Grinding Optimization Spotlight (Part2)

Learn how you can get real-time feedback on your Mill Feed Dilution with our Hematite ’21 release. 

Hematite is the very latest release of our suite of decision intelligence applications built specifically for the mining industry.  The applications all use Scientific AI that allow you to run your mine and plant at its optimum state at all times. Allowing you to maximise net metal production in the most sustainable, efficient and safe way.

We are so proud of the work our team has completed over the last six months to get this release shipped to our customers that we wanted to share a series of deeper dive spotlights into some of the new product capabilities, this time taking a closer look at another new capability offered in our Grinding Optimization app.

Kevin O'Kane

The Grinding Circuit Optimization Challenge

The operation and performance of Grinding Mills, especially SAG Mills, are significantly influenced by the density of the mill charge. The decision of what feed dilution ratio to target on its feed or how much water to feed into the mill compared to the solids feed is critically important.

The challenge is that most Operators do not have live feedback on whether their water flow rate (or dilution ratio) is producing the right density in the Mill. They often only get feedback on this when they take a sample, after which they adjust their water feed rate or dilution ratio setpoint.

This uncertainty means that mills operate at wrong densities for a long time, limiting the Grinding Circuit’s throughput, negatively impacting grind size and/or unnecessarily driving up the cost of production ($/t).

 

The IntelliSense.io Solution

Our Grinding Application solves this challenge with its Mill Charge Density Virtual Sensor and its accompanying Alert.

The Grinding Application’s Mill Charge and Material Transport Models track the flow of ore, water and slurry through the mill in real-time – taking into account how the ore breaks to mix into the slurry. It uses a combination of first-principle and empirical modelling to provide Virtual Sensors of both the Density and %Solids in the mill’s charge, as well as in the discharge. These give Operators immediate feedback on whether their dilution ratio or water flow rate is correct for the current ore type.

These real-time Virtual Sensors are delivered in brains.app as Alerts, which compares the Virtual Sensors with desired/optimal operating bands (minimum & maximum limits) for the particular mill. Whenever the Mill Charge Density moves outside of these ranges, Operators are alerted to correct their mill feed dilution by adjusting their control system setpoint (or changing the mill feed water directly).

 

The Value Delivered

This direct feedback on mill feed dilution allows Operators to:

  • Make immediate corrections, instead of waiting for the next sample before they are able to change the dilution
  • Run at more stable mill charge densities, leading to increased throughput, a more stable P80 and reduced specific power usage
  • Prevent unnecessary mill stoppages (caused by Mill Overloads)

The below image shows the SAG Discharge %Solids Alert being triggered because the Virtual Sensor shows that this %Solids value has decreased below its minimum value.

This Virtual Sensor can also be used historically by Metallurgists to do root-cause analysis, and see which operating regions/strategies gave the best performance over a given period or for a given ore type.

Mill Feed Dilution

An Operator Dashboard showing the Mill Charge Density Virtual Sensor being activated.

Want to know more? We would be delighted to provide a live demo tailored to your requirements.

The IntelliSense.io Application Portfolio