Table of Contents
There are presently in use many grinding mill control schemes and concepts. In this case, one operating method of controlling the particle size of the cyclone overflows for proper metallurgical results and controlling the level of the cyclone feed sump and new feed rate to obtain lowest unit cost of finished (ground) material which is usually at the maximum grinding mill throughput. This system stresses simplicity of field instrumentation and control strategy. Both of these factors lead to better operator understanding and acceptance. The low capital investment of this system also makes it more attractive at the Buick Concentrator when compared to other more complicated systems utilizing mass flow measurement and a computer to calculate circulating load and infer particle size. Furthermore, it has the advantage of utilizing an actual measurement for control, rather than a difficult-to-maintain inferred value.
Plant Description
The mill treats an average of 1,800,000 TPY of ore containing an average of 8.5% lead and 3.5% zinc. The lead concentrate produced is 75% lead with a 96% lead recovery while the zinc concentrate produced is 57% zinc with a 88.5% zinc recovery. A copper concentrate is not produced.
The crushing plant includes a 42″ x 70″ Nordberg gyratory primary crusher, a 13″ x 84″ Allis-Chalmers secondary crusher and a 10″ x 84″ tertiary crusher. The tertiary crusher operates in closed circuit with two 8″‘ x, 16′ screens.
The grinding circuit consists of a 12′ x 16′ rod mill operating in open circuit feeding 20″ cyclone classifiers operating in closed circuit with a 14′ x 15’ ball mill. The cyclone overflow goes to lead rougher flotation which is followed by zinc rougher flotation. There are two stages of lead cleaning and three stages of zinc cleaning. Both lead and zinc concentrates are then thickened, filtered, and conveyed to a loadout building to be shipped to various smelters for refining.
For manual or automatic control of the grinding circuit, the ability to readily adjust certain variables is essential if operating objectives are to be achieved and system dynamics are to be maintained within the boundaries of successful operating capacities. To maintain continuous operation at desired throughput and slurry densities, it is essential that ore feed rate and grinding circuit water flow rates be readily controllable. To guarantee good control for most operating circumstances, all circuit dynamics and operating procedures need to be examined so relevant variables can be incorporated into the control scheme.
Generally, the tendency in size sensitive ore processing facilities is to grind finer than necessary, so product liberation during flotation can be maintained at a high level. However, this tendency to overgrind has the disadvantages of increasing the grinding cost per ton of ore milled, producing slimes and reducing grinding circuit throughput. Therefore, one of the prime factors to be considered in a size sensitive mineral liberation process should be the size of particles produced by the grinding section. Equally important is the need to maintain a minimal unit cost of ground material, which is often at a high throughput, with best operating conditions usually being a compromise between high throughput and fine grind. Generally, these two parameters cannot be the only considerations, since many other factors such as mine output, rod or ball mill capacity, flotation capacity, etc., have an important bearing on the control philosophy.
It is essential that cyclone overflow particle size be closely controlled, since flotation efficiency and maximum grinding circuit throughput are both dependent upon the final product size.
Commissioning
Final commissioning was completed during an intensive on-site effort initiated after satisfactory operation of each element of the control system had been achieved. Due to the nature of the tests conducted, the cooperation and involvement of process operators was necessary and their help and suggestions directly contributed to the eventual success of the project.
The primary steps executed to complete the commissioning were:
- Open loop responses of feed control loop and rod mill water control loop were recorded. As an example, the rod mill feed was manually increased by a known percentage of the full operating range, and rod mill feed as indicated by the belt scale was recorded before and after the change.
- Process gain delay and rise times were measured from recorded responses.
- Open loop response of the particle size control loop was recorded when sump water addition rate was altered.
- Loop gain, delay and rise times were measured.
Following completion of the control system commissioning, a series of tests were performed to determine whether grinding circuit performance and design objectives had been achieved. These tests served the dual purposes of completing the control system commissioning, and determining whether the control system possessed the required dynamic capability for control over the entire operating range, and would maintain control during normal operator adjustments, and other potential upsets.
Ideally, the circuit should be operated with the sump nearly full, because this situation represents maximum mill loading and probably maximum throughput which, at this facility can be regarded as best operating performance. However, under manual control, random circuit variations hamper the operator’s ability to maintain a high sump level, and to reduce-the possibility of overflowing the sump, he is likely to operate the circuit with a much lower sump level.
Since the flow versus total dynamic head characteristic of the cyclone feed pump is the major factor influencing the relationship between sump level and feed rate in this case, use of a pump with a “flatter” curve (i.e., a pump with larger changes in flow for changes in total dynamic head) could solve the problem. Another approach would be to use a variable speed pump to maintain sump level constant with pump speed controlling feed rate.
Ideally, changes in head grade which are accompanied by changes in ore hardness, would be automatically accommodated by changes in throughput, and operation at the highest sump level would assure maximum throughput. However, external “constraints” such as the limited flotation capacity mentioned above, and rod mill feed equipment limitations frequently require throughput to be controlled at lower levels. Therefore, the ability to select any sump level setpoint is a necessary operating requirement.
Each control loop was tested for response to a step change in one of the circuit variables, for the purpose of checking the speed of response and to check whether the loop had been optimally tuned. Ideally, the closed loop step response should have exhibited a slight initial overshoot, so that the new setpoint value could be attained as quickly as possible. As examples of closed loop responses.
Particle size control loop speed of response and ability to maintain tight control were tested by reducing the rod mill feed rate by a large step (16%) following by a large increase (26%). The cyclone overflow particle size did not deviate more than 1% from the 59% -200 mesh setpoint throughout the test.
Response to changes in sump level setpoint during normal operation was recorded. Originally, a response to one large setpoint change was considered, but owing to the initial excessive increases in rod mill feed rate, which could exceed the maximum feed for the mill, two smaller setpoint adjustments were used. By this means, the rod mill feed rate was limited to acceptable operating levels.
