No shortcuts to successChris Fleming reports from the vanguard of metallurgy
Chris Fleming reports from the vanguard of metallurgy
By Eavan Moore
December 01, 2013
Chris Fleming has spent much of his 40-year career working on new metallurgical processes and, over this time, he has learned how to properly move projects forward – from the lab testing phase on to large-scale pilot plant testing and commercial plant consulting. Fleming has pioneered gold extraction techniques at Mintek in South Africa, led Lakefield Research in Ontario, and served as vice-president of global metallurgy after Lakefield’s acquisition by SGS. Recently retired from management, Fleming continues his hands-on involvement as a senior metallurgical consultant. He shared the secrets to successful pilot plant design and technology development in an interview with CIM Magazine.
CIM: What advice do you frequently find yourself giving to clients on pilot projects?
Fleming: It is very important to do comprehensive bench-scale laboratory testing prior to piloting. The pilot plant should never be used to develop the flowsheet. The prime purpose of a pilot plant is to prove a flowsheet by operating continuously for a sufficiently long period to bring all streams to a physical and chemical steady state.
Although it can be seen to be self-serving, my most frequent advice to clients is “don’t take any shortcuts.” Metallurgical test work is very expensive, but the cost is trivial compared to the cost of oversizing or undersizing the commercial plant, or slow ramp-up, or metallurgical underperformance due to the negative effect of recycle streams, or poor rheology, et cetera.
CIM: What shortcuts would someone be tempted to take?
Fleming: The most damaging shortcut a mining company can take is not doing enough mineralogical and metallurgical work to fully understand and quantify the variability of their deposit, and the impact of this variability on the metallurgical performance of their commercial plant. This is particularly dangerous in the sizing of mills. Ore hardness and mill throughput can vary by a factor of two or three across a deposit, and if the mill is designed for a constant feed there will be problems. I think if you were to survey the industry, you would see that’s where the biggest problems lie – getting your design throughput through the mill consistently.
CIM: Are there common culprits when small-scale results do not translate as predicted to the larger scale?
Fleming: There are many culprits, but the two most common, and probably most damaging, are failure to properly consider the impact on metallurgical efficiency of species in recycle streams by incorrectly designing and operating the pilot plant, and incorrectly sizing equipment by incorrectly scaling up from the pilot plant.
CIM: How do you do a better job of scaling up? What aspects of pilots are easily scaled up, and which are much more difficult?
Fleming: Mass, water and chemical balances scale up well. Energy balances are more difficult but can be easily quantified from theoretical studies.
The most difficult aspect of scale-up is equipment sizing. Rates of reactions are generally faster in pilot plants because of the higher energy input per unit volume in small reactors. If they were scaled up linearly, the commercial plant would be undersized. So scale-up correction factors have to be applied, and here the temptation may be to be overly conservative, with the result that equipment is oversized.
Also, the importance of understanding slurry rheology cannot be overemphasized. The temptation is to maximize slurry density in a plant and thereby minimize equipment size, but there is an optimum slurry density beyond which mass transfer is negatively impacted.
CIM: What can be learned from pilot plant installations now that could not be learned 10 years ago?
Fleming: I think the greatest advance in the last 10 to 20 years has been in recognizing the importance of understanding ore variability and ensuring the feed to pilot plants properly reflects this variability so that the commercial plant is designed to be able to handle variable feeds.
CIM: The economic downturn has given the industry a new cost-cutting focus. Does SGS have to find ways of cutting its own costs when projects are shelved?
Fleming: We do everything we can to retain our key people during slow periods – metallurgists, technicians and pilot plant operators – and do a pretty good job of that, I think. Running pilot plants properly takes a lot of experienced operators, and it is obviously important to keep these people busy even when we are not running pilot plants. This wasn’t always possible at SGS or Lakefield, and the profitability of the company was quite closely tied to pilot plant activity. But we have a very big metallurgical group now, and our pilot plant folk can be gainfully utilized in other activities in between pilot plants.
CIM: What are the most promising developments you see in hydrometallurgy?
Fleming: The successful and widespread commercialization of pressure oxidation has been the most significant advance for hydrometallurgy in the 40 or so years since the commercialization of copper solvent extraction. High acid pressure leaching of laterites has had a more checkered start but the process works well. Bacterial oxidation of sulphides has had a slower start than pressure leaching, but there are several commercial plants in the world. The current widespread interest in developing rare earth resources in the Western world has initiated a flurry of hydrometallurgical testing activity in met labs, but the technical challenges are enormous and we are still quite a long way from producing individual rare earth metals by solvent extraction outside of China.
CIM: What contributions to metallurgy are you proudest of?
Fleming: The work I did on modelling the carbon-in-pulp (CIP) and carbon-in-leach (CIL) processes with Mike Nicol in the 1980s, when we were both at Mintek in South Africa, was important. The CIP process wasn’t well understood at that time and had never been modelled before. Although our model was superseded by better, more mathematically rigorous models, ours is still widely quoted and used today because of its simplicity and relevance.
Since immigrating to Canada in 1990, I think my most important team contributions have been in the early developments of the thiosulphate leaching process for refractory preg-robbing gold ores, the development of processes to recover cyanide from gold plant tailings (Augment, Hannah and SART) and the development of a process to treat low grade base metal/PGM concentrates (Platsol).
CIM: What are the most important lessons you have learned?
Fleming: An early lesson and reality check for me was the fact that mining companies and the investment community don’t like new technology. Of the cyanide recovery processes I have helped develop, only SART has found commercial application, and there is still no plant using the Platsol process.
For rapid and successful commercial implementation, a new process has to be simple to understand, easy and safe to operate, mechanically robust and tolerant of feed changes and, most important of all, it must be significantly better than a well-understood conventional process. CIP and CIL qualified on all counts, and we have seen the result.
CIM: Do the promising processes that you have mentioned meet those criteria?
Fleming: The first requirement is that there is a real need, and a real need arises because conventional technology doesn’t do the job anymore. In the case of pressure oxidation and bacterial leaching, the need arose because more and more gold ore bodies are refractory in nature. The processes are not simple, but they’re relatively easily understood and relatively easily engineered to make robust operations.
There’s undoubtedly a need to develop rare earth resources outside of China. But I’m not sure that many of the companies trying do it realize just how difficult the task is that they are getting into. We’ve been doing a lot of testing work here, and I’m sure other laboratories have as well. All I know is that it’s extremely complex.
Combining geostats, AI and VR to find gold deposits
Jean-Philippe Paiement explains how SGS has combined historical geological data with machine learning and virtual reality tools to identify and visualize drilling targets