Global demand for flat steel is on the rise. After years of decline and reductions in capacity, one executive calls the rebound “historic.” As a result, operators of hot strip mills are looking for solutions to help them improve their availability and throughput. The key to doing that is reducing downtime.
There are two kinds of downtime: planned (when you perform scheduled maintenance) and unplanned (when you fix equipment that’s failed). By monitoring the condition of individual assets, you can reduce both. If you know which equipment is healthy and which is failing, you can service only the machines that need it—which is essential when it comes to reducing planned maintenance stops. And early detection of faults helps maintenance teams schedule inspections, repairs or replacement well before these ailing assets fail, reducing unplanned downtime.
There are many ways to monitor asset health. In this article, I’m going to tell you why we believe our system, SAM4, is hands-down the best choice for the specific challenges of the hot strip mill. In short:
- SAM4 is incredibly accurate and provides weeks to months of advance notice.
- SAM4 is easy to use (and the only viable option in harsh conditions).
- SAM4 starts providing value from the start.
Across our installed steel industry base, SAM4 has consistently detected over 90 percent of developing faults up to 7 months before failure. And this is for both electrical and mechanical failures. Because SAM4 captures and analyzes electrical data, it will see electrical faults earlier than any other technology. Because it captures both current and voltage, it will see mechanical failures across the full drive train. From bearing damage to a bent shaft to broken rotor bars: SAM4 will detect these and other developing faults at least several weeks, and often several months, in advance, alerting maintenance teams well before the asset fails.
SAM4 will also detect destructive operating conditions. Consider a misaligned roll: while it might keep running without failure for a year, continuing to operate it that way will aggravate other failures down the line—as well as potentially affecting product quality. The insights into these conditions offer the opportunity to deal with them before they lead to more advanced failures or product discarding.
In addition to preventing unplanned downtime and loss of product, SAM4’s detection accuracy and operational insights enable the shift to a condition-based maintenance approach. Traditionally, the line is stopped every one to two weeks to replace a predetermined number of motors and rolls based on the calendar. A much more efficient approach is to only perform maintenance on rolls because they’ve developed a fault or when performance decreases, but not before.
Because SAM4 analyzes electrical signals from inside the motor control cabinet, there is no need to install any sensors directly on or even near the equipment in the field. In fact, this is the only viable way to collect condition monitoring data: the ambient temperatures of the HSM are so high that most sensors will simply burn up in a matter of days.
A second element is SAM4’s service model, which marries the best of man and machine. Continuous monitoring produces far more data than a human analyst can process in real time. But most of that data reflects healthy operation. SAM4 uses finely tuned machine-learning algorithms to automatically sift through the data to find anomalies. Our data scientists take over as needed from there, supported by a team of maintenance engineers and machine experts who communicate relevant, actionable information directly to the team on the HSM floor. This approach provides all the benefits of human expertise in a way that scales infinitely, in practical terms.
Third, SAM4 is easy to use because it monitors the most relevant components of the mill, and presents the data in the system of your choice. Whether it’s a motor, pump, press or roll: as long as it’s powered by an AC motor, SAM4 can monitor it. The comprehensive health, performance and energy consumption of connected assets is available in SAM4’s online dashboard, or integrated into your own monitoring system.
Sensor installation for the entire line can be completed within one month of signing the contract. Once installed, the data starts flowing, kicking off a 2- to 3-week learning cycle in which SAM4’s algorithms create a model of the machine’s baseline behavior that is finely tuned to the specific parameters of that machine and its process in the customer’s specific hot strip mill environment. After that, the system is fully trained and operational. But right from the start, SAM4 starts looking for known patterns that signal developing anomalies, based on its vast and growing library of “fingerprints of failure.”
As the demand for steel continues to rise, so do the demands we place on the availability of mill equipment. SAM4’s proven technology offers market-leading failure detection accuracy based on a plug & play system that is easy to install and scale. Implemented in a matter of weeks, you’ll likely see benefits before the year is over. Please reach out if you’d like to know more.