When faced with a loss-in-weight feeder performance problem, the feeder itself will be the initial focus of troubleshooting scrutiny. But, what if the feeder checks out, yet problems persist?
In this case, the underlying causes must derive from elsewhere, whether within the operating environment, the equipment up- or downstream, or the process materials.
Loss-in-weight feeder performance and machine-status displays include input and output parameters that range from feed rate to motor-drive commands, span settings and alarm limits. The displays allow users to monitor and manage feeder operation and parameter trending points to external conditions that may limit performance.
Display and trending of this type illuminate the categories of possible problem causes. Diagnostics further sharpen skill sets that allow users to address problems expeditiously, reduce unplanned downtime and improve process efficiency.
Off on the horizon
Focusing on harder-to-diagnose performance-related problems, thoughtful troubleshooting is needed because the causes of less-than-optimal performance find refuge in odd, unexpected places.
Starting with the feeder, this article will first identify its operating principles. A loss-in-weight feeder operates by continuous weighing of the feeder itself, its hopper and the charge of material to be fed. Feeder weight declines as material discharges, and feeder speed is constantly adjusted to derive the desired gravimetric and subsequent weight loss rate.
This approach is accurate even at very low rates and provides complete material containment and material-handling flexibility. Any feed device may be used that is suited to its material and rate, including screw or vibratory feeders for solids, or pumps or valves for liquids.
Given how it works, a loss-in-weight feeder calls for two accommodations: periodic recharge of the supply hopper and isolation from the process environment. Accordingly, a loss-in-weight feeder performs a balancing act. It interacts with the process by receiving and discharging material, yet it remains isolated for maximum weigh accuracy.
It strikes this balance through combined measures in feeder design, application and implementation. They include mounting, connection, material supply and process environment considerations.
In operation, a loss-in-weight feeder continually supports a simplified control loop as it strives to drive mass flow error to zero. The time it takes to complete one loop represents the interval over which weight loss is measured and feeder speed adjusted. This simplified control loop is used as a template to identify and model where the causes of feeding problems arise.
The goal is to control flow rate so that trending of measured mass flow is a prime indicator of performance problems. However, the weight measurement itself, when trended, will help most to narrow the diagnostic possibilities. Once the feeder is eliminated as problem cause, the user should examine the surrounding process environment.
The operating environment
Likely the first alert to a feeding problem will be a control system alarm. When used properly, alarms guard against feeding woes, detecting violations of feeder motor speed limits, weight signal integrity and feed rate deviations. Alarms signify that something bad is happening but not what’s causing the problem.
Alarms are triggered by events that cross an alarm limit. Its cause may be a condition that lingers long enough for diagnosis, or it may be an isolated event that is not diagnosed. By capturing parameter change over time, trending can make it possible to correlate feeder events and conditions with those of the external process environment, even without a triggered alarm.
The most useful trending parameter from the process environment is measured weight. For example, consider a loss-in-weight feeder operating at a modest 6 kilograms/hour rate and a low accuracy of 1 percent. It must maintain its average per-second discharge rate between 1.665 and 1.668 grams with a weighing system that is the only support of a relatively massive operating assembly, which includes the feeder, hopper and material and is physically connected to the upstream and downstream.
Most modern loss-in-weight feeding systems are designed to combat performance-erosive influences faced in process environments. However, given the need to reliably discern exceedingly small weight changes in hostile process surroundings, trending analysis can overcome system white noise to reveal sources of process influence and allow problem resolution.
Each of the scenarios below suggest a different cause or narrows the set of possibilities. Today’s sophisticated feeder weighing systems usually use a low-pass filter to screen out most environmental contamination, but in practice, even with a problem-free feeder, net hopper weight trend lines will reveal small but acceptable bumps, noise and other irregularities.
Scenario 1: Isolated, short duration
A passing plant worker bumping a feeder or placing a cup of coffee on it can cause isolated, short-duration weight disturbances. Most loss-in-weight feeders are programmed to recognize and ignore a brief disturbance. They quickly determine if they should compensate for any resulting excess or shortfall in discharge. Harm inflicted by isolated, infrequent disturbances is typically not significant.
Scenario 2: Regular occurrence and duration
For feeders in a disturbance-prone environment, the cumulative effect of ongoing disturbances degrades overall feeder performance. If weight-trending displays this disturbance pattern, the likely cause is shock or vibration transmitted from nearby equipment or the plant structure. If multiple sources are in play, trace patterns may first appear random, but closer inspection should reveal its composite character. The frequency and duration of disturbance should help direct the user to the offending equipment, but the remedy will depend on the situation.
Scenario 3: Random occurrence and duration
When analysis finds disturbances that seem random in occurrence or duration, the troubleshooting is more difficult. Without direct evidence as to cause, the user must eliminate potential causes so only the actual cause remains. Is the feeder being buffeted by rogue air currents? Do disturbances still occur when the rest of the process is shut down? Is something odd going on inside the feeder? Are the feeder’s process connections okay? Fortunately, a continuing series of random disturbance is relatively rare.
Scenario 4: Correlated with refill
A loss-in-weight feeder often needs to return to "weigh-ready" as soon as possible after refill, allowing operation to resume. Refill disturbs feeder weight, so settling time is required to allow the scale to stabilize.
Several external factors can cause disturbance following refill. The refill system itself is not weighed and is considered part of the process environment. Any unintended post-refill leakage from the refill device, such as less-than-complete shutoff, will corrupt the feeder weight measurement until the leakage ceases. In one example, a refill device positioned some distance from its hopper due to limited headroom caused a post-refill weight disturbance because the length of the transit caused the flow to protractedly trail off. In this event, the user should check the refill device for proper operation and confirm positive shut-off.
