Upon his arrival in 1997 at Falconbridge Ltd. (today Xstrata Process Support) as superintendent of mineral processing, Norman O. Lotter was given an open
opportunity when it came to improving the efficiency of existing concentrator operations. Fast forward ten years, Lotter and his team successfully adapted
the existing high-confidence flotation testing (HCFT) into a hybrid approach with complementary technologies. This providing a platform to quantify the
full potential of an ore body in a concentrator and to specify an optimum flowsheet that would deliver this potential. As a member of this season’s CIM
Distinguished Lecturers Program, this leader in process mineralogy guides us through the learning process behind the research of “Modern flowsheeting
CIM: What was your approach to improving concentrator efficiency?
Lotter: I put together two five-year plans where one would build upon the other. The first plan (1997-2002) was aimed at the dimension of time; in other
words, building the process mineralogy team to address performance improvements for existing concentrators that were in production. The second step
(2002-2007) focused on the dimension of space — looking at the entire ore body from a more predictive point of view, using modern processing mineralogy.
This naturally resulted in working with viable new mine projects, and this is where the much larger value was realized.
CIM: How did adapting the existing HCFT lead to an improvement in efficiency?
Lotter: HCFT originated in South Africa in the late eighties and early nineties, and was based on platinum mineral processing. The adaptations we made
required establishing different sampling requirements and modifying existing sample preparation protocols. We embedded this approach into a standard
practice. This means starting from representative sample material, and then identifying and reducing measurement errors in flotation testing so that you
will have a 95 per cent level of confidence in the metallurgical results that you report.
This approach also ensures proper characterization of the ore body from true samples in a manner that easily identifies the processing implications. In
particular, it captures variations in the ore body and when properly translated into the flowsheet design, minimizes the risk that unexpected responses
will develop in concentrator operations.
CIM: How do you capture variations in the ore body?
Lotter: We achieve this by analyzing and testing several drill core samples from a mineral resource. Individual assays of metal grade in the drill core
will provide some idea of bulk ore grade, but it is insufficient for predicting the metallurgical performance of the ore resource. By stratifying the
sampling of drill core, more accurate sampling equations result and, consequently, a more representative sample of ore resource will be achieved. The
flotation testing of several samples of drill core now adds measurements that quantify that variation. In a new mine, we absolutely prefer to work from
drill core as the sample material.
CIM: Why is working from drill core a better approach?
Lotter: In taking the conventional approach when sinking a pilot shaft down the middle of the ore body, you are essentially extracting a couple of hundred
tons of ore from that particular point. You have no guarantee that the rest of the ore body has similar characteristics. But the drill core comes from
holes through the full space of the ore, across a grid. They expose the full variation of the ore far better than the pilot shaft. Another drawback to the
pilot shaft is that it is more costly and time consuming. Drill core sampling will save millions of dollars and at least one year to 18 months in the
project’s timeline. Drill core is produced in the normal course of exploration activities; all we ask is that the geologists receive additional funding and
schedule to do a little bit more drilling — it’s a marginal cost compared to the conventional practice.
CIM: What do you think are the best practices in variability testing?
Lotter: The best practice of process mineralogy will first look at drill core testing and divide it into Geomet Units (ore types). We will then
characterize samples according to these units. This is one of the key pieces in modern process mineralogy, in that the characterization of the Geomet Unit
is not only geological and mineralogical, it is also metallurgical — part of that description derives from flotation tests. Characterization of these units
will result in very accurate mineral processing prediction. This also helps in identifying problematic ores. Run-of-mine sampling identifies the existence
of a metallurgical problem, but characterization of the individual Geomet Unit will help you pin down the culprit.
CIM: You mentioned earlier that the first step in your plan involved building a process mineralogy team. What was in place before then?
Lotter: After a couple of months of visiting many of Falconbridge’s operations and projects in 1997, it was clear to me that there was an opportunity to
demonstrate the value of process mineralogy. Geologists and mineral processors were (traditionally) regarded as two separate disciplines. In addressing the
specific needs of the projects, this convention needed to be changed to bring out the full flowsheeting needs. This was achieved by appointing a small
multi-disciplined team in the technology centre at Sudbury.
CIM: How did creating a cohesive team benefit the operation?
Lotter: There was a fair amount of on-the-job cross-training between mineral processors and geologists. It’s an approach that was put together with a
deliberate vision. It encouraged representatives of both groups to work and talk together and, quite naturally, cross-disciplinary information flowed.
People have inquiring minds and want to make a contribution. This exchange helped us to ask the right questions — a key step at the beginning of the
process, leading us to success. We became a balanced team; a strong team with different skills that worked together very easily.