Moving toward mining asset predictive maintenance
Ask mining executives about their top concerns and predictive maintenance lands in the top three. Due to an emerging ISO 55000 asset management standard, predictive maintenance has taken on new significance for mining companies. Despite their increased interest in predictive maintenance, most mining companies still operate on a preventative or planned maintenance strategy, resulting in unplanned downtime and preventable, costly repairs. Moving from preventive to predictive maintenance can enable mining companies to improve their overall equipment effectiveness (OEE) from 50-75% to 90% or more. But what is predictive maintenance, and how can mining companies embrace this strategy?
Exploiting data for asset insight
Predictive maintenance is an asset maintenance strategy that relies on real-time operational data to anticipate and predict when an asset will fail. The strategy enables companies to address asset problems before failures occur, reducing downtime and repair costs. With predictive maintenance, companies conduct asset maintenance only when necessary.
The key to implementing predictive maintenance is real-time operational data, which provides clues such as asset engine speed, exhaust temperature, or cylinder pressure, to predict asset failure. Implementing an enterprise-level operational data infrastructure can help mining companies reduce maintenance activity time by 20-50%, increase equipment availability by 10-20%, and decrease maintenance costs by 5-10%, according to Deloitte Analytics.
OSIsoft has helped many mining companies implement predictive maintenance through its technology, the PI System. The PI System captures real-time operational data coming from sensors, manufacturing equipment, and other devices, and transforms it into rich, real-time insight-connecting sensor-based data to systems and people. By using data to detect anomalies, the PI System enables companies to predict machine reliability.
Predictive maintenance in action
Syncrude, a major Canadian crude oil producer, implemented the PI System to employ predictive asset maintenance for 131 haul trucks and five shovels. This equipment is used over long distances around oil sand ore and in other harsh conditions that cause deterioration. Using the PI System, the company gained insight into the deterioration process and the potential for the vehicles to fail. As a result, the Syncrude saved CA$20 million in just one year.
Similarly, Barrick Gold, the world's second largest gold mining company, used the PI System at its Cortez mine in Nevada. The company glued Wi-Fi, battery-powered sensors onto its equipment to collect real-time data and vibration information. With this data, Barrick Gold can use analysis tools to predict future equipment failure. In one instance, after only 36 days of implementation, the company saved approximately $600,000 in potential downtime by identifying the potential failure of a scrubber.
Not just for asset maintenance
Real-time operational data contains a wealth of information that mining companies can use to generate major operational improvements. In addition to asset maintenance, mining operations can use the PI System to improve mining processes, environmental compliance, health and safety, quality and KPI, and reporting.
For example, Syncrude saved a lot of money in cost avoidance by improving reliability on his haul trucks and improved the safety of its people during the dumping procedures. Barrick Gold used the PI System in a sustainability project at its Goldstrike mine, where the technology helped the company reduce environmental deviations by 45% and fan trips by 61%. Australia's BHP implemented the technology to collect around five million data points (and growing) across 32 sites. This data has allowed the mining and metals company to deploy more than 120 solutions to improve operations. Last, ArcelorMittal Mining in Canada used the PI System to collect data throughout its supply chain. As a result, the company shipped an additional three million tons of product and grew revenue by $C120 million by identifying planning issues and making adjustments across the whole supply chain.
Moving toward predictive maintenance first requires access to real-time operational data presented in context. Context lets users know if that data has been captured during normal operations, maintenance, emergency stoppage, etc. The PI system allows you collect a nearly limitless number of data points, and presents them in context, making it easy to understand what's happening with your assets, implement further analysis, and perform predictive maintenance.