The Hidden Connections: How Controls, Power Systems, and Environment Affect Rotating Equipment Reliability

When rotating equipment starts vibrating excessively or experiencing repeated failures, it’s natural to suspect a mechanical problem. But the root cause isn’t always what you might expect. Electrical disturbances, control system behavior, and even changing environmental conditions can all play a role.

We talked with Six D Testing & Analysis Senior Consultant, Gene Smiley about some of these less obvious causes, why conventional troubleshooting can miss them, and how testing and analysis can help engineers get to the root of the problem.

When we talk about electromechanical interaction in rotating equipment, what exactly do we mean?

Rotating equipment doesn’t operate in a vacuum. It’s connected to structures, foundations, piping, valves, reservoirs, ductwork, lubrication systems, and sometimes other rotating equipment. It’s electrically connected to a three-phase power distribution system shared with other equipment, and it’s connected to a control system.

Why should engineers be concerned about it?

These systems operate in an environment where temperature, humidity, electromagnetic interference, and contamination often change. This can affect equipment behavior, introducing vibration, thermal issues, electrical disturbances, and other problems that can shorten equipment life or affect production quality.

Rotating equipment can perform well during initial testing but develop problems once in operation. How much of this can be attributed to interactions between the mechanical system and its controls?

Sometimes it’s the dominant factor. Other times, it’s irrelevant. The important thing is that equipment doesn’t necessarily operate under the same conditions in the field as it did during testing. Once installed, it’s interacting with the electrical supply, controls, driven equipment, and surrounding structures, each with its own dynamics.

For example, a pump may perform well during testing, but once connected to a control system, differences in response times or operating characteristics can introduce vibration or instability that wasn’t anticipated during design. That’s why a successful factory test doesn’t necessarily guarantee trouble-free operation in the field.

What are some of the more common control system issues that can contribute to excessive vibration, torsional oscillations, or premature component failure?

Control systems have their own dynamics and response times, and those need to be compatible with the equipment they’re controlling. Think about how a typical control system works. You start with a setpoint, established manually or automatically. A sensor measures some actual operating condition, perhaps taking 100 readings per second. The controller compares those measurements to the setpoint at regular intervals and calculates the difference, or error signal.

Based on that error, the controller determines whether an adjustment is needed. That might mean changing a valve position to regulate flow or adjusting a motor drive’s speed or torque.

The challenge is that the mechanical system doesn’t respond instantly. A sudden change in speed or torque can produce a fast or slow transient response, sometimes followed by either slow or rapid oscillations before the system settles.

Meanwhile, the controller continues trying to make adjustments. The sensor sampling rate, controller update rate, control-loop tuning, and mechanical response time all influence how the overall system behaves. If those dynamics aren’t properly coordinated, the result can be excessive vibration, torsional oscillations, or instability.

Can a control system designed to improve equipment performance actually make a mechanical problem worse?

Absolutely. In fact, that’s one of the more interesting challenges we encounter with rotating equipment. Consider a control system that’s continually adjusting motor speed to maintain a desired operating condition. If the mechanical system hasn’t fully responded to the previous adjustment before the controller makes another correction, the controller can end up chasing the error signal.

Instead of stabilizing the equipment, each correction may reinforce the previous oscillation. Under certain conditions, the oscillations can grow rather than decay, potentially creating damaging torque fluctuations and mechanical stresses.

And this isn’t just a theoretical concern. Poorly tuned or unstable control systems can cause serious mechanical damage, sometimes destroying the very equipment they were intended to protect. A controller may be performing exactly as programmed, but if its timing or tuning isn’t compatible with the equipment’s dynamics, it can enable the very instability it’s supposed to prevent.

When a rotating system begins experiencing vibration or repeated component failures, how can engineers determine whether the root cause is mechanical, electrical, control-related, or some combination of the three?

First, you need to understand the machine’s normal operating states and permitted transitions. Then review the control system data stored in the plant historian. Does the problem occur when a state change is commanded? Does vibration change as expected during a normal transition? Or does it consistently appear in one operating state? Those relationships can help determine whether the control system is involved.

From there, additional monitoring, sometimes using temporary sensors to collect higher-resolution measurements at specific locations, can help narrow down the root cause.

What role does field testing play in identifying electromechanical interaction problems that might not be apparent through conventional vibration monitoring or design simulations?

Conventional vibration sensors measure symptoms, but they don’t necessarily provide context. What operating state was the machine in? Did the vibration occur during a transition? What else was happening at the time?

Design simulations or digital twins can help, but their accuracy depends on what’s included. Does the model account for the control system? Are boundary conditions such as piping and foundations accurately represented?

Until those assumptions are validated, there’s still uncertainty. Field testing helps establish what’s actually happening and whether the model reflects real operating conditions.

When investigating these problems, what measurements and analysis techniques are most useful for understanding the relationship between the mechanical system and its controls?

It’s a sleuthing process, just like when you go to the doctor with a specific complaint.  What does he do? He collects context data. Beyond your immediate symptoms, twenty questions about timing and circumstances of when it started, does it hurt all the time, do you have symptom XYZ at the same time, what’s your blood pressure, is there a fever, what does your blood work look like, and so on.

