Managers and engineers responsible for mining operations regularly face tasks including:
- Determining realistic daily and monthly production volumes.
- Testing mine plan feasibility.
- Evaluating the outcome of operational improvements.
- Quantifying return on investment.
- Justifying fleet requirements.
They have to make decisions and commitments while working in a complex mining system, where no component is isolated. Typical constraints of mines include:
- Many interdependencies and overlapping activities.
- Interaction between equipment units.
- Changing cycle times of processes.
- Spatial limitations, yield and give-way logic of moveable equipment units.
- Complex layouts.
- Limited capacity of bunkers and conveyor systems.
How do managers plan mining processes while considering all of the above? Traditionally, they make assumptions that are never close to reality, such as:
- Average haulage distance. The cycle time of haulage is different, point to point and time to time, under the influence of uncontrollable parameters.
- Average dump time. If the system overflows or there are queues in front of the ore passes, then the dump time changes.
- Percentage of time a conveyor is stopped due to overflow. It is never constant, and it cannot be specified by a single number.
Simulation allows mine planners to model processes as they are and get rid of these assumptions.
With the help of Amalgama and one of the Big Four consulting firms, the largest European potash producer simulated the mining process using AnyLogic. A big potash mine is 8 x 8 km in size. 900 thousand tons are mined per month on three underground levels. The mine has 21 km of conveyor belts taking the ore through the system to the skip hoist. The ore is mined with borers, which continuously crush the rock and load it to their attached ore buffers. This ore is then dumped into a dump truck and the dump trucks make runs between borers and ore passes.
20 x (Borer + Dump Truck)
The capacity of an ore pass is three tons, but a dump truck carries 22 tons of ore. After the first three tons, the dumping speed depends on the current load of the conveying system underneath the ore pass. Since the conveyor can already be loaded with ore from other upstream borers, the system may be constrained. To get rid of this constraint, mine planners were going to change the equipment configuration by adding Mobile Ore Loaders, or MOLs. In the TO-BE scenario, MOLs played the role of a buffer between dump trucks and ore passes. The dump truck quickly dumped ore into the MOL and returned to the borer while the MOL continued dumping ore into the conveyor system.
Adding Mobile Ore Loaders
By adding buffering capacity, mine planners hoped to lower the dump truck cycle time. The main question was whether usage of MOLs would allow them to get rid of one borer while keeping the production volume. The borer carried high operational expenses and required having a maintenance team in the mine, so its removal would significantly lower the operational expenses. Additional questions included which borer to remove and where to use the five MOLs.
Amalgama’s simulation developers created an AnyLogic simulation model of the mine to answer these questions. The model included the whole mining process from drilling to hoisting, precisely as the plant was laid out.
The model was very detailed, and all the processes were simulated with minimal simplification, making the model very accurate.
The first experiment with the model was how the mine system would behave if the external constraint of conveyor speed was removed. This experiment helped find three borers that had a production rate limited by internal constraints such as their own performance, maintenance intervals, buffer size, etc.
As a result, three borers were chosen to be candidates for removal with minimum influence on the mine production rate, since MOLs would only remove constraints caused by the speed and capacity of conveyors.
The effect of removing each of these three borers was studied with simulation. These experiments showed that removing borer #65 reduced production the least.
Then several scenarios were run to determine where to send the five MOLs to maximize ore production. Five borers were chosen to send the MOLs to. This scenario showed only a 1.02% decrease in production volume, which was negligible. At the same time, this scenario showed a significant decrease of OPEX, since one borer had been removed from the mine.
The simulation model of the underground mine provided operational improvements to Europe’s largest potash producer. These improvements allowed operational costs to be cut while keeping the same production volume. Once the project was finished, the mining simulation model has been used continuously for monthly production planning, identification of potential process bottlenecks, and the evaluation of proposed changes.
Project presentation by Andrey Malykhanov, Amalgama