Condition monitoring systems today allow mine operators to synthesize and analyze huge amounts of real-time data on equipment health in a customized format. The challenge for companies, however, is managing all that information.
Asked what he considers the most helpful advance in technology in the last several years, Justin Harkness, surface maintenance planner at Newmont’s Carlin Trend in Nevada, answers: “Most definitely real-time condition monitoring.”
Although most major equipment manufacturers now offer machine health monitoring – from sensors to software – a number of third parties have entered the game. Some sell very specific systems – BMT WBM, for example, focuses its health monitoring on stress and vibration shovels and draglines. But others seek both to bring all the data from different vendors under one umbrella, and to streamline it for busy personnel. In doing so, they are exploring – and sometimes running up against – the outer limits of advanced data handling.
Newmont uses the onboard diagnostics system and fleet management software provided by Caterpillar to receive machine health alerts, view historical data, and make maintenance decisions based on real-time information across its operations. Numerous sensors integrated into mobile equipment report temperature, pressure and other data through CAT’s Vital Information Management System (VIMS), which sends that information wirelessly to its MineStar equipment management platform.
The biggest benefit, from Harkness’ point of view, is being able to catch emerging problems early enough to schedule maintenance downtime when people and parts are available. In an unplanned shutdown, not only does it take time for maintenance personnel to come by but the machine also cools down while it waits and has to be brought back to normal temperatures and loads to recreate its symptoms. With MineStar, some parameters can be viewed remotely while the machine is still in production.
For example, he says, “exhaust temperature differentials from left bank to right bank can indicate a few different things, and trending specific parameters to other machines in the same fleet can highlight subtle differences. It becomes even more interesting from time to time to compare these same operating parameters to same-model machines at different sites with different haul profiles.”
With a detailed picture of machine condition, Harkness notes, maintenance can also be tailored to the part. An engine, for example, might have an expected life of 17,000 hours but its current condition could warrant extending its life. A 13,000-hour engine, running in a questionable condition with a less-than-acceptable history, might be a candidate for replacement before the end of its expected life.
A diversity of suppliers
The global mining heavyweights of the world have generally adopted real-time health monitoring; the list includes Teck, Rio Tinto, BHP Billiton and Vale. Monitoring is not exclusively the domain of big companies – see Dundee Precious Metals’ Chelopech mine (p. 64), profiled in this issue. But according to Bart Winters, product director for Honeywell’s asset management solutions, the technology is still in its early adopter phases. “The thought leaders are the ones that are doing it the most, they are the ones that can see the value,” he says. “We see deployments at larger customers across multiple sites, but smaller sites are also starting to see the value and rolling out programs.”
Those who do adopt it have a fair amount of choice in how they ultimately view and use their data. The information collected by original equipment manufacturers’ (OEM) systems can move almost as easily through third-party systems as through the company’s proprietary platform, given a little cooperation between suppliers. “We have interfaces to all the major onboard health systems,” says Glen Trainor, vice-president of sales and marketing at Wenco International Mining Systems, whose products include ReadyLine asset health management software. “We have partnerships with third party devices as well. Wenco’s philosophy is that we are willing and able to interface to anything that’s on board.”
Wenco’s system shares broad similarities in data use with Honeywell Process Solutions’ Mobile Equipment Management (MEM) and Modular Mining’s MineCare. First, rules set by the company compare values to normal ranges or to one another and issue alerts as needed to an employee acting as fleet monitor. For instance, the temperatures on the left and right wheel motors might be too high, or they may differ from each other in a way that suggests a problem. If there is an issue, an alarm will pop up on the screen. Second, the fleet monitor can view trends for that parameter and request a real-time feed. And third, the system stores historical data in case the mine needs to investigate an event after the fact.
Winters says, from his experience, mines tend to want to minimize the amount of information distracting the machine operator in the cab and centralize workflow. He recalls a specific incident on site while having a conversation with a fleet monitor: “We were looking at the system and all of a sudden an alarm comes in that the driver is driving with the parking brake on. The fleet monitor says, ‘Give me a second, I have to call this guy on the radio.’ He gets on the radio, and he hears the alarm beeping in the truck, in the cab. He says, ‘Hey, Joe. You’re driving with the parking brake on.’ Joe says, ‘Thanks.’”
