The business of bulk material handling comes with humongous operational data available; true, even before the advent of the term "Internet of Things." This data can provide incredible value for bulk handling sites but only with the right analytical approach. Learn more in "Making Bulk Material Handling Completely Smart with IoT."
Some would argue that this data and the ability to perform analysis on operations has been available for years, yet the multitude of software platforms installed at a bulk material handling site provides contradictory results.
Until recently, the integration of data from multiple sources has been a challenge. Once the data can be integrated and cleansed, there are tremendous use cases that can answer such questions as:
- Why is the bulk terminal performing the way it is?
- How do we measure the difference between the daily plan of operations and actual performance?
- Is it possible to identify the root causes of the difference between planned and actual performance?
Learn more in "Using Predictive Analytics Software to Enhance Throughput."
Maintaining Efficient Operating Plans
Bulk handling sites develop operating plans to organize work orders or task lists. These operating plans vary by duration: long term (monthly), short term (week or shorter), and daily. Activities at the site such as reclaiming, stacking, shiploading, car dumping, scheduled maintenance, and train loading are defined by a work order that includes these attributes:
- start time
- end time
- product and quality
- tonnage (or speed)
- equipment
- stockpile
- 'unload from' or 'load to'
No matter how optimal the plan is, there are always surprises: inclement weather, unscheduled equipment failure, train delays, ship delays, and other external factors that can cause deviation from the plan. Because operations within a site are so inter-dependent, delays will propagate upstream and downstream. With advanced analytics, we can answer such questions as:
- Could a minor downstream delay result in a major upstream delay or tonnage deviation?
- Could a minor upstream delay cause a significant impact on downstream operations?
By measuring the difference between actual and planned performance, we can determine the effectiveness of our operating plans. This can be achieved by looking at metrics such as adherence to schedule and throughput. This measure helps us understand the work order types that do work and work order types that always result in deviation. Knowledge of these deviations can be incorporated in planning.
Factors Affecting Terminal Capacity and Performance
Taking a typical OEE (overall equipment effectiveness) approach, a terminal’s performance at the system level is measured by tonnage, performance, and availability. Performance depends on certain inherent factors that don't change frequently as well as an efficient operating plan. The inherent factors include: the number of products coming through the terminal, equipment and stockpile mapping, and the train/ship schedule pattern. The capacity of the terminal depends on these factors plus an efficient operating plan.
A good (and optimal) plan is one that utilizes equipment efficiently and delivers to ships the correct tonnage of the correct product and quality within the correct time window. For example, a good plan will minimize the number of stockpiles used in an effort to minimize setup time. A good plan is where deadtonnes left in stockpiles are minimal. A good plan's inherent quality is that it would bring the yard to a steady state. Identifying the steady state of a yard is a challenging task and comes with experience of rigorous data analysis.
Configurable Adherence Settings
Part of measuring the adherence to a plan is determining to what degree a difference is significant. Is a 1 minute difference between actual train arrival and scheduled train arrival significant? Or is only a larger difference significant, say, 10 minutes? Is a deviation from plan of 1,000 tonnes of interest to operators and decision makers? Different organizations will vary in their answers based on specific individuals and corporate policy.
Verifying adherence to schedule and adherence to tonnage is straightforward, but verifying the assignment of equipment and stockpiles requires further reflection. Example assignments include:
- car dumper to train
- train-stacker-stockpile
- reclaimer-stockpile-shiploader
- shiploader and berth to ship
A work order specifies the equipment and stockpiles used. An operator could change one or both of these during the execution of the work order. Such deviations in work orders can indicate a lack of robustness in the plan. Pertinent system level parameters include:
- throughput (tonnes)
- number of work orders executed
- number of inbound tonnes
- setup time incurred due to changes in stockpile-equipment assignment
- throughput outloaded directly without going to stockpile
- inventory distribution of stockpiles at the conclusion of operating plan execution
- performance = (actual throughput rate / plan throughput rate)
- availability = (total time - unscheduled downtime - scheduled downtime) / total time
- adherence to maintenance window
The right analytics approach compares planned and actual performance and includes the best and worst performance in the history of the site using the system level parameters listed above.
Improving Performance with Eka's Analytics Solution
Eka's Commodity Analytics Cloud platform delivers industry-specific apps that cover P&L, procurement, risk, and supply chain. Unlike generic business intelligence tools, Eka's advanced analytics platform has been built with the specific needs of commodities companies in mind.
Part of Eka's Commodity Analytics Cloud platform is Eka's Bulk Terminal Performance app, which enables site operators to compare planned and actual performance of a site from various perspectives.