drilling because it’s lower cost per metre and faster than diamond drilling. These are both significant factors but they may not take account of the value of forgone data, including reliable grades, representative mineralogical information and the need to undertake specialist metallurgical and geotechnical drilling, when this data could be obtained more systematically and routinely during the course of initial resource evaluation. The lost opportunity involved could increase the amount of drilling required and resource risk inherent in the project, and the time required to collect data required for different levels of study. In whole project terms, cost savings may not be realised.

Understanding how mineralisation is structured also contributes to an understanding of resource continuity, which is arguably the single most important factor in geological modelling and resource evaluation.

Data Quality

Effective sampling underpins and helps to ensure data quality, essential in the minimisation of resource risk. Data quality assurance requirements extend beyond assays, to bulk density data, understanding rock mass characteristics and variations in these characteristics, mineralogy, other physical characteristics of all rock materials that could be disturbed or affected by mining, and sample locations in three-dimensional space.

The quality and consistency of geological logging is also very important. The issue is achieving consistency between geologists who will, naturally, place different emphasis on different attributes, based on their previous experience. A structured approach to logging helps to achieve consistency by prompting geologists to provide specific observations in their descriptions. A structured logging system does not need to use codes. Significant advances in text search and interpretation software have been made which assist in making logs more readily interpreted for computer-based analysis and geological modelling. Coded logging systems need to be carefully designed, usually for specific prospects, so that they do not attempt to deal with the plethora of rock types, mineralisation and alteration assemblages and styles evident in mineral deposits, which invariably reflect a complex combination of geological conditions.

All geological software features rules that may be applied to help to ensure the basic integrity of drilling data. A database with integrity, however, may be far from validated. Validation is achieved by putting data to use: viewing new drill holes in the context of previous, surrounding drilling that will identify inconsistencies beyond those covered by simple integrity checks that need to be further investigated.

There is also a need to distinguish between observations and interpretations. Competently recorded observations are facts that should only be changed after a thorough review of samples or core, either directly or with the aid of high quality sample and core photographs. Observations, for example, include downhole depths, lithology, mineralogy, grainsize, texture, structural features and discontinuities. Interpretations include stratigraphy and geological domains, which will (and should be expected to) evolve as knowledge of the mineralisation being investigated develops.

Technologies

There are a few, readily available, robust technologies that are available to explorers and miners to maximise the value of drilling data:

• As mentioned above, flexible text search capabilities may be used to prepare logs for further analysis or rapidly identify features of interest recorded in logs.
• Accurate drill hole surveying. Drill hole locations should be surveyed. Simple, handheld GPS data may be adequate for initial, reconnaissance drill holes, which can always be more accurately surveyed later. Drill hole coordinates and elevations collected by a licensed surveyor are a means of assessing the reliability of terrain models that will become important in resource evaluation.
• Reliable down-hole surveying is just as important in determining where mineralisation sampled by drilling is located. Every drill hole deviates to some extent.

• Collection of oriented core should be considered essential. Understanding both the location and orientation of geological features removes an entire layer of uncertainty in geological interpretations. Reliable, effective tools have been collected that can be included in stored core as a permanent record of the observations collected. It’s perfectly possible to collect reliable orientation data from both inclined and vertical drill holes.
• Use a drilling method that maximises sample integrity to minimise sample or core loss. Record core and RC sample recovery. The use of triple-tube drilling helps to minimise core disturbance during both drilling and core presentation, improving the quality of samples and fundamental geotechnical data collected from core in the field.

• Collect high quality core photographs. Lighting is everything. Photographs of wet core should be free of reflections, especially where use of computer-based machine learning techniques is being considered for the collection of lithological and rock mass quality information.
• My view is that high quality core or non-cored drilling sample photographs are essential as drilled materials begin to deteriorate as soon as samples are recovered and should not be considered to be a permanent reference.
• Image resolution is a secondary consideration to producing images free of reflections, removal of which can become the greatest contributor to the cost of applying these techniques.

The advent of hyperspectral mineralogical data collection has been game-changing in recent years by improving the consistency of geological data, as well as providing comprehensive mineralogical data that may be used to classify mineralisation and associated rocks for resource evaluation, metallurgical and environmental management and remediation purposes. Good data may be collected by all analytical laboratories from coarse materials during sample preparation at low cost.

Down-hole geophysical logging offers quantitative collection of important data including structure (acoustic televiewer, microresistivity and dipmeter logs are all proven technologies), bulk density, hydrogen density (which is a proxy for moisture content), mineralogical variations, grades (in an increasing range of settings), and rock mass properties. The minerals industry has been slow to adopt technologies that are critical tools for minimising resource risk in oil and gas exploration and reservoir characterisation and in other energy resource evaluation applications, including coal and geothermal energy which employ similar drilling techniques to the minerals sector.

An important feature of a number of the technologies and techniques mentioned above is that they can, partly, bridge the gap between data provided by non-cored versus cored drilling.

Planning is key.

Exploration, in particular, is a high technical risk activity but that should not preclude planning for success, deciding when to decide to collect data that will be needed to take a project through exploration to studies focusing on resource development, which should be early in the course of the project.

Explorers frequently take the approach of, following a promising intersection (which establishes a project of merit), stepping out to test for an extension to the revealed mineralisation, and gradually infilling the project area over time. Projects of merit gather momentum quickly by attracting additional attention and resources, which may create inertia in project management if an effective plan is not in place.

A potentially productive response to a project being recognised as having merit is to drill to assess mineralisation continuity at an early stage. This provides detailed information on geological and mineralisation variability that, in turn, helps to determine the level of drilling required to achieve desired levels of confidence and progress the project efficiently. Surprisingly few explorers do this.

Does this really matter?

There are surprisingly few studies in the public domain of the benefits of this approach. Two anonymous examples follow.

A company was evaluating a shale-hosted, stratiform base metals project using predominantly RC drilling. The sentiment was that they were dealing with a potentially large resource but scoping studies showed convincingly that the grades were marginal for open cut mining. Adopting cored drilling (HQ3 core) with good, routine core photography and down-hole geophysical logging allowed drilling to be oriented more parallel to bedding (the deposit was actually folded), collection of basic geotechnical data to replace assumptions and acquisition of comprehensive, quantitative bulk density data. Resistivity and magnetic susceptibility logging clearly defined strong lithostratigraphic controls on mineralisation development. Recoverable grades increased by one-third when sampling was changed from regular one-metre samples downhole to lithologically based sampling. The geophysical density logs showed that laboratory measurements were highly conservative. Specific gravity was being measured on pulverised samples using volumetric glassware, but consistent results are hard to achieve in a high throughput, production focussed environment like a commercial assay lab being put under pressure on both costs and turnaround. Changing sampling protocols to honour controls on mineralisaion resulted in the project’s NPV more than doubling, dispelling doubt regarding the economic feasibility of the project. This also formed the foundation for a re-assessment of selective mining options that ultimately led to a ten-fold increase in NPV.

Studies of a structurally controlled, moderately to steeply dipping, lode-style, zinc-lead-silver deposit used routinely collected, basic, rock mass properties observations (RQD and visually estimated rock strength) to better understand variations in rock mass properties throughout the deposit. Oriented core was able to be used to demonstrate that a thickened, apparently moderately dipping section of the deposit was an artefact produced by reverse faulting, which was able to be incorporated in the deposit’s geological model due to oriented core demonstrating a flattening of the deposit being unlikely. There were concerns that intense shearing in the mineralised zone hangingwall would contribute to excessive dilution during mining. Collection of samples using both mineralisation style and fracture intensity, however, enabled areas where a pillar could be left between stopes and the sheared ground enabled this risk to be effectively managed and contrinuted to a decision to develop the project. Interestingly, the analysis of data that contributed to this decision was made by a new owner of the project, after the former owner that collected the data, decided to divest rather than proceed with the development of the project.

Last Words

There is a strong case for treating arguably any project, and certainly, one that has reached project of merit status as a case where we need to plan for success. Think carefully about the observations of mineralisation and the enclosing host rocks acquired during drilling and what data needs to be collected to describe them in an effective, objective manner that facilitates examination of all potentially feasible development options. Scoping studies have an important role in demonstrating a project’s potential. Pre-feasibility studies provide a platform for the assessment of multiple, potential development options, the best of which then form the basis for feasibility studies.

A former colleague conied the term re-feasibility study to describe the too frequent need to repeat pre-feasibility studies, even after proceeding to feasibility and realising that the option being assessed was demonstrably not the best option for the project. There are cases where this is a consequence of advances in mining and metllurgical technologies or quantum shifts in the market for the commodities to be produced. There are far more cases, however, where this is a consequence of feasibility studies becoming too focussed on limited options too early. The cost of this is significant in terms of capital invested in studies and the need for more, study team morale and investor confidence, which are important but often overlooked considerations. Good planning from the outset provides a means of avoiding this.

We frequently hear and read expressions like “the voice of the orebody” and ”the rocks speak” which are pretty good metsphors, generally, for the message presented here.

I’d go one step further. The rocks speak, but we need to learn their language and listen”. There is a conversation to be had that will benefit our profession.

We work in an environment where, usually, exploration and mining tenure provides mineral resource asset security, which should facilitate more open sharing and discussion of situations like these, to the benefit in terms of both technical competence and much-needed investor public confidence in resource stewardship by our profession and industry.