Mining Goes Digital – Mueller et al. (Eds)
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
426
Generating pushbacks using direct block mine production
scheduling algorithm
C. Aras, K. Dagdelen & T. Johnson
Colorado School of Mines, Golden, CO, USA
ABSTRACT: Traditional mine production scheduling relies on generating pushbacks with
price parametrization and aggregating the blocks inside these pushbacks into benches to generate long term annual production schedules. The incremental fashion of obtaining pushbacks
fails to incorporate the operational requirements such as multi capacity, multi destinations,
blending requirements, truck hours and stockpiles. Since many production schedule plans
highly depend on the design of pushbacks, poor designs will prevent the schedules from
achieving a maximum NPV or even obtaining a feasible solution. In this paper, block by
block yearly schedules will be generated by solving the mine production scheduling problem
under the operational requirements and the resulting yearly schedules will be grouped into
phases for the pit design. Then, the schedules based on the traditionally generated phases will
be compared against the schedules obtained by direct block scheduling guided phase designs.
1 INTRODUCTION
Mine production can be defined as a complex system formed by its elements which are the
operational mining system, the blocks to be mined and functions with a goal of maximizing
the discounted cash flows of the mining operations. The boundary within which the system
functions is expressed by an ultimate pit which contains all the blocks that are profitable to
be mined considering only pit slope constraints. This assumes all the resource capacities are
unlimited, the mills are free from any kind of blending requirements and the time value of
money is ignored. The actual mine production is free of such unrealistic assumptions, therefore understanding the interactions between the mining blocks within this system becomes
significant to achieve the objective of maximizing the profit. The factors that delineate these
interactions between the blocks and the operational mining system can be commodity price,
mining cost, processing cost, grade of the blocks, mineral resource classification categories,
cycle times, material type, tonnages mined and processed in a year, total available truck hours,
certain blending requirements depending on the material type, multiple processing destinations with different ore recoveries, stockpiles, pit slope angles based on the rock formations,
cutoff grades and time value of money. Traditionally, some assumptions are made to reduce
the complexity of the system. One of these assumptions is that designating blocks to the most
profitable destination results in neglecting factors that influence the block interactions. The
decision based on the value at the destination will underestimate the potential profit. It can be
illustrated as follows. Let us assume there are two possible processing destinations for an ore
block, and that only one more ore block can be sent to a destination where the highest recovery can be achieved. If there are two ore blocks where the lower grade ore block is overlaying
the higher-grade ore block, sending the lower grade ore to a destination where less ounces will
be recovered will create an opportunity for a high-grade ore to be processed at a destination
where more ounces could be recovered. Another common assumption is made based on cycle
times. If there are multiple waste dumps to send a waste block, the operator makes a choice
based on the lowest cycle time. The fact is waste dumps have limited capacities, therefore if
a waste block is required to be mined, the optimizer should be able to send it to an available
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
426
Generating pushbacks using direct block mine production
scheduling algorithm
C. Aras, K. Dagdelen & T. Johnson
Colorado School of Mines, Golden, CO, USA
ABSTRACT: Traditional mine production scheduling relies on generating pushbacks with
price parametrization and aggregating the blocks inside these pushbacks into benches to generate long term annual production schedules. The incremental fashion of obtaining pushbacks
fails to incorporate the operational requirements such as multi capacity, multi destinations,
blending requirements, truck hours and stockpiles. Since many production schedule plans
highly depend on the design of pushbacks, poor designs will prevent the schedules from
achieving a maximum NPV or even obtaining a feasible solution. In this paper, block by
block yearly schedules will be generated by solving the mine production scheduling problem
under the operational requirements and the resulting yearly schedules will be grouped into
phases for the pit design. Then, the schedules based on the traditionally generated phases will
be compared against the schedules obtained by direct block scheduling guided phase designs.
1 INTRODUCTION
Mine production can be defined as a complex system formed by its elements which are the
operational mining system, the blocks to be mined and functions with a goal of maximizing
the discounted cash flows of the mining operations. The boundary within which the system
functions is expressed by an ultimate pit which contains all the blocks that are profitable to
be mined considering only pit slope constraints. This assumes all the resource capacities are
unlimited, the mills are free from any kind of blending requirements and the time value of
money is ignored. The actual mine production is free of such unrealistic assumptions, therefore understanding the interactions between the mining blocks within this system becomes
significant to achieve the objective of maximizing the profit. The factors that delineate these
interactions between the blocks and the operational mining system can be commodity price,
mining cost, processing cost, grade of the blocks, mineral resource classification categories,
cycle times, material type, tonnages mined and processed in a year, total available truck hours,
certain blending requirements depending on the material type, multiple processing destinations with different ore recoveries, stockpiles, pit slope angles based on the rock formations,
cutoff grades and time value of money. Traditionally, some assumptions are made to reduce
the complexity of the system. One of these assumptions is that designating blocks to the most
profitable destination results in neglecting factors that influence the block interactions. The
decision based on the value at the destination will underestimate the potential profit. It can be
illustrated as follows. Let us assume there are two possible processing destinations for an ore
block, and that only one more ore block can be sent to a destination where the highest recovery can be achieved. If there are two ore blocks where the lower grade ore block is overlaying
the higher-grade ore block, sending the lower grade ore to a destination where less ounces will
be recovered will create an opportunity for a high-grade ore to be processed at a destination
where more ounces could be recovered. Another common assumption is made based on cycle
times. If there are multiple waste dumps to send a waste block, the operator makes a choice
based on the lowest cycle time. The fact is waste dumps have limited capacities, therefore if
a waste block is required to be mined, the optimizer should be able to send it to an available