Monitoring and controlling hydrocyclones using vibration data

This invention relates to monitoring and control of mining equipment. More particularly, this invention relates to a system for monitoring vibration of hydrocyclones and controlling operation of the hydrocyclones based on vibration information.

Hydrocyclones are used in mining processes to separate two streams of material into overflow and underflow streams. Many process variables must be controlled to create a dynamic condition for excellent separation. It is not unusual for roping, plugging, or other undesirable conditions to occur due to materials or speeds or other conditions.

Classification is a critical component of any minerals milling circuit. Hydrocyclones efficiently separate mill discharge, returning coarse material to the mill and sending fine materials to downstream processing, such as flotation. Due to process abnormalities, hydrocyclones may plug or develop an underflow condition referred to as “roping.” Whenever a hydrocyclone is roping or plugged, large amounts of coarse, misplaced material may end up in hydrocyclone overflow and downstream float cells. Once roping occurs, the separation process becomes unsatisfactory. When plugging occurs, the process stops altogether, requiring excessive manpower to clean out and restart the process.

In prior systems, operators had no indication of roping before the problem occurred. To determine which of many hydrocyclones may be roping, an operator had to climb to the top of a group of hydrocyclone units (also referred to herein as a “pack” of hydrocyclones) to see which unit was not operating properly.

One attempt to address the roping detection problem involves attaching ultrasonic sensors to an underflow portion of the hydrocyclone. As the underflow approaches roping, the underflow hits progressively lower on the hydrocyclone splash skirt and eventually misses the splash skirt altogether. The corresponding changes in ultrasonic signal strength are processed to aid in recognition of roping and to generate an alarm when roping occurs. Unfortunately, ultrasonic energy in the 40 kHz frequency range is substantially attenuated by elastomeric liner materials, interfaces, gaps and boundaries, and distance, all of which are present in hydrocyclone applications. Furthermore, all of these attenuation conditions can change from time to time during normal operation.

What is needed, therefore, is a reliable method to detect the onset of roping, plugging, and other adverse conditions and to immediately notify the control room as to which hydrocyclone unit is experiencing a problem. In particular, a detection method is needed for monitoring both the overflow and the underflow conditions of a hydrocyclone using vibration analysis. The desired method performs vibration analysis in sonic frequency ranges, as opposed to the ultrasonic frequency ranges of previous applications.

The above and other needs are met by a hydrocyclone vibration measurement system that includes a set of upper vibration sensors, a set of lower vibration sensors, and a vibration analysis system. In a preferred embodiment, the set of upper vibration sensors include at least one vibration sensor attached to an overflow portion of each hydrocyclone in a hydrocyclone pack. The upper vibration sensors generate a plurality of analog overflow vibration signals. The set of lower vibration sensors include at least one vibration sensor attached to an underflow portion of each of the hydrocyclones in the hydrocyclone pack. The lower vibration sensors generate a plurality of analog underflow vibration signals.

The vibration analysis system includes an analog-to-digital converter and one or more overall vibration processing channels. The analog-to-digital converter samples the analog overflow vibration signals and underflow vibration signals at a fixed sampling rate and converts the analog overflow and underflow vibration signals into digital overflow and underflow vibration signals. The one or more overall vibration processing channels process the digital overflow and underflow vibration signals to generate overflow and underflow scalar vibration values representing overall vibration measured at the overflow and underflow portions of the hydrocyclones in the hydrocyclone pack.

In some embodiments, one or more band pass filters are provided to band pass filter the digital overflow and underflow vibration signals to generate band pass limited digital overflow and underflow vibration signals.

In some embodiments, the vibration analysis system is operable to interface with a distributed control system to provide control information to the distributed control system based at least in part on the overflow and underflow scalar vibration values.

In some embodiments, the upper vibration sensors are disposed such that a sensing axis of each upper vibration sensor is substantially perpendicular to a central flow axis of the hydrocyclone to which it is attached, and the lower vibration sensors are disposed such that a sensing axis of each lower vibration sensor is substantially perpendicular to the central flow axis of the hydrocyclone to which it is attached.

In some embodiments, the vibration analysis system includes a comparison module that compares one or more of:

the overflow scalar vibration values measured during a first time period to the underflow scalar vibration values measured during the first time period; the overflow scalar vibration values measured during the first time period to the overflow scalar vibration values measured during a second time period that is different from the first time period; the underflow scalar vibration values measured during the first time period to the underflow scalar vibration values measured during the second time period; the overflow scalar vibration values measured during the first time period to baseline overflow scalar vibration values measured during a time period when one or more of the hydrocyclones were operating in a normal condition; and the underflow scalar vibration values measured during the first time period to baseline underflow scalar vibration values measured during the time period when one or more of the hydrocyclones were operating in a normal condition.

