Advancing Together With Barrick Gold

Mining Barrick predicts the future… of mine equipment maintenance

New Asset Health tool receives data from several sources, including fleet management software, equipment work order information, oil analysis data, and sensors installed on equipment

Cody Bakker, a Reliability Engineer at Barrick’s Cortez mine in northern Nevada, had recently participated in his first meeting for an Asset Health tool that would let him focus on what he does best: making sure mining equipment purrs along the way it’s supposed to. After finishing his shift, Bakker wanted to learn more about this tool, which monitors mining equipment and predicts when it will require maintenance. So he stopped by the C0deM1ne with one of his colleagues to get a demonstration.

“We figured, ‘Let’s swing in’ because we wanted to get access to it, learn more about it and how to use it,” Bakker says. “When we got there, we realized it wasn’t fully functional yet, but since we were there, we gave a lot of feedback on what we wanted and what we’d eventually like it to do.”

Instead of spending hours gathering data, it’s all going to be fed into one dashboard

Currently, Bakker has to track down information on maintenance work performed on various equipment parts, such as wheel motors that power the movement of the massive CAT795 haul trucks used at Cortez. He spends hours verifying that the information is accurate and then more time preparing spreadsheets to input that data. He also needs to develop summary reports for the maintenance department.

The Asset Health tool will make Bakker’s job much easier because all of this information will be automatically fed directly into the tool from several different data sources, including fleet management software, equipment work order information, oil analysis data and sensors installed on equipment. The tool will analyze this information, allowing it to predict when parts are likely to fail by drawing on a library of common causes for those failures.

This will reduce the amount of time that technicians and their supervisors will have to troubleshoot, ensure that replacement parts are ordered in advance of a failure and help get equipment back into action quicker.

“I think it’s great what they’re coming out with so far,” Bakker says. “Instead of spending hours gathering data it’s all going to be fed into one dashboard where you can drill down into information and be able to see trends and everything at a quick glance.”
 

Members of the Asset Health tool user council—a group of frontline employees and supervisors who test and make suggestions to improve digital solutions—share their thoughts on what they’ve seen of this tool so far.


Health alerts

Eventually, the tool will help maintenance teams produce custom reports and set alarms that will give them advance warning when a motor is running hotter than normal, for example. Technicians will be able to compare how parts perform against identical equipment on site and across the company to see if any are behaving differently and what that might mean in terms of proactive maintenance.

“Think about how a person’s health would be monitored: you might collect readings on health indicators such as their blood pressure, cholesterol levels, weight, and temperature,” says Jami Dwyer, the Product Owner for the Asset Health project. “The Asset Health tool will similarly allow us to monitor our truck’s health by analyzing measurements such as oil samples, filter pressures, and temperatures from the truck in near real-time. If the truck is healthy, we can look at more background data to extend the life of a particular component even further.”

Maintenance teams will also be able to monitor operator performance to ensure operators aren’t abusing equipment or running it outside of acceptable operating ranges.

The Asset Health project team has targeted some of the largest maintenance costs at a mine, starting with haul trucks and their engines. On average, haul truck engines are changed every 18,000 hours, or roughly every three years. They cost upwards of $360,000 each, so the goal is to extend their life as long as possible, but if they fail, they can do a lot of collateral damage to other systems, such as the engine block and the radiator. This could more than double the repair costs and cause significant downtime, Dwyer says.


From predictive to prescriptive

The real test of the tool’s success will be how much maintenance work can be shifted from unplanned to planned. At Cortez, the balance is approximately 50/50, but Dwyer and her team are aiming for the stars in hopes of hitting the moon. They ultimately want to achieve an 80/20 split to align with industry best practice. Unplanned maintenance work causes delays and increases operational costs.
“The tool aims to go beyond predictive to prescriptive maintenance,” Dwyer says. “We’re going to build a library of recommendations for what to do about specific problems so our technicians can fix them as quickly and efficiently as possible.”

Currently, the Asset Health project team is feeding terabytes of information into the consolidated data platform to generate dashboards for maintenance teams to analyze and make better, faster decisions. The consolidated data platform acts as a conduit for information across Barrick’s myriad databases, some of which feed into the Asset Health tool.

“It’s not just a matter of making sure that the data is accurate, or chasing and putting out fires but it’s also looking ahead to the future and saying, ‘Based on the type of work we’ve been performing on all of this, do we believe we can meet our goals?’ ” Bakker says.