Metal Detectors

Metal Detection technology finds the needles in its haystack

Jun 06, 2018

When faced with a major challenge, Syncrude’s Research and Development (R&D) team leaves no stone unturned in its search for solutions.

This was the case about three years ago, when Mildred Lake Mine Replacement (MLMR) project members approached Research Engineer Darcy Daugela about metal detectors on the facility's conveyor system. Simply put, they weren’t working well and key hydrotransport equipment was being damaged by loose metal, also known as tramp metal.

 

“We were in this situation where we have a lot invested in this system, world experts haven’t found a solution, and our own expert team has been working on the problem for months and is running out of options,” says Darcy. “Of course, Syncrude has never backed down from a challenge like that.”

 

When it first opened in 1997, the North Mine delivered oil sand ore to primary crushers where rocks, large oil sand lumps and stray objects were rejected or screened out before going to slurry preparation and hydrotransport to the Extraction facility. However, the system also shed around three per cent of the usable ore.

 

The reject system was eliminated as technology in the North Mine evolved. A second set of crushers called ‘sizers’ was added to reduce lump size for slurry preparation. This change captured the lost ore and was supposed to end the reliability and reprocessing issues associated with the rejects system. But the new system relies on metal detection and rejection technology to be successful.

 

That’s because the sizers are designed to crush rock and ore, not metal. Tramp metal is a reality in the mining environment where items like plates and bucket teeth can wear off and fall into the ore feed. Significant damage can occur when a large metal item gets caught in the sizers or passes through to the hydrotransport line.

 

The commercial metal detection technology Syncrude used at that time involved a transmitter positioned above the conveyor belt that sends out an electromagnetic pulse. Any metal near this pulse induces a weak electromagnetic echo that is detected by a receiver under the belt and sounds an alarm. It automatically opens a gate to reject the metal, then resets the system.

 

Unfortunately, all metal near the transmitter, including the conveyor structure, produces an echo. This creates background magnetic noise and is a major challenge for modern mining metal detectors. The high background noise leads to many false alarms.

 

Initial tests found that the background noise was drowning the signal. The cost of false alarms, repairs to the crushers and conveyor belt, and system downtime was high. Darcy’s team looked at what had been done in the global metal detection field, going back to the Second World War and the use of magnetometers to detect German U-boats in open water. Magnetometers detect small changes in the earth’s magnetic field, called anomalies. This is referred to as Magnetic Anomaly Detection, or MAD. 

 

“We used this idea and put magnetometers under the conveyor belt,” says Darcy. “We found the ore and stationary metal did not change the magnetic field, but moving metal on the conveyor creates a signal.”

 

It was a start. The team also discovered that magnetic fields can change due to passing vehicles, items close by, or rotating parts on the conveyor. They placed two sensors side by side and found that changes in the magnetic field affected both sensors at the same time. Then they put one sensor in front of the other and found the sensors picked up changes caused by metal on the moving conveyor rather than other objects nearby. They had a solution.

 

Darcy’s pitch to design and build a full-scale prototype was approved immediately. Six months after being tasked with the problem, the team had a prototype collecting data.

 

That prototype is now the primary means of metal detection for MLMR, and the return on investment is significant. Prior to its introduction there were 30 tramp metal incidents per month with an average of 2.5 outage hours per incident. Today, with the new technology in place along with initiatives to reduce metal in the feed and increase the life of the crusher teeth, the system is experiencing less than three incidents per month.

 

The new Syncrude metal detection technology has opened the door to many exciting future applications. It not only detects metal in a high metal environment, it can also detect a metal object moving on the other side of a thick metal wall.

 

Proof once again that when necessity calls, Syncrude’s R&D team is up to the challenge.