Based on one or more of these comparisons, the comparison module generates control information that may be used in initiating control or making decisions about control of the hydrocyclone.

In some embodiments, the comparison module compares the overflow scalar vibration values in a frequency range of 0 kHz to 4 kHz measured during the first time period to baseline overflow scalar vibration values in the same frequency range measured during a time period when one or more of the hydrocyclones were operating in a normal condition.

In some embodiments, the comparison module generates control information for controlling operation of the hydrocyclone when energy levels of the overflow scalar vibration values in the frequency range of 0 kHz to 4 kHz measured during the first time period are higher than energy levels of the baseline overflow scalar vibration values in the same frequency range measured during a time period when one or more of the hydrocyclones were operating in a normal condition. The control information may comprise an alert message that informs an operator regarding a roping condition that may be occurring in the underflow from one or more of the hydrocyclones, or the control information may comprise a control signal that initiates a shutdown of one or more of the hydrocyclones.

In some embodiments, the comparison module compares the underflow scalar vibration values in a frequency range of 0.5 kHz to 2.5 kHz measured during the first time period to baseline underflow scalar vibration values in the same frequency range measured during a time period when one or more of the hydrocyclones were operating in a normal condition.

In some embodiments, the comparison module generates control information for controlling one or more of the hydrocyclones when energy levels of the underflow scalar vibration values in the frequency range of 0.5 kHz to 2.5 kHz measured during the first time period are higher than energy levels of the baseline underflow scalar vibration values in the same frequency range measured during a time period when one or more of the hydrocyclones were operating in a normal condition. The control information may comprise an alert message that informs an operator regarding a roping condition that may be occurring in the underflow from one or more of the hydrocyclones, or the control information comprises a control signal that initiates a shutdown of one or more of the hydrocyclones.

In some embodiments, the comparison module generates control information for controlling operation of one or more of the hydrocyclones when

energy levels of the underflow scalar vibration values measured during the first time period are less than energy levels of the overflow scalar vibration values measured during the first time period, or energy levels of the underflow scalar vibration values measured during the second time period which occurs after the first time period are greater than energy levels of the overflow scalar vibration values measured during the second time period.

Such a transition of the energy levels of the underflow scalar vibration values from less than to greater than the energy levels of the overflow scalar vibration values may indicate a plugged condition of one or more of the hydrocyclones. The control information may comprise an alert message that informs an operator regarding the plugged condition of one or more of the hydrocyclones, or the control information may comprise a control signal that initiates a shutdown of one or more of the hydrocyclones.

In another aspect, the invention provides a method for measuring and analyzing vibration associated with a hydrocyclone. A preferred embodiment of the method includes the following steps:

(a) sensing vibration associated with an overflow portion of the hydrocyclone using an upper vibration sensor attached to the overflow portion; (b) the upper vibration sensor generating an analog overflow vibration signal indicative of the vibration sensed in step (a); (c) sensing vibration associated with an underflow portion of the hydrocyclone using a lower vibration sensor attached to the underflow portion; (d) the lower vibration sensor generating an analog underflow vibration signal indicative of the vibration sensed in step (c); (e) sampling the analog overflow vibration signal and underflow vibration signal at a fixed sampling rate and converting the analog overflow and underflow vibration signals into digital overflow and underflow vibration signals; and (f) processing the digital overflow and underflow vibration signals to generate overflow and underflow scalar vibration values representing overall vibration measured at the overflow and underflow portions of the hydrocyclone.

In some embodiments, the method includes providing control information to a distributed control system based at least in part on the overflow and underflow scalar vibration values.

Other embodiments of the invention will become apparent by reference to the detailed description in conjunction with the figures, wherein elements are not to scale so as to more clearly show the details, wherein like reference numbers indicate like elements throughout the several views, and wherein:

FIG. 1 depicts a cutaway view of a hydrocyclone;

FIGS. 2A and 2B depict placement of sensors on a pack of hydrocyclones according to a preferred embodiment;

FIG. 3 depicts a vibration sensing and analysis system according to a preferred embodiment;

FIGS. 4A and 4B depict vibration spectra indicating the presence and absence of roping in the underflow of a hydrocyclone;

FIGS. 5A, 5B and 5C depict vibration waveforms indicating the presence and absence of roping in the underflow of a hydrocyclone;

FIGS. 6A and 6B depict vibration waveforms indicating the presence and absence of roping in the underflow of a hydrocyclone;

FIGS. 7A-7F depict vibration waveforms indicating the effects of plugging in a hydrocyclone; and

FIGS. 8A-8D depict vibration waveforms indicating the effects of roping in a hydrocyclone.

