The unique light signatures of high-grade ore (left) and low-grade ore from a clay lens could be the answer to quickly identifying and separating the two | Photo courtesy of Imperial Oil
Despite being known as the oil sands, Alberta’s bitumen-rich ore is more than just plain sand. Rather, the soil is largely a mix of silts, sands and clay,
and the variability can present serious challenges to the surface mine operator. Ron Myers, manager, facilities and environment research at Imperial Oil,
says the variation in the mix being mined can have enormous impact on the extraction plant.
That is where the project led by Benoit Rivard, professor and associate chair of Earth and Atmospheric Sciences at the University of Alberta, offers much
promise. The underlying science is pretty straightforward: shine a very narrow wavelength of light at an ore sample, record the intensity of the light
being reflected, and repeat the process across a range of wavelengths. The result is a spectrographic “fingerprint” of the sample which, Rivard says, “can
be informative in regard to the material makeup,” helping to anticipate and avoid costly interruptions at the extraction stage.
A fine solution
“If you go from what’s a good processing ore to suddenly hitting a clay lens with a very high fines content, and that goes into the plant without any
warning, you could upset extraction to the point where your bitumen recovery percentage goes from the mid-90s to as low as 50 per cent,” Myers explains.
The financial impact of a major recovery drop can be enormous, with no cost-effective ways to address the situation readily available. Myers points out
that the clay lenses or poor processing ores can be small enough to slip through, even when the mine geologists are taking core samples a mere 50 metres
apart. The obvious answer — decreasing the spacing between core samples — becomes prohibitively expensive very quickly.
Rivard says a unique aspect of the oil sands makes the current spectroscopy project possible. “The reason we’re able to see how much there is of each
particle size is that the fines are dominated by a certain type of mineral — clay minerals,” he explains. “The sands are dominated principally by quartz
and feldspars.” Each has a unique signature that allows one to be distinguished from the other and appropriate processing adjustments made.
Rivard says that the researchers are not examining grain sizes directly but are able to extrapolate from the mix of mineral types in the sample. In other
contexts, says Rivard, a single mineral type would be spread across multiple grain sizes, making the technique inapplicable.
Of course it is one thing to show that something works in the lab and entirely another to implement it in the field. What Imperial wants from the project
is a way to provide real-time ore composition and particle size readings from the mine face, says Myers. These can then be used to adjust the processing
conditions at the extraction plant or froth treatment facility, ensuring optimal recovery levels and smooth operation.
Bringing the lab to the mine face
Of course, putting the spectroscopic equipment at the mine site presents a number of significant challenges, says Rivard. The first of these is the limited
time available to examine a given sample. The ore is constantly moving from the dragline to the processing plant and if the sensors cannot keep up with the
pace, the data they produce will arrive too late to do any good. The issue, he explains, is that the less time a sensor spends collecting reflected light
at any given wavelength, the higher the chance of measurement errors creeping in.
The situation is going to be further exacerbated by the uneven nature of the chunks of ore being sampled, Rivard adds. “The ore will be blocky, with
different sizes of blocks. When you shine a light on it, there will be shadows; some parts may be well illuminated while others are in shadow,” all of
which will alter the amount of reflected light and decrease the accuracy of the measurements.