Improper hopper venting is another potential cause of post-refill weight disturbance. Proper venting permits air displaced by incoming material to escape, and facilitates material de-aeration and settling. Venting may be passive or active.
If it is passive, air is left to exit on its own and is only impeded by the aperture provided and any resistance presented by any sock or filter.
An improperly sized vent or clogged filter delays complete venting, temporarily pressurizing the hopper and inducing stress on flexible connections. The worst case is if the feeder hopper pressure is enough to force material out the discharge. This condition leads to a perceived weight disturbance, feed rate error or abnormal motor speed trend line. To fix this, the user can simply clean or replace the filter and, if needed, increase vent size.
Active venting and dust collection use suction to encourage air exit and forces the compromise weighing. If active venting is too aggressive, low pressure in the hopper can induce stresses on flexible connections that will register as weight disturbances and transmit vibrations from the process environment.
One possible refill-related cause is internal to the feeder. A loss-in-weight feeder’s weighing system is not available to control feed rate during the brief refill operation because that point material is being quickly added to the feeder. To avoid discharge stream interruption, feeder speed is typically held constant during refill at the rpm measured just before refill, temporarily placing the feeder into a volumetric operating mode. After the refill and its ensuing settling delay, the feeder re-enters gravimetric operation, and speed can vary as required.
Depending on the process material’s compressibility, this traditional approach may or may not generate a sensed weight disturbance as the feeder returns to gravimetric operation. For a readily compressible material for which density changes appreciably as a function of headload, feeder speed just before refill is somewhat higher than it should be after refill when the material being fed was compressed due to the applied weight of newly added material.
As a result, when feeder speed is held constant at this higher speed, progressive overfeeding occurs during refill, and feeder speed is abruptly reduced when gravimetric operation resumes. However, some feeders avoid this shortcoming by memorizing the feeder’s recent weighing history and using that information to smoothly reduce feeder speed during the short refill period.
If the feeder does not have this capability and an immediate post-refill weight disturbance is seen, look at the feeder’s speed trend line to see if there is a significant difference between speed values just before entering refill and just after the apparent disturbance has passed. If a difference is seen and other possibilities have been eliminated, consult a feeder supplier.
Any disturbance over a long period will cause a feeder upset. Again, a feeder control system will ignore some brief disturbances, but a controller will act on any long-term disturbance outside certain limits and the feeder’s speed will change accordingly. Once again, weight trending allows the user to identify the problem.
Scenario 5: Constant
A weight measurement contaminated with extraneous input can seem intimidating. Once possible internal causes such as electronic noise, static, or binding of scale flexures or pivots are eliminated, it becomes increasingly clear that a direct route must be influencing the weighing environment. The diagnosis is often rather simple, and its solution is usually apparent.
In one common scenario, poor mounting occurs when feeders are installed without adequately considering the transmission of shock or vibration through the feeder base or other supports. In another, flexible inlet, venting and discharge connections or electrical wiring and cabling have a stiff or stressed installation.
A final potential cause is inletting or discharging to pressure or suction, or feeder purging. Because the feeder inlet and discharge experience a pressure differential, a net force is applied to it. Theoretically, if that force is constant, no problem occurs because a loss-in-weight feeder operates off of sensed differences in weight, not absolute weight.
However, even small variations in differential pressure can influence the feeder’s sensitive weighing system. Case-by-case consultation is required to resolve this type of contamination with pressure-balanced inlet, discharge and vent connections.
The material being fed is the only part of the external process that crosses the feeder’s defensive line. Common difficulties include bridging, arching and other flow-through-the-feeder problems such as material caking, clumping or buildup on the feed screw, tube or agitator if it is used.
These problems should be anticipated and addressed during testing or pre-op shakedown. However, actual process conditions change, and the character of the material can vary. These changes can cause old flow problems to return or new ones to emerge. While such problems will probably cause alarm conditions, monitoring feeder performance variables through display trending can help identify and diagnose emerging concerns.
Typical material related problems include arching, bridging and some other forms of hopper blockage. As the feeder empties material below the blockage, feed rate falls to zero, hopper net weight remains constant, feeder speed maxes out in its attempt to dose material that is no longer available, and weight loss per revolution drops to zero. In a trend line pattern typical of material buildup, weight loss per screw revolution declines more than expected over time as material builds up on the metering elements.
In response, feeder speed increases to compensate for the efficiency reduction. If buildup stabilizes and is not severe, feed rate and hopper weight remain on track. However, too much buildup will eventually trigger an alarm condition related to feeder speed or violation of weight loss per revolution limits. This condition could arise from any of several causes including a different material supplier, changes in storage or transport practices or mistakenly introducing the wrong material.
Feeder speed increases in step-like fashion to adjust for the sensed reduction in weight loss per revolution, and the opposite scenario would occur if density increased. If the change in material properties or handling characteristics is too great, the feeder may not be able to accommodate and an alarm would sound.
Troubleshooting is tough, especially when the cause of a problem may not lie where it manifests itself. Putting a feeder’s display trending capabilities to work can provide valuable clues to determining some of the more elusive causes of feeding performance problems.
Editor’s Note: This article was adapted from a Coperion K-Tron technical paper, "Using Feeder Trending as an Early Warning System."
John Winski is director of sales for Coperion K-Tron in Sewell, New Jersey. He has 28 years of experience at Coperion K-Tron and holds a degree in electronics engineering technology from Temple University.