FIG. 1 depicts an exemplary hydrocyclone unit 10 , which may also be referred to herein as a cyclone. When the unit 10 is operating properly, the exiting slurry is dispersed from an underflow portion of the hydrocyclone in a cone shaped pattern as depicted within the “normal” region of the underflow end. When roping, occurs the exiting slurry in the underflow streams out in a tight pipe-shape flow pattern, which is shown in FIG. 1 as the “rope” region.

FIG. 2A depicts a cross section view of pipe flange sections of the overflow portions 10 a 1 - 10 h 1 of eight hydrocyclones 10 a - 10 h in a hydrocyclone pack 11 . An accelerometer 12 a - 12 h is attached to each of the overflow portions 10 a 1 - 10 h 1 . The eight accelerometers 12 a - 12 h are referred to herein as the upper set of accelerometers.

FIG. 2B depicts a cross section view of pipe flange sections of the underflow portions 10 a 2 - 10 h 2 of the same eight hydrocyclones 10 a - 10 h in the hydrocyclone pack 11 . An accelerometer 14 a - 14 h is attached to each of the underflow portions 10 a 2 - 10 h 2 . The eight accelerometers 14 a - 14 h are referred to herein as the lower set of accelerometers.

As shown in FIGS. 2A and 2B , the eight accelerometers in each set are spaced apart around the pack at an equal spacing of approximately 45 degrees. It will be appreciated that although the exemplary embodiment includes eight underflow accelerometers and eight overflow accelerometers, other numbers of accelerometers could be used, and other radial spacings could be implemented depending on the number of hydrocyclones in the pack. Thus, the invention is not limited by the number or angular placement of accelerometers.

In a preferred embodiment, the sensor mounting pads are welded to the hydrocyclone flanges so that when the accelerometers are mounted, they all measure radial vibration. This orientation is preferably the same for the underflow and overflow mounting locations on each hydrocyclone. Thus, the sensing axis of each accelerometer is preferably disposed perpendicular to the surface of the cylindrical flange section of the hydrocyclone as indicated by the arrows.

As shown in FIG. 3 , the accelerometers 12 a - 12 h and 14 a - 14 h are connected through junction boxes 16 a - 16 b to a vibration analysis system 18 , such as a CSI Model 6500, which is interfaced with a distributed control system (DCS) 20 and an operator display device 22 .

The vibration analysis system 18 includes an analog-to-digital converter (ADC) 24 that over-samples the vibration signals from the two sets of accelerometers, such as at 50.2 kHz, 104.4 kHz or 204.8 kHz, to generate digital vibration data. In a preferred embodiment, the digital vibration data from each accelerometer is processed in a separate parallel processing channel of the vibration analysis system 18 . To simplify FIG. 3 , only two processing channels are depicted—one for processing data from the upper (overflow) accelerometers and one for processing data from the lower (underflow) accelerometers.

In the overflow processing channel, an overall vibration processor 28 a generates overall vibration scalar values indicating overall vibration energy sensed by the overflow accelerometers. For some data analysis and plotting operations, a band pass filter 26 a filters the overall overflow vibration data to generate band pass limited data in the 0-4 kHz band. In the underflow processing channel, an overall vibration processor 28 b generates overall vibration scalar values indicating overall vibration energy sensed by the underflow accelerometers. For some data analysis and plotting operations, a band pass filter 26 b filters the overall underflow vibration data to generate band pass limited data in the 500-2500 Hz band.

While the 0-4 kHz and 500-2500 Hz frequency bands have been determined to be appropriate for the hydrocyclone pack examples described herein, other hydrocyclone packs may exhibit vibration energy in slightly different frequency ranges due to structural make-up, configuration, and the materials being processed. Thus, one skilled in the art will appreciate that inventions described herein are not limited to any particular embodiment involving any particular frequency ranges of interest.

The overall vibration values at the output of the overall vibration processors 28 a and 28 b may be converted to frequency spectrums based on Fast Fourier Transform (FFT) processing for plotting on the operator display 22 . Alternatively, the overall vibration values may be plotted in time domain format on the display 22 .

In a preferred embodiment, the vibration data at the output of the overall vibration processors 28 a and 28 b are provided to comparison module 30 that includes logic for comparing scalar vibration energy levels from the upper and lower sets of accelerometers to detect detrimental conditions, such as roping and plugging. In a preferred embodiment, control signals from the comparison module 30 are provided to the distributed control system 20 for use in controlling the operation of the hydrocyclone, as described in more detail hereinafter.

