It can be challenging to maintain high throughput in gold processing operations. There are many stages in the operations that can hinder the flow of ore through the system, and each presents its own challenges. The first step to improving system performance is identifying actual constraints and understanding the reasons that constraints occur. Here, we outline the most common causes of throughput bottlenecks. We also discuss how AI and other technologies can address those causes.
Variability in ore hardness, along with mineral and clay content, can affect both the downstream grinding and the recovery processes. Clay bearing ore, for example, can be problematic when fed to either crushers or tumblers, producing slimes which can destabilize flotation circuits and lead to decreased throughput.
Typically, variability in ore and/or mineral content requires operators to make many decisions reactively due to the lack of proper monitoring or predictive capabilities. Knowledge of the characteristics of constituent minerals, and how they impact processing sequentially, is prerequisite to having a more stable throughput. Source: 911Metallurgist
Flaws in processing are more common in the handling of refractory or complex ores. Inadequacies such as the under sizing of grinding mills, insufficient tank volume, or excessive recirculation can lead to constraining bottlenecks.
Due to the high costs and the significant operational disruptions that can be caused when a salient flaw in design is addressed, process execution or operational optimizations are typically used to accommodate design constraints.
While some processing bottlenecks are caused by fixed system constraints, others are dynamic.
Ore characteristics, feed blends, and operating conditions can continuously change the limiting equipment in your system.
The primary crusher is the first equipment where the system can be limited in some cases. There are cases where the grinding circuit, thickener, or flotation section can limit the system. Dynamic bottlenecks can be hard to keep track of when there is a lack of operational data, as well as analytics that can show changing process limitations.
Source: MinAssist
Many bottleneck issues can be identified prior to production. There is commonly a lack of geological or metallurgical data in many feasibility studies. In these cases, throughput assumptions can be unrealistically optimistic.
When the ore variability is not accounted for, the processing plants underperform the expectations. This leads to poor mine planning, sub-optimal operational efficiency, and poor returns.
AI in the planning domain can incorporate new operational and geological conditions when planning production.
Source: ResearchGate
Limited visibility across multiple processing stages can lead to undetected bottleneck issues, even causing production to be impacted.
Tools that employ digital twins of industrial processes with real-time monitoring and AI for further augment process visibility. Mining teams can use these tools to understand how a process behaves under various conditions to identify constraints and adjust the process to avoid loss in throughput.
Source: Simio
NTWIST uses AI diagnostics and advanced process modeling to provide clients with alerts to throughput constraints.
The operational data we process can be used to circumvent ore variability, bottlenecks, equipment constraints, and plant design limitations. Our data is unique in that it is used to create answers.
Track throughput trends to spot and respond to challenges earlier. Maximize the potential of your plant with NTWIST’s AI-driven operational tools.
Bottlenecks limit not only production but also the efficiency of actual work done and how well resources are utilized. Addressing the true nature of the limitations of a process and using advanced technologies, the throughput of a process in mining can be locked in to provide greater value over time.
A bottleneck in a gold processing plant is any process, piece of equipment, or operational constraint that limits the throughput of the plant.
Ore variability (for example, changes in the hardness of the ore or the mineralogy of the ore, as well as the clay content of the ore) can negatively affect the grinding efficiency and recovery, among other delays in the processing.
Bottlenecks can occur from crushers, grinding mills, thickeners, flotation cells, and material handling equipment and operations.
AI can scan real-time operational data to help predict, identify, and solve the limitations of throughput and suggest ways in which a plant can be better utilized.
Real-time data enables workers to see emerging bottlenecks and other processing limitations before they negatively impact production as well as help optimize process controls.
Yes. Under sizing equipment, inadequate processing capacity, and poorly designed processing flows can create permanent processing limitations that negatively impact processing.
References911Metallurgist. (n.d.). Gold Extraction & Recovery Processes. Retrieved from https://www.911metallurgist.com/blog/gold-extraction-recovery-processes/
McKinsey & Company. (n.d.). Refractory gold ores: Challenges and opportunities for a key source of growth. Retrieved from https://www.mckinsey.com/industries/metals-and-mining/our-insights/refractory-gold-ores-challenges-and-opportunities-for-a-key-source-of-growth
MinAssist. (n.d.). 10 areas where money is lost in mineral process operations. Retrieved from https://minassist.com.au/10-areas-where-money-is-lost-in-mineral-process-operations/
ResearchGate. (n.d.). The Hidden Flaw in Feasibility Studies: Why Mining Projects Underestimate Throughput Risk. Retrieved from https://www.researchgate.net/publication/389143417_The_Hidden_Flaw_in_Feasibility_Studies_Why_Mining_Projects_Underestimate_Throughput_Risk
Simio. (n.d.). The Benefits of Process Digital Twin Technology in the Mining Industry. Retrieved from https://www.simio.com/benefits-process-digital-twin-technology-mining-industry/