Similar to the airline industry, where the reliability approach was first defined, maintenance professionals in the mining industry often do not have the privilege of observing their equipment while it is operating. This is changing. Thanks to the introduction of wireless technology worldwide, mobile maintenance crews can observe how their equipment operates better than ever before. At Teck's Fording River operations, an open-pit coal mine in southeast British Columbia, the mobile maintenance crews are using wireless technology to proactively address some of the worst contributors to equipment failures.

If you boil maintenance reliability strategies down to their essence, the primary goal is to efficiently and effectively manage the P-F curve for all equipment failure modes. Whether this involves upfront reliability design, supply chain management, planning and scheduling, or predictive and preventative maintenance tasks, the focus is on the P-F interval with the goal of lowering the costs to the product user.

At Fording River, we are trying to move as far up the failure curve as possible to eliminate secondary damage that our highest impact failure modes incur. Following the introduction of wireless technology, we began using a software application called mobile equipment monitor developed by Honeywell Process Solutions, a thirdparty automation control solutions provider. The mobile equipment monitor sends equipment data to our desktops in real-time, across all equipment types, ranging from haul trucks to shovels and drills. Throughout the past eight months, we have been testing this software to see if it can help us move up the P-F curve.

For our haul truck fleets, one of the major causes of equipment downtime is frame cracking. With the freeze/thaw cycle that can occur in the Canadian Rocky Mountains almost year-round, the haul roads are continually heaving up and down. This creates large potholes that have the potential to cause significant damage to the trucks' frames and other major components. Repairs can take several days to complete and often require the removal of other components to perform the work. Effective work identification, planning, scheduling and execution can significantly lessen the impact these failures have on the bottom line. Eliminating these failures from happening in the first place - true predictive maintenance - would have the largest positive impact on our bottom line.

The root cause of impact failures is a combination of truck speed, payload and road conditions. If you remove any one of these three, the problem goes away. Obviously, stopping production isn't a viable option, so the focus has shifted to road conditions. With the real-time strut pressure data and overall payload readings we receive from the mobile monitor, combined with GPS coordinates, we have been able to successfully pinpoint the locations of our "bad actor" sections of roads. This has allowed us to intelligently provide our road maintenance crews with a priority list of which sections of haul roads are costing our operations the most lost production.

Some in the industry say real-time strut data and payload readings are not necessary to maintain haul roads, and as Figure 1 shows, they are correct. I would argue, however, that the benefit of strut data and payload readings is their value as informational tools that enable the operations foreman to effectively prioritize and dispatch road crews to strategically tackle daily road repairs.

Figure 1: Example of a haul road

The benefits to this approach are threefold. First, production is obvious, as you can see from the before and after information in Figure 2. In this example, haul trucks passing through a section of road after it was repaired were averaging almost five seconds faster travel times through the repaired section while maintaining a similar strut pressure reading. This means trucks can travel through the haul road section at a greater - but still safe - speed without causing damage to the frame and other components. Using time stamps that correlate to physical observations, the dashed lines in Figure 2 represent the start and end of the haul road in question. Throughout the course of a year and a large fleet of trucks, this results in millions of dollars of increased production.

Figure 2: Before and After cycle times/speeds/payload readings

Cycle times through the poor section of road were measured across several units, both before and after the repair work, in order to gain statistical significance. With poor road sections, operators have to slow down in advance of a pothole and then speed up again afterwards. In our test, data was collected from 100 metres before to 100 metres after an identified rough section to account for the loss in production.

Table 1: Cycle times through poor section of haul road

The second benefit is decreased maintenance costs. As the frame spikes are lessened by the smoother roads, the truck's frame and other components are not overstressed. This decreases welding time on maintenance days and increases the life of other major components, such as suspension, steering and the sensitive onboard electronics. The third benefit is improved operator health, safety and morale, as repeatedly driving over rough roads can take a physical toll on operators. Although instructed to always drive to conditions, operators can still be exposed to jarring while driving over rough roads. Reducing these occurrences would certainly benefit our operators' health and wellness.

Since the data and examples noted in this article were from a pilot test of the software, the road prioritization was done manually, but with considerations made for a production environment. The key challenge was how to properly manage and respond in a timely manner to the large amount of data collected. Recognizing this, the Honeywell system allows the user to build logic sets that run continuously against all streaming data. As you can see in Figure 3, we've developed an approach to autonomously monitor road conditions and prioritize based on lost production.

Figure 3: Automated data management logic

We start by eliminating all of the strut pressures that are less than 120% of the haul truck's accepted payload. Then we split the mine site into a large grid, made up of 100 metre X 100 metre sections. The payload readings that are more than the specifications are then clumped together by grid location. The last step is to apply a formula combining the frequency and severity to filter the worst road locations to the top of the road work "to-do" list. Road crews are then dispatched at the beginning of every 12-hour shift to work on the areas that are most impacting the site. The road criticality is then recalculated daily to ensure high-impact areas are addressed.

In summary, by adopting the strategy of monitoring strut data and payload readings to effectively prioritize and dispatch road crews to strategically tackle daily road repairs, we have seen quantifiable improvements across production, maintenance and operator health. This strategy will continue to be developedand enhanced at Fording River so we can expand upon these gains. As we move forward, continuous improvement should drive the application of this system into areas yet unforeseen.

David Hengen, CMRP, was a reliability engineer for Teck Resources and has since moved to Australia to work as a reliability engineer with a multinational mining company. In his current position, he collaborates with Operations and Maintenance personnel to reduce chronic failures and ensure the most cost-effective maintenance strategy is being utilized. He holds a B.Sc degree in Mechanical Engineering from the University of Alberta in Canada.