Asked whether the full potential of these systems is realized by their users, Winters replies in the negative: “These systems have to continuously be able to justify themselves. They always have to demonstrate where they are adding value.” To that end, he thinks it is important that MEM includes a record of equipment failures avoided, what he calls “documentation of saves.” When someone detects a condition using MEM, they can record a save post-investigation that is reviewed and approved by management, then used to make regular reports on just how much value the system has added.
Newmont does not currently measure cost savings specific to real-time monitoring in its maintenance department, according to Harkness. “It can be seen as a contributor to measured KPIs such as mean time between failure and availabilities, as well as planned versus unplanned maintenance percentages,” says Harkness. “When a machine problem is identified in a subtle or potential state of becoming a larger issue, and parts orders are accurately identified and complete with manpower set aside to perform the repair before it becomes an unexpected down, cost savings in maintenance are a direct result and that can be seen and measured.”
Merging the offline and online
Colin Donnelly, director of product management at DINGO, thinks that the existing real-time monitoring systems are missing something. “They haven’t created sensors to indicate how many wear particles or contaminants are circulating within your engine,” he points out. “People are neglecting the machines – and the parts of the machines – that can’t talk.”
Most mines collect oil samples and inspect magnetic plugs and filters to check their machinery, but they often neglect to make effective use of that information, in his view. Large amounts of data are collected only to languish in file cabinets or emails.
To help companies extract the value from this data and improve decision-making, DINGO created a third-party condition monitoring software product, Trakka, that takes that information – whether it comes from the mine or third-party vendors – and translates it into corrective maintenance actions that feed into an email-like inbox checked every day. “We’re just repackaging the data into a platform that shows people if you pay attention, there’s a large amount of money at the other end,” says Donnelly. DINGO’s services include staff support who analyze the data and recommend actions.
He says he believes DINGO’s Trakka system stands out because it includes a process for following up on maintenance activity to make sure it was successful. “If the task didn’t physically improve the condition, then it was either the wrong task or it was done incorrectly, and we need to do something else,” he explains. “We actually keep those things open a lot longer, because we want confirmation that the task resolved the issue and the equipment returned to a normal operating state. That’s where I think it complements some of the enterprise resource planning systems that pretty much every customer has in place to manage their maintenance work.”
Donnelly says it was that functionality that led a Canadian customer to combine its use of Trakka and MEM. The Honeywell software displays real-time equipment health data, but some of the same information can be viewed in a Trakka window using the MEM-assigned tags for each value.
Understanding big data
Vendors agree they need to work on taking more of the load off mine operators. There is widespread agreement that the growth of “big data” – so much information it demands more sophisticated analytical tools – is driving software development in this field.
Trainor says that predictive modelling, or the use of information from multiple sources to develop patterns that predict maintenance needs, will develop further in future. Wenco is trialling more predictive systems in two mines in Australia this year and next.
At Honeywell, Winters says he envisions defining similar types of operation and comparing them across a fleet; for example, taking the exhaust gas temperature when a truck is hauling uphill underload. Even further in the future, Winters would like to see more advanced diagnostics and, ultimately, machine learning, in which observations made during investigations could be automatically incorporated into the diagnostic process the next time a similar issue crops up.
Third-party developers – and, according to them, users – would have an easier time of it if data came out in less proprietary formats. Wenco started stripping its systems of all proprietary aspects 15 years ago, and today its customers use its public API to plug monitoring data into enterprise-level software if they so wish.
“It’s problematic today,” says Winters. “The operating companies are pushing the OEMs to standardize on the data that’s coming off so that they can deploy additional tools to make use of that data.” That effort is being led by the Global Mining Standards and Guidelines Group’s Onboard Technology and Connectivity Working Group. In the meantime, third-party providers can cite numerous happy customers who, with typical mining ingenuity, are willing to reconcile the seemingly disparate for the sake of greater productivity.
By Eavan Moore
See more at: http://www.cim.org/en/Publications-and-Technical-Resources/Publications/CIM-Magazine/2014/September/technology/Come-together#sthash.U8IALJhP.dpuf