The stockyard plays a vital role in the mine-to-ship supply chain. Operations at a stockyard are unique, complex, and dynamic – efficiencies or the lack thereof at this crucial stage can make an enormous difference to a company’s bottom line. Central to this is the challenge of overcoming what is sometimes called the short term scheduling problem.
The overarching goal of stockyard operations is simple: load ships with the tonnage requested and at the right quality specification, and do this as quickly as possible. But like many things in life this is easier said than done. In fact, the reality of ensuring efficient usage of resources and space within a stockyard to achieve this goal can be a logistical nightmare.
Meeting Stockyard Challenges to Maximise Throughput
To start with, demand and quality specifications for each outbound product are unique, and each ship needs to be loaded with the right products at the right quality. Adding to the complexity of this is the fact that the contents of trains coming from a mine differ in quality, and the content is only known accurately post-departure and not before. One option here is to identify trains that will result in the right quality when stacked together, and arrange the stacking process like so. Another is to blend material from different stockpiles during reclaiming, but this means more reclaimers with a lower average throughput. Either way, the scheduler will want to stack inbound material in such a way that the difference between the average quality and target quality is minimised at any given time.
»»» Download brochure on Eka's bulk handling for mining platform. «««
This challenge is made yet more complex by the fact that train and ship schedules are a given, and relatively inflexible. Trains need to be able to quickly dump their cargo so that rolling stock of empty trains is available at mines for the next load. Ships have a very limited time window during which they can be loaded.
Of course, the stockyard itself has a finite capacity, with limited resources. These input-output gymnastics must be achieved via a limited number of car-dumpers, shiploaders, reclaimers, stackers, ship berths, stockpiles and so on, all within a tight timeframe to ensure high throughput. Then there are the wider constraints: for instance, operating constraints such as the time taken to move material from stockpile A to stockpile B, or safe operating limits of equipment all need to be factored in. Learn more in "Making Bulk Material Handling Completely Smart with IoT."
It’s a jigsaw puzzle of the highest order: but unlike a jigsaw it is dynamic – the state of affairs on the ground and the objectives are constantly moving. Ships can be delayed. Shiploading windows are determined based on tonnage, ship schedule, and tide forecast and it is imperative shiploading windows are not missed. Trains may arrive late, or with the wrong quality. Equipment breaks down. The potential variables are too long to list. All these variables mean that decisions made within any planning and scheduling framework will frequently need to be revisited at short-notice, and arrangements recalibrated to accommodate the disturbance in such a way that the impact on throughput and cost is minimised. This is made more difficult still by the fact that decisions are inter-related in complex ways. This, in short, is the ‘short term scheduling problem.’
Solving the Trilemma
How to tackle such a fiendish problem? There is a trilemma here: the ideal aim is to maximise accuracy and throughput while minimising cost. But at first glance focusing on any two factors will hurt the third. You can minimise cost and maximise throughput, but you’ll sacrifice accuracy. You can maximise accuracy while minimising cost, but at the expense of throughput. And you can simultaneously maximise accuracy and throughput by throwing traditional resources at the problem (more manpower, more machinery) – but then your costs balloon. All three are unacceptable from a business point of view.
The magic sauce for unravelling the trilemma and squaring the circle lies – as with many of these sorts of problems – in technology. This becomes plain when one considers the deeply interrelated nature of decisions in short-term scheduling alongside the number of decision combinations involved in planning and optimising tasks. To illustrate, take a 24 hour period and imagine there are 2 ships, 25 trains a day, 24 stockpiles, 3 stackers, 3 reclaimers, 2 shiploaders, 2 car-dumpers, and 5 products. That makes for a gargantuan 1,036,800 potential stacking and reclaiming decisions. Scale-wise, that’s a job for a computer, not a human.
More specifically, it’s a job for modern, sophisticated real-time stockyard optimisation systems. These use successive linear programming based on the ‘Mixed Integer Programming’ approach. In English, this means that – taking all of the discussed factors as inputs along with stockyard layout – they are able to optimise schedules, resources and stacking/reclaiming methods to maximise throughput at the same time as maximising quality on an automated basis (at a small fraction of the time and cost that doing it manually or through less powerful systems would require). A good system will be able to produce an optimal stacker, reclaimer, car-dumper and shiploader schedule in under 30 seconds, and will continue to do so at hourly intervals as conditions change.
Leveraging the Benefits
Once such a system is in place, it can deliver myriad business benefits. It provides automated planning for the next 24 hour period – a godsend given that the quality of commodities on inbound trains next day and the rest of the week is an unknown. A good system will optimise the stacking and reclaiming approach in such a way that room is made to accommodate the next day’s trains. It will also allow for the automatic creation of an alternative plan should equipment break down, and incorporate this possibility into planning (e.g. stacking product in two stockpiles with two reclaimers to minimise risk of disruption in the event of a breakdown). And it will go a long way to helping keep inventory below stockyard capacity: high inventory has a major negative effect on throughput.
Optimisation aside, another major business benefit is that such systems allow for users to quickly run ‘what if’ scenarios. Users can select for different stacking and reclaiming methods, and change a whole range of variables from equipment available to ship-loading speed and so on. They can then compare and contrast the results. This makes for a powerful tool when it comes to longer-term planning. A stockyard operator will be able to quickly see, for instance, the impact of a forecast reduction in equipment available in the yard.
Ultimately, however, the only benefit that counts at the end of the day is the benefit to the bottom line. And a good stockyard optimisation system can deliver this in spades. Even what appear to be minute improvements in efficiency and scheduling can translate into huge savings. Given that a reclaimer reclaims 10,000 tons of iron ore per hour; adding just 5 minutes more productivity per machine means enhancing throughput by 800 tons per machine per day. Assuming an iron price of $50 per ton, which makes for a daily generation of $40K per machine – a stockyard optimisation system will very quickly pay for itself.
Eka’s Commodity Analytics Cloud
Eka's Commodity Analytics Cloud is an advanced analytics solution that brings commodity specific analytics to all business users. The platform delivers industry-specific apps that cover supply chain, P&L, procurement, and risk. Unlike generic business intelligence tools, Eka's advanced analytics platform has been built with the specific needs of commodities companies in mind.
