De-bottlenecking a diamond treatment plant
The following case illustrates an example of where we helped a large diamond mine identify exactly what was causing their diamond recovery process to run substantially slower than expected.
A large diamond ore treatment plant was designed to treat 1400 tonnes of ore per hour, but it had only averaged about 970 tonnes per hour over the previous 5 years. Mine management wanted to know why this was and what had to be done to fix the problem.
After interviewing various metallurgists and managers at the plant, it became evident that everyone had a different theory as to why the plant was under performing. The truth was, nobody really knew. Given this lack of certainty, our team took a very data centric approach to answer the manager's question. For our analysis, we downloaded detailed data from the plant data historian, which contained data sampled every few minutes over the prior year. Our approach to finding the answer using the data was:
Determine which section of the plant was the "bottleneck" (i.e. the system constraint)
Determine the major factors influencing the bottleneck
Rank the factors in order of importance
Isolate the root causes for each of the causal factors influencing bottleneck performance
After performing the analysis, we determined unambiguously that the process "bottleneck" was a section of the plant that was responsible for "secondary crushing" of the ore. We also determined that plant performance could be improved by between 25% and 37% simply by changing various control system and plant settings, modifying operator behaviours, adding another stockpile and making some minor plant operating software changes. With the exception of the addition of a stockpile, none of these items required any significant capital expenditure. The analysis provided concrete steps that if put in place, would ensure that the plant would achieve its designed operating capacity.