Tonnage control under maximum and instantaneous power constraints

In the mining industry, there are many applications for ore to be conveyed without posing a hazard to either personnel or equipment and without overflow or overload. In certain applications, there is a strong incentive to maximize conveyor throughput to minimize loading time, for example, to reduce hourly docking fees for boat loading.

A common installation is for multiple hopper-fed variable speed feeders to dump onto a long, fixedspeed transfer conveyor on which a belt scale measures throughput in tonnes per hour (t/h). The main constraint for high throughput is the power draw on all the conveyor motors.

The standard PID tonnage controller measures the belt scale weight and manipulates the feeder rates. The distance from the belt loading point to the scale, combined with belt speed (m/s), determines the transportation time lag. This lag is used to determine the difficulty of closed loop control; the longer the lag, the harder it will be to control tonnage, which means controller settings will have to be more conservative to minimize the risk of overloading the transfer conveyor belt. At startup, the time to reach the requested nominal tonnage will be directly proportional to the lag. For example, for a 700-metre distance between the belt loading point and the scale, at a speed of 3.5 m/s

(at nominal tonnage), the estimated lag will be over three minutes. In closed loop control, it would take the tonnage PID controller approximately 30 minutes to stabilize, which would most likely result in a conveyor belt overflowing with ore and an overloaded motor.

Variable feeder operation combined with this lag-induced slow reaction time results in a lower average flow than the calculated nominal rate.