Before collecting more symptom data, like vibration, ask what information is already available and whether it provides enough operating context to interpret existing vibration and other measurements of the symptoms. For example, the plant’s DCS (distributed control system) records operating states, process conditions, and commanded changes in its historian. That information can be correlated with detailed measurements from permanent or temporary vibration sensors. Smart motor drives, smart circuit breakers and even cameras can also provide valuable information about electrical power conditions and other changes in energy demand or the environment.

Once we have enough data, we can analyze time-frequency relationships using MATLAB or other tools. Spectrograms (vibration vs some 3rd axis operating parameter like speed or load), coherence, correlation, and phase analysis help us understand how different parts of the system interact and eliminate potential root causes. The key is understanding the operating context behind the measurements, not simply collecting more data.

Can you walk us through an example where what initially appeared to be a mechanical problem turned out to involve the control system or an unexpected interaction between components?

Sure. A few years ago, I worked on a freshwater supply system for a large city that sat along a river.  Several large pumps moved water through one-meter-diameter pipes to a treatment plant about four kilometers inland and several tens of meters higher in elevation. The pump drive motors were experiencing repeated winding failures, and we were called in to find out why.

Moving that much incompressible water through a long pipeline creates tremendous inertia and the potential for catastrophic pressure surges during startup. To reduce that risk, operators had implemented a soft-start strategy, increasing motor voltage in small steps over two to three minutes.

That strategy seemed reasonable, but our testing revealed an unexpected problem. Each voltage step was large enough to excite the motor-pump system’s torsional resonances. Essentially, the system was being struck with a torsional hammer hundreds of times during every startup. Those electromagnetic forces stressed the motor windings, causing them to physically move within their slots and dramatically shortening motor life.

So, a control strategy intended to protect the hydraulic system was actually contributing to motor failures. Without investigating what was happening during startup, it would have been easy to keep replacing damaged motors without ever addressing the real problem.

Are these issues something engineers can anticipate and address during design and commissioning, or are they typically discovered only after equipment is in service?

Many can be anticipated, but a design or digital twin is only as comprehensive as the experience and assumptions behind it. Think about a relatively simple motor-driven fan. You can model the motor and fan, but what other conditions matter? Contamination can accumulate on fan blades, increasing unbalance. Foundation stiffness can also change seasonally, from rigid frozen soil in winter to more flexible in other seasons.

Those changes can significantly affect natural frequencies, vibration amplification, mode shapes, stress distribution, and motor-fan alignment.

Can you give an example?

I’ve encountered “behavior not considered in the design model” multiple times even on very large systems. One involved a 120-inch diameter tunnel ventilation fan where an unusual axial umbrella mode of the fan cage on the shaft caused high axial vibration during commissioning.

Another involved a large forced-draft fan at a coal-fired generating station in the upper Midwest. In that case, large, out-of-phase lateral vibrations at the fan’s inboard and outboard bearing pedestals contributed to a significant diagonal foundation crack. That problem varied seasonally, meaning the heavy concrete foundation’s boundary conditions weren’t constant.

These examples illustrate an important limitation of modeling. A simulation can be technically accurate based on the information provided, but if changing environmental conditions or structural behavior aren’t accounted for, the predicted performance may be very different from what happens in the field.

What’s one of the biggest misconceptions engineers and equipment operators have when troubleshooting rotating equipment problems?

I think it would be jumping to the assumption that a problem must originate within the equipment itself. Engineers are often surprised when the root cause traces back to design assumptions, overlooked requirements, or something elsewhere in the system.

We’ve talked about mechanical examples involving fans, but electrical supply systems can create similar problems. For example, an unbalanced three-phase voltage supply can produce significant vibration at twice the line frequency. In a two-pole motor-driven system, that can easily be mistaken for mechanical misalignment. You could spend considerable time trying to correct a mechanical problem that actually originates in the electrical supply.

Another example involves shaft currents. Improper grounding, winding problems that create asymmetric magnetic fields, or high-frequency content from a variable-frequency drive can cause electrical currents to pass through motor bearings.

These currents can damage the bearing surfaces, creating a characteristic pattern called fluting. As the bearing surface damage progresses, vibration increases and bearing life decreases. Replacing the bearings may temporarily address the symptoms, but it won’t solve the problem if the underlying electrical condition remains.

What advice would you offer to help prevent these problems before they lead to costly failures?

The best advice I can offer is to involve different areas of expertise early in the process. Designers, digital twin developers, analysts, installers, operators, and maintenance personnel all bring valuable perspectives. Each sees the equipment from a slightly different angle, and those perspectives can help identify potential problems that might otherwise be overlooked.

It’s also important to challenge the original design assumptions and operating requirements. Are the conditions the equipment will experience consistent with what was anticipated? Have the electrical, mechanical, and control system dynamics been properly evaluated?

Addressing these questions during design and commissioning can help prevent equipment failures, lost production, and potential safety concerns.

Contact Us