In some embodiments, the comparison module 30 compares baseline and historical vibration levels and compares vibration signals from the upper and lower accelerometer sets at the same and at different times to make decisions about the control of the processing conditions for hydrocyclone operations. Analysis of the vibration data provides information that adds insight and value to a plurality of process measurements coming from other sources. The vibration data is preferably combined with interdependent process inputs to define a reliable process control for hydrocyclones used in mining operations.

In some embodiments, selective decimation of the oversampled data provides significant information about material movements inside the hydrocyclone pack. In a preferred embodiment, the vibration analysis system 18 executes software for analyzing the oversampled data using methods described in U.S. Patent Publication No. 2014/0324367, entitled “Selective Decimation and Analysis of Oversampled Data,” (“the '367 Publication”) which is incorporated herein in its entirety by reference. The selective decimation methods are preferably applied to discern and characterize conditions within the hydrocyclones, including roping and plugging and precursors of those and other hydrocyclone processing conditions.

In some embodiments, selective decimation is implemented in the vibration analysis system 18 by collecting oversampled data, such as at a 50.2 kHz, 104.4 kHz or 204.8 kHz sampling rate for a block of time, such as for 10 seconds or 100 seconds or some other regular oversampled interval. The entire data block is analyzed as a sampling interval using the methods disclosed in the '367 Publication. Frequency analysis of information within each sampling interval preferably focuses on resonant frequencies of the hydrocyclone structure to detect changes in amplitude and to detect modulation as the process vortexes add and subtract and line up with those natural resonant frequencies.

In some embodiments, software executed by the vibration analysis system 18 analyzes selective decimation statistics for all or a portion of the sampling intervals. Attribute analysis of data within sampling intervals is particularly helpful for distinguishing between operating states and process changes. Such attribute analysis methods and methods for diagnosing material movement in a process by selective decimation are disclosed in the '367 Publication.

In some embodiments, a bump test is used to determine one or multiple resonant structural features in the vicinity of the overflow accelerometers 12 a - 12 h and the underflow accelerometers 14 a - 14 h for a given hydrocyclone. A bump test is a preferred technique because it is typically conducted using negative linear averaging such that when a series of impacts (such as twelve) is performed, the enormous amount of vibration energy from most other sources, such as process materials moving through the operating hydrocyclone, is subtracted leaving only the signature of the resonant response to the bump test impacts. Based on the resonant frequencies, damping and response information, the vibration analysis system 18 provides information to an operator about where to expect highest, lowest and intermediate multiplication of vibration frequency information detected by the accelerometers. Such information may be provided as a scalar value via a Modbus interface.

Table 1 below sets forth process characteristics and related physical responses, along with expected sensor responses and lag times.

As shown in the band pass limited vibration spectrum of FIG. 4A , vibration energy in the 1 kHz to about 4 kHz region appears to be the primary energy band of consideration for overflow analysis. In this three-dimensional data plot, the upper (rear) data set exhibits higher energy levels in the 1-4 kHz range, which is indicative of the presence of roping. The lower (front) data set exhibits normal baseline energy levels associated with the absence of roping.

As shown in the band pass limited vibration spectrum of FIG. 4B , about 0.5 kHz to about 2.5 kHz appears to be primary region of interest for the underflow analysis. In this three-dimensional data plot, the upper (rear) data set exhibits higher energy levels in the 0.5-2.5 kHz range, which is indicative of the presence of roping. The lower (front) data set exhibits normal baseline energy levels associated with the absence of roping.

The upper portion of FIG. 5A depicts an example of overall vibration based on data collected near the overflow of a hydrocyclone, and the lower portion depicts an example of vibration data collected near the underflow of the same hydrocyclone. After a first pump speed increase, the average overflow vibration energy level increases to about 27 mg and the average underflow vibration energy level increases to about 160 mg. After a second pump speed increase, the average overflow vibration energy level increases to about 40 mg, and the average underflow level hardly increases at all, although the peak-to-peak underflow amplitude level is somewhat greater due to the fact the process changed slightly. The underflow vibration energy pattern of FIG. 5A after the second speed increase is indicative of a normally operating (non-roping) hydrocyclone.

The upper portion of FIG. 5B depicts an example of overall vibration based on data collected near the overflow of a hydrocyclone, and the lower portion depicts an example of vibration data collected near the underflow of the same hydrocyclone. After the first pump speed increase, the average overflow vibration energy level increases to about 35 mg and the average underflow vibration energy level increases to about 148 mg. After the second pump speed increase, the overflow vibration energy level increases to about 51 mg and the underflow vibration energy level increases to about 217 mg. Note the significant peak-to-peak amplitude increase in the underflow vibration energy after the second pump speed increase. This is an indicator of roping in the hydrocyclone that began after the second speed increase. Based on the roping indication, the operator (or the DCS automatically) may open up one or more other available hydrocyclones in the pack to distribute flow so that the particular slurry is not as concentrated. Alternately, the operator (or the DCS automatically) may adjust the pump speed or adjust the density of the slurry going to the pack.

FIG. 6A depicts examples of overall vibration based on data collected near the overflows of three hydrocyclones in a pack. FIG. 6B depicts vibration data collected near the underflows of the same three hydrocyclones. The significant peak-to-peak amplitude fluctuations in the underflow and overflow vibration energy in the time range between the dashed lines are indicators of roping in hydrocyclone number 7 . No roping is indicated in the data from hydrocyclone numbers 2 and 6 . The increased vibration amplitude in the overflow of hydrocyclone number 2 is most likely due to condition changes in the flow rate. Generally, when one hydrocyclone experiences roping, other hydrocyclones can have changes in their flow rates.

FIGS. 7A-7F depict example plots of vibration data collected during a series of events related to the operation of six hydrocyclones in an eight hydrocyclone pack, one of which experiences a plugging event. In this example, the hydrocyclones (designated as 2 - 1 through 2 - 6 ) are controlled by a human operator rather than by control logic. The problematic hydrocyclone and the other hydrocyclones in the same pack are evaluated to differentiate the problematic hydrocyclone from the others. The plots depict vibration data collected from an overflow accelerometer (grey) and an underflow accelerometer (blue) on each hydrocyclone. The open or closed state of the hydrocyclone is depicted by the green trace. As expected, the plugging event has an impact on all the hydrocyclones within the pack.

FIG. 7B depicts overall vibration based on data collected for hydrocyclone 2 - 2 , which show the vibration levels from the underflow and overflow accelerometers “invert” relative to the readings from the other cyclones. This indicates that hydrocyclone 2 - 2 is plugged. FIG. 7A and FIGS. 7C-7F show jumps in the vibration levels measured by the underflow sensors (blue) on all of the neighbouring hydrocyclones, although none of the others “invert.” Based on the plugging indication, the operator (or the DCS automatically) may close the valve to the plugged hydrocyclone and open another hydrocyclone, if available. The plugged hydrocyclone would then need to be flushed or cleaned out. Generally, not all hydrocyclones in a pack are running at the same time. Therefore, extra hydrocyclones may be available to absorb the load generated when a problem occurs.

As indicated by FIG. 7D in this example, the operator mistakenly closed hydrocyclone 2 - 4 in response to the plugging condition of hydrocyclone 2 - 2 . A few minutes later, the operator reopened hydrocyclone 2 - 4 and then closed hydrocyclone 2 - 2 .

As indicated by FIGS. 7E and 7F , the underflow vibration levels of the normally operating hydrocyclone 2 - 5 and the closed hydrocyclone 2 - 6 are also affected by the plugged condition of hydrocyclone 2 - 2 .

FIGS. 8A-8D depict example plots of vibration data collected during a series of events related to the operation of four hydrocyclones in a pack ( 2 - 1 through 2 - 4 ), one of which experiences a roping event. As with the previous example, the vibration levels of the problematic hydrocyclone and the other hydrocyclones in the pack are evaluated. This is another example in which the hydrocyclones were controlled by a human operator. As shown in FIG. 8C , the overflow and underflow vibration levels for hydrocyclone 2 - 3 increase dramatically and their levels converge, clearly indicating a roping event. Vibration levels of the other two open hydrocyclones ( 2 - 1 and 2 - 4 ) show the effects of the roping of unit 2 - 3 , but those levels do not converge. An analysis of this data indicates conclusively that unit 2 - 3 is the problem unit in this example. Based on the roping indication, the operator (or the DCS automatically) may open up one or more other available hydrocyclones in the pack to distribute flow so that the particular slurry is not as concentrated. Alternately, the operator (or the DCS automatically) may adjust the pump speed or adjust the density of the slurry going to the pack.

The foregoing description of preferred embodiments for this invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obvious modifications or variations are possible in light of the above teachings. The embodiments are chosen and described in an effort to provide the best illustrations of the principles of the invention and its practical application, and to thereby enable one of ordinary skill in the art to utilize the invention in various embodiments and with various modifications as are suited to the particular use contemplated. All such modifications and variations are within the scope of the invention as determined by the appended claims when interpreted in accordance with the breadth to which they are fairly, legally, and equitably entitled.