Education, News

How new methods are helping to develop rock solid knowledge

Understanding the rock in a quarry and what can be made with it are fundamental knowledge that can make or break businesses. Steve Franklin, founder of Eltirus, explains the new methods being developed that can help gain a better understanding of both rock type and variability.

Most geological assessment comes from either the mapping of rocks that we can see (either on the surface or in a face) and from drilling (using methods such as chip and diamond core ) holes to obtain samples of rocks below the surface.

Samples (either from the surface or sub-surface) are then subjected to a range of tests to better understand their degree of weathering, composition, and physical characteristics.

One of the primary limitations of any physical sampling is that you are only able to sample a part of the deposit and it can be difficult to ascertain how much sampling is required.

By this, I mean, that while you may be able to determine with some accuracy the characteristics of a rock at each sample point, a geologist then has to determine what is between the sampling points by extrapolation.

In essence, if I know what is at point A and at point B (which is say 250m away), I can predict that what is in between will be of a similar set of qualities.

As you might imagine, this may or not be true and is primarily dependent on how consistent the deposit is – in other words, the more variability within the deposit, the more sampling that would be required to accurately describe it.

For example, you might completely miss say a dyke, simply because no drill hole intercepted it. Just because it is not in the model, doesn’t mean it doesn’t exist. As one geologist famously noted, “the best understanding you have of any deposit is when it is fully extracted.”

So, what other methods exist, particularly ones that can help us to better understand more variable deposits?

Hyperspectral survey

We recently had the opportunity to visit PlotLogic in Brisbane and find out about the ground-breaking work they are doing with hyperspectral survey.

The development of hyperspectral survey dates back to the early 1980s, when NASA started to build the first airborne hyperspectral imagers, such as the Airborne Imaging Spectrometer (AIS) and the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). These sensors demonstrated the potential of hyperspectral imaging for earth observation and space exploration as they could provide detailed spectral information that was not possible with other types of sensors.

A hyperspectral survey is a method of collecting and analysing information from across the electromagnetic spectrum with the purpose of identifying materials that are not visible to the human eye.

Hyperspectral surveys are performed by measuring the absorption of light as a function of wavelength, using imaging spectrometers or hyperspectral imagers. These devices can capture images in many narrow spectral bands, ranging from visible light to near and short-wave infrared light. The resulting images are combined to form a three-dimensional hyperspectral data cube, which can be processed and analysed to extract spectral signatures of different materials.

While hyperspectral survey methods have been known for a while, PlotLogic does something quite different with the technology to produce some amazing results. Their approach is to combine LiDAR and hyperspectral imaging technology with advanced machine learning algorithms to deliver highly accurate ore and material characterisation.

So successful has this technology been in the mining industry, that it enabled BHP Iron Ore to extend the life of one of their major operations by a further five years (approximately 85mt) – a remarkable result in anyone’s language.

By scanning a face and analysing the results with their AI based software, a picture of different material types can be created without drilling or sampling. For example, if you had lower quality or deleterious materials in a face, the PlotLogic system can identify them for you.

The system can also be used to scan drill core or be placed over a conveyor belt to determine material quality in real time.

At this stage, we are still exploring use cases in the aggregate industry, however we believe that the technology may have application in the determination of pyritic or other deleterious materials in hard rock quarries and identification of high silt/clay areas and organics in sand. It may also find an application in the cement industry identifying problem chemistry such as high sodium equivalent materials.

We look forward to continuing to work with PlotLogic to determine what kind of applications there may be in the aggregate, sand and cement industry. They are particularly interested in finding a quarry in the Brisbane area that is willing to participate in field trials – if you are interested, please contact me.

Borehole imaging

Borehole Imagining. Picture: Supplied

One of the other technologies we have had some success with is borehole imaging. By this, I mean running a sensor down an RC (or blast hole) and taking a 360-degree photo of the borehole from top to bottom (or if filled with water, using sonic imaging). This approach can be used as a cost-effective alternative to drilling diamond holes.

Reverse Circulation (RC) or blastholes holes can be drilled, the chips logged, and downhole imaging used to gain a better idea of the structure of the material. In many cases the quality of the imagery is such that geotechnical logging can be conducted from the data.

To date, a primary issue is visualising the data. We have long been waiting for the ability to wrap the imagery around a borehole in our geological modelling software to see the data in a more representative setting.

We have also trialled a combination of downhole logging with measure while drilling data (MWD) to determine what could be gained from it. The combination of imaging with drill rotation speed, feed pressure and other variables can help us gain a more comprehensive understanding of what is being drilled. This said, more work needs to be done to better understand the correlation between these two data sets. For example, just because penetration rate is low does not mean that the material is hard – it could just be that it is actually soft and the driller has slowed down to ensure the bit doesn’t become blocked.

While it is still early days, we expect to see an approach taken whereby blastholes are logged using hyperspectral downhole imaging and combined with measure while drilling performance data to provide near real-time analysis of likely rock type and deleterious materials before the shot is fired.

Further advances

For any that use Propeller (aka Trimble Stratus) you may be aware that there is functionality to add sub-surface designs.

Initially introduced for the construction industry to show underground services, we are having some good results with showing geological models and block models. Stay tuned for more information as we investigate this new functionality. •

For more information, contact

Send this to a friend