GEOVIA Whittle: Simultaneous optimization | Part 3

Advanced SIMO module enables mine planners to create optimal long term schedules for the operation of open pit mines.1 Advanced SIMO uses the ProberB engine that was developed by Whittle Development Pty Ltd. The method to find the optimal schedule is decomposed into three steps. Following step 1 and 2, discover now the last step "Finding an optimal solution" and download the full article.

By D. Place, G. Whittle, and N. Baxter


STEP 3: FINDING THE OPTIMAL SOLUTION

The iterative approach described previously does not guarantee that the solution found is the optimal solution. Figure 5 illustrates the concept in which starting from different initial feasible schedules can lead to different local maximums, and that approach could return a local maximum instead of the optimal solution.


Figure 5 Global maximum vs local maximums




Using a method analogous to Monte-Carlo method, the procedure is repeated a large number of times starting from different initial feasible solutions. Like a mountain climber reaching the summit of one peak, only to find that they were surrounded by other higher points of the mountain, this method places many mountain climbers randomly through the range and the final solution is obtained from the climber who summits at the highest peak. Care has been taken to ensure that the full solution space is randomly sampled so that all local maximums are found. The iterations could be left to run for a long time, however the rate at which a better maximum is found decreases quickly, and the algorithm is stopped when the chance of finding a significantly better solution becomes largely improbable. (See Figure 6.)


Figure 6 Maximum NPV as a function of iteration number




CONCLUSION

The GEOVIA Whittle Advanced Simultaneous Optimization approach is able to find better outcomes compared to traditional methods. We have shown that the optimal mine plan for an open pit problem can be reduced to finding the optimal schedule through the simultaneous optimization of blends, stockpiles, and processing strategies within an iterative approach. The algorithm consists of iterating through the space of possible schedules to find the near optimal schedule that obeys the mining and processing constraints. In a comparison study (Cf. Internal document: Advanced SIMO vs Milawa and SPCO, G. Whittle, 2016), it has been demonstrated that the schedule out of Advanced SIMO optimization was worth a total of \$713 million or 41.4 percent additional NPV, whereas the legacy tools Milawa and SPCO showed only \$487 million or 28.4 percent. This study was performed on the Marvin Copper & Gold example used in the Money Mining Seminar from Whittle Consulting Pty Ltd. We expect to find greater returns as more and more users apply the Advanced SIMO to their mining problems.

▶ Download the full article​​​​​​​

Access all the best practices already available:

 

best practice ​​​​​​​GEOVIA Whittle