Page 96 - Advances In Productive, Safe, and Responsible Coal Mining

80 Advances in Productive, Safe, and Responsible Coal Mining

when η¼1. Given that all panels have the same barrier pillar width, mining four
panels leaves 240ft of coal in barrier pillars compared with 180ft of coal in barrier
pillars when mining three panels. In addition, the solution when η¼1 leaves out
310ft of coal at the end of the strip compared with 410ft in all other solutions. Note
that coal left at the end of a strip will likely be mined, making the last panel wider than
the recommended width. However, even then, the practice will be suboptimal as it will
lead to mining a wider panel than anticipated. Thus, solutions that leave very little coal
are to be preferred, which is how the model behaves.
It can also be observed that solutions mine strips with more differing panel widths
when recovery is the predominant factor (i.e., relatively higher values of η). These solu-
tions deemphasize production rate twice. First, these solutions tend to use lower produc-
ing panels more, which will lower production rates during mining. Second, the
production rate during mining will be even lower because the mine has to switch panel
sizes frequently, which negates efficiency gains from repetition. The ability to recognize
this tendency for lower production rate mine plans when production rate is completely
removed from consideration is a benefit of the proposed dual optimization approach.

5.4 Conclusions and recommendations

This chapter presented an approach to optimize coal recovery and production rate as a
function of panel dimensions. Discrete-event simulation (DES) is first used to estab-
lish the relationship between production rate and panel width. An optimization model
is then formulated that maximizes coal recovery and production rate. The coal recov-
ery problem was shown to be similar to the cutting stock problem and modeled by
adapting the cutting stock problem. The model used production rate indexes for panel
widths derived from DES results. The optimization problem is solved using CPLEX’s
integer programming solver, which is based on the branch-and-cut algorithm. Time
study data from an underground room-and-pillar coal mine in Southern Illinois, the
United States, are presented to illustrate how to determine the relationship between
production rate and panel width using the DES model. Having used those production
rates to determine production indexes, an instance of the optimization problem involv-
ing 10 mining strips is solved as a case study. Optimal panel widths used in such strips
are examined when considering all six panel widths from the same room-and-pillar
coal mine modeled in the DES case study. The case study results show that a dual
optimization approach that maximizes recovery and production rate is beneficial
because, in addition to accounting for management’s dual objectives, it leads to solu-
tions with more consistent mine plans (i.e., fewer panel widths are used in the mine
plan). Also, results show that an analyst should carefully choose the ratio in the model
that specifies the relative importance of recovery and production rate to obtain results
that are most useful.
The authors recommend that future research explore how to solve the optimization
problem using the branch-and-price algorithm. This will overcome the limitation of
the current branch-and-cut algorithm, which is computationally expensive when there